. EEG?

x . .afé$«

. 31:91,..u1
: . :3.»
I‘ll“... Adah «Hawk-(mu...
”a 9.3%. a «a
gﬁﬁz a? , .
mu sunny“? 3
1... hmannz .. s
.
.

.2

3m. ,
. V...)
,1 P0was...

(I:

53.5..

. :

pl .ﬁmmlrn‘
thfde1 9!

.m‘ .. . .t I!

. ;
let!-
’5.
2

2 331219;:

oﬂur {All 'I‘:

-. ’1

x V

6.1.“.- .x. m1.“

{$513 .13»? _
.s t V n u

‘ a. human. km”... :.

"mute. if

.453.

.. . 3%,...
.33.. 1.2113}:
x. .

g

71 .
‘ . V i‘
“5”, mks}?

If t
1 y .23..

v 1.. t: at: I
3 $£.Qu.i3!hr\ti.i.§.§
3: 1.2.5)! .1 .

. r _

.§.§..‘....,.§m. ‘
. .9! .5...
I 3.2;.
gmmé 2»
15:77.13”.
$.41:
a

u.
13.4.5122 ii.
$3.13..

NV:..D\,II..:1.

{Iii-0.7!.
n: .H‘ittf.
. I

 

Haaaiiga

.:.

 

.. nx5.§§§§%¥
l'. .

III.

Hum. 73

«£3 ,

1 ' LIBRARY
3 00 ; Michigan State
University

This is to certify that the
thesis entitled

The Relationship between Childhood Sexual, Emotional, and
Physical Abuse to Juvenile Delinquency and Adult Criminality

presented by

Stacey-Ann Marie Gordon 3

has been accepted towards fulﬁllment
of the requirements for the

MS degree In Criminal Justice

/Z/ /AZ

Major ofe’ssor’s nature
a a

Date

 

 

MSU is an Afﬂnnatlw Action/Equal Opportunity Institution

 

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

 

ﬁﬁ'q

W5; 9 i '3

 

MAY {3 4 2088

 

my 2 Ming

 

111001171
. FOU-

 

i'! !\ ' '

 

a; §.r~
.|

 

 

 

 

 

 

 

 

 

2/05 p:/ClRC/DaleDue.indd-p.1

 

THE RELATIONSHIP BETWEEN CHILDHOOD SEXUAL, EMOTIONAL, AND
PHYSICAL ABUSE TO JUVENILE DELINQUENCY AND ADULT CRIMINALITY

By

Stacey-Ann Marie Gordon

A THESIS

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

MASTER OF SCIENCE
Department of Criminal Justice

2006

. EDI"

£HU.in'IIﬂllulLrtvi,ii.O-Iﬁ .? r1” F blunt» .Ll-Iihrhlb :

art.

ABSTRACT

THE RELATIONSHIP BETWEEN CHILDHOOD SEXUAL, EMOTIONAL, AND
PHYSICAL ABUSE TO JUVENILE DELINQUENCY AND ADULT CRIlyﬂNALITY

By
Stacey-Ann Marie Gordon

There are several purposes of the research study. The purposes of the study are to
assess 1) the relationship between the type of childhood abuse, whether physical, sexual,
or emotional, and juvenile delinquent behavior; 2) the relationship between the type of
childhood abuse and adult criminality; and 3) the effects of intervening variables such as
age, gender, race, child’s ﬁnal placement, the perpetrator’s relationship to the victim,
alcohol or drug use and approximated socioeconomic status.

Univan'ate, bivariate, and multivariate analyses are conducted. Logistic regression
techniques were used to examine the relationships between the independent variables and
each dependent measure. Ofﬁcial records for 593 abused youth re analyzed and revealed
the following: 1) a consistent and inverse relationship between age and juvenile
delinquency, 2) a consistent and direct relationship between age and adult criminality, 3)
a consistent relationship between gender and juvenile delinquency, 4) a consistent
relationship between gender and adult criminality, 5) a relationship between race and
juvenile delinquency, and 6) a relationship between race and adult criminality. The
variables child’s ﬁnal placement, the perpetrator’s relationship to the victim, SES, drug
or alcohol use, and abuse types are insigniﬁcant to the analysis of both juvenile

delinquency and adult criminality.

To Laurice Millicent Gordon

iii

ACKNOWLEDGENIENTS

The author wishes to express sincere appreciation to Dr. Charles Corley, Dr.
Homer Hawkins, and Dr. Vincent Hoffman for their superb guidance and assistance in
the preparation of this thesis. In addition, special thanks to Mr. Mengie M. Parker whose
familiarity with statistical techniques was helpful in the completion of this thesis. Thanks
also to my family and friends whose constant support, encouragement, and love helped

with the ﬁnal product.

iv

TABLE OF CONTENTS

LIST OF TABLES ....................................................................................... iii
INTORDUCTION ........................................................................................ 1
CHAPTER I ................................................................................................ 6
The Statement of the Problem .................................................................... 6
Research Hypotheses ................................................................................ 8
CHAPTER II .............................................................................................. 10
Review of the Literature ......................................................................... 10
Physical Abuse .................................................................................. 26
Emotional Abuse ............................................................................... 27
Sexual Abuse ..................................................................................... 28

The Effects of Abuse ......................................................................... 30
Social Learning Theory: Differential Association ............................... 33
CHAPTER III ............................................................................................ 36
Methodology .......................................................................................... 36
Measures and Scales .......................................................................... 40
Independent Measures .............................................................. 40

Dependent Measures ................................................................ 44

Study Limitations ................................................................................... 45
CHAPTER IV: Data Analysis and Findings ................................................. 48
Description of Findings .......................................................................... 48
Descriptive Statistics .......................................................................... 48
Bivariate Statistics ............................................................................. 52
Juvenile Delinquency .............................................................. 52

Adult Criminality .................................................................... 56

Regression Analysis .......................................................................... 60
CHAPTER V: Discussion and Conclusion ................................................... 72
Findings and Discussion ..................................................................... 72
Conclusion ......................................................................................... 80

APPENDICES ............................................................................................ 85

Appendix I .................................................... _ ..................................... 86
Appendix II ........................................................................................ 87
Appendix III ....................................................................................... 88
REFERENCES ........................................................................................... 89

vi

LISTS OF TABLES

Number Page
1. Sutherland’s Principles of Differential Association 34
2. Frequency Distribution of Independent measures 50
2a. Descriptive statistics for independent measures 51
2b. Descriptive Statistics for Independent measures 51
3. Frequency Distribution of Dependent measures 51
3a. Descriptive Statistics for Dependent measures 52
4. Pearson’s Correlation Coefﬁcients for Juvenile Delinquency 52
4a Crosstabulation of Juvenile Delinquency and Independent measures 53
5. Pearson’s Correlation Coefﬁcients for Adult Criminality 56
5a. Crosstabulation of Adult Criminality and Independent measures 57
6. Logistic Regression Analysis on Juvenile Delinquency - Model 1 62
7. Logistic Regression Analysis on Adult Criminality - Model 2 63
8. Logistic Regression Analysis on Juvenile Delinquency - Model 3 64
9. Logistic Regression Analysis on Adult Criminality - Model 4 65
10. Logistic Regression Analysis on Juvenile Delinquency - Model 5 66

vii

1 l. Logistic Regression Analysis on Adult Criminality - Model 6
12. Logistic Regression Analysis on Juvenile Delinquency - Model 7

l3. Logistic Regression Analysis on Juvenile Delinquency - Model 8

14. Summary of Tested Hypotheses

viii

68

68

69

70

INTRQDUCTION
The purpose of the study is to examine the relationship between childhood abuse

to that of a juvenile delinquent or a subsequent adult criminal charge. In today’s society
and with the ever-changing times children must not only be protected from the monsters
that lure and prey outside their homes, playgrounds, and/or school grounds but also from
those that reside within the home. Furthermore, youths face many obstacles that might
prevent or slow down healthy and positive growth patterns without factoring in those
problems experienced within the home environment.

In the early 1960’s the child abuse phenomenon began to receive increased focus
on the national level. That focus continually grew because of the increasingly large
reports of alleged child abuse incidents made to Child Protective Services (CPS). Child
Protective Services reports that during the period 1980 through 1993 there was a 155%
increase in allegations of child maltreatment (Snyder, 1998, as cited in Bartollas &
Miller, 2001). As such an increased focus is placed on the study of child abuse and
related issues. i

By 2003, 2.9 million referrals were made to Child Protective Services of which
approximately 906,000 were found to be substantiated cases of child abuse and neglect
(U S Department of Health & Human Services, 2003: 41; 2006: 46). Both of the reports
state that 4.9% of child victims are emotionally or psychologically abused, 9.9% are
sexually abused and 18.9% suffer from physical abuse. Additionally, 60.9% of child
victims are reportedly neglected by a parent or primary caregiver (US Department of

Health & Human Services, 2003: 22; 2006: 41).

Childhood abuse has become a more prominent and aggressive predictor of
juvenile delinquency because of its importance as an underlying factor of deviant
behavior and its implications for subsequent adult criminal involvement. That is, child
abuse is a very consistent as well as a damaging risk factor for offending behavior
(Stewart, Dennison, & Waterson, 2002). Although, no single factor accounts for a
youth’s development of delinquent or criminal behavior(s) the literature establishes the
importance of childhood abuse as a risk factor for subsequent delinquency (Smith &
Thornberry, 1995; Stewart et al., 2002) and adult criminality (Widom, 1989).
Additionally, the juvenile justice ﬁeld has placed more emphasis on the distinctive
relationship between childhood victimization‘, juvenile delinquency and subsequent adult
criminal behavior.

The Deﬁnitign 21' Child Agug

Despite extensive study on the topic a noted problem with child abuse and neglect
research is its conceptual and operational clarity. Following the 1974 Child Abuse
Prevention and Treatment Act (CAPTA) a very broad deﬁnition is applied to characterize
child abuse and neglect (Barnett, Manly, & Cicchetti, 1993),

“Child Abuse and Neglect means the physical or mental injury, sexual abuse,

negligent treatment, or maltreatment of any child under the age of eighteen by a

person responsible for the child’s welfare under circumstances which indicate the

child’s health or welfare is harmed or threatened thereby” (Public Law 93-247

section 2).

 

‘Wcﬁmizaﬁmmmemntemofﬂusmedaindudesmlymofcmdhoodphysim sexuaLor
emotional abuse. Inthisanalysisonly theaforementionedactsofcommissionare considered. Neglect, the
act of omission is excluded from the data amlysis. The term victimimtion is used interchangeable with
maltreatmentandabuse.

From its inception, the child abuse concept has been plagued by either inadequate or
incomplete deﬁnitions that are used to describe and measure the phenomenon.

Barnett et al (1993) gives a short illustration of the development of child abuse
deﬁnitions throughout the years. Child abuse is characterized by four major deﬁnitions:
the medical-diagnostic, the sociological, the legal, and the ecological perspectives.

The medical-diagnostic deﬁnition approaches the issue of child abuse as a
symptom of pathology. Under that perspective, emphasis rests on providing treatment to
the parent following the assumption that they suffer from a disorder. In short, the focus
rests solely with an assessment in order to provide treatment. In response to the medical-
diagnostic the sociological deﬁnition surfaced which views child maltreatment as an act
that is culturally inappropriate or that falls outside of the norms of the society. Following
that deﬁnition child maltreatment was simply a “social judgment” (Barnett et al., 1993:
19). What does and does not constitute child abuse is greatly guided by cultural norms
and practices. For instance, the use of spanking might be more acceptable in other
Western societies than in the United States. For example, the cultural context in Jamaica
allows for hitting a child in school for it is socially acceptable. That social norm is far
different from that in the United States.

The legal deﬁnition emphasizes the need to focus on universal guidelines and
standards for identifying and making judicial decisions. The last perspective, the
ecological deﬁnition, attributes maltreatment to and places emphasis on the environment
or ecosystem. Just as there are varied conceptual deﬁnitions and emphases for child
maltreatment there are also different operationalizations. Therefore, deﬁnitions of

maltreatment change based on the approach used. For example, based on the legal

approach emphasis is placed on obtaining physical evidence of the abuse whereas the
ecological viewpoint places importance on the impact of the environment.

There are generally three deﬁnitional problems relevant to child maltreatment: 1)
deﬁning the same maltreatment differently, 2) overlapping operational deﬁnitions for one
type of maltreatment, and 3) the failure to identify other maltreatment types other than ,
the ones being studied (Zuravin, 1991: 100). To overcome the problem of deﬁnition ﬁve
principles are applied in formulating a maltreatment classiﬁcation system: the use of
subtypes, conceptual clarity, the measurability of operational translations, the purpose of
deﬁnition, and the age and developmental stage of the child (Zuravin, 1991:102-104).

Using maltreatment subtypes is an effective way for ensuring that an adequate
conceptual and operational deﬁnition of abuse is used (Barnett et al., 1993). Instead of
categorizing abuse generally it is measured by the subtypes of physical, sexual, or
emotional abuse and neglect (Zuravin, 1991 : 102). Those subtypes are ﬁrrther delineated
by other categories of abuse. For example, physical abuse is measured by burning,
choking, and hitting.

Conceptual clarity is ensured by including all the criterion or behavior that ﬁts the
speciﬁc abuse type as well as its subtype(s) (Zuravin, 1991: 102). In order to achieve
clear concepts there should be at least one criterion that makes one type of abuse
distinctively different from another (Zuravin, 1991: 103). That ensures that more than one
type of abuse does not Conceptually deﬁne the same maltreatment.

All behaviors used to describe the abuse concept must be both observable and
measurable (Zuravin, 1991: 103). Behaviors must be tangible and measurable in terms of

the duration, severity, and ﬁ'equency of the abuse. As mentioned earlier, the purpose of

the deﬁnition of abuse must be stated. The purpose of deﬁnition guides the direction of
the research whether it is law and policy, treatment or another (Zuravin, 1999: 104).
Lastly, the age and developmental age of the child is considered when developing
classiﬁcation systems. For instance, behaviors considered inappropriate when a child is
an infant differs from those when the child is an early teen (Zuravin, 1991: 105; Barnett
et al., 1993: 42). For example, leaving a four year old at home unattended constitutes
neglect and more speciﬁcally failure to supervise. For a child twelve and over that
behavior is no longer conceptualized as neglect.

This thesis examines the relationship between childhood abuse, juvenile
delinquency, and adult criminality and more speciﬁcally assesses what are the
predominant predictors of such behavior(s). Univariate, bivariate, and multivariate
techniques are used in the data analysis procedures. Frequency runs are used to describe
the characteristics of the data Chi-square analyses and correlations are used to examine
the bivariate relationships and associations between the independent and dependent
variables. Multivariate logistic regression techniques are used to examine juvenile

delinquency and adult criminality as a function of the various independent variables.

CHAPTER 1
W

The study proposes to assess the relationship between three of the most prominent
social problems in the United States namely childhood abuse, juvenile delinquency, and
adult criminal behavior.

The youthful population is faced with several impediments during the transition
from childhood to adolescence and again from the teenage years to adulthood.
Additionally, youths constantly face social problems and contexts such as, poverty,
isolation, economic stress, and dilapidated environments to which are added the extra
strain of familial struggles and dysfunctions. A prevalent element within the family
dysfunctions category is the ever-increasing problem and occurrences of child abuse. In
the context of this thesis, victimization is a measure of the occurrences of child abuse
whether physical, emotional, or sexual.

Child maltreatment is a rather broad concept inclusive of different types of abuse
and neglect such as sexual, physical, and/or emotional abuse as well as physical,
emotional, legal, moral, and/or educational neglect. Childhood maltreatment throughout
this paper refers to childhood sexual, physical, or emotional abuse. The study proposes to
assess the impact of the different aforementioned abuse types on a juvenile delinquent
charge as well as an adult criminal charge. It also assesses the nature of that relationship,
that is, whether a certain type of abuse versus another is more indicative of juvenile
delinquency or adult criminality. ‘

The study objectives are to 1) establish if there is a relationship between the type

of abuse either physical, sexual, or emotional and juvenile delinquent behavior, 2) to

determine if a relationship exists between abuse type and adult criminal behavior 3) to
determine if there are intervening variables that affect or account for the relationship
between child abuse and delinquency and that of abuse and criminality, and 4) to
establish what characteristics of age, gender, race, placement, SES, and alcohol or
substance use relates to both the risk of becoming a juvenile or an adult offender.

It is important to understand the intricate relationship between childhood abuse,
juvenile delinquency, and adult criminality because of its theoretical and policy
implications. Grave importance is attached to linking juvenile delinquent career types to
adult criminality because of its policy implications and theory development (Kempf-
Leonard, Tracy, & Howell, 2001). Policy is guided by both informed qualitative and
quantitative research studies that accurately examine the relationship in question. It is
imperative to engage in sound evidence-based and data-driven decision-making policies
to both strengthen theoretical frameworks and to reestablish the validity of prior research.
The study also has implications for theory development where existent theories can be
tested and reinforced or new ones formulated.

If in fact a direct relationship is identiﬁed between childhood abuse, juvenile
delinquency, and adult criminality then interventions and preventions can be targeted
within the home environment or wherever else it is necessary and appropriate. Due to the
recurrent link between abuse, delinquency, and criminality, efforts must be made to
provide for juveniles programs based on a continuum of care by addressing all aspects
and inﬂuences of youth life. Such programs could include elements that are geared
towards the development of prosocial behaviors or teaching appropriate family

management practices. Speciﬁcally, programs must be geared towards addressing the

individual factors of the youth involved, peer inﬂuences, school factors, community
characteristics, and family inﬂuences (Hawkins, Herrenkohl, Farrington, Brewer,
Catalano, Harachi, & Cothern, 2000). Individual factors such as aggressive and
antisocial behaviors as well as the inﬂuences of delinquent peers and/or siblings must be
addressed. Additionally, other aspects of life must be factored into prevention and
intervention processes, such as academic failure, community or neighborhood
disorganization and child maltreatment.2

Childhood abuse is like a plague that results in the mental, physical, and
emotional disﬁgurement of youth. Various agents of the juvenile justice system question
whether the physically, mentally, sexually, or verbally abused child becomes a delinquent
youth and subsequently an adult offender due to the abuse(s) experienced. However, the
question remains what causes or accounts for the differences between youths who have
been abused who lead criminal lifestyles and those who do not? Essentially, what
intervening variable(s) accounts for the differences in offending or non-offending
behavior(s) among youth who experience childhood maltreatment?
Rﬂrch Hymthgg

The primary hypothesis is that childhood abuse increases the risk of a juvenile
delinquent charge and an adult criminal charge.

H1: There is a relationship between gender and juvenile delinquency.

H2: There is a relationship between gender and adult criminality.

H3: There is a relationship between race and juvenile delinquency.

H4: There is a relationship between race and adult criminality.

 

2 Theaforementioneddoes notincludeall theyouthful inﬂuencestlnt mightshapedelinquentorcriminal
behavior(s) but instead was sinrply an illustration of the variety of inﬂuences present. The exclusion of an
inﬂuence suchasparentalcriminality, poverty, orpoorfamilypracticesdoes notdiscountitsimportance.

H5: There is a relationship between age and juvenile delinquency.

P15: There is a relationship between age and adult criminality.

H7: There is a relationship between SES and juvenile delinquency.

H32 There is a relationship between SES and adult criminality.

H9: There is a relationship between physical abuse and juvenile delinquency.
H10: There is a relationship between physical abuse and adult criminality.

H11: There is a relationship between emotional abuse and juvenile delinquency.
H12: There is a relationship between emotional abuse and adult criminality.
H13: There is a relationship between sexual abuse and juvenile delinquency.
H14: There is a relationship between sexual abuse and adult criminality.

H15: There is a relationship between a child’s ﬁnal placement and juvenile
delinquency.

H16: There is a relationship between a child’s ﬁnal placement and adult
criminality.

H17: There is a relationship between drug or alcohol use and juvenile delinquency.
H13: There is a relationship between drug or alcohol use and adult criminality.
H19: There is a relationship between the perpetrator’s relationship to the victim
and juvenile delinquency.

H20: There is a relationship between the perpetrator’s relationship to the victim

and adult criminality.

CHAPTER 2
W

Youths experience many obstacles in their daily existence ranging from the
problems at home to coping with the pressures of adolescence. As previously mentioned,
one recurring and ever-growing obstacle to positive youth development which has been
extensively researched is the experience of child abuse. Additionally, wide debate exists
regarding the causes and correlates of both juvenile delinquent and adult criminal
behavior.

Several studies investigate the relationship between childhood abuse, juvenile
delinquency, and adult criminality (Brown, 1984; Widom, 1989; Smith & Thomberry,
1995; Hamilton, Falshaw, & Browne, 2002; Stewart et al., 2002). Others examine the
more speciﬁc relationship between physical abuse and subsequent adult behavioral
problems (Caetano, Field, & Nelson, 2003). Prior research concludes that childhood
abuse is a correlate of juvenile delinquency and adult criminality (Widom, 1989; Stewart
et al., 2002). The literature review will ﬁrst report on juvenile delinquency, adult
criminality, and childhood abuse. Secondly, the review will examine the literature as it
relates to childhood physical, sexual, and emotional abuse. The section ends with a brief
overview on the effects of childhood abuse.3
Jugnile Delinquent and Adult Criming Behavior

A widely recognized and highly accepted notion is that the role of the youthﬁrl
offender in society has evolved (F erro, 2003 ). That is, juvenile and youth crimes that

were existent in the early to mid 19905 differ substantially from that of the youth of the

 

3Theefl’ectsofchildhoodabuseareonlybrieﬂyaddressedbeeausetheintentionofthispaperisnotto
examineorassesstheimpactorconsequencesofabuseinsteaditissimplytoexploretherelationship.

10

latter 19903 and present day. Additionally, the dynamics of juvenile life and crimes have
changed signiﬁcantly where several issues are compounded creating a downward ripple
effect for youths involved. That is, youths of today are not only overwhelmed by the
problem of peer pressures, but also by issues such as abuse and other social situations,
which are age inappropriate.

Present day juveniles are labeled as more volatile, engaged in more heinous
criminal acts, and are constantly exposed to age inappropriate behaviors. Ferro (2003)
noted that the substantial increase in juvenile crime is attributed to a myriad of factors
including: the growing popularity and abuse of illegal drugs, the increased availability of
ﬁrearms, a downsized economy, and the anticipated growth in the juvenile population. In
addition to the aforementioned sources of juvenile crime, the assertion still remains that
youthful delinquent and adult criminal behaviors have taken on new dimensions.

The study of juvenile crime and delinquency is not new but instead has been an
integral part of society for many years (F erro, 2003). However, the issue of delinquent
behavior is often complicated by the fact that youths enter the juvenile justice system
with many mental and psychological problems (Brown, 1984). The relationship between
victimization and subsequent mental and emotional problem behaviors is established
(Browne & Finkelhor, 1986). Consequently, youths that enter the juvenile justice system
pose a tremendous challenge because it becomes more difficult to address their social
needs when they have other psychological needs that also have to be met. However, it is
insufﬁcient to address the offense committed by the juvenile, by the application of harsh
sanctions, whilst ignoring what could possibly be the underlying problem or source of the

dysﬁrnction.

ll

As such, the problem of juvenile delinquency or adult criminality cannot be
effectively addressed without assessing the correlates of those behaviors. If a child’s
delinquent behavior stems from the abuse(s) experienced within the home environment
then identifying the presence and addressing the effects of the ill treatment can only alter
resultant behavioral problem(s). Similarly, the effects of such experiences on adult
behaviors must be assessed.

Bivariate analysis reveals both a strong and signiﬁcant correlation between
childhood maltreatment and the likelihood of ofﬁcial delinquency (Smith & Thomberry,
1995). The Smith and Thomberry (1995) study concludes that 45% of the abused versus
31.7% of the non-abused youth in the sample engage in official delinquent behaviors.
Similarly, one study examines the direct pathway from childhood maltreatment to that of
juvenile offending among a sample of Australian youth (Stewart et al., 2002). The
analysis found that youth with a maltreatment experience are more likely than those non-
maltreated to engage in delinquent behavior(s), 17% versus 10% respectively. Similarly,
Widom ( 1989) in her study assesses the relationship between abuse, neglect, and violent
criminal behavior. Her analysis ﬁnds that those subjects who report childhood abuse had
higher criminal rates than those who did not.

Childhood Abug and 53° Relation to Dgingueng and Criminality

There are several problems associated with the study of childhood abuse that are
relevant to both the issues of delinquency and criminality: the measurement and
deﬁnition of abuse, the use of official data sources only, and establishing temporal order.
Research cannot conclusively state the percentage of youth who have contact with the

juvenile justice system that have been abused. That lack of consensus is due in part to

12

both the inaccuracy of data and the difﬁculty in collection methods (Osofsky, 2001;
Jewell, 1999: 17).

It was only in the year 1968 that all then 50 of the United States moved towards
enacting a law for ofﬁcially reporting suspected incidents of child abuse4 (Jewell, 1999:
3). For example, before the implementation of the National Incident-Based Reporting
System (NIBRS) both the Uniform Crime Reporting (UCR) and the National Crime
Victimization Survey (NCVS) systems stated inaccurate numbers of child abuse (Jewell,
1999: 3; Finkelhor & Ormrod, 2000). In addition to the problem of the under and/or over
reporting of victimizations is the fact that essential variables, such as age, were often
excluded from the analysis where offenses against children under the age of 12 were not
counted (F inkelhor & Ormrod, 2000).

However, throughout the years researchers use different operational deﬁnitions
for childhood abuse. A common problem with abuse deﬁnitions is that some are more
inclusive than others. For example, instances of assaults might be included in the
measurement of sexual abuse. Similarly, the issue of conceptualization often arises where
sexual abuse includes both sexual molestation and a youth’s exposure to inappropriate
sexual content such as pornographic materials That dilemma necessitates the need for the
juvenile justice ﬁeld to further study the dynamics of abused and/or neglected youth in
order to adequately provide system services. It is also important in order to establish
approximate and accurate numbers of abused youth which, aides with the policy of
providing juveniles with a continuum of services thereby increasing their ultimate

successﬁrl re-entry and reintegration.

 

4Theactmandatingthereportingofchildabuseincidentswasnotsignedintolawuntil1974.

l3

The importance of establishing the temporal order of delinquency and abuse is
addressed throughout the literature. Loeber, Kalb, and Huizinga (2001) address whether
delinquency increases a juvenile’s chances of victimization or does the latter serve as a
risk factor for delinquent behavior. Although the authors speciﬁcally address serious or
violent victimization’ the article does lend some support for the notion that the
relationship between childhood abuse and juvenile delinquency is established. However,
the temporal sequencing of victimization and delinquency is ensured by the use of a
target population of youth aged birth through 11 (W idom, 1989). Additionally, the
longitudinal nature of a research study also validates temporal order unlike cross-
sectional studies (Stewart et al., 2002). The present study ensures temporal sequencing
because youth are sampled ﬁom birth through age eleven thereby allowing the abuse to
occur before a delinquent offense.

Smith and Thomberry (1995) examine the association between childhood
maltreatment and juvenile delinquency speciﬁcally through ofﬁcial and self-reported
data. The authors emphasize the limitation of using official reports of child abuse because
its correlation with juvenile delinquency could result from the tendency for more formal
recognition for some families versus others. That is, official action might be taken against
some families more so than others where a higher social standing might decrease the
likelihood for formal recognition of the abuse. Although there is an apparent class bias in
the use of official data that alone does not discount the validity of the observed

relationship (Brown, 1984).

 

5 The analysis of victimimtion in this thesis focuses solely on the physically, emotionally, and sexually
abusedchild Caution mustbeusedwhendrawing inferences fromtbe Loeberetal (2001)studybecause it
involved a broader deﬁnition of victimization inclusive of offenses such as assaults and burglaries.

14

Malamute

There are gender differences in the commission of juvenile delinquent and adult
criminal behavior(s). Similarly, there are gender differences in the experience of
childhood abuse. For the calendar year 2003 52% of child victims of abuse were females
versus 41% who were males (US Department of Health & Human Services, 2006: 41).
However, there are instances where no substantial differences in the incidences of abuse
for males or females are reported. Smith and Thomberry (1995) report that 13% of males
in the sample experience abuse similar to 14.1% of females. In their analysis no
signiﬁcant gender differences are found for the incidents of abuse. Similarly, Stewart et
al (2002) and Hamilton et a] (2002) report no signiﬁcant gender differences in the
experience of maltreatment.

Nevertheless, research contends that males are signiﬁcantly more criminal than
females (Siegel, 2003: 68; Piquero, Gover, MacDonald, & Piquero, 2005). Kilpatrick and
Saunders (2000) found that males are more likely than females to commit a delinquent
act, 17.7% versus 6.7% respectively. Similarly, males report a higher ﬁequency of adult
criminal behavior than females, 38.1% versus 13% respectively (Widom, 1989). The
literature also examines gender differences for youths who have been abused and/or
neglected. The preliminary report from the Youth Development Series6 concludes that in
both the Denver and Pittsburgh research sites males are the most likely to be victims
(Loeber et al., 2001).

Research also states that males with a history of abuse or neglect are at a greater

risk for adult criminal behavior (W idom, 1989; Stewart et al., 2002). On the same note,

 

‘TheYouthDevelopment SeriespresentsﬁndingsfromtheProgram ofResearchontheCausesand
Correlates of Delinquency. The study sites include Rochester, New York; Denver, Colorado; and
Pittsburgh, Pennsylvania. The Loeber et al (2001) analysis included only the Denver and Pittsburgh sites.

15

Widom (1989) reports that 38.1% of abused males versus 13% of maltreated females
later engaged in adult criminal behavior(s). In that study, older black males with a history
of abuse or neglect are at an increased risk for displaying criminal behavior. Similarly,
Stewart et al (2002) reports that 25% of abused males in comparison with 11% of abused
females are later delinquent. Therefore, males who are abused are more likely than
females to later engage in criminal activities. Research also suggests that a youth’s
chances of entering the juvenile system and proceeding totthe adult criminal system are
greatly increased by the experience of the three forms of maltreatment: physical, sexual,
and emotional abuse (Osofsky, 2001).

Although much is reported with regards to gender differences in juvenile
delinquency the gender gap has not been thoroughly assessed because until recently study
samples typically have been strictly male (Malinosky-Rummel & Hansen, 1993).
Additionally, abuse samples are not always equally represented by gender; for example,
most studies of sexual abuse have targeted female subjects and only a small number of
males (Kendall-Tackett, Williams, & F inkelhor, 1993). However, one of the most
substantial and noted differences between juvenile delinquent females and males is that
the former intemalizes whereas the latter extemalizes (Kendall-Tackett et al., 1993;
Elliott, 1994; Feldman, 1997; Caetano et al., 2003). That is, females are more prone to
suppress their feelings whereas males are more likely to act outwardly.

The hypothesis that males are more likely than females to engage in delinquent
acts is because males extemalize their feelings (Caetano et al., 2003). There are
distinctive patterns of criminal behavior for females and males; however, other

researchers have not entirely adopted the theory of internalizing and externalizing

l6

behaviors (Hamilton et al., 2002). That disagreement possibly increases the necessity for
the youthful and adult systems to treat male delinquent and criminal subjects separately
ﬁ'om females because the literature has established that the same approach should not be
used for both. Additionally, the increased involvement of girls in the juvenile justice
system is well noted (Pepi, 1997; Acoca, 1999). Furthermore, research concludes that
females who experience early instances of physical, emotional, or sexual abuse are at an
increased likelihood to enter the juvenile justice system (Osofsky, 2001).

Age

The general premise is held that there is an inverse relationship between
victimization and age as well as crime (Siegel, 2003: 67). Following that principle an
increase in age indicates a decrease in child victimization, as well as juvenile
delinquency, and adult criminality. With respect to child abuse, youth within the age
group birth through three are most likely victims representing 16.4% of maltreatment
cases in the year 2003 (US Department of Health & Human Services, 2003: 23).
Conversely, those within the age ranges 16 through 17 and 12 through 15 are the least
likely to experience victimization representing only 5.9% and 10.7% respectively (U S
Department of Health & Human Services, 2003: 23).

The assumption is held that younger children are more likely to be victimized
because of two main reasons. Firstly, their age makes them defenseless to such acts and
secondly because most of their time is spent within the conﬁnes of the home
environment. Overall, the literature states that there is an inverse relationship between
age and delinquent involvement as well as adult criminality (Siegel, Welsh, & Senna,

2003: 67). The premise is that individuals age-out of deviant behavior.

17

Bag

The relationship between race and child abuse has been examined where
contrasting ﬁndings are reported. Research has shown that youth of minority status are
more likely to be the victims of youthﬁrl crime than any other racial grouping (Finkelhor
& Ormrod, 2000). Similarly, Caetano et al (2003) maintain that occurrences of childhood
abuse are greater for Blacks and Hispanics as compared to Whites. According to the US
Department of Health and Human Services’ (2003) report Paciﬁc Islanders, American
Indians or Alaska Natives, and African-Americans have the highest victimization rates
recorded at 21.4%, 21.3%, and 20.4% in that order.

Additionally, other studies conclude on a larger scale that Whites are less likely
than both Blacks and Hispanics to be victimized (Paschall, F lewelling, & Ennett, 1998;
Loeber et al., 2001; Piquero, Macintosh, & Hickman, 2002). However, contrasting
ﬁndings are reported. For example, Smith and Thomberry (1995) found no signiﬁcant
racial differences for the abused group reporting percentages for Whites, African-
Amcricans, and Hispanics: 16.4%, 14%, and 8.7% respectively. However, Morales
(1998: 16) states that there is no association between child abuse and race but instead the
strongest correlate for child abuse is poverty.

Research also reports that Whites are less likely than those of minority status,
speciﬁcally Blacks and Hispanics, to be juvenile offenders (Paschall et al., 1998; Piquero
et al., 2002). Similarly, Kilpatrick and Saunders (2000) assessed the prevalence of both
familial and nonfamilial violence on adolescents and conclude that Non-Caucasians are
more likely than Caucasians to commit a delinquent offense, 18.1% versus 9.9%

respectivelnyonversely, Huizinga and Elliott (1987) found no statistically signiﬁcant

18

racial patterns or differences in the commission of delinquent activities . However, in that
analysis, minorities recorded higher rates for more serious offenses than Whites.

On a similar note, one study found that of the abused and neglected group Blacks
(39%) are more likely than Whites (24.4%) to have a subsequent criminal record
(Widom, 1989). Overall, the likelihood for system involvement is greater for African-
American youths than for Whites (Williams, Ayers, Outlaw, Abbott, & Hawkins, 2001).
In that study, race explained 3% of the variance in juvenile justice system involvement
for delinquency at both ages 15 through 16 and again at 18 through 19.

Th; Pemetggtor’s Relationship t2 the Child Victim

In looking at childhood abuse it is important to note that most crimes against
youth are not committed by strangers but instead by an individual that the child is
familiar with (Finkelhor & Ormrod, 2000; Osofsky, 2001). Additionally, most instances
of childhood abuse are committed by adults (Schmalleger, 2003: 591).

F inkelhor and Orrrrrod (2000), in their article characterizing crimes against
juveniles, found that adult offenders are responsible for 55% of all reported juvenile
victimizations. Additionally, family perpetrators committed 20% of the offenses against
youth. For example, in the instance of sexual abuse usually both the victim and the
offender are familiar with each other (Barlow, 1990: 205). Furthermore, 83.9% of child
abuse perpetrators are parents (whether biological, step, or adoptive) whereas only 13.4%
are non-parental perpetrators (US Department of Health & Human Services, 2003: 24).
Similarly, Hamilton et al (2002) reported that 49% of offenses against juveniles are at the

hands of family members. That ﬁnding is important because the literature states that

19

abuse by a family perpetrator is more detrimental than that by a stranger (Browne &
Finkelhor, 1986; Kendall-Tackett et al., 1993).

Using a multi-State data set comprised of approximately 200,000 child abuse
perpetrators the 2002 National Child Abuse and Neglect Data System (NCANDS) study
reports that there is a gender difference in terms of child abusers (Shusterrnan, Fluke, &
Yuan, 2005). That study reports that 46% of perpetrators are male whereas the remaining
54% are females. Additionally, male perpetrators are found to be older than females.
Similarly, the US Department of Health and Human Services (2003) reports similar
percentages where 58.2% of perpetrators are females and 41.8% are males.

Although violent crime is beyond the scope of this study research reports that
only 11% of violent crime child perpetrators are strangers (Finkelhor & Ormrod, 2000).
Additionally, simple assault is the most prominent crime committed against juveniles
while sexually based crimes are the least committed offense against youths by strangers.
However, the majority of all youthful victims are the sufferers of the latter offense, where
more than 70% of all reported sexual offenses are committed against young victims
(F inkelhor & Ormrod, 2000).

Research shows that most abuse incidents occur within the home environment
where youths from abusive households demonstrate a higher probability for displaying
subsequent delinquent behaviors. As such, the family dynamic and environment is
important for understanding the relationship between childhood abuse, juvenile
delinquency, and adult criminality because behaviors are learned not only through
socialization but also from direct or indirect observation. In an abusive environment such

violent acts (physical, sexual, or emotional abuse) are reinforced as positive and socially

20

acceptable behaviors. As a result, the susceptible child copies and practices such social
responses. Additionally, the susceptibility to a variety of abuses in adulthood also
increases as a result of childhood victimization (Brown, 1984; Mullen, Martin, Anderson,
Romans, & Herbison, 1996).

An individual’s exposure to violence, in this instance abuse, is predictive of his or
her own subsequent delinquent behavior (Nofziger & Kurtz, 2005). It is hypothesized that
the exposure to such violent acts and behaviors increases the vulnerability for juvenile
delinquency. That assertion is based on the ideology that violence is a learned response or
behavior (Siegel et al., 2003: 121). Juveniles who are victimized by violence are also at
an increased risk for violence in later adulthood (Nofziger & Kurtz, 2005). Although the
study does not address violent crime it is important to note that prior victimization is also
indicative of future violent behavior because there is a strong relationship between ones
exposure to violence and subsequent offending behavior.

Similar to the varying and adverse effects that result from physical, emotional, or
sexual abuse there are differences in who are the perpetrators. Someone within the
youth’s immediate environment usually carries out the acts of both physical and
emotional abuse whereas sexual abuse is typically carried out by a male outside of the '
home (Mullen et al., 1996). For example, only 27% of sexual abuse perpetrators are
parents whereas 75.9% resided outside of the home (US Department of Health & Human
Services, 2003: 64).

Sgimgngmig Sgtus (SES)
Youth that are subjected to social and economic hardships and family disruptions

are more prone to childhood abuse than those in organized environments (Mullen et al.,

21

1996). A strong correlation is noted between poverty and child abuse and/or neglect as
well as the severity of the maltreatment (Pelton, 1994: 74). Although social situations
such as poverty, unemployment, and welfare are risk factors for abuse that does not mean
that child abuse is inevitable in families with a low SES and nonexistent in those with a
higher SES.

The Hawkins et al (2000) research study assesses the strengths and duration of
changeable risk and protective factors for youth violence. The study also addresses the
importance of identifying and targeting risk and protective factors in order to provide
adequate intervention and prevention programs. That meta-analysis of 66 studies, reports
that the potential for increased involvement in crime or violent behavior(s) is often ﬁreled
by poverty (Hawkins et al., 2000).

The importance of the home environment has been continually addressed
throughout the content of this thesis. Childhood abuse is a direct signiﬁer of the reduction
or breakdown in the quality of family life, which directly affects the subsequent
behaviors of the youth (Siegel et al., 2003: 196). The reduction in the quality of family
life can result from economic hardships, poor family management practices, parental
substance abuse, and such other factors (Hawkins et al., 2000).

Based on the level of signiﬁcance attached to the home environment a fairly
logical assumption can be made. Ifa parent or primary caregiver engages in poor family
practices such as neglect or failure to provide or being verbally abusive or emotionally,
unavailable to the youth’s needs then that both weakens the family structure and strains
the social environment. For example, in a home plagued with a history of chronic

substance abuse the child is forced to assume inappropriate age responsibilities such as

22

those required of the parent. Additionally, poor family relationships and practices foster
maladaptive behaviors towards the vulnerable and impressionable child.

Broken families7 might also exhibit characteristics, conducive to abuse, such as
social disorganization, violent or aggressive behavior, and/or substance abuse. Wells and
Rankin (1991) conducted a meta-analysis of the impact of broken homes on juvenile
delinquency. Across the 50 studies included the prevalence of juvenile delinquency rests
between 10% and 15% higher for those youths ﬁ'om broken homes in comparison to
those from intact environments. Additionally, the study concludes that there are no
signiﬁcant differences in the impact of broken homes by gender or race.

Crime as traditionally being viewed as a lower-class phenomenon. The view has
been solidiﬁed because of speciﬁc characteristics that relate to criminal behavior such as
poverty and disorganized communities that are ever present in lower class
neighborhoods. Weis (1987) examines the correlation between social status and juvenile
delinquency using the Seattle Youth study (SYS) data set for the years 1978 and 1979.
Using a disproportionate stratiﬁed sample the researcher over-samples for males, blacks,
lower SES, and official delinquents. The study reports no relationship between social
class and self-reported or ofﬁcial measures of delinquency. Using an ecological measure
of class social status was calculated by the median income of the census tract in which
the juvenile lived. The study concludes that the association between ofﬁcial delinquency
and SES was weak with a reported gamma value of -. 17. Therefore, a weak and negative

relationship exists between one’s SES and delinquent involvement. Furthermore, the

 

7 'I'heassurnptionisnotthatbroken familiesareexclusivelyfromalow SESbutinsteadthatsuchafamily
environment contributes to the likelihood of abuse, and delinquent behaviors.

23

author concludes that regardless of the measure or scale used for either SES or
delinquency only a weak relationship was observed (W eis, 1987).

According to theory, a low SES is a strong correlate of juvenile delinquent
behavior. However, empirical ﬁndings suggest that there is only a weak relationship
between juvenile delinquency and a low SES. Wright, Caspi, Mofﬁtt, Miech, and Silva
(1999) reexamine the relationship between SES and juvenile delinquency guided by the
general hypothesis that it is a matter of causation and not correlation. They conclude that
the causal link between SES and juvenile delinquency is indirect, and exists only because
of the presence of intervening variables.

Similar to the problem of a concise deﬁnition of abuse social class is often not
easily deﬁned (Brown, 1984). Brown notes a limitation in the social class variable when
drawing causal inferences with abuse. That is, as noted before, there is an apparent class
bias where cases involving those in the upper strata are not reported. In his analysis
Brown (1984) concludes that child maltreatment and juvenile delinquency are correlated
independemly of social class where there is a weak inverse relationship with abuse. In
contrast to the Brown ﬁndings, Smith and Thomberry (1995) report a signiﬁcant
difference in maltreatment by social class where 19.5% of those of lower class status in
comparison to 8.2% of the middle and upper classes report instances of abuse.

Child’s Fing Placemtnt

Frame (2002) examines the relationship between a child’s placement in foster
care to child and family characteristics as well as child abuse and neglect reporting
histories. In that analysis, reasons for the child’s removal from the home include: 80.8%

neglect, 13.6% physical abuse, 2.1% sexual abuse, and 1% emotional abuse. Similar to

24 -

the Stewart et al (2002) study those sexually abused were most likely to be placed outside
of the home environment (Frame, 2002). In the Stewart et al (2002) analysis both
sexually and emotionally abused youth were 35% more likely to be placed outside of the
home. Physically abused youth were 41% more likely to be placed in out-of-home
placements.

Stewart et al (2002) report that 26% of abused youth versus 13% of non-abused
youth who were placed outside of the home subsequently offended. The literature
contends that placements outside of the home prove more detrimental to the youth than
in-home placements. Wall (2005: 67) states that group care placements are greatly
associated with the exhibition of more aggressive and delinquent behaviors than any
other placement types combined.

There are a number of pathways through which child victims of abuse become
delinquent and/or criminal. A child’s ﬁnal placement after the dissolution of a
dependency petition ﬁled with the Court is one possible link. waever, the majority of
abused youth are not removed from their home environment (Frame, 2002; Wall, 2005:
29). Wall (2005) reports that following an allegation of child abuse against a parent or
primary caregiver 90% of abused youth remain in in-home placements.

Drug or Algghgl use

The literature has debated the causal link between drug and/or alcohol use and
involvement in crime. An even greater problem rests with the association between
chronic substance abuse and delinquent as well as criminal behavior(s). Increasingly, the

importance of identifying and addressing substance abuse is heightened because of the

25

widespread use among juveniles. Despite efforts on the part of the government and its
war on drugs incidents of both drug and alcohol use have risen (Flowers, 1999: 3).

Flowers (1999: 7) notes that emphasis should be placed on the role of alcohol or
drugs in criminal behavior because of its far-reaching social and economic consequences.
That is, the use or abuse of drugs or alcohol leads to increased criminal involvement in
order to satisfy an addiction. Additionally, there is a direct correlation between substance
abuse and criminal involvement not only for drug crimes but also for violent, economic,
and sexual offenses (Flowers, 1999: 32).

Peers, family, personality, school, and delinquency are noted as the correlates of
drug use for juveniles8 (Flowers, 1999: 61).Parental drug or alcohol abuse is highly
associated with child abuse (Morales, 1998216). The literature outlines that child abuse is
also a manifestation of parental abuse (Flowers, 1999: 153) and that drug or alcohol use
increases delinquent or criminal involvement.

Physical Abuse

Physical abuse is the most easily documented as well as regularly studied type of
maltreatment (Barnett, Manly, & Cicchetti, 1993: 35). Research states that physical abuse
is more predictive of ﬁrture offending than both sexual and emotional abuse (Stewart et
al., 2002). One possible explanation for that ﬁnding is that physical abuse is more
detectable than both sexual and emotional abuse. That is, physical abuse typically results
in bruises on the child victim whereas both emotional and sexual abuse has no visible
signs, marks, or bruises. However, that does not mean that physical abuse always results

in bodily injuries whereas other types of abuse do not.

 

8Peersarerecordedasthestrongestpredictorsfordruguse.

26

Although a speciﬁc association does not exist between race and child abuse
studies note that a relationship exists. The prevalence of childhood physical abuse is not
greatest among Blacks and Hispanics but instead among Whites (Caetano et al., 2003).
With respect to gender, Kilpatrick and Saunders (2000) conclude that males are more
likely than females to be physically abused 21.3% versus 13.4% respectively. No clear
difference exists for abuse perpetrators. That is, males represent 22% of perpetrators in
comparison to 18% of females (Shusterrnan et al., 2005). Additionally, 22% of males in
comparison with 18% of females are associated with childhood physical abuse.9

The experience of physical abuse during childhood along with having witnessed
parental violence both serve as risk factors for criminal behavior in subsequent
adulthood, speciﬁcally in the Caetano et al (2003) analysis. Both Widom (1989) and
Smith and Thomberry (1995) report that those who are physically abused are more likely
to commit violent crimes later in life. On the other hand, Brown (1984) reports that the
experience of physical abuse has no correlation with delinquent behavior. Furthermore,
physical abuse is very commonly associated with later aggressive and delinquent
behaviors. However, in the Wall (2005) study physically abused youth are not found
more aggressive than those who had experienced other forms of maltreatment.
Emotional Abuse

There is a noted paucity of research on childhood emotional abuse, which is
usually studied alongside both physical and sexual maltreatment (Spertes, Yehuda,
Wong, Halligan, & Seremetis, 2003). Additionally, emotional abuse is noted as the most

problematic and difficult subtype to deﬁne because it is not identiﬁable by physical

 

9Inthatanalysis36%ofmalesversus66%offemalesarreassociatedwithchildneglect.

27

marks and bruises, unlike sexual or physical maltreatment, and as such might not come to
the attention of government authorities.

Emotional abuse is potentially the most harmful type of maltreatment of the three
studied (Jewell, 1999: 114). That is, emotional abuse attacks the youth internally where
the child fosters negative thoughts and feelings of self-worth. The Brown (1984) study on
the correlation between social class, childhood maltreatment, and juvenile delinquency
found that the experience of emotional abuse showed a positive correlation with
delinquent behavior. Therefore, emotional abuse is directly related to the likelihood of
delinquent behavior.

Sequ Abgge

The research literature reports substantial declines in child abuse cases. F inkelhor
and Jones (2004) provided several key explanations for the decline in child sexual abuse
cases throughout the 19905. They report explanations such as a real decline in sexual
abuse incidents, changes in Child Protective Services (CPS) data collection methods or
deﬁnitions, and the increased conservatism within CPS. The decline is also attributed to
the increased focus on the implementation of prevention and intervention programs and
the ability of the system to identify criminal sexual behavior (F inkelhor & Jones, 2004).

There are noted gender differences in the commission and incidents of sexually
based crimes committed against juveniles. There are also contrasting research ﬁndings
with regards to the most likely victims of sexual offenses.10 Studies report that young
females are more likely than males to be the victims of sexual abuse (Pepi, 1997;

Kilpatrick & Saunders, 2000). Kilpatrick and Saunders (2000) report that females are

 

'°Verylittleempiriealresearchlnsbeendonewithmalesintlnstudyofsemalabusemrownedt
Fimtelhor, 1986).

28

more likely than males to be the victims of sexual abuse, 13% versus 3.4% respectively.
Additionally, males“ are more often times than females found to be the perpetrators of
sexual abuse against youth (Shusterman, et al., 2005). Shusterrnan et al (2005) report that
26% of male perpetrators are sexually abusive versus only 2% of females. Conversely,
there are no differences between the incidences of sexual abuse for males and females
when both instances of touching and non-touchingl2 forms of the abuse are included in
the analysis (Jewell, 1991: 58).

The Krahé et al (1999) study highlights the importance of assessing the causal
link between victimization in childhood and subsequent revictimization. A direct causal
relationship is observed between sexual abuse at infancy and victimization as well as
further ill treatment during adolescence. However, no direct relationship exists between
childhood sexual abuse and future abuse or victimization in adulthood (Krahé et al,
1999). The Krahé et al study found that the nonvictimized sample had fewer experiences
of unwanted sexual occurrences than both the control and comparison groups (1999).
They conclude that childhood sexual abuse is a substantial risk factor for subsequent
victimization in later stages of life.

When looking at the many dimensions of childhood victimization, that is, the
physical, emotional, and sexuaL it is important to emphasis that each type may lead to a
speciﬁc or differing pattern of individual behavior. That analysis proves difficult because
abuse types usually are not studied in isolation. The type of abuse experienced impacts or

inﬂuences the likelihood of offending behavior. For examme, the type of abuse is related

 

” The eategory male in that instance refers to non-parent perpetrators only.
‘2 Non-touching instances of sexual abuse include exhibitionism whereas contact includes masturbation,
erotic fondling or kissing, oral stimulation, etc.

29

to the likelihood of offending behavior for youth who experience physical abuse but not
for youthful victims of either sexual or emotional abuse (Stewart et al., 2002).

However, the range of harm associated with emotional, physical, and/or sexual
abuse is unknown but the general consensus is that childhood abuse correlates with
subsequent problems in adulthood (Starr, MacLean, & Keating, 1991: 9; Mullen et al.,
1996). That necessitates the need for studying the nature of the relationship between the
types of abuse at infancy, juvenile delinquency, and adult criminality.

The Effﬂ of Child Abuse

The speciﬁc effect of physical, emotional, and sexual abuse is not mutually
exclusive (Mullen et al., 1996). However the general consensus is that abuse is associated
with later problems in adulthood. Nevertheless, many researchers assert that there are
speciﬁc outcomes associated with a precise type of abuse (Brieke & Runtz, 1990; Mullen
et al., 1996). For example, physical abuse is generally associated with the exhibition of
aggressive behaviors whereas emotional abuse is typically believed to result in or to be
associated with the exhibition of low self-esteem and conﬁdence, and sexual abuse is said
to result in maladaptive sexual behaviors in later adolescence and adulthood (Spertes et
al., 2003).

While some child maltreatment effects are immediately identiﬁable there are
others that are not. Due to the ﬁequency, duration, and severity of the abuse experienced
there can be short- or long-term consequences in both the child’s physical and
psychological well-being. However, there are many noted short- and long-term effects
associated with the experience of childhood abuse (Mullen et al., 1996; Caetano et al.,

2003).

30

However, other researchers maintain that there is a difficulty in deciphering the
speciﬁc effects of each type of abuse where consequences are not mutually exclusive. For
example, Hamilton et al (2002) examine the link between recurrent maltreatment and
offending behavior and conclude that there is a difficulty in determining the effects of
speciﬁc types of abuse because some maltreatment occurs in unison. For example,
physical abuse does not necessarily result in bodily injuries but also in more severe
trauma that might even lead to a handicap. Additionally, the act of sexual abuse might
involve physical injuries to the victim. That poses a difficulty in research study samples
where the effects of speciﬁc abuse types are not easily distinguishable. For example,
nonsexually abused children are reportedly sexualized in the Kendall-Tackett et al (1993)
study, where 17% of those physically abused exhibit sexually inappropriate behaviors.
Mullen et al (1996) also states that physically abused youth have sexual problems as well
as a lowered self-esteem. That study ﬁnding contradicts the general belief that physically
abused youth exhibit aggressive and not sexually inappropriate behaviors.

A direct relationship is noted between emotional abuse and neglect with adult
emotional and somatic ﬁrnctioning (Spertes et al., 2003). The Spertes et al (2003)
ﬁndings states that psychological or verbal abuse is used to predict psychopathology as
well as negative moods. The authors conclude that it is not only overt forms of abuse
(physical and sexual) that impact adult development and health.

Mullen et al (1996) report that the impact of emotional abuse varies by the gender
of the abuser. For example, emotional abuse by a female results in adult psychopathology
whereas abuse by a male leads to difficulty in sexuality in adulthood. Browne and

Finkelhor (1986) examine the long and short-term effects of childhood sexual abuse on a

31

sample of female victims. The identiﬁed short-term or initial effects include fear, anger,
hostility, guilt, and shame. The noted long-term effects of childhood sexual abuse include
but are not limited to depression, self-destructive behaviors, revictimization in adulthood,
and promiscuity. ’3

The effects of sexual abuse are longstanding where the consequences are often
displayed in adulthood. Research states that women who are sexually abused as children
are more likely to be victims of sexual aggression in later life compared to those who are
not exposed to that type of social condition (Krahé, Scheinberger-Olwig, Waizenhofer, &
Kolpin, 1999).

In their research study Krahé et al (1999) assert that childhood experiences of
sexual abuse among the sample of female adolescents constitutes a risk factor for
subsequent revictimization. Additionally, those individuals with a child history of sexual
abuse score higher on the sexual revictimization scale than those with no prior instances
of abuse (Krahé et al., 1999). Further effects are noted where the authors maintain that
feelings of worthlessness among the women sampled predated, followed, or co-existed
with the sexual abuse incident(s). Those feelings of worthlessness existent in childhood
are indicators for a greater risk of being sexually revictimized in adolescence (Krahé et
al., 1999).

The relationship between child abuse to that of both juvenile and adult criminality
is a greatly debated topic. However, research shows that not all youths who are abused
lead delinquent (Hamilton et al., 2002) or even criminal lifestyles (Jewell, 1991: 60;

Widom, 1989; Margolin, 2005; Wall, 2005). Nevertheless, youth exposed to such

 

‘3 Anin-depthanalysis ofthe effects ofphysical, emotional, orsexualabusewillnotbeconductedinthis
study.

32

conditions display an increased likelihood for deviancy. Margolin (2005) reports that
70% of abused youth do not follow the intergenerational cycle and become abusive adults
or even report ofﬁcial delinquency. Similarly, Smith and Thomberry (1995) conclude
that the majority of abused youth do not later offend. Furthermore, no substantial
differences are found in the Widom (1989) study where 26% of the abused group versus
21.1% of the non-abused had a criminal record.

min Th : Diffe nti I A iati 11

Several theoretical frameworks are used to explain both the juvenile delinquency
and adult criminality concepts. Social Learning theory, more speciﬁcally Differential
Association, is used as the theoretical basis for understanding the link between child
abuse, juvenile delinquency, and adult criminality. Nevertheless, no single theory can
adequately address and/or explain the child abuse phenomenon (Egeland, 1991).

The early childhood period is characteristic of many fundamental changes and
learning where youth experience good and/or bad behavioral inﬂuences. The Social
Learning supposition falls under the broader category of behavioral theory, which states
that learning experiences shape behavior. Social Learning theory holds that an
individual’s delinquent or criminal tendency is fostered within their learning
environments such as the school or the home. Contrary to the Cesare Lombroso line of
reasoning, that people inherit criminal traits, social learning theorists believe that all
behaviors are learned.

Learning theory places great emphasis on the paucity of good parenting skills
possessed by the parent that ultimately results in bad practices by the child. Under the

learning theory perspective child abuse occurs because of a parent(s) or primary

33

caregiver’s inability to discipline the child in an appropriate manner or by using a method
that is not harmful.

The application of Social Learning theory follows the reasoning that behavior is
acquired through associations. Therefore, those acts learned in abusive environments
shape an individual’s subsequent behaviors. Speciﬁcally, it explains why an abused child
displays the violence learned inside the home, by leading a delinquent or criminal
lifestyle, for the youth is socialized in that particular antisocial manner. Ultimately, the
youth learns abusive behaviors that manifest themselves in the form of delinquent acts

instead of prosocial interactions and socializations.

Table 1: Sutherland’s Principles of Differential Association

 

Criminal behavior is learnt.

Criminal behavior is learnt in interaction with others in a process of communication.

The principal part ofthe learning ofcriminal behavior occurs within intimate personal groups.

When criminal behavior is learned, the learning includes (a) techniques of committing the crime,

which are sometimes very complicated, sometimes very simple, and (b) the speciﬁc direction of

motives, drives, rationalimtions, and attitudes.

5. The speciﬁc direction ofmotivesanddrives is learnedfromdeﬁnitions ofthe leplcodesas
favorable or unfavorable.

6. A person becomes delinquent beeause of an excess of deﬁnitions favorable to violations of law
over deﬁnitions unfavorable to violations of law.

7. Differential associations may vary in frequency, duration, priority, and intensity.

8. The process of learning criminal behavior by association with criminal and anticriminal patterns
involves all the mechanisms that are involved in any other learning. 7

9. While criminalbehaviorisanexpressionofgeneralneedsandvalws, itisnot explainedby those

genemlncedsanalwsdnmnmcﬁnﬁmlbemorismermressionofﬂwmmeneedsmd

values.

PPS“?

 

 

Source: Edwin Sutherland, Principles of Criminology 4Eli Edition. (Chicago: J.B. Lippincott, 1947), pp.
6-7 as cited in Sigel et al., 2003: 121.

 

Edwin Sutherland’s theory of Differential Association states that delinquent or
criminal behavior is learned through associations. The theory does not assert that it is the
association with criminal versus noncriminal individuals that produces nonconforming
lifestyles. Instead it is the interactions with individuals who are tolerant of or who

condone violations of the law, such as delinquent peers or siblings, versus those

34

 

individuals who do not (Barlow, 1990: 75; Albanese, 2001: 120; Schmalleger, 2003:
116). That is, delinquency or criminality results because ideas that justify crime outweigh
those that prohibit criminal behavior. Therefore, if a child is exposed to more
prodelinquent than prosocial behaviors the result is an increased involvement in or
possibility for delinquency and criminality. However, if a youth’s prosocial deﬁnitions
outweigh antisocial ones then the probability for both delinquency and criminality
decreases.

Sutherland’s theory has several basic principles that are outlined in Table 1 (page,
34). The theory is criticized for its very general deﬁnitions, its exclusion of other
important variables such as personality traits and psychological variables, and the lack of
an explanation for why individuals respond differently to the same situation (Barlow,
1990: 76; Siegel, 2003: 223). However, despite several criticisms of Sutherland’s theory
its importance to the ﬁeld of studying delinquent and criminal behaviors is well

established.

35

CHAPTER 3
W

Both empirical and theoretical evidence suggests that child victims of abuse are at
an increased likelihood for deviancy and ﬁnther behavioral problems in adulthood. The
study is intended to reiterate and evaluate the association between childhood
victimization, juvenile delinquency, and subsequent adult criminality.

The study will examine 1) if a relationship exists between the type of abuse and
juvenile delinquency, 2) if a relationship exists between the type of abuse and adult
criminality, and 3) if the relationship between child abuse, juvenile delinquency, and
adult criminality is affected by characteristics such as race, gender, age, ﬁnal placement,
drug or alcohol use, the perpetrator’s relationship to the victim, and socioeconomic
status.

The deﬁnition of child abuse is often plagued with difﬁculties where studies use
diverse conceptual and operational deﬁnitions to measure and describe the same
maltreatment (Egeland, 1991: 47; Mullen et al., 1996). For example, proper
differentiation must be made between abuse versus physical discipline or an accidental
versus an intentional incident (Jewell, 1999: 3). As such,

“Child maltreatment should be conceptualized and operationalized in a way that

reﬂects a comprehensive understanding of this form of socially opprobrious

behavior” (Brown, 1984: 260).

Nevertheless, theoretical clarity is often times absent from the conceptualization of
childhood abuse. Childhood victimization (measured by physical, sexual, or emotional

abuse) is broadly deﬁned as occurrences of ill-treatment toward a child by a parent,

36

primary caregiver, or any other individual. Demographic data and information is
collected regarding ofﬁcially reported delinquent behaviors and adult criminal
involvement. Information is obtained from both ofﬁcial sources of juvenile data and
police records.

$221.:

The database used is, “Childhood Victimization and Delinquency, Adult
Criminality, and Violent Criminal Behavior in a Large Urban County in the Northwest
United States, 1980-1997.” Research participants include a matched control group and a
sample of abused and/or neglected youth. The abused and/or neglected sample was made
dependents by the Superior Court of a large urban county in the Northwestern region of
the United States during the years 1980 through 1997.

Participants included in the study are those ages birth through 11 years during the
January 1, 1980 to December 31, 1984 period that had a dependency petition ﬁled with
the Court. Dependency petitions are ﬁled because an interested partyM alleges abuse
and/or neglect against the juvenile’s parent or primary caregiver. A ﬁled dependency
petition marks the courts initial involvement with the youth. Following a dependency
petition the juvenile is under the jurisdiction of the court because of the alleged abuse
and/or neglect. Juvenile dependency records are obtained from the county courthouse.

The initial sample of abused and/or neglected youth consists of 2,262 ﬁled

dependency petitions. A ﬁnal sample of 877 abused and/or neglected youth are included

 

MAnyindividualeant’rleadependencypetitionwiththejuvenilecourt.Aninterestedpartyincludes,butis
not limited to, theyouth’s grandparent or non-custodialparentaswell asChild Protective Services.
However,anymdividmlmatiswncemedwimmewelfamofmechﬂdismgardedasmhnereﬂedpany.

37

in the database. 1.5 Birth records are then collected from the Department of Health, for the
period January 1, 1969 through December 31, 1984, to obtain a control group matched on
gender, age, approximated socioeconomic status, and ethnicity.

With a matched pool of participants, child abuse and neglect registries are then
examined to exclude those youths in the control group with a history of abuse and/or
neglect. The ﬁnal sample of matched controls consists of 877 participants matched on the
characteristics of age, gender, ethnicity, and approximated socioeconomic status. The
ﬁnal sample included in the database comprises of a matched one-to-one control group of
877 cases and 877 abused and/or neglected subjects.

In this study the researcher was interested only in instances of physical, sexual,
and emotional abuse not cases of neglect. Cases are ﬁltered to exclude subjects with an
instance of neglect only (284) and those with no incident of abuse (877). A ﬁnal sample
of 593 abused youth is included in the analysis.

Pmedure

The Maltreatment Classiﬁcation Coding Scheme, ‘6 (MCS) originally developed
by Barnett, Manly, and Cicchetti, is modiﬁed and recoded to collect information on the
type of abuse and/or neglect reported that preceded the dependency petition ﬁled with the
court (1993). English and Widom (2003) modiﬁed that classiﬁcation scheme in order to

illustrate the particular types and severity levels of child maltreatment present in the

 

‘5 An initial sample of2,262 dependency petitions was obtained 187 eases were thm excluded from the
sample of abused and/or neglected children due to reasons such as, the child’s death, the unavailability of
the child’s dependency record or the child’s relocation out of the study area A further 2,075 children were
excluded from the study because their dependency record could not be located, the child resided outside of
thecountyofirrterestorthechildwasnotbominthestateofinterest.
“theMameamthodmngmemcrudedmunAppmdixsecumismemomnedmmdcveroped
by English and Widom (2003 ).

38

sample because the Barnett et al classiﬁcation scheme had no subtypes or measures of
severity for abuse. The measure for the severity of maltreatment ranges from low to high.

For both the MCS and the modiﬁed version child maltreatment types include:
moral, legal, or educational neglect, emotional abuse, neglect, sexual abuse, and physical
abuse. 17 However, the modiﬁed version further categorizes the type of abuse into
subtypes. For example, in the Barnett et a1 (1993) classiﬁcation scheme physical abuse
has no subtypes. However, in the modiﬁed version there are nine subtypes for physical
abuse based on the severity of the maltreatment.18

Data is provided for both juvenile and adult arrests and encounters with the law.
Juvenile arrest data is obtained through court records for both the matched control and the
abused and/or neglected groups. For all matched controls and the abused and/or neglected
subjects adult criminal arrest records are collected from local, county, state, and federal
law enforcement records.l9 Records are collected for the period January 1, 1980 through
December 31, 1997.

Both descriptive and inferential statistics are used in the analysis of data. The
general distribution and skewness of the data is examined through descriptive statistics.
Frequencies are run with the measures of central tendency as the primary descriptives.
That is, the mean, median, or mode, whichever is the most appropriate measure based on

the level of data analyzed.

 

l7Onlyreportsandmeasuresofphysierrl,sexual,andemotionalabusearweusedintheanalysisof
maltreatmentforjuveniles.Neglectconsﬁnnesthemcstprommenttypeofchﬂdabusebmﬁisnotincluded
intheanalysis.
18TheMaltreatmeutCodingSchemewillbediscussedingreaterdetailundertherespectiveheadingsof
childabuse.
'9Allroutinetraﬂicoﬂ’enseswereexcludedfmmthecormtofofﬁciallyrecordedoﬂ‘ensesforallstudy

participants.

39

Relationships are tested using the logistic regression technique. A series of
multivariate analyses are ﬁrst conducted to explore the associations between childhood
abuse, juvenile delinquency, and adult criminality. Multiple logistic regression is
employed as the appropriate technique for examining the effects of the predictor variables
on both the dichotomous outcome variables: delinquency and criminality. Signiﬁcance
levels, percentages, and odds ratios are also reported. Odds ratios are used because it
provides an estimate of the likelihood of a subject having a speciﬁc dependent variable
(delinquency or criminality) among those who exhibit a particular characteristic (sexual,
physical, or emotional abuse) relative to those who do not possess that trait.
Independent Variables
Gender

In the logistic regression models females are coded 0 and males 1. The
differences in gender for the abused sample are explored for juvenile delinquency and
adult criminality. The expected relationship is that males are both more likely to have a
juvenile delinquent as well as an adult criminal charge.

Race

Race serves as an identiﬁer variable. It characterizes ethnicity into three
categories, White, Black, and Other. The racial category Other consists of those subjects
who identify as Native Americans, American Asians, or American Paciﬁc Islanders.
Racial differences are examined for both adult criminality and juvenile delinquency. The
expected relationship is that Whites are less likely to exhibit both delinquent and criminal
behaviors. In the logistic regression models race is recoded into a dichotomous measure

with 0 (Whites) and 1 (N on-Whites).

40

Child’s Final Placement

This variable captures the residential placement of the abused child after the
dissolution of the dependency petition ﬁled with the Court. A child made dependent by
the court could be placed with the biological parent(s), a group home, adopted parent(s),
foster care, or some other residential placement. The relationship between ﬁnal
placement and subsequent delinquency as well as criminality is examined. The expected
relationship is that those youth placed outside the home are at an increased likelihood for
both delinquent and criminal activities.
Perpetrator’s Relationship to the Victim

This variable represents whether the perpetrator was a family member, a non-
family member, or a stranger. The expected relationship is that most perpetrators are
family members and not strangers to the victim. That relationship is hypothesized
because youths spend most of their time in the home environment. The variable was
recoded into a dichotomous measure for use in the logistic regression models where
family perpetrators (0) were compared to non-family perpetrators (1).
Socioeconomic Status

Socioeconomic status (SES) data are obtained from the Census Bureau of the
United States. Socioeconomic status is measured using census variables grouped
according to the ﬁve broad categories: (a) the percent of single female headed
households, (b) the median family income, (c) the percentage of individuals in the
household over 25 years with a high school diploma, (d) the percent on welfare, and (e)
the percent that lived below the poverty level. Socioeconomic status includes the youth’s

census tract of residence at birth. For this study the interval level variable mean family

41

income is used as a proxy for SES. The expected relationship is that those ﬁ'om a lower
SES are more likely to engage in delinquent and criminal behavior. It is also expected
that a positive relationship exists between a low SES and the occurrence of child abuse.
Age

This variable represents the number of years living.
Alcohol or Drug use

A dichotomous variable is used to measure the use of drugs or alcohol. The
variable does not measure the frequency, duration, or severity of drug or alcohol use. The
variable is simply a measure of whether or not a subject has used or has not used drugs or
alcohol. The variable is coded 0 (no drug or alcohol use) and 1 (drugs or alcohol use).
Physical abuse

Physical abuse includes any act that leads to the purposeful bodily disﬁgurement
of the youth involved (Siegel, 2003: 206). Physical abuse is a dichotomous variable based
on ofﬁcial records of child abuse only. The variable is coded 0 (no reported incident of
physical abuse) and 1 (one or more physical abuse incident).

The modiﬁed Maltreatment Classiﬁcation Scheme (MCS) is used to measure
instances of physical abuse among the youth sampled. Nine subtypes are identiﬁed to
describe physical abuse. Those subtypes are based on the location of the maltreatment on
the child’s body. Locations included the head, torso, buttocks, and limbs. Other measures
of physical maltreatment include, punching, choking, burning, violent handling, shaking
and the nondescript. The type of physical abuse is categorized into severity levels one (no
marks indicated) through six (permanent disability or death). For a detailed list of the

subtypes for physical abuse see appendices I & II.

42

Sexual abuse

Childhood sexual abuse is deﬁned as the unwanted occurrences of and the
exploitation of youth through the acts of rape, incest, or molestation (Siegel et al., 2003,
206). Instances of sexual abuse also include fondling the child, the child’s exposure to
indecent sexual material, and/or sodomy. The measure of sexual abuse includes both non-
contact and contact instances of maltreatment.

The modiﬁed version of the MCS was also used to measure instances of sexual
abuse. Sexual abuse is measured by a dichotomous variable based on oﬁicial records
only. The variable is coded as 0 (no sexual abuse) and 1 (a sexual abuse incident). The
measure of sexual abuse includes ﬁve severity levels ranging from instances where the
caregiver exposes the child to explicit sexual stimuli (severity level 1) to forced
intercourse with the child (severity level 5). For a detailed list of sexual abuse subtypes
and severity levels see appendices I & II.

Emotional abuse

Emotional abuse includes instances where the child is constantly criticized and
their emotional needs are rejected or ignored by a parent(s) or primary caregiver. It also
involves the inability of the parent/primary caregiver to provide for the child. Emotional
abuse is a dichotomous variable measured based on officially reported incidents only.
The variable is coded 0 (no emotional abuse) and 1 (one or more instance of emotional
abuse).

The modiﬁed MCS is used to measure the type and severity of emotional
maltreatment. Emotional maltreatment is typiﬁed by 27 codes and clustered into ﬁve

categories of severity. It is operationalized by the number of types of alleged emotional

43

abuse incidents reported. Using the modiﬁed MCS the types of emotional abuse include
the absence of age appropriate socializations for the child, calling the child derogatory
names, binding the child’s hands and feet for moderate time periods and the use of
conﬁnement and isolation, and threats of suicide or abandonment in the child’s presence.
Severity levels range ﬁ'om the caregiver expects inappropriate levels of responsibility
from the child (severity level 11) to instances where the caregiver conﬁnes the child for
extended periods that exceed eight hours (severity level 55). For a detailed list of severity
levels and subtypes for emotional abuse see appendix III.
Dependent Variablee
. Juvenile Delinquency

As noted by Whitehead and Lab (2006) the deﬁnition of juvenile delinquency
changes based on its purpose, that is, whether it be criminal law, status offenses, or social
or criminological.20 In the context of this paper, both the criminal law and status offense
deﬁnitions are applicable. Therefore, juvenile delinquency is deﬁned as any behavior that
if committed by an adult would be characterized as a crime. The term juvenile
delinquency is used only because of the age of the offender. That is, delinquency is
merely a substitute for criminality where the only distinction between the two types of
behavior is age. Additionally, delinquency is also deﬁned as any offense for which only a
juvenile can be charged such as curfew violations, running away, or truancy.

The juvenile delinquency variable measures the occurrences of delinquent

behavior by the youth that come to the attention of the juvenile justice system. An ofﬁcial

 

2° The social or criminologieal deﬁnition is the view of juvenile delinquency which takes into consideration
both the legal and status offense deﬁnitions of delinquent behavior. Tint perspective does not view juvenile
delinquency as a dichotomy. Since the researcher is limited to a dichotomy for juvenile delinquency the
social or criminologieal perspective is not applicable.

44

measure of delinquency is used in the analysis. Arrest data includes criminal histories,
excluding traffic offenses, for subjects at four levels of law enforcement: local, county,
state, and federal. Some juvenile charges include: a crime against property, drug/alcohol
offense and ﬁreworks violation. That variable measures only the occurrence of juvenile
delinquency and not the type, duration, or severity of acts committed. The dichotomous
delinquency variable is coded 0 (no juvenile charge) and 1 that represents (one or more
juvenile charge) before the age of 18.
Adult Criminality
An adult criminal charge is deﬁned as any offense that is committed by an

individual above the age of 18 for which the individual is arrested. Adult criminality is
measured by the occurrences of ofﬁcially reported criminal offenses at the local, county,
state, and federal levels. Adult charges include but are not limited to gambling, homicide,
assault, weapons violation, prostitution, and drug or alcohol offense. The dichotomous
variable adult criminality is coded 0 (no adult charge) and 1 (adult charge).
Limitetions of the Study

There are several noted limitations of the present study. The study will not
determine or evaluate the eﬁ‘ect(s) of each type of abuse on youth. The main purpose of
the study is to determine if a relationship existed between the type of abuse and a
delinquent or criminal charge. Due to very general and dichotomous measures of abuse
caution must be used when making inferences about the nature of the relationships. That
is, the analysis does not consider the frequency, duration, or severity of the type of abuse.
For instance, the nature of the relationship between physical abuse and juvenile

delinquency might differ if those factors were included in the analysis. The speciﬁc

45

nature of the relationship between each type of abuse and delinquency as well as
criminality is beyond the scope ’of the study.

Additionally, in the analysis, all types of abuse are measured singly limiting the
explanatory power of maltreatment to both juvenile delinquency and adult criminality.
Therefore, a different relationship might exist between the occurrence of both sexual and
emotional abuse to juvenile delinquency versus an instance of physical and emotional
abuse. That limitation is important because the research literature clearly establishes that
most abuse occurs in unison and not isolation.

The scope of the study is also limited because relevant variables are omitted ﬁ'om
the analysis such as parental mental health and substance abuse. The literature shows that
there is an apparent link between a parent’s mental health and child abuse. Mullen et al
(1996) conclude that a parent’s poor mental health21 increases a child’s risk factors for
abuse. In their analysis, both a father’s as well as a mother’s poor mental health increased
the odds of both physical and emotional abuse. Similarly, Flowers (1999: 153) states that
child abuse often times is viewed as an indication of parental substance abuse. If included
in the analysis the nature of the relationships between child abuse and juvenile
delinquency as well as adult criminality could be better explained.

The data collection procedure relies solely on ofﬁcial records for childhood abuse
and both juvenile and adult arrest information. That heavy reliance on official data is
problematic because not all instances of child abuse are reported, investigated, and
further substantiated. Adding to the problem is the fact that the child victim might not
have been made a dependent of the Court. Furthermore, not all delinquent or criminal

behaviors come to the attention of government authorities. Individuals in the abused

 

2‘ In the Mullen et al (1996) study the measure ofparent’s mental health included drug or alcohol use.

46

group could have committed delinquent and/or criminal act(s) that did not come to the
attention of the law. For instance, Huizinga and Elliott (1987) used longitudinal data ﬁom
the National Youth Survey22 (NY S) to examine juvenile offender prevalence, incidence,
and arrest rates by race. The analysis found that, through both self- and official reports,
that 84% of juvenile offenders are not officially arrested or haVe a crime reported against
them (Huizinga & Elliott, 1987). Therefore, not all incidents of juvenile delinquency or
adult criminality led to formal recognition.

The study is also limited to general and dichotomous measures of both juvenile
delinquency and adult criminality which do not take into consideration severity,
ﬁ'equency, or duration. A very cautious interpretation must be taken when making
inferences from the study because of the lack of controlling for those variables in the
analysis (Kendall-Tackett et al., 1993). The measure of both juvenile delinquency and
adult criminality is limited because they quantify only one deviant incident. If the
frequency, severity, and/or duration are included in the analysis it is expected that the
relationship between juvenile delinquency and child abuse would change substantially.
The same is expected for drug or alcohol use. Additionally, a relationship could be
reported if a measure of violent crime is used instead. The literature has not focused on
the impact of drug or alcohol use relative to a juvenile delinquent charge but instead as it

relates to violent crime (Flowers, 1999: 36).

 

22 Using a probability sampling technique the NYS sampled youths aged 11 through 17 years old in the
United States timing the year 1976.

47

CHAPTER 4

DATA ANALY§I§ AND FINDINQS

The initial sample of abused and/or neglected study participants consisted of 877
subjects. A ﬁrrther 284 participants were excluded from the analysis because of a report
of child neglect. Of the total sample of abused and/or neglected subjects 67.6% reported
an experience of sexual abuse or one or more instances of physical or emotional abuse.
Therefore, the ﬁnal count of the sample of abused participants included in the present
study was 593. All respondents had a juvenile dependency petition ﬁled with the Superior
Court during the period January 1, 1980 through December 31, 1984.

The age range to be included in the sample during the data collection period was
from birth through 11 years old. For those abused subjects age on March 1, 1999 was
recorded. The median age recorded was 23. The mean age recorded was 23.60 with a
standard deviation of 2.80. The minimum value for age wasl9 and the maximum was 30.
Therefore, the sample ranged from late adolescents to early adulthood. Females
comprised 56.2% of the sample while males accounted for the remaining 43.8%. With
respect to race, 77.0% of the respondents were White, 21.1% Black, and 1.9% comprised
the Other category.

or the abused subjects 10.7% were placed outside of the home23 following the
dependency petition ﬁled with the Court whereas 89.3% remained within the home
environment. The subjects’ natural parent(s), that is, 74.2%, committed the majority of
abuses recorded against study participants. Other perpetrators of abuse included: a step,

foster, or adoptive parent (12.1%), a parent’s non- or live in partner (6.8%), or a

 

23 Placement outside of the home includes kin care.

48

grandparent, sibling, or other non- or household member (6.8%). That ﬁnding goes to
further dispel the myth that most occurrences of child abuse occur outside of the victim’s
home environment. Instead, most often someone that the child is familiar with carries out
the abuse.

Only 11.3% of the sample of abused youth had any reported involvement with
drugs or alcohol whereas the remaining 88.7% did not. That low percent representing a
juvenile’s involvement with illegal substances is possibly due to the fact that information
on drug or alcohol use was collected by ofﬁcial means only. Therefore, the possibility
exists that the variable was underreported.

Mean family income was widely dispersed. Income was used a proxy for
socioeconomic status with a range ﬁ'om $10, 845 to $60, 299 {std = 6378.7, mean ($24,
128.53), and median ($24, 455)}. The premise was that those from a low SES (income
level) have an increased likelihood of both juvenile delinquent and adult criminal
behavior(s). Several outlier values for this variable caused the mean income level to be
greater than the modal recorded. That is, modal categories, $15, 180 and $18, 688,
recorded were below the stated mean and median.

Child abuse was measured by the incidents of physical, sexual, or emotional
maltreatment. A combined measure of maltreatment was not used in the analysis. Of the
‘ sample 42.2% reported one or more occurrence of physical abuse. However, the majority
of the sample (57.8%) did not report an incident of physical abuse. Approximately a half
of the sample (49.1%) had experienced one or more instance of emotional abuse whereas
the remaining 50.9% did not. The ﬁ'equency distribution for the ﬁnal measure of

maltreatment, sexual abuse, revealed that 35.6% of the respondents had a sexually

49

inappropriate encounter while the remaining 64.4% did not (see Tables 2, 2a and 2b,

 

 

pages 50 and 51).
Table 2: Frequency distribution of independent measures
Characteristics of the Sample
f %
Age
19 18 3.0
20 77 13.0
21 82 13 .8
22 67 11.3
23 65 11.0
24 51 8.6
25 66 11.1
26 53 8.9
27 57 9.6
28 26 4.4
29 30 5.1
30 1 .2
Gender
Male (1) 260 43.8
Female (0) 333 56.2
Race
White (0) 435 77.0
Blackll) 119 21.1
Other (2) 11 1.9
Drug/Alcohol involvement
No (0) 526 88.7
Yes (1) 67 11.3
Final Placement
In home ( 1) 484 89.3
Out-of-home (2) 58 10.7
Perpeuators relationship to the victim
Natural parent(s) (l) 434 74.2
Step/foster/adoptive parent (2) 71 12.1
Parent's non/live in partner (3) 40 6.8
Grandparent/sibling/other 40 6.8
non/household member (4)
Physical Abuse
No (0) 343 57.8
One or more (1) 250 42.2
Emotional Abuse
No (0) 302 50.9
One or more (1) 291 49.1
Sexual Abuse
No (0) 382 64.4
Yes (1) 211 35.6

 

50

Table 2a: Descriptive Statistics for independent measures

 

 

Measure Range Median Mean SD
Age 19 - 30 23 23.60 2.80
Mean family income 10. 845 - 60. 299 24. 455 24. 128.5 6378.7

 

Table 2b: Descriptive Statistics for independent measures

 

 

Measure Range Median Mode _S_ll

Gender 0-1 0 0 .4_9
Race 0-2 0 0 it;
Drug/alcohol involvement 0-1 0 0 g
Final Placement 0-1 1 1 .3_1
Perpetlator‘s relationship to the victim 1-4 1 1 .8_9
Physical Abuse 0-1 0 0 .4_9
Sexual Abuse 0-1 0 0 ﬂ
Emotional Abuse 0-1 0 0 .551

 

The majority of the sample of abused youth did not have an ofﬁcially recorded
juvenile delinquent or adult criminal charge, 82.0% and 59.7% respectively. Only 18.0% of

the abused group had a juvenile charge whereas 40.3% recorded an adult charge (see Tables 3

and 3a, pages 51 and 52).

Table 3: Frequency distribution of dependent measures

 

 

Clmracteristics of the Sample
n 0/o

Juvenile Delinquency

No delinquent charge (0) 486 82.0

Delinquent charge (1) 107 18.0
Adult Criminality

No criminal charge (0) 354 V 59.7

Criminal charge (1) 239 40.3

 

51

Table 3a: Descriptive Statistics for dependent measures

 

 

Measure Range Median Mode SD
Juvenile Delinquency 0 — 1 0 0 .38
Adult Criminality 0 - 1 0 0 .49

 

Bivariate Analysis (Juvenile Delingueng)

A signiﬁcant correlation was found between age and delinquent involvement.
There was an inverse correlation between age and a delinquent charge with a Pearson’s r
value of -.095 (n = 593, p < .021). That is, as age increases involvement in delinquency

also decreases (see Table 4, page 52).

Table 4: Pearson’s Correlation Coefﬁcients for Juvenile Delinquency

 

 

(N = 593)
Variable Correlation Coefﬁcient Sig
Age -.095* .021
Mean Family Income -.028 .502

 

* Correlation is signiﬁcant at the .05 level (2-tai1ed).

A signiﬁcant relationship (phi = .310, p < .000) was reported between gender and
subsequent involvement in delinquent behavior(s). A signiﬁcant lambda (2.) value of .155
was observed. Therefore, the prediction error with regards to delinquency decreases by
15.5% if the subject’s gender is known. In the sample, 68.5% of all males and 92.5% of
all females had no recorded delinquent offense. The remaining 7.5% of all females and
31.5% of all males had an ofﬁcial delinquent charge. That is, 63.4% of those with no
recorded delinquent charge were female and the remaining 36.6% were male. For
participants with a delinquent offense 23.4% were females whereas 76.6% were males

(see Table 4a, page 53).

52

The relationship between race and a juvenile delinquent charge was signiﬁcant
with a Cramer’s V value of .197 (n = 593, p < .000). For those with a delinquent charge,
62.4% were Whites and 37.6% were Blacks. Those in the Others category recorded no
juvenile offense. Similarly, only 14.5% of all Whites and 31.9% of all Blacks sampled
had committed a juvenile delinquent offense. Therefore, in terms of proportionality a

greater percentage of Blacks had a delinquent charge (see Table 4a, page 53).

Table 4a: Crosstabulation of juvenile delinquency and independent measures

 

 

Juvenile Delinquency
Yes No
f ”/o f %
Gender
Male 82 76.6 178 366*
Female 25 23.4 308 63.4
Race
White 63 62.4 372 80.2“
Black 38 37.6 81 17.5
Other 1 1 2.4
Drug/Alcohol involvement
No 91 85.0 435 87.5
Yes 16 15.0 51 10.5
Final Placement
In home 78 83 .9 406 90.4
Out of home 15 16.1 43 9.6
Perpetrator‘s relationship to the victim
Natural parents 86 81.1 348 72.7
Step/foster/adoptive 8 7.5 63 13.2
Parent's non/live in partner 5 4.7 35 7.3
Grandparent/sibling/other 7 6.6 33 7 .3
Non/household member
Physical Abuse
No 56 52.3 287 59.1
Yes 51 47. 199 40.9
Emotional Abuse
No 40 37.4 262 639*
Yes 67 62.6 224 46.1
Sexual Abuse
No 86 80.4 296 60.9"
Yes 21 19.6 190 39.1

 

* Signiﬁcant at p < .05

53

There was a signiﬁcant relationship between sexual abuse and a delinquent charge
(Cramer’s V = .156, p < .000). Only 19.6% of participants with an instance of sexual
abuse had a recorded delinquent charge while the remaining 80.4% who experienced
sexual abuse were not delinquent. Of the sample 39.1% had a sexually abusive incident
but no delinquent charge whereas 60.9% had a joint distribution of no abuse or juvenile
delinquency. Only 3.5% (21) of the total sample had a joint distribution of delinquency
and sexual abuse (see Table 4a, page 53).

A signiﬁcant relationship was noted between emotional abuse and delinquency
(phi = .127, p < .002). For participants who recorded one or more instance(s) of
emotional abuse 62.6% also had a delinquent charge. Conversely, 37.4% of the subjects
with a delinquent offense recorded no emotionally abusive incident. The joint distribution
of no delinquent offense and no emotional abuse was represented by 53.9% of the
sample. A signiﬁcant and symmetric lambda value of .068 (p < .009) was recorded.
Therefore, if the experience of emotional abuse is known the error in predicting a
juvenile delinquent charge decreases by 6.8% (see Table 4a, page 53).

Bivariate analysis did not reveal a signiﬁcant relationship between a child’s
placement and subsequent delinquency (phi value = .080, p < .063). The majority
(90.4%) of subjects in in-home placement had no delinquent offense. The remaining
9.6% placed outside the home had no delinquent charge. Conversely, 16.1% of
participants who were placed in out-of-home care subsequently offended whereas 83.9%
placed in home had a delinquent charge. Only 2.8% of the sample had a joint distribution

of delinquency and out-of-home placement (see Table 4a, page 53).

54

An insigniﬁcant relationship existed between income and juvenile delinquency. A
Pearson’s r value of -.028 (n = 593, p < .502) was reported. Although insigniﬁcant there
was an inverse association between mean income and a juvenile delinquent charge.
Therefore, an increase in income leads to a decrease in the likelihood of delinquency
though the relationship was statistically insigniﬁcant (see Table 4, page 52).

Bivariate analysis revealed an insigniﬁcant relationship between the perpetrator’s
relationship to the victim and a delinquent juvenile charge (Cramer’s V = .082, p < .267).
For those subjects with a delinquent offense 81.1% of the perpetrators of abuse was' a
natural parent(s). Other recorded percentages included: 7.5% by a step, foster, or adoptive
parent, 4.7% by a parent’s non- or live-in partner, and 6.6% by a grandparent, sibling, or
other non- or household member (see Table 4a, page 53).

There was an insigniﬁcant relationship between drug or alcohol use and
delinquency with a phi value of .054 (p < .871). Four-hundred and thirty-ﬁve (89.5%)
study subjects had a joint distribution of no drug or alcohol use and no juvenile
delinquent offense. The remaining 10.5% reported drug or alcohol use but no ofﬁcially
delinquent offense. For subjects with a juvenile delinquent charge 85.0% had no drug or
alcohol use whereas the remaining 15.0% did. Only 2.7% (16) of abused youth had a
joint distribution of delinquency and drug or alcohol use (see Table 4a, page 53).

A statistically insigniﬁcant relationship existed between physical abuse and
juvenile delinquency (phi = .052, p < .203). For those with no incident of physical abuse
59.1% recorded no delinquency. The remaining 40.9% had a joint distribution of no
juvenile charge and one or more instance of physical abuse. A joint distribution of

delinquency and physical abuse was recorded for 47.7% of the sample. Only 8.6% (51) of

55

the total sample experienced both a juvenile delinquent and at least one physical abuse
event (see Table 4a, page 53).

In summary, there was a statistically signiﬁcant relationship between gender,
race, age, emotional abuse, and sexual abuse relative to a juvenile delinquent charge. On
the other hand, a signiﬁcant relationship was not noted between a child’s placement, the
perpetrator’s relationship to the victim, drug or alcohol involvement, mean family
income, or physical abuse.

Biva ’ te All is Adult riminali

A signiﬁcant correlation was found between age and an adult criminal charge.
There was a direct relationship based on Pearson’s r value .134 (n = 593, p < .001). That
is, an increase in age also increases criminality, a ﬁnding contrary to a juvenile
delinquent charge. A signiﬁcant relationship existed between mean family income and
adult criminality. Based on the Pearson’s r value (-.133, p < .001) there was an inverse
association between income and criminality. Therefore, an increase in income was

statistically signiﬁcant to a criminal charge (see Table 5, page 56).

Table 5: Pearson’s Correlation Coefﬁcients for Adult Criminality

 

 

(N = 593)
Variable Correlation Coefficient Sig
Age .134* .001
Mean Family Income -.133" .001

 

* Correlation is signiﬁcant at the .01 level (2-tailed).

Bivariate analysis revealed a signiﬁcant relationship between gender and
criminality (phi = .334, p < .000). For those with no recorded criminal charge 69.8%
were females whereas the remaining 30.2% were male subjects. In terms of participants

with an adult offense 36.0% were females whereas 64.0% were males. A signiﬁcant

56

lambda value of .226 was recorded. Therefore, if a subject’s gender was known the
prediction error for criminality decreases by 22.6% (see Table 5a, page 57).

There was a signiﬁcant relationship between a subject’s race and a subsequent
adult criminal charge (phi = .203, p < .000). Of those with an adult charge 67.0% were
Whites, 31.3% Blacks, and 1.8% Others. Additionally, 34.5% of all Whites, 58.8% of all
Blacks, and 36.4% of all Others recorded an adult criminal charge. A lambda value of

.059 (p < .053) was recorded but not signiﬁcant (see Table 5a, page 57).

 

 

Table 5a: Crosstabulation ofadult criminality and indemdeut measures
Criminality
Yes No
f % f %
Gender
Male 153 64.0 107 30.2“
Female 86 36.0 247 69.8
Race
White 150 67.0 285 836*
Black 70 31.3 49 14.4
Other 4 1.8 7 2.1
Drug/Alcohol involvement
No 207 86.6 319 90.1
Yes 32 13.4 35 9.9
Final Placement
In home 190 89.2 294 89.4
Out of home 23 10.8 35 10.6
Perpetrator's relationship to the victim
Natural parents 188 79.7 246 70.5
Step/foster/adoptive l9 8. 1 52 14.9
Parent‘s non/live in partner 14 5.9 26 7.4
Grandparent/sibling/other 15 6.4 25 7.2
Non/household member
Physical Abuse
No 122 51.0 221 62.4“
Yes 117 49.0 133 37.6
Emotional Abuse
No l 12 46.9 . 190 53. 7
Yes 127 53.1 164 46.3
Sexual Abuse
No 175 73.2 207 58.5“
Yes 64 26.8 147 41.5

 

* Signiﬁcant at p < .05.

57

Bivariate analysis revealed an insigniﬁcant relationship (phi = .003, p < .953)
between ﬁnal placement and subsequent criminality. Of those with an adult criminal
charge 89.2% had been placed inside the home whereas 10.8% were in out-of-home
placement. For those with no adult charge 10.6% were placed outside of the home
whereas the remaining 89.4% remained in the home. Only 4.2% (23) of the sample had a
joint distribution of adult criminality and out-of-home placement (see Table 5a, page 57).

For those with no adult charge 62.4% did not report an incident of physical abuse
whereas 37 .6% did. For those subjects with an adult criminal charge 51% had no
incident(s) of physical abuse. A signiﬁcant relationship was noted between criminality
and abuse (phi = .113, p < .006). For study participants with one or more instance of
physical abuse 49% had a criminality charge. Only 19.7% (117) of the total sample had a
criminal charge and at least one instance of physical abuse (see Table 5a, page 57).

A signiﬁcant relationship exists between adult criminality and the experience of
sexual abuse (Cramer’s V = .151, p < .000). For those with no adult oﬁ‘ense 41.5%
reported sexual abuse whereas 58.5% did not. Conversely, for those study subjects with
an adult charge 73.2% reported no sexual abuse. The remaining 26.8% reported a
sexually abusive incident. Only 10.8% (64) of the total sample had a joint distribution of
sexual abuse and adult criminality (see Table 5a, page 57).

An insigniﬁcant association exists between the perpetrator’s relationship to the
victim and adult criminality (Cramer’s V = .114, p < .054). For those study subjects with
a criminal charge 79.7% of the perpetrators of abuse was a natural parent(s). Other

recorded percentages included: 8.1% by a step, foster, or adoptive parent, 5.9% by a

58

parent’s non- or live-in partner, and 6.4% by a grandparent, sibling, or other non- or
household member (see Table 5a, page 57).

An insigniﬁcant relationship existed between drug or alcohol use and adult
criminality (phi = .054, p < .186). Similar percentages were reported for the joint
distribution of drug and alcohol use and adult criminality, as were those for delinquency
and drug use. For those with no adult charge 90.1% had no prior drug involvement while
the remaining 9.9% did have drug use. For subjects with an adult charge 86.6% recorded
no drug use. On the other hand, the remaining 13.4% did have drug involvement. Only
5.4% of the total sample had a joint distribution of criminality and drug use. (see Table
5a. page 57)

Of the sample 53.1% had a joint distribution of criminality and one or more
instance of emotional abuse. On the other hand, 46.9% had both a criminal charge and no
instance of emotional abuse. For those with no criminal offense 53.7% had no emotional
abuse. An insigniﬁcant relationship was noted between emotional abuse and criminality
(phi = .067, p < .104). A statistically insigniﬁcant relationship was also noted by the
lambda value of .028 (p < .332). Of the total sample 21 ..4% recorded both an adult charge
and one or more instance of emotional abuse (Table 5a, page 57).

Bivariate analysis revealed a signiﬁcant relationship between age, income, gender,
physical abuse, and sexual abuse to that of an adult criminal charge. However, an insigniﬁcant
association was noted for a child’s ﬁnal placement, the perpetrator’s relationship to the
victim, drug or alcohol use, and one or more instance of emotional abuse.

Bivariate analyses were conducted to determine the association between the

various independent variables and each dependent measure. Several signiﬁcant

59

relationships were identiﬁed throughout the study for both juvenile delinquency and adult
criminality. For both dependent measures age, gender, race, and sexual abuse were
signiﬁcant in the analysis. There was an inverse correlation between age and a juvenile
delinquent charge whereas a direct relationship existed between age and an adult criminal
charge. In terms of gender, males were more likely than females to have both a
delinquent and a criminal charge. Similar percentages were reported for juvenile
delinquency and adult criminality by race. In terms of proportionality Blacks had more
instances of both delinquency and criminality. The majority of the sample that had been
sexually abused did not have a delinquent or a criminal charge.

Insigniﬁcant relationships for delinquency as well as criminality were also noted
for a child’s ﬁnal placement, the perpetrator’s relationship to the victim, and drug or
alcohol use. In terms of a child’s ﬁnal placement the majority of delinquent and criminal
charges were committed by those abused youth placed in-home. For both delinquency
and criminality natural parents represented the majority of child abusers. Only a small
percentage of the sample had a delinquent or criminal charge relative to drug or alcohol
use. At the bivariate level, emotional abuse was signiﬁcant for a delinquent charge but
insigniﬁcant for a criminal charge. Additionally, socioeconomic status and physical abuse
were insigniﬁcant to a delinquent charge but signiﬁcant to an adult criminal charge.
Regmeion Ana_l!sis

Two main regression models were speciﬁed in the analysis. 24 Model 1 measured
the predictability of the various independent variables on juvenile delinquency and Model

2 for adult criminality.

 

24'l‘heuiainregressionmodelsforjuveniledelinquencyandadultcriminalitywerespeciﬁedbyphysielrl
abumbeeausemeassrmpﬁmistmwsumgaassodaﬁmerdstsbaweenmmrypeofamrelativeto

60

The main model for juvenile delinquency (Model 1) correctly classiﬁed 83.4% of
the cases. There was a statistically signiﬁcant relationship between the independent
variables of age, race, and gender relative to a juvenile delinquent charge. The model
shows that an increase in age resulted in a decline in the log odds of a delinquent charge
by .907 (p < .050), holding all other variables constant. The log odds of a juvenile
delinquent charge are positively related to race25 and gender. That is, the log odds of a
delinquent charge are higher for males than females based on the unstandardized (B =
1.825, p < .000) and standardized (b = 6.202, p < .000) betas." Similarly, non-Whites
were more likely to have a delinquent charge (B = 1.090, b = 2.973, p < .000).

The proxy used for SES shows no statistically signiﬁcant relationship with
delinquency based on the unstandardized (B = .000, p < .671) and standardized (b =
1.000, p < .671) beta values. That is, when mean family income increased by one unit the
odds of delinquency was unchanged, holding all other variables constant. There was no
statistically signiﬁcant relationship between a child’s placement, the perpetrator’s
relationship to the victim, and drug or alcohol use to delinquency. That is, as a
perpetrator’s relationship moved ﬁ'om a natural parent to a stranger the odds of

delinquency decreased, holding all other variables constant. Similarly, one or more

 

bmhmmﬁemoumalmﬂuwunmmmmnmodekwnsiaedofmemdepmdanmeaaueszﬁml
placement, mean family income (SES), age, perpetrator’s relationship to the victim, physical abuse, drug or
alcohol use, gender, and race.

25 Whites are used as the reference group in all speciﬁed regression models.
26Standardizedbetameasureshowwelleachindependentvaliablesprcdictsthedependentvariablewhen
allotherindependentvariablesareheldconstant Standardizedbeta(b)represemstheoddsratio
(likelihood) of an outcome occurring relative to the reference category. Values are standardized with a
meanoanndastandarddeviationofl. Anoddsratiovaluegreaterthan 1.00indieatesanincreaseinthe
likelihood of the event Odds ratios less than 1.00 indicate a decreamd likelihood of the event occurring
Standardized betas are used when they are multiple independent variables measured in different units It
allows for the comparison ofeﬁ’ects. Unstandardizedbeta (B) isalsousedfor makingapredictionbutthe
independentvariablesarekeptintheiroriginal units. Forexample,thevariableageismeasruedinyears
thereforetheunstandardizedcoeﬁicientisalsoinyears.

61

incident of physical abuse and drug or alcohol use both decreased the log odds of
delinquent involvement whereas out-of-home placement increased the odds (see Table 6,
page 62). In the main model for juvenile delinquency gender had the greatest explanatory
power because of a larger Wald statistic (38.624).
Table 6: logistic Regression Analysis on Juvenile Delinquency - Model 1
Main Juvenile Delinquency Model - Model 1 (N = 593)
B SE Wald b Sig

 

Constant -.296

Final Placement .415 .386 1.157 1.514 .282
Mean Family Income .000 .000 . 180 1.000 .671
Age -.097 .050 3.845 .907 050*
Perpetrator's relationship to -. 176 .327 .291 .838 .590

the victim

Physical Abuse -.449 .260 2.973 .638 .085
Drug or Alcohol involvement -.452 .362 1.558 .636 .212
Gender 1.825 .294 38.624 6.202 000*
Race 1.090 .308 12.479 2.973 .000“

 

* Correlation is signiﬁcant at the .05 level (2-tailed).

The main model for adult criminality (Model 2) correctly classiﬁed 69.4% of the
cases. The model produced similar results to that of a delinquent charge. There was a
statistically signiﬁcant and positive relationship between age, gender, and race with an
adult criminal charge. The model shows that an increase in age indicates an increase in
the log odds of criminality, holding all other variables constant, based on unstandardized
(B = .135, p < .000) and standardized (b = 1.145, p < .000) betas. Furthermore, the log
odds of an adult criminal charge are higher for males (B = 1.565, b = 4.784, p < .000) and
non-Whites (B = .8766, b = 2.401, p < .000).

No statistically signiﬁcant relationship was noted between socioeconomic status
and adult criminality baSed on the unstandardized (B = .000, p < .060) and standardized
(b = 1.000, p < .060) beta values. A child’s placement, drug use, and the perpetrator’s

relationship to the victim also had no signiﬁcant relationship with an adult criminal

62

charge. Drug or alcohol use as well as out-of-home placement were associated with a
decrease whereas non-parental abuse increased the log odds of criminality. Furthermore,
one or more incident of physical abuse resulted in a decrease in the odds of adult
criminality, holding all other variables constant. However, those ﬁndings had no
statistical signiﬁcance. In the model, gender had the greatest explanatory power with a
recorded Wald value of 53.5 10 (see Table 7, page 63).

Table7zlogisticRegressionAnalysisonAdultCriminality-Mode12
Main Adult Criminality; Model — Model 2 (N = 593)

3 SE Wald b Sig

 

Constant -3.298

Final Placement -.064 .335 .036 .938 .849
Mean Family Income .000 .000 3.541 1.000 .060
Age .135 .038 12.325 1.145 000*
Perpetrator‘s relationship to .036 .235 .023 1.036 .880

the victim

Physical Abuse -.382 .204 3.498 .683 .061
Drug or Alcohol involvement -.321 .304 1.1 16 .725 .291
Gender 1.565 .214 53.510 4.784 .000“
Race .876 .257 11.609 2.401 .001"

 

* Correlation is signiﬁcant at the .05 level (2-tailed).

Four additional regression models were utilized in the subgroup analysis for both
juvenile delinquency and adult criminality. Model 3 speciﬁed the effect of sexual abuse
on a juvenile delinquent charge and Model 4 its effect on an adult criminal charge. Model
5 assessed the relationship between emotional abuse and juvenile delinquency. Model 6
looked at the relationship between emotional abuse and adult criminality.

The independent variables: age (B = -.101, b = .904, p < .043), race (B = 1.033, b
= 2.808, p < .001), and gender (B = 1.768, b = 5.861, p < .000) remained signiﬁcant in
the analysis of juvenile delinquency when sexual abuse was added. In Model 3, an

increase in age indicated a decrease in the log odds of juvenile delinquency, holding all

63

other variables constant. Also, being male or non-White increased the odds of a
delinquent charge.

In Model 3, no change was reported in the signiﬁcance of the other independent
variables. Though of no statistical importance, ﬁnal placement and perpetrator’s
relationship to the victim both recorded a positive relationship to delinquent involvement.
Out-of-home placement and non-parental abuse indicated an increase in the log odds of a
juvenile delinquent charge. Socioeconomic status (B = .000, b = 1.000, p < .621)
indicated that with an increase in the independent variable by one unit the log odds of a
juvenile delinquent charge was unchanged, holding all other variables constant. Sexual
abuse had no statistically signiﬁcant relationship to a delinquent charge based on beta
values (B = .516, b = 1.675, p < .127). Though no statistically signiﬁcant association was
noted the presence of sexual abuse indicates an increase in the log odds of juvenile
delinquency, holding all other variables constant. The greatest explanatory power again

rested with gender with a Wald statistic of 35.289 (see Table 8, page 64).
Table8: IngisticRegressionAnalysisonJuvenile Delinquency-Mode”
Sexual Abuse Model -— Model 3 (N = 593)

B SE wind b Sig

 

Constant -.895

Race 1.033 .312 10.973 2.808 .001‘
Fiml Placement .340 .386 .777 1.406 .378
Mean Family Income .000 .000 .245 1.000 .621
Age -. 101 .050 4.087 .904 .043“
Perpetrator's relationship to .011 .334 .001 1.011 .974

the victim

Sexual Abuse .516 .338 2.326 1.675 .127
Drug or Alcohol involvement -.403 .363 1.228 .669 .268
Gender 1.768 .298 35.289 5.861 000*

 

" Correlation is signiﬁcant at the .05 level (2-tai1ed).

Model 4 assessed the relationship speciﬁed by sexual abuse on an adult criminal

charge. The overall test of the model correctly predicted 70.2% of the cases of adult

64

criminality. Three statistically signiﬁcant relationships were found in that model. A
statistically signiﬁcant and positive relationship existed between race, age, and gender to
criminality. The model shows that non-Whites and males had a greater likelihood of
criminality than Whites and females. Additionally, an increase in age by one unit
indicated an increase in the log odds of a criminal charge, holding all other variables
constant.

SES (B = .000, b = 1.000, p < .089) had no signiﬁcant effect on an adult criminal
charge. There was no signiﬁcant relationship between a child’s placement, the
perpetrator’s relationship to the victim, and drug or alcohol use to an adult criminal
charge. Sexual abuse had no signiﬁcant effect on an adult criminal charge. Although the
relationship was insigniﬁcant the data suggests that out-of-home placement, an incident
of sexual abuse or drug use decreased the log odds of criminality. On the other hand, non-
parental abuse increased the odds of adult criminal behavior, holding all other variables

constant (see Table 9, page 65).

Table9:LogisticRegresﬁonAnalysisonAdultCriminaIity—Model4

Sexual Abuse Model - Model 4 (N = 593)

 

3 SE wind b Sig

 

Constant -3.482

Race .914 .261 12.291 2.494 .000*
Final Placement -. 104 .334 .098 .901 .755
Mean Family Income .000 .000 2.896 1.000 .089
Age .132 .038 11.831 1.141 001*
Perpetrator‘s relationship to .039 .243 .026 1.040 .873

the victim

Sexual Abuse -.066 .236 .079 .936 .778
Drug or Alcohol involvement -.340 .305 1.240 .712 .265
Gender 1.625 .221 53.869 5.079 000*

 

* Correlation is signiﬁcant at the .05 level (2-tailed).

Model 5 assessed the impact of emotional abuse on a juvenile delinquent charge.

The model correctly classiﬁed 82.5% of the cases of juvenile delinquency. The

65

independent variables age, gender, and race remained signiﬁcant in the analysis of
emotional abuse. An inverse and signiﬁcant relationship exists between age and juvenile
delinquency. Therefore, an increase in age indicated a decrease in the log odds of a
delinquent charge, holding all other variables constant (B = -.099, b = 906, p < .048). A
positive and signiﬁcant relationship was noted between race and delinquency as well as
gender and delinquency. The log odds of a delinquent charge are greater for males (B =
1.817, b = 6.152, p < .000) and non-Whites (B = 1.051, b = 2.860, p < .001).

Based on the unstandardized (B = .000, p < .559) and standardized (b = 1.000, p <
.559) beta values SES had no effect on a delinquent charge, holding all other variables
constant. Therefore, with an increase in income by one unit the log odds of a delinquent
charge remain unchanged. A positive but statistically insigniﬁcant relationship existed
between ﬁnal placement and delinquency (B = .343, b = 1.410, p < .373). A negative and
insigniﬁcant relationship was noted for perpetrator’s relationship, drug or alcohol use,
and emotional abuse. Though insigniﬁcant, emotional abuse decreased the log odds of a
delinquent charge, holding all other variables constant (see Table 10, page 66).

Table 10: Logistic RegressionAnalysisonJuvenileDelinquency-Models
Emotional Abuse Model — Model 5 (N = 593)

 

B SE Wald b Sig

 

Constant -.482

Race 1.051 .312 11.364 2.860 .001*
Final Placement .343 .385 .794 1.410 .373
Mean Family Income .000 .000 .342 1.000 .559
Age -.O99 .050 3.915 .906 .048*
Perpetrator‘s relationship to -.022 .333 .004 .978 .947

the victim

Drug or Alcohol involvement -.386 .365 1.118 .680 .290
Gender 1.817 .294 38.198 6.152 000*
Emotional Abuse -.361 .273 1.755 .697 .185

 

* Correlation is signiﬁcant at the .05 level (2-tailed).

66

Model 6 speciﬁed the relationship between emotional abuse to that of adult
criminality. The model correctly classiﬁed 69.8% of the adult criminality cases. Similar
to the analysis of juvenile delinquency, that of criminality shows that the independent
variables (age, gender, and race) remained signiﬁcant in the analysis. All three variables
had a positive relationship with an adult criminal charge. An increase in age by one unit
resulted in an increase in the log odds of criminal behavior (B = .132, b = 1.141, p <
.001), holding all other variables constant. Based on the beta values (B = 1.619, b =
5.049, p < .000) in Model 6 males were at an increased likelihood for criminal behavior
than females. Also, non-Whites had a signiﬁcantly greater log odd of adult criminality
than Whites (B = .910, b = 2.485, p < .000).

Several insigniﬁcant relationships were produced in Model 6. A positive yet
statistically insigniﬁcant relationship was produced for ﬁnal placement, perpetrator’s
relationship to the victim, and emotional abuse. Out-of-home placement decreased the
log odds of criminal behavior (B = -. 103, b = .902, p < .095). Non-parental abuse was
found to increase the odds of an adult criminal charge. The data also suggests that a one
unit change in emotional abuse, that is, having one or more incident also increased the
log odds of criminal behavior (B = .073, b = 1.076, p < .120). Based on the
unstandardized (B = .000, p < .083) and Standardized (b = 1.000, p < .083) beta values the
data suggests that SES had no effect on a criminal charge holding all other variables
constant. Therefore, with an increase in mean family income by one unit criminality

remained unchanged (see Table 11, page 68).

67

Table 11: Logistic Regression Analysis on Adult Criminality - Model 6
Emotional Abuse Model - Model 6 (N = 593)

E SE Wald b Sig

 

Constant -3.546

Race .910 .258 12.443 2.485 000*
Final Placement -. 103 .334 .095 .902 .758
Mean Family Income .000 .000 3.005 1.000 .083
Age .132 .038 11.873 1.141 001*
Perpetrator‘s relationship to .040 .239 .029 1.041 .866

the victim

Drug or Alcohol involvement -. 347 .307 1.277 .707 .258
Gender 1.619 .215 56.568 5.049 000*
Emotional Abuse .073 .212 . 120 1.076 .729

 

* Correlation is signiﬁcant at the .05 level (2-tailed).

The two ﬁnal regression models consisted of those variables that were found to be
signiﬁcant throughout the analysis. In both Model 7 and Model 8 the independent
variables included are age, race, and gender.

In Model 7 the overall test of the model correctly classiﬁed 80.9% of the cases of a
juvenile delinquent charge. In Model 7 (see Table 12, page 62) an increase in age
indicated a decrease in the log odds of a juvenile delinquent charge, holding all other
variables constant. Males were also found to have an increased likelihood of a delinquent
offense. Contrary to prior regression models, in Model 7 Non-Whites had a decreased log
odds of delinquency (B = -.999, b = .368, p < .000), holding all other variables constant.

Table 12: Logistic Regression Analysis on Juvenile Delinquency — Model 7
Significant Effects Model - Model 7 (N = 593)

3 SE Wald b Sig

 

Constant .386

Age ' -.096 .044 4.645 .909 031*
Gender 1.805 .262 47.500 6.078 000*
Race -.999 .257 15.145 .368 000*

 

* Correlation is signiﬁcant at the .05 level (2-tai1ed).

68

In Model 8, the overall test of the model correctly classiﬁed 69.9% of the cases of an
adult charge. In Model 8 an increase in age indicated an increase in the log odds of
criminal behavior, holding all other variables constant. Males were had an increased
likelihood of adult criminal behavior in comparison to females. Contrary to prior
regression models, in Model 8 non-Whites had a decreased likelihood for adult
criminality (see Table 13, page 63).

Table 13: Logistic Regression Analysis on Adult Criminality - Model 8

Signiﬁcant Effects Model — Model 8 (N = 593)

 

B SE Wald b Sig

 

Constant -3.643

Age .138 .035 15.308 1.148 000*
Gender 1.590 .196 65.988 4.900 000*
Race -1.026 .223 21.103 .359 000*

 

* Correlation is signiﬁcant at the .05 level (2-tailed).

The variables race, age, and gender were signiﬁcant in all the regression models.
In the absence of abuse types ﬁ'om the analysis the signiﬁcant effects regression models
showed that Whites had a greater likelihood for both a juvenile delinquent and an adult
criminal charge. That ﬁnding is inconsistent with the literature which clearly establishes
that minority groups are more delinquent and criminal (Paschall et al., 1998; Piquero et
al., 2002). However, in all other regression models speciﬁed by abuse types non-Whites
had increased log odds for both delinquent and criminal behavior.

Throughout the analysis, an increase in age by one unit indicated a decrease in the
log odds of juvenile delinquency. On the other hand, an increase in age resulted in an
increase in the odds of an adult criminal charge. The latter study ﬁnding is inconsistent
with the research literature. Males were consistently found to have an increased

likelihood for delinquent as well as criminal behavior throughout the regression analyses.

69

All other independent variables remained insigniﬁcant throughout the analysis.
For all regression models socioeconomic status was unrelated to both a delinquent and
criminal charge. In both of the main models (physical abuse) and the sexual abuse models
out-of-home placement increased the log odds of delinquency and decreased the
likelihood of a criminal charge. On the other hand, in relation to the emotional abuse
models out-of-home placements increased both the odds of delinquency and criminality.

Drug or alcohol use consistently indicated a decreased likelihood of both a
delinquent and criminality charge. Non-parental abuse was consistently associated with
an increased likelihood of criminality in all models. For both physical and emotional
maltreatment non-parental abuse decreased the odds of delinquency. For sexual abuse
stranger abuse increased the likelihood of an adult criminal charge. Physical and
emotional abuse decreased the likelihood of a juvenile delinquent charge whereas sexual
abuse increased the odds. Both physical and sexual abuse decreased the likelihood of
criminality where emotional abuse increased the log odds.

Table 15 provides a summary of whether or not the stated hypotheses were
supported throughout the analysis.

Table 15: Summaey of Tested Hypetheses

 

Hypothesis 1: Supported
There is a relationship between gender and juvenile delinquency.

Hypothesis 2: Supported
There is a relationship between gender and adult crinrinality.

Hypothesis 3: Supported
There is a relationship between race and juvenile delinquency.

Hypothesis 4: Supported
There is a relationship between race and adult criminality.

Hypothesis 5: Supported
There is a relationship between age and juvenile delinquency.

70

Table 15 (cont’e)

Hypothesis 6: Supported
There is a relationship between age and adult criminality.

Hypothesis 7: Not Supported
There is a relationship between SES and juvenile delinquency.

Hypothesis 8: Not Supported
There is a relationship between SES and adult criminality.

Hypothesis 9: Not Supported
There is a relationship between physical abuse and juvenile delinquency.

Hypothesis 10: Not Supported
There is a relationship between physical abuse and adult criminality.

Hypothesis 11: Not Supported
There is a relationship between emotional abuse and juvenile delinquency.

Hypothesis 12: Not Supported
There is a relationship between emotional abuse and adult criminality.

Hypothesis 13: Not Supported
There is a relationship between sexual abuse and juvenile delinquency.

Hypothesis 14: Not Supported
There is a relationship between sexual abuse and adult criminality.

Hypothesis 15: Not Supported
There is a relationship between ﬁnal placement and juvenile delinquency.

Hypothesis 16: Not Supported
There is a relationship between ﬁnal placement and adult criminality.

Hypothesis 17: Not Supported
There is a relationship between drug or alcohol use and juvenile delinquency.

Hypothesis 18: Not Supported
There is a relationship between drug or alcohol use and adult criminality.

Hypothesis 19: Not Supported
There is a relationship between the perpetrator’s relationship to the victim and juvenile delinquency.

Hypothesis 20: Not Supported
There is a relationship between the perpetrator’s relationship to the victim and adult criminality.

 

71

CHAPTER 5

W
Finding; ens! Discussion

The purpose of the study was to examine the relationship between childhood
abuse relative to juvenile delinquent and adult criminal behaviors. The study objectives
were to l) establish if a relationship existed between the type of abuse whether physical,
emotional, or sexual and delinquent behavior, 2) to determine if a relationship existed
between abuse type and adult criminal behavior, 3) to establish what characteristics of
age, gender, race, placement, SES, and alcohol or drug use related to juvenile
delinquency or adult criminality.

The theoretical framework applied in the study was Edwin Sutherland’s
Differential Association perspective. That ideology, guided by the belief that behaviors
are learned, followed the general premise that delinquent and/or criminal behaviors were
acquired through a youths’ associations or exposure to abuse. That is, behaviors are
learned through social acts that are observed in the home environment.

The database used was, “Childhood Victimization and Delinquency, Adult
Criminality, and Violent Criminal Behavior in a Large Urban County in the Northwest
United States, 1980-1997.” Participants included a matched control group and a sample
of abused and/or neglected youth made dependents by the Superior Court. For this study,
the researcher was interested in only instances of physical, sexual, or emotional abuse not
cases of neglect. Cases were ﬁltered to exclude subjects with an instance of neglect only
(284) and those with no incident of abuse (877). A ﬁnal sample of 593 abused youth was

included in the analysis.

72

Univariate, bivariate, and multivariate techniques were used in the data analysis
procedures. Frequency runs were used to describe the data characteristics. Chi-square
analyses as well as correlations were used to examine bivariate relationships and
associations between the independent and dependent variables. Multivariate logistic
regression techniques were used to examine juvenile delinquency and adult criminality as
a ﬁrnction of the various independent variables.

Hypothesis one stated that there is a relationship between gender and juvenile
delinquency. That hypothesis was supported in all regression models as well as bivariate
analyses which showed a statistically signiﬁcant relationship between gender and
delinquent behavior. All statistical results demonstrated that males were signiﬁcantly
more delinquent than females. That study ﬁnding is consistent with the literature, which
' states that males are more delinquent than females (Siegel, 2003: 68).

For example, the Kilpatrick and Saunders (2000) study illustrated that the causal
pathway for juvenile delinquent behavior differs as a ﬁrnction of gender. In that analysis,
the prevalence of prior delinquency was greater for males than females. Males were more
likely than females to commit a delinquent act, 17.7% versus 6.7% in that order
(Kilpatrick & Saunders, 2000).

The second hypothesis stated that there is a relationship between gender and adult
criminal behavior. That hypothesis was also supported by the data ﬁndings. Similar to the
ﬁndings for a juvenile delinquent charge, bivariate and multivariate data analysis proved
that a statistically signiﬁcant relationship exited between gender and adult criminality.
Males were again proven to be more criminal than females a ﬁnding that is consistent

with the literature (Siegel, 2003: 68; Piquero et al., 2005). In the Widom (1989) research

73

study, males reported a greater ﬁ'equency of adult criminality than females, 38.1% versus
13% respectively. Overall, females were less likely than males to have an adult arrest
(W idom, 1989).

Hypothesis three stated that there is a relationship between race and a juvenile
delinquent charge. That hypothesis was supported in the study. A signiﬁcant association
was found between race and juvenile delinquency at both the bivariate and multivariate
levels. The data suggested that non-Whites had a greater likelihood than Whites to be
delinquent. That ﬁnding is consistent with the literature that non-Whites are more likely
to be delinquent than Whites (Paschall et al., 1998; Piquero et al., 2002). However, the
signiﬁcant effects regression model of juvenile delinquency produced contradictory
ﬁndings. In that model Whites were at increased odds for juvenile delinquent behavior.

Several studies have assessed the relationship between race and juvenile
delinquency. The research literature states that Whites are less likely than those of
minority status, speciﬁcally Blacks and Hispanics, to be juvenile offenders, 18.5 versus
9.9% respectively (Paschall et al., 1998; Piquero et al., 2000; Kilpatrick & Saunders,
2000).

Hypothesis four stated that there is a relationship between race and an adult
criminal charge. That hypothesis was supported. The data suggests that a signiﬁcant
relationship existed between race and criminal behavior where Non-Whites had an
increased likelihood of adult criminality (Widom, 1989). However, the Signiﬁcant effects
regression model of adult criminality was contradictory to the research literature. The
likelihood of system involvement is greater for minority youth than for Whites (Williams

et al., 2001). Additionally, for those abused and/or neglected youth Blacks (39%) were

74

more likely than Whites (24.4%) to have a subsequent adult criminal record (W idom,
1989)

Hypothesis ﬁve stated that there is a relationship between age and a juvenile
delinquent charge. That hypothesis was consistently supported throughout the analysis.
At both the bivariate and multivariate levels an inverse and signiﬁcant relationship was
noted between age and a juvenile delinquent charge. That ﬁnding is consistent with the
research literature, which states that delinquent involvement decreases with age because
people age-out of such behaviors (Siegel et al., 2003: 67). Developmental criminology is
guided by the premise that people age-out of delinquent as well as criminal behavior.
That is, with an increase in age the likelihood for involvement in such behaviors
decreases.

Hypothesis six stated that there is a relationship between age and adult
criminality. That hypothesis was supported in the analysis where a statistically signiﬁcant
relationship was noted between age and an adult criminal charge. However, the direct
relationship found between age and adult criminality is inconsistent with the research
literature. The literature states that an increase in age is associated with a decrease in
adult criminal involvement (Siegel et al., 2003: 67). However, in the present analysis, an
increase in age indicated an increase in the likelihood of adult criminality.

Hypothesis seven stated that there is a relationship between a subjects’s
socioeconomic status and a juvenile delinquent charge. The hypothesis was not supported
in the analysis. The study, at the multivariate level, did not support the hypothesis but
consistemly stated that the relationship between an increase in SES and juvenile

delinquency was unchanged. However, at the bivariate level an insigniﬁcant and inverse

75

relationship was noted between SES and a juvenile delinquent charge. The latter ﬁnding
is consistent with the research literature which states that poverty and a low SES are
directly correlated with juvenile delinquency (Wells & Rankin, 1991).

Hypothesis eight stated that there is a relationship between SES and adult
criminality. That hypothesis was supported at the bivariate level where an insigniﬁcant
and inverse association was found. However, at the multivariate level, no relationship
was established between SES and an adult criminal charge. That ﬁnding is inconsistent
with the research literature which states that a lower SES is associated with the increased
odds of criminal involvement (Hawkins et al., 2000).

Hypothesis nine stated that there is a relationship between physical abuse and a
juvenile delinquent charge. That hypothesis was not supported in the study. The
relationship between physical abuse and a juvenile delinquent charge was insigniﬁcant in
the bivariate analysis. At the multivariate level, physical abuse was associated with a
decrease in the log odds of a juvenile delinquent charge. However, that ﬁnding was
statistically insigniﬁcant. The study ﬁnding was also inconsistent with the literature
which states that those who experience physical abuse tend to display more aggressive
and delinquent behaviors (Stewart et al., 2002)

Hypothesis ten stated that there is a relationship between physical abuse and adult
criminality. That hypothesis was not supported in the analysis. At the bivariate level there
was a signiﬁcant relationship between an adult charge and one or more instance of
physical abuse. However, in the multivariate analysis physical abuse was insigniﬁcant.
Although insigniﬁcant, physical abuse decreased the likelihood of an adult criminal

charge. That ﬁnding is inconsistent with the literature which states that physical abuse is

76

directly associated with crime (W idonr, 1989; Smith & Thomberry, 1995; Caetano et al.,
2003).

Hypothesis eleven stated that there is a relationship between emotional abuse and
juvenile delinquency. In t the present study that hypothesis was not supported. At the
bivariate level, there was a signiﬁcant relationship between emotional abuse and a
juvenile delinquent charge. In the multivariate analysis, though insigniﬁcant, emotional
abuse was associated with decreased log odds of juvenile delinquency. That ﬁnding is
inconsistent with the research literature which states that emotional abuse is positively
correlated with juvenile delinquency (Brown, 1984). Additionally, the harmful effects of
emotional abuse have only recemly been studied separately ﬁom other abuse types
(Spertes et al., 2003).

Hypothesis twelve stated that there is a relationship between emotional abuse and
an adult criminal charge. That hypothesis was not supported. At both the bivariate and
multivariate levels no statistically signiﬁcant relationship was found. Emotional abuse
was insigniﬁcantly correlated with the increased likelihood of adult criminality. That
ﬁnding is consistent with the literature that emotional abuse is linked to adult behavioral
problems (Spertes et al., 2003).

Hypothesis thirteen states that there is a relationship between sexual abuse and a
juvenile delinquent charge. That hypothesis was not supported in the analysis. At the
bivariate level a signiﬁcant relationship was found between sexual abuse and a juvenile
delinquent charge. At the multivariate level, an insigniﬁcant relationship existed where
the log odds of juvenile delinquency increased with the experience of sexual abuse. The

latter ﬁnding is consistent with the literature.

77

Hypothesis fourteen stated that there is a relationship between sexual abuse and
an adult criminal charge. That hypothesis was not supported in the analysis. At the
bivariate level there was a signiﬁcant relationship between sexual abuse and an adult
criminal charge. In the multivariate analysis only an insigniﬁcant relationship was found.
The experience of sexual abuse was associated with the decreased likelihood of adult
criminality. That ﬁnding is inconsistent with the literature which states that youth
sexually abused in childhood have both the increased likelihood of revictimization as
well as the involvement in criminal behaviors (Krahé et al., 1999).

Hypothesis ﬁfteen stated that there is a relationship between a child’s ﬁnal
placement and a juvenile delinquent charge. That hypothesis was not supported in the
study. At both the bivariate and multivariate levels no signiﬁcant relationship was found.
Though insigniﬁcant, out-of-home placements increased the log odds of juvenile.
delinquency. That ﬁnding is consistent with the literature (Wall, 2005).

Hypothesis sixteen stated that there is a relationship between a child’s ﬁnal
placement and an adult criminal charge. That hypothesis was not supported in the
analysis. At both the bivariate and multivariate levels no signiﬁcant relationship was
found. Though insigniﬁcant, out-of-home placements increased the log odds of adult
criminality. That ﬁnding is inconsistent with the research literature (Wall, 2005). The
relationship between placement and delinquency as well as criminality is moderated by
other factors such as the type and duration of placement as well as the age of the child
when ﬁrst placed (Frame, 2002).

Hypothesis seventeen stated that there is a relationship between drug and/or

alcohol use and a juvenile delinquent charge. That hypothesis was not supported. At both

78

the bivariate and multivariate levels an insigniﬁcant relationship was found. Though
insigniﬁcant, drug use was associated with the decreased odds of juvenile delinquency.
That ﬁnding that is inconsistent with the literature which states that drug/alcohol use is
directly related to delinquent behavior (Flowers, 1999: 3).

Hypothesis eighteen stated that there is a relationship between drug use and an
adult criminal charge. That hypothesis was not supported. At both the bivariate and
multivariate levels an insigniﬁcant relationship was found. Though insigniﬁcant, drug use
was associated with decreased odds of adult criminality. That ﬁnding is inconsistent with
the literature (Flowers, 1999: 7).

Hypothesis nineteen stated that there is a relationship between the perpetrator’s
relationship to the victim and juvenile delinquency. That hypothesis was not supported in
the analysis. At both the bivariate and multivariate levels an insigniﬁcant relationship
was found between perpetrator’s relationship to the victim and a juvenile delinquent
charge. At the multivariate level abuse by a stranger decreased the log odds of juvenile
delinquency. However, that relationship was statistically insigniﬁcant. The research does
not consistently state that abuse by a parent versus that by a stranger increases delinquent
involvement. Although abuse by a family member is believed to be more harmﬁrl to the ’
child victim than that by a stranger it does not discount its relation to juvenile
delinquency (Browne & Finkelhor, 1986; Kendall-Tackett et al., 1993).

Hypothesis twenty stated that there is a relationship between the perpetrator’s
relationship to the victim and an adult criminal charge. That hypothesis was not
supported. At both the bivariate and multivariate levels only an insigniﬁcant relationship

was noted. At the multivariate level abuse by a Stranger increased the log odds of adult

79

criminal behavior. The research states with no consistency that abuse by a parent is more
indicative of adult criminality than that by a stranger.
Conclusion

The causal link between child abuse and subsequent involvement in deviant
behavior(s) is well established throughout the research literature (Loeber et al., 2001).
However, the research consistently states that not all abused youth eventually engage in
delinquent and/or criminal behaviors (Widom, 1989). Nevertheless, it is necessary to
understand how the developmental pathway differs for an abused youth and another who
has not experienced child maltreatment. Juvenile delinquency is clearly noted as one of
the possibly long-term consequences of child maltreatment (Widom, 1989). Similarly, the
link has been established between child abuse and future adult criminality. Therefore, it is
necessary to gain an accurate understanding of the victim-to-offender pathway in order to
explore a youths’ movement along that developmental sequence (Hamilton et al., 2002).

Several research hypotheses were consistently supported throughout the data
analysis. A relationship was found between gender, race, and age relative to a juvenile
delinquent as well as an adult criminal charge. Those relationships were expected
because the literature has established that those demographic variables are related to the
likelihood of both juvenile delinquency and adult criminality independently of childhood
abuse. The directions of those relationships were also consistent with the research
literature. That is, the literature consistently states that males, Non-Whites, and younger

individuals have a greater likelihood of both juvenile delinquent and adult criminal

behavior(s).

80

On the other hand, several of the research hypotheses were not supported by the
data. No consistent relationship was found between SES, physical, sexual or emotional
abuse, child’s ﬁnal placement, drug or alcohol use, and the perpetrator’s relationship to
the victim relative to juvenile delinquency and adult criminality.

The hypotheses that SES was related to juvenile delinquency and adult criminality
were not supported in the present research study. Research has used several measures to
scale SES; the measure used in this study was mean family income. However, mean
family income was used only as a proxy for SES and as such might not have been the
most appropriate measure. In the present analysis, the relationship between SES and both
juvenile delinquency and adult criminality remained unchanged.

None of the three abuse types, used in the study, were consistently found to be
related to juvenile delinquency or adult criminality. That ﬁnding of no relationship
between abuse type and each dependent measure can be attributed to the dichotomous
measure of child abuse used. Additionally, the measures of abuse types used did not
account for the severity or duration of the incident. Both physical and emotional abuse
considered ﬁequency whereas the measure of sexual abuse did not. It is expected that a
relationship would exist between child abuse and juvenile delinquency as well as adult
criminality if those factors were accounted for in the analysis. For example, the length of
time of the abuse impacts the likelihood of deviant or criminal behavior. Therefore, it is a
logical assumption that a single occurrence of abuse differs in relation to its impact on
ﬁrture behavior than abuse which occurs with a greater ﬁequency.

A child’s ﬁnal placement after the dissolution of a dependency petition ﬁled with

the Court showed no relationship to either juvenile delinquency or adult criminality. One

81

possible explanation for the lack of a relationship is that out-of-home placement was not
subdivided into categories. For instance, the expectation is that those youth placed in'
group home placements would have an increased likelihood for both juvenile delinquent
and adult criminal behavior(s) as compared with those in other out-of-home placements.
Similarly, no relationship existed between drug or alcohol use and juvenile delinquency
or adult criminality. The measure of drug or alcohol involvement used did not consider
the frequency of use. It is expected that chronic substance abuse has a greater impact on
juvenile delinquent or adult criminal behavior than one time drug use. However, the data
was limited to a dichotomy that simply recorded whether a youth had ever used drugs
and/or alcohol and not the extent of use. The perpetrator’s relationship to the victim was
also unrelated to either a juvenile delinquent or an adult criminal charge. That association
was unfounded possibly because a dichotomy was used to measure the perpetrator’s
relationship. The unsupported hypotheses can all be attributed to the fact that all
measures of data were from ofﬁcial sources only.

The present research aides with ﬁxture studies in one major facet. The study
ﬁndings in relation to race, age, and gender add validation to prior research. It solidiﬁes
the consistent ﬁndings that Non-Whites, younger youth, and males have an increased
likelihood for both juvenile delinquency and adult criminality. However, based on the
limitations of the research data ﬁiture studies must make accommodations for more
speciﬁc and accurate measures. That is, subtypes of abuse must be used when assessing
the relationship with both juvenile delinquency and adult criminality. For example,
studies must assess not only the relationship between subtypes of abuse but also ﬁrrther

divisions of those categories. Additionally, it is beneﬁcial for ﬁrture research to more

82

accurately gauge SES. Other measures of SES can include whether the youth lives below
the poverty line, the reliance on welfare, or the educational gain of the parent(s). Lastly,
to circumvent the shortcomings of ofﬁcial sources self-reported data can be used to
compliment ofﬁcial data.

The response to the problem of childhood abuse and its relationship to juvenile
delinquency and adult criminality involves acquiring a proﬁle of the most likely victim.
With an established proﬁle, at initial contact with the juvenile justice system, the abused
child can be recognized and treated. Based on the study ﬁndings, the most complete
proﬁle must be characterized using the elements of race, age, and gender as it relates to a
juvenile delinquent charge. That is, non-Whites, younger youth, and males are the most
likely to become involved in juvenile delinquent activities. Therefore, any adequate
response to the problem must prioritize based on those factors. The same proﬁle is also
applicable to adult criminal involvement.

Despite the importance and validity of the current study ﬁndings they are not
groundbreaking because prior studies have researched the relationship between child
abuse and juvenile delinquent as well .as adult criminal involvement. However, the
research study does highlight the speciﬁc relationship when abuse types are studied in
isolation. That is, it assessed the relationship between each abuse type and juvenile
delinquency as well as adult criminality. Although, the research literature states that most
incidents of abuse do not occur in unison (Hamilton et al., 2002) that does not discount
the need for or validity of the present study. However, a possible explanation for the
insigniﬁcant association between child abuse types and both of the dependent measures

was the fact that abuse types were Studied singly.

83

The issue of child abuse must be addressed because generic programs are
insufficient to address the needs of that special population. That is, no general program
model can be used for all youths, that is, whether they have been categorized as mentally
retarded, behavioral challenged, abused or neglected, delinquent, and/or another.
Specialized programs must be in place to respond to the psychological difﬁculties
associated with child victimization. That necessitates the application of more front-end
programs and policies to combat the psychological and physical challenges that child
' abuse adds to the problem of offending behavior. Additionally, it aides in the process of
allowing the youth to move beyond the abuse and to develop socially so that they can be
successfully reintegrated into their community and society while building or rebuilding
trust and conﬁdence in their social surroundings and situations.

Special care must be given to youths that fall under the child abuse category.
Importantly though, abused juveniles should not be leﬁ feeling helpless and scarred
because of the ill-treatment received. It is important that the direct care staﬁ‘ within the
juvenile justice system not view the victims of abuse as scarred and helpless or allow that
feeling to transfer to the youth. On the other hand, every attempt must be made to avoid
or prevent the stigma associated with child abuse (Browne & F inkelhor, 1986).
Therefore, the juvenile justice system, because youth might progress from there to the
adult criminal system, must be equipped to identify and deal with the intricacies of child

abuse.

84

Appendices

85

Apmndix I

 

Maltreatment Coding Scheme“
Physical Abuse 9 subtypes Head Choking
> Torso Burns
Buttocks Shaking
Limbs Nondescript
/ Sexual Abuse Violent Handling
Subtypes

Neglect ———* Failure to Provide \
Food

Lack of Supervision Hygiene
Clothing
Shelter
Emotional Abuse Medical
Moral/Legal/Educational Neglect
Severity Codes
Detail 27 Types
A Lack Supervision
Environment
Substitute Care

*Modiﬁed Barnett, Manly, Cicchetti ( 1993) Coding Scheme

86

Aneendix H

Maltreatment Coding Scheme*
Severity Coding for:

Severity is coded on a scale of 1 (low) through 5 or 6 (high)

Physical Abuse to the Head/Face/Neck
Severity 1 = No marks indicated
Severity 2 = Minor marks
Severity 3 = Numerous or non-minor marks
Severity 4 = Emergency Room or medical treatment
Severity 5 = Hospitalization for more than 24 hours
Severity 6 = Permanent Disability or Death

Sexual Abuse
Severity l = Caretaker exposes child to explicit sexual stimuli or activities
Severity 2 = Caretaker makes direct requests for sexual contact with child
Severity 3 = Caretaker engages child in mutual sexual touching
Severity 4 = Caretaker attempts to (or actually) penetrate(s) child
Severity 5 = Caretaker has forced intercourse with child

* Modiﬁed Barnett, Manly, Cicchetti Coding Scheme 1993

87

Aemndix HI

Maltreatment Coding Scheme“
Severity for Emotional Maltreatment

(Severity for emotional maltreatment is coded using 27 diﬁ‘erent codes. These codes are
clustered in groups or levels of severity.)

Severity 11 = Caregiver expects inappropriate level of responsibility

Severity 12 = Caregiver undermines the child’s relationships

Severity 13 = Caregiver belittles or ridicules child

Severity 14 = Caregiver ignores child’s bids for attention I
Severity 15 = Caregiver uses fear or intimidation as a method of discipline .

Severity 21 = Caregiver does not permit age appropriate socialization

Severity 22 = Caregiver places child in a role-reversal

Severity 23 = Caregiver thwarts child’s sense of maturity and responsibility

Severity 24= Caregiver rejects or is inattentive to child’s need for affection

Severity 25= Caregiver allows child to be exposed to caregiver’ S extreme but non violent
marital conflict

Severity 31 = Caregiver blames child for marital problems

Severity 32 = Caregiver has inappropriate or excessive expectations for the child

Severity 33 = Caregiver makes serious threats to injure the child

Severity 34 = Caregiver calls child derogatory names

Severity 35 = Caregiver binds child’s hands and feet for moderate periods of time

Severity 36 = Caregiver exposes child to extreme, unpredictable, or inappropriate
behavior

Severity 37 = Caregiver demonstrates a pattern of negativity or hostility toward the child

Severity 41 = Caregiver threatens suicide or abandonment in ﬁont of child

Severity 42= Caregiver blames the child for suicide or death of another family member
Severity 44= Caregiver conﬁnes and isolates the child (between 5 and 8 hours)
Severity 45 = Caregiver binds child or places child in conﬁnement for less than 2 hours

Severity 51 = Caregiver makes a Suicidal attempt in front of child

Severity 52 = Caregiver makes a homicidal attempt or realistic threat of homicide against
the child without actual physical harm to child

Severity 53 = Caregiver abandons child for 24 hours or more

Severity 54= Caregiver uses extremely restrictive methods to bind a child or places child
in conﬁnement for 2 or ore hours

Severity 55= Caregiver conﬁnes child for extended periods (more than 8 hours)

* Modiﬁed Barnett, Manly, Cicchetti (1993) Coding Scheme

88

References

89

REFERENCES

Acoca, L. (October 1999). Investing in girls a 21‘ century strategy. Juvenile Justice, .4
(1). Retrieved September 16, 2005, ﬁ'om
www.ncjrs.gov/html/ojjdp/jjjournal1099fmvest3.html

Albanese, J. S. (2001). Brief edition criminal justice. Needham Heights, MA: Allyn &
Bacon.

Barlow, H. D. (1990). Introduction to criminology (5“ ed.). Glenview, IL: Scott,
F oresman & Company.

Bartollas, C. & Miller, S. J. (2001). Juvenile justice in America (3“! ed.). Upper Saddle
River, NJ: Prentice Hall.

Brieke, J. & Runtz, M. ( 1990). Differential adult symptomatologies associated with three
types of child abuse histories. Child Abuse and Neglect, 14, 357-364.

Brown, S. E. (1984). Social class, child maltreatment, and delinquent behavior.
Criminology, 22 (2), 259-278.

Browne, A & F inkelhor, D. (1986). Impact of child sexual abuse: A review of the
research. Psychological Bulletin, 99(1), 66-7 7 .

Caetano, R, Field, C., & Nelson, S. (2003). Association between childhood physical
abuse, exposure to parental violence and alcohol problems in adulthood. Journal
of Adolescent Research, 18 (3), 240-257.

Egeland, B. (1991). A longitudinal study of high-risk families: Issues and ﬁndings. In R
H. Stark & D. A Wolfe (Eds). The ejects of child abuse and neglect: Issues and
research (pp. 33-56). New Yorlg NY: The Guilford Press.

Elliott, D. (1994). Serious violent offenders: Onset, developmental course, and
termination. Criminology, 32(1), 1-21.

90

English, D. J ., & Widom, C. S. Childhood victimization and delinquency, adult
criminality, and violent criminal behavior in a large urban county in the northwest
United States, 1980-1997 [Computer File]. ICPSR version. Seattle, WA: State of
Washington Department of Social and Health Services, Oﬁice of Children’s
Administration Research [producer], 2002. Ann Arbor, MI: Inter-university
Consortium or Political and Social Research [distributor], 2003.

Feldman, C. (1997). Childhood precursors of adult interpartner violence. Clinical
Psychology: Science and Practice, 4, 307-334.

F erro, J. (2003). Library in a book juvenile crime. New York, NY: Facts on File.

F inkelhor, D. & Jones, L. M. (January 2004). Explanations for the decline in child sexual
abuse cases. Juvenile Justice Bulletin. Washington, DC: US Department of
Justice, Ofﬁce of Justice Programs, Office of Juvenile Justice and Delinquency

Prevention.

Finkelhor, D. & Ormrod, R (June 2000). Characteristics of crimes against juveniles.
Juvenile Justice Bulletin. Washington, DC: US. Department of Justice, Ofﬁce of
Justice Programs, Oﬁice of Juvenile Justice and Delinquency Prevention.

Flower, R B. (1999). Drugs, alcohol, and criminality in American society. Jefferson, NC:
MaFarland & Company, Inc.

Hamilton, C. E., Falshaw, L. & Browne, K. D. (2002). The link between recurrent
maltreatment and offending behaviour. International Journal of Oﬂ'ender Iberapy
and Comparative Criminology, 46 (1), 75-94.

Hawkins, J. D., Herrenkohl, T.I., Farrington, D. P., Brewer, D., Catalano, RF ., Harachi,
T.W. & Cothern, L. (April 2000). Predictors of youth violence. Juvenile Justice
Bulletin. Washington, DC: US Department of Justice, Ofﬁce of Justice Programs,
Ofﬁce of Juvenile Justice and Delinquency Prevention.

Huizinga, D. & Elliott, D. (1987). Juvenile offender prevalence, incidence, and arrest
rates by race. In 1.6. Weis, R D. Crutchﬁeld, & G. S. Bridges (Eds). Juvenile
delinquency readings (2" ed.) (pp. 98-103). Boston, MA: Pine Forge Press.

91

 

Jewell, D. L. (1999). Confronting child maltreatment through recreation (2“‘1 ed).
Springﬁeld, IL: Charles C Thomas Publisher Limited.

Kendall-Tackett, K. A, Williams, L. M., & Finkelhor, D. (1993). Impact of sexual abuse
on children: A review and synthesis of recent empirical studies. Psychological

Bulletin, 113(1), 164-180.

Kilpatrick, D. G. & Saunders, B. E. (2000). Prevalence and consequences of child
victimization: Results from the National Survey of Adolescents, Final Report.

Krahé, B., Scheinberger—Olwig, R., Waizenhofer, E. & Koplin, S. (1999). Childhood
sexual abuse and revictimization in adolescence. Child Abuse & Neglect, 23 (4),
383-394.

Loeber, R, Kalb, L. & Huizinga, D. (August 2001). Juvenile delinquency and serious
injury victimization. Juvenile Justice Bulletin. Washington, DC: US Department
of Justice, Ofﬁce of Justice Programs, Ofﬁce of Juvenile Justice and Delinquency
Prevention.

Malinosky-Rummel, R & Hansen, D. (1993). Long-term consequences of childhood
physical abuse. Psychological Bulletin, 114, 68-79.

Morales, A (1998). The rising incidence of child abuse. In B. J. Grapes (Eds). Child
abuse: Contemporary issues compmion (pp. 14-18). San Diego, CA: Greenhaven

Press, Inc.

Mullen, P., Martin, J., Anderson, J., Romans, 8., & Herbison, G. (1996). The long-term
impact of the physical, emotional, and sexual abuse of children: A community
study. Child Abuse & Neglect, 20(1), 7-21.

Nofziger S. & Kurtz, D. (2005). Violent Lives: A lifestyle model linking exposure to
violence to juvenile violent offending. Journal of Research in Crime and

Delinquency, 35 (3), 251-266.

Osofsky, J. D. (October 2001). Addressing youth victimization. Coordinating Council on
Juvenile Justice and Delinquency Prevention Action Plan Update. Washington,

92

 

DC: US Department of Justice, Oﬂice of Justice Programs, Ofﬁce of Juvenile
Justice and Delinquency Prevention.

Paschal], M., Flewelling, R & Ennett, S. (1998). Racial differences in violent behavior
among young adults: Moderating and confounding effects. Journal of Research in
Crime and Delinquency, 42 (1), 3-26.

Pelton, L. H. (1994). Controversial issues in child welfare. In B. J. Grapes (Eds). Child
abuse: Contemporary issues companion (pp. 74-82). San Diego, CA: Greenhaven
Press, Inc.

Pepi, C. (1997). Children without childhoods: A feminist intervention strategy utilizing
systems theory and restorative justice in treating female adolescent offenders.
Women & Therapy, 20(4), 85-100.

Piquero, A, Macintosh, R, & Hickman, M. (2002). The validity of a self-reported
delinquency scale: Comparisons across gender, age, race, and place of residence.
Sociological Methods and Research, 30 (4), 492-529.

Piquero, N., Gover, A, MacDonald, J ., & Piquero, A (2005). The inﬂuence of
delinquent peers on delinquency: Does gender matter? Youth and Society, 36 (3),
25 1-275.

Schmall er, F. (2003 ). Criminal justice today: An introductory text for the 21 ” century
( ed.). Upper Saddle River, NJ: Prentice Hall.

Siegel, L. (2003). Criminology (8th ed.) Belmont, CA: Thomson Wadsworth.

Siegel, L., Welsh, B., & Senna, J. (2003). Juvenile delinquency: Theory, practice, and
law (8th ed.). Belmont, CA: Thomson Wadsworth.

Smith, C. & Thomberry, T. P. (1995). The relationship between childhood maltreatment
and adolescent involvement in delinquency. Criminology, 33(4), 451-482.

Starr, RH., MacLean, D. J. & Keating, D. P. (1991). Life-span developmental outcomes
of child maltreatment. In R H. Stark & D. A Wolfe (Eds). ﬁre eﬂects of child

93

abuse and neglect: Issues and research (pp. 1-3 2). New York, NY: The Guilford
Press.

Stewart, A, Dennison, S. & Waterson, E. (October 2002). Pathways ﬁom child
maltreatment to juvenile oﬂending. Australian institute of criminology: Trends
and issues in crime and criminal justice. Retrieved December 1, 2005 from
http://www.aic.gov.arr/publications/tandi/tandi241t.html.

Wall, A E. (2005). Problem behaviors in maltreated chilcb'en and youth: Irﬂuential
child peer, and caregiver characteristics. Ph D Dissertation. The University of

North Carolina at Chapel Hill.

Wells, L.E. & Rankin, J. H. (1991). A meta-analysis of the impact of broken homes. In
J .G. Weis, R D. Crutchﬁeld, & G. S. Bridges (Eds). Juvenile delinquency
readings (2“ ed.) (pp. 124-130). Boston, MA: Pine Forge Press.

Widom, C. S. (1989). Child abuse, neglect, and violent criminal behavior. Criminology,
27(2), 251-272.

Williams, J. H., Ayers, C. D., Outlaw, W. 8., Abbott, R. D. & Hawkins, J. D. 2001. The
effects of race in juvenile justice: Investigating early stage processes. Journal for
Juvenile Justice and Detention Services, 16 (2), 77-91.

Wright, B., Caspi, A, Mofﬁtt, T., Miech, R., & Silva, P. (1999). Reconsidering the
relationship between SES and delinquency: Causation but not correlation. In J .G.
Weis, R D. Crutchﬁeld, & G. S. Bridges (Eds). Juvenile delinquency readings
(2nd ed.) (pp. 115-123). Boston, MA: Pine Forge Press.

Zuravin, S. J. (1991). Research deﬁnitions of child physical abuse and neglect: Current
problems. In R. H. Stark & D. A Wolfe (Eds), Ihe eﬂects of child abuse and
neglect: Issues and research (pp. 100-128). New York NY: The Guilford Press.

94

rfujyjiinjj;I113]