Ilml'll’lill]lll‘lllln‘illll‘lll

3 1293 01565 4373

 

 

This is to certify that the
thesis entitled

Fa-i 1y Process Vari ables and Jyveni 1e Delinquency

presented by

Ti nothy Robert Sahr

has been accepted towards fulfillment
of the requirements for

Master of Science degree in Criuinal Justice

ﬂax/Z4

 

 

 

Major pr essor

Date Decedwer 11, 1996

MS U is an Afﬁrmative Action/Equal Opportunity Institution

 

LIBRARY

Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove thin checkout from your record.
TO AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

 

    
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Affirmative Actior'VEquel Opportunity IMKWOH
1

 

FAMILY PROCESS VARIABLES AND JUVENILE DELINQUENCY

BY

Tim R. Sahr

A THESIS

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

MASTER OF ARTS

Department of Criminal Justice

1997

ABSTRACT

FAMILY PROCESS VARIABLES AND JUVENILE DELINQUENCY
By

Tim R Sahr

The researcher’s purpose in this study was to examine the inﬂuence of family
process variables (i.e., family communication. family conﬂict, family discipline, and
family activities) on youths' level of delinquency. A secondary analysis was done of
data from the National Survey of Children. which was a longitudinal study conducted
from 1976 to 1987. The present study was focused on data collected In 1981, with
a sample of 1,423 children and an 82% response rate.

The study findings indicated that family process variables such as family
communication, level of attention the child receives, family discipline. family leisure
time, and amount of supervision all had significant roles in predicting juvenile
delinquency. Family communication seemed to have the greatest inﬂuence. with

family discipline also having more of an effect than the other variables.

i would like to thank and dedicate this thesis to my wife, Lisa, to my mother
and father, Carol and Gary, and to my family and friends, without whose love and

support this achievement would not have been possible.

TABLE OF CONTENTS

LIST OF TABLES ................................................ vi
Chapter

I. THE PROBLEM ....................................... 1

Introduction .............................. '. ........... 1

Problem Statement and Purpose of the Study ............... 2

Overview ............................................ 5

II. LITERATURE REVIEW ................................. 6

Introduction .......................................... 6

Research on the Family and Delinquency ................... 6

Social Control Theories ................................ 10

The Broken Home Hypothesis .......................... 14

Family Process Variables .............................. 17

Ill. DESIGN AND METHODOLOGY ......................... 25

Introduction ......................................... 25

Research Design ..................................... 26

Procedures ......................................... 27

Description of the Sample .............................. 30

Variables ........................................... 31

Dependent Variable ................................. 31

Independent Variables ............................... 32

IV. RESULTS .......................................... 34

Introduction ......................................... 34
Relationship Between Family Process Variables and

Status Offenses .................................... 35
Relationship Between Family Process Variables and

Property Offenses .................................. 42

iv

Relationship Between Family Process Variables and

Personal Offenses .................................. 55
Relationship Between Family Process Variables and
Alcohol and Drug Offenses ........................... 62
SUMMARY, CONCLUSIONS. LIMITATIONS.
AND RECOMMENDATIONS ............................ 81
Summary ........................................... 81
Conclusions ......................................... 84
Limitations and Recommendations for Further Research ...... 86
REFERENCES ............................................ 89

LIST OF TABLES
Association Between Family Process Variables and Status
Offenses: Frequency of Running Away From Home ............... 36

Association Between Family Process Variables and
Property Offenses: Damage to Property ........................ 44

Association Between Family Process Variables and
Property Offenses: Theft .................................... 50

Association Between Family Process Variables and Personal
Offenses (How Often the Juvenile Physically Hurt Someone) ........ 56

Association Between Family Process Variables and Alcohol
and Drug Offenses (How Often the Juvenile Drank Alcohol) ......... 63

Association Between Family Process Variables and Alcohol
and Drug Offenses (How Often the Juvenile Used Marijuana) ....... 69

Association Between Family Process Variables and Alcohol

and Drug Offenses (How Often the Juvenile Used Crack/
Cocaine) ................................................ 75

vi

CHAPTER I
THE PROBLEM

Inimdustinn

In American society. delinquency is seen as an ongoing problem that
seems to be worsening. This Image is further reinforced by the media, which
give the public daily reminders of increased gang activity. assaults. thefts,
carjackings, and even murders. Many of these crimes are attributed to juvenile
offenders. However, in reality, the rate of juvenile delinquency has remained
about the same, but juveniles are being incarcerated for longer periods of time
(T rojanowicz 8. Morash, 1987). Nonetheless. this fact still does not address the
real essence of the problem, which is how to prevent and handle juvenile
delinquency.

Many treatment programs such as social casework, reality therapy,
psychoanalysis, transactional analysis. vocational guidance. and crisis and family
intervention have been implemented in an effort to combat delinquency. Yet,
high recidivism rates suggest that these programs, as a whole, have been
unsuccessful. Likewise, programs that have been successful have failed to have

a signiﬁcant long-term effect (T rojanowicz & Morash, 1987).

2

Much research has been conducted in an effort to address this problem.
and it has led to many different ideas concerning the source of delinquency.
These ideas include opportunity theory, ieamed association, social control
theory, labeling theory, and the broken home hypothesis. Speciﬁcally.
proponents of the broken home hypothesis have suggested that a breakdown or
disruption of the family unit can lead to difficulties for the child. which in turn may
lead the youth to engage in delinquent behavior (Grove 8 Crutchﬁeld, 1982).

Ifthe broken home hypothesis is correct, the family unit is a fundamental
starting point in the prevention of juvenile delinquency. The primary function of
the family is to socialize the child (Nye. 1958). Any failure (intentional or not) by
the family to do so could result in an array of problems for the child. Essentially,
then, it is from the family that the child ultimately acquires his or her values, self-

esteem. and sense of morality (Nye, 1958).

WWW

Previous researchers on the causes of juvenile delinquency were
preoccupied with stmctural factors of the family and their effect on delinquency
(Cernkovich 8 Giordano, 1987). These studies dealt primarily with single-parent
families and broken homes (Chilton 8 Markle, 1972; Glueck 8 Glueck, 1950;
Mastueda 8 Heimer, 1987; Rankin, 1983). These researchers suggested that
single parents are incapable of providing enough care. guidance, and discipline

for their children, and that delinquency is the inevitable result.

3

For many years. the broken home was believed to be one of the main
causal factors contributing to delinquency (Wilkenson. 1974). However, in recent
years, researchers such as Wells and Rankin (1986) have raised questions
about previous studies that examined solely broken homes in conjunction with
delinquency. Other researchers such as Cemkovich and Giordano (1987)
suggested that too much emphasis has been placed on the structural aspects of
the family, whereas little. if any, attention has been given to family process and
interaction dynamics (functional variables).

The amount of attention given to stnictural factors ofthe family is puzzling,
given the suggestion in past research that structuralism is not the sole cause of
delinquency (Cernkovich 8 Giordano, 1987). Nye (1958) indicated thatthe family
structure itself was not related to delinquency. but that the relationships and
interaction patterns within the family were key variables contributing to
delinquency.

However, for reasons that remain unknown. emphasis in delinquency
research Still is placed on structural factors as they relate to delinquent behavior.
Yet. such research has failed to yield any significant and conclusive ﬁndings.
This stagnation has contributed to a lack of inclusion of family variables relative
to delinquency research, and even less understanding of their relationship to
delinquency (Cernkovich 8 Giordano, 1987).

Some investigators have conducted research and presented findings on

family function variables (Canter, 1982; Cemkovich 8 Giordano. 1987; Patterson

4
8 Dishion, 1985; Van VoorhiS, Cullen, Mathers, 8 Garner, 1988). However, most
studies on family processes and interaction relative to delinquency have omitted
a number of family variables and/or have had methodological shortcomings.
Researchers such as Canter (1982), Johnson (1985), and Norland, Shover,
Thornton, and James (1979) examined family process variables and
delinquency, but they asked subjects only one or two questions pertaining to the
topic. For example, Norland et al. studied intrafamily conﬂict and delinquency,
but they used only one statement to measure family conﬂict: 'There is a lot of
tension and conﬂict in my home.” in a Similar vein, Canter measured family
involvement using only one question: “How much have your parents inﬂuenced
what you’ve thought and done?" Although these types of questions are a step
in the right direction, this kind of measurement simply is not comprehensive
enough to provide significant insight into the subject (Cernkovich 8 Giordano,
1987)

Some researchers have suggested that psychological and social
interaction within the family must be emphasized over family structure. There is
an indication that family is just as important as, If not more important than, peer
groups and social environments in terms of inﬂuence on delinquency. However,
this topic has not been given enough attention in research, and as a result there
is insufficient information concerning this aspect of juvenile delinquency.

To address the problem discussed above, this writer focused on family

process variables in relation to juvenile delinquency. Speciﬁcally, the

5

researcher’s purpose in this study was to examine the inﬂuence of family process
variables (i.e., family communication, family conﬂict, family discipline, and family

activities) on youths' level of delinquency.

Qxemim

Chapter I contained an introduction to the study, a statement of the
problem, and the purpose of the Study. Chapter II contains a review of literature
on topics pertinent to the study. The chapter begins with an overview of research
on family and delinquency. The social control theories and the broken home
hypothesis are discussed In detail. A discussion of family process variables
concludes the chapter. The research design and methodology are described in
Chapter III. The sample is described, as are the three waves of the National
Survey of Children, data from which were used in the present research. The
dependent and independent variables are discussed, and the data-analysis
procedures are set forth. Results of the data analyses are presented in Chapter
IV. Chapter V contains a summary of the study, conclusions, and recommenda-

tions for further research.

CHAPTER II

LITERATURE REVIEW

inimdumian

This study was undertaken to examine family interactions and processes
and to evaluate their effect on youths’ delinquency. The subject of family and
delinquency is not a new one. However, only recently has attention been diverted
from the structural aspects of the family to its functional or process aspects. The
relationship between the family and delinquent behavior is examined in this
literature review. The chapter begins with an overview of research on the family
and delinquency. Next. social control theories and the broken home hypothesis
are considered in greater detail. The chapter concludes with a discussion of

family process variables.

8 I II E 'l | D I'

In the study of family and delinquency, there have been primarily two
schools of thought. The ﬁrst one, structuralism, deals with quantitative factors
such as parental absence, family size, and birth order. The second school of
thought, functionalism, contends that qualitative factors such as parent/child

interaction, style of discipline, and marital happiness inﬂuence delinquent

7
behavior (Rosen, 1985). The primary approach of most previous researchers has
been to explore the family structure and its inﬂuence on delinquency.

One of the first major investigations of family and delinquency was Glueck
and Glueck’s 1950 study, entitled unravelingJuyenﬂLDelinquency. The
research, concluded during the 1940s, consisted of matched samples of 500
ofﬁcial delinquent youths and 500 ofﬁcial nondelinquent youths. The delinquent
sample was drawn from two correctional facilities, whereas the nondelinquent
sample was drawn from the public school system. The sample groups were
matched in terms of race. age, intelligence, and socioeconomic status. The
researchers collected large amounts of data regarding the subjects’ social,
psychological, and biological characteristics; family life; school performance; and
work experience.

The composition of the data set is important because it is one of the most
thorough and informative data sets ever collected, and the fact that the data have
not been fully used is puzzling (Laub 8 Sampson, 1988). What Glueck and
Glueck found was that family processes were the most powerful predictors of
delinquency and other juvenile problems. The authors identiﬁed several family
process variables, such as parental affection, supervision, attachment, and
discipline, as powerful predictors of delinquency. They also suggested that these
factors were the most important family correlates of serious, persistent

delinquency.

8

Glueck and Glueck found that a higher percentage of delinquents than
nondelinquents came from broken homes. This ﬁnding, coupled with the family
process variables, indicates that the family may be the most important inﬂuence
on a child during his or her development, and that the most effective preventive
measures (discipline, affection, supervision, communication, and family
cohesiveness) exist within the family unit. The researchers' emphasis on the
family as a major source of delinquency was and still is not a popular concept in
contemporary delinquency theories (Hirschi, 1983). Other researchers, such as
Bordua (1962), have suggested that modern sociological thinking tends to view
delinquency as a stress reaction to class structure, rather than to a breakdown of
the family unit. .

Although Glueck and Glueck amassed an impressive sample group and
data set, many of their techniques have been severely criticized, rendering their
ﬁndings questionable. To begin with, the researchers believed that mesomorphic
body type (biological make-up) was a major predictor of delinquency. (However,
their research showed that family process variables were directly related to
delinquency [Laub 8 Sampson, 1988]). This idea has long since been
discredited, and it is often ridiculed. This, then, led to another criticism of the
study. which was that It placed too much emphasis on the relationship of
nonsociological variables to crime. This would lead one to believe that the
investigators’ analysis of the other variables was insufficient and that the

conclusions drawn from the analysis were ill conceived.

9

In addition to their biological etiology, many of Glueck and Glueck’s
methods also have been questioned—for instance, conditions that existed before
the youths were institutionalized. In addition, the fact that the researchers failed
to take into account the basic rate of delinquency in the population of interest
added to the Skepticism regarding their ﬁndings (Laub 8 Sampson, 1988). For
instance, Glueck and Glueck claimed to have had great success with a
delinquency scale that they developed, which was based on family discipline,
affection, and cohesiveness. However, they failed to take into account the base
rate of delinquency in the population of interest. Because their sample design
had 50% delinquents and 50% nondelinquents, this proportion must then be
indicative of the general population (in this case, Boston, which had an estimated
delinquency rate of 10% in the general population), or the results will be useless
(Laub 8 Sampson, 1988).

Criticisms of the Glueck and Glueck study may have inﬂuenced future
researchers to focus on structural aspects of the family rather than the family
process variables that Glueck and Glueck identiﬁed. The discrediting of Glueck
and Glueck’s research is unfortunate because, although the investigators were
harshly criticized for their emphasis on body type, their sampling and matching
procedures, and selection of prediction items, their study served as a basis for
reestablishing the central role of family dynamics in the study of delinquency

(Rosen, 1985).

10
SociaLQQntmllhecries

After the Glueck and Glueck study, a major move in the research on family
and delinquency included development of the social control theory. In all
probability, it has been the most widely accepted and discussed theory
concerning family variables and delinquency. The emphasis of this theory is that
children develop ”bonds” between themselves and their families and/or their
communities. It Is the strength of these bonds that deterrnlnes whether or not a
child succumbs to delinquent behavior. Nye (1958) and Hirschi (1969) stressed
the importance of the bonds formed by the child to the family and the larger
society.

Nye (1958) identiﬁed social control as consisting of four different types of
control: direct control (based on inﬂuencing the child's behavior through the
threat of punishment or coercion), indirect control (based on the child’s
attachment to and Identiﬁcation with conventional persons, such as parents),
intemallzed control (autonomous patterns of conformity withinthe individual’s own
personality, self-concept, and conscience), and control over opportunities and/or
deviant activities (whereby compliance is a result of restricted choices). Inherent
in direct control is the threat of punishment for bad behavior and the promise of
rewards for good behavior in order to gain compliance with the accepted norms.
Indirect control is the belief that the youth will refrain from delinquency as a result
of modeling his or her behavior after that of a signiﬁcant other, or that the child will

not want to disappoint his or her role model with bad behavior. Intemalized

11

control is unique because it allows the child to develop his or her own set of
norms and gives the youngster the ability to monitor his or her own behavior.
Control over opportunity and/or deviant activities involves direct restraints on the
child, which can assume many forms (laws, folkways) and be enforced by a
variety of social agencies (police, teachers, peers). Of all these types of social
control, Nye believed that the family was the strongest and most important agency
of direct control (Wells 8 Rankin, 1986).

Two important aspects of Nye's theory are that the bond between child and
family/community is part of the youth’s personality and, more important, that
parental monitoring is effective only within the actual physical limits of the parent’s
supervision. Hirschi (1969) later built and expanded on these concepts. The
relevance of Nye's theory is that it establishes the family as an integral part of the
child's social development and behavior, and it laid the foundation for Hirschl’s
research.

Hirschi (1969) did not examine why juveniles commit crimes, but rather
why they do not. He concluded that it is the ties (bonds) that are developed
between the youth and his or her environment that suppress delinquency. Hirschi
believed that there are four bonding elements that inhibit delinquent behavior.
These elements are attachment, involvement, belief, and commitment. Each of
these components is positively related to conformity and is thought to have

independent effects on delinquency.

12

The ﬁrst two ties, attachment and involvement, are directly related to family
process variables. Attachment is the strength of the child’s ties to his or her
parents, school, and peers. Youths who are not attached to others are thought
to be insensitive to other people’s opinions and are not bound by others’ norms
and values. This may leave such youths open to drift Into delinquent behavior.
Those who identify more closely with their parents are more likely to consider their
parents’ feelings and opinions when and if they contemplate performing a
delinquent act. The element of involvement is based on the premise that if a
youth is involved in conventional activities (with his or her family, and so on), he
or She will not have time to participate in delinquent behavior.

Whereas Nye (1950) placed the bonds within the child’s own conscience.
Hirschi (1969) placed the conscience of the bond within the child’s signiﬁcant
others. In other words, Nye believed that the child could internalized the bonds,
and Hirschi thought that the bonds had to be implemented through external
forces. Although Hirschi agreed with Nye that direct parental controls are
effective only when the child is under the parent's direct physical supervision, he
expanded the concept and explained that parental attachments are more
Important than parental supervision. Hirschi contended that as long as the
parents are "psychologically' present, attachment will prevent the child from
drifting into delinquency. In other words, attachment helps the child develop a

conscience that distinguishes between right and wrong. Even though the child’s

13
parents are not there to guide him or her, the child intemalizes right from wrong
and is inhibited from participating in delinquent behavior.

The last two ties, belief and commitment, relate to the individual’s belief in
community norms and values, and a commitment to conventional goals within the
community. Other researchers have concurred with Hirschi’s ﬁndings regarding
these ties. Bahr (1979) also found that direct controls were ineffective in curbing
delinquency, but discovered support for indirect controls. Children violated rules
only when they had no reason to conform (Bahr, 1979). if a child is strongly
attached to the parent and the parent is not involved in deviant acts, a reason or
motivation for conforming exists. The youth intemalizes the rules. As a result, the
youth refrains from breaking the rules, not because to do so would be illegal, but
rather because he or she has no motivation to do so. Hence, when a Strong
attachment exists between parent and child, and the child intemalizes the rules,
direct controls are not as important because the child’s behavior is guided by his
or her conscience (Bahr, 1979).

Although some importance has been attached to the ”psychological
presence“ of the parents as described by Hirschi (1969) and Bahr (1979), the
question then arises: How is that psychological presence established, and how
is it measured? Cemkovich and Giordano (1987), Patterson and Dishion (1985),
Van Voorhis et al. (1988), and Wells and Rankin (1988) found that direct parental
controls can be effective and that the rejection of such controls by Nye, Hirschi,

and Bahr was premature (Rankin 8 Wells, 1990). Their rejection of direct

14

parental controls could account for the attention that recent researchers have
given to broken homes in relation to delinquency. Presumably, the absence of
one or both parents could lead to a weak bond with the child and produce a
Situation in which the child would have a difﬁcult time refraining from delinquent

behavior.

W

The assumption underlying the broken home hypothesis is that children
need guidance and discipline when they are growing up in order to be properly
socialized. Providing the necessary guidance and discipline is the primary
responsibility of the child’s family. It is then logical to assume that a deprivation
occurs when the child loses one of his or her biological parents. Jackson (1957)
stated that the American family has the greatest inﬂuence on children and that
any disruption in the family can be devastating for the child. The biggest
disruption, of course, would be the loss of a parent, through death, divorce,
separation, or desertion. This type of loss can result in loss of affection, economic
hardship, and/or loss of a role model for proper behavior. This reduction in
parental control could make socialization more difﬁcult and could lead the child
into delinquent behavior (Rankin, 1983).

The concept of broken homes contributing to delinquency can be traced
to Glueck and Glueck’s (1950) research, in which they found that a higher

proportion ofjuvenile delinquents than nondelinquents came from broken homes.

15

Later, Browning (1960), Gold (1963), and Slocum and Stone (1963) found similar
results. Monahan (1957) found that most recidivists came from broken homes
and suggested that children who resided in broken homes were charged with
delinquency offenses more often than children in the general population.
Similarly, Rankin (1983) found a high correlation between broken homes and
running away, truancy, and auto theft. He also discovered that broken homes
were differently related to different types of delinquency.

Although the above-mentioned researchers made a strong case for the
broken home hypothesis, the theory has received strong criticism throughout the
years. Stem (1964) stated that the actual breakup between parents is preceded
by disorganization, tension, confusion, and chaos. Because these negative
factors existed before the formal separation, the broken home could not have
been the major factor contributing to delinquency.

Chilton and Markie (1972) found that more children who come into contact
with police agencies and juvenile authorities arefrom disrupted families, and that
children from disrupted families often are charged with more serious offenses
than those from whole families. Interestingly, Johnson ( 1986) found that family
structure was not related to self-reported delinquency, but did have a strong
association with ofﬁcial delinquency. This leads one to believe that the police and
other ofﬁcials are more willing to formally sanction delinquents from lower-class,
broken homes than those from higher-class and/or whole families. Youths from

higher-socioeconomic-status homes were merely warned and released or given

16

some other type of informal sanction, whereas children from lower-class, broken
homes were handled through the system.

It seems that the association between broken homes and delinquency is
as much a reﬂection of the attitudes and decisions ofjuvenile justice personnel
as it is a reﬂection of the delinquency-producing tendencies of broken homes
(Shoemaker, 1984). This concept leads into another criticism of the broken home
hypothesis: It relies too much on ofﬁcial data and not enough on self-reported
data. Because the hypothesis depends on so much official data, critics argue that
it is Simply reactional rather than cause and effect. In other words, the law treats
children from broken homes differently than it does those from intact families
(Browning, 1960).

Other criticisms of the broken home hypothesis are (a) that researchers of
the theory have failed to study the complex interrelationships of family members
within the home, (b) that there is a virtual absence of any systematic conceptual
speciﬁcation and corresponding empirical measurement of the broken home as
a sociological variable, and (c) that many of the variables used in the existing
research were oversimpliﬁed (Wells 8 Rankin, 1986). Parental discipline and
supervision, communication between child and parent, family activities, the family
members’ feelings toward each other, the parents’ interest in their children’s
activities, how the children feel about their parents (or parent) as role models, and
how affection is displayed within the home are just some of the variables that

have been overlooked in previous research.

17

Without looking at these variables, one has no idea what is actually going
on in the home, and by concentrating solely on structural variables (such as
family size), all that one sees is the outside shell of the family unit. Because
growing up is a social and psychological process, it is only logical to examine
social and psychological variables when attempting to identify aspects of the
socialization process that lead to delinquent behavior.

It is becoming evident, then, that the strength of the relationship between
a child and his or her parents and the rest of the family plays an important role in
the child’s development. Whether or not the child comes from a broken home is
inconsequential in the development of delinquent behavior. What is important is
the strength and quality of the relationships within the family structure, for it is the
state of those relationships with which the child ultimately identiﬁes, and not

whether both parents live at home.

E 'l E I I . l I

Instead of being the result of one or two simple factors, delinquency is
caused by multiple variables that are complex and interrelated (Wells 8 Rankin,
1990). One of these variables is relationships between members of the family,
which are constantly maturing and changing. Because delinquency is not the
result of any single variable, it is also logical to state that it is not subject to any

Single theory or hypothesis.

18

Wells and Rankin (1990) stated that delinquency was the result of a
combination of many different factors and theories. As an example they
discussed how social control and parental discipline could interact with each other
to produce varying degrees and types of delinquency. Although their study
focused primarily on direct parental controls and delinquency, they did
acknowledge that the importance of family process variables (relationships and
interactions within the family) has largely been overlooked , and they stressed that
researchers Should begin to direct their attention to this aspect of delinquency.
Their ﬁnding that direct parental controls (discipline) and attachment consistently
have been related to different types of delinquency seems to support this idea.

In the past, research dealing with family process variables was largely
downplayed, despite strong evidence that the family plays a major role in juvenile
delinquency (Grove 8 Crutchﬁeld, 1982). Many people believe that researchers
simply do not see the family as an important enough factor and view divorce as
a normal occurrence in modern families (Wilkenson, 1974). Other researchers
may be afraid of 'overpsychologizing" delinquency, but because delinquency is
part of the socialization process, which is indeed psychological, that line of
thinking does not seem logical.

In an attempt to bring the family back into delinquency research, Grove
and Crutchﬁeld (1982) examined the family process variables from a different
perspective. Instead of questioning children about their attitudes and beliefs

concerning their parents, the researchers examined the family interaction

19

variables from the parents’ point of view. Starting with Parsons and Bales’s
(1955) axiom that an intact home with a good marriage would be associated with
a lack of delinquency and that the lack of a harmonious home and good family
relationships would have a strong association with delinquency, Grove and
Crutchﬁeld went on to examine the literature supporting this premise. They found
that most of the research was merely a description of how juveniles who were
already labeled delinquents perceived their parents. The juveniles were
questioned after their delinquent behavior had been established. Thus, their
responses may have been distorted by a lack of information about their parents.
The youths also had already acquired a delinquent identity, which could have
affected their feelings and attitudes concerning their environment, as well (Grove
8 Crutchﬁeld, 1982).

The ﬁndings from Grove and Crutchﬁeld’s study indicated that the
strongest indicator of delinquency was the parents’ own perceptions of their
children. The researchers interpreted this ﬁnding in two ways. First, parents who
do not get along with their children and do not like the way they behave probably
act in ways that promote delinquency. Second, children who misbehave tend to
promote negative feelings and actions on the part of their parents. These ﬁndings
also coincided with those from other studies stressing that quality relationships
between parent and child are essential building blocks in the socialization
process. Speciﬁcally, it has been found that home quality is the most important

factor in preventing delinquency (Van Voorhis et al., 1988), that interactions within

20

the family unit are more important than structural factors in eliminating negative
behavior (Haskell 8 Yablonsky, 1982), that a reanalysis of the Glueck and Glueck
study of 1950 also indicated that family process variables are the most important
correlates of delinquent behavior (Laub 8 Sampson, 1988), and that parents who
act as positive role models, providing strong leadership and high self-esteem, are
good models for socialized behavior and insulate their children from delinquency
(McCord, 1991 ). The fact that Grove and Crutchﬁeld’s study had results Similar
to those of other studies, even though they used different methods, lends
credence to this study and strengthens the ﬁndings of previous studies as well.

Focusing, then, on parents’ role in providing a healthy interactional
environment for their children, as well as their responsibility for being good role
models, it is apparent that they also must have strong social Skills in order to
monitor their children’s behavior. Patterson and Dishion (1985) stated that poor
parenting Skills, along with deﬁcits in the child’s social skills, increase the chances
that the child will associate with delinquent peers, which in turn may lead to
delinquency.

Patterson and Dishion stated that delinquent behavior develops in two
stages. The ﬁrst stage occurs during preadolescence; it starts with a breakdown
in the family management procedures. This process damages the child’s social
and academic skills, which in turn increases the child’s antisocial behavior. This
leads to the child’s rejection by normal peers and academic failure. In the second

stage, during adolescence, the child’s poor social skills, coupled with the parents’

21
poor parenting skills, place the youth at risk for contact with deviant peer groups,
which could lead the child to engage in high rates of delinquent behavior
(Patterson 8 Dishion, 1985).

Although Patterson and Dishion leaned In the direction of Cloward and
Ohlin’s (1960) opportunity theory, their main premise is clear: For a child to resist
delinquency, he or She must have strong parents. These parents must know how
to be good role models, effective personal managers, and good communicators
and listeners; provide support and guidance; monitor their children’s behavior;
and administer discipline fairly and effectively. Parents’ failure to socialize their
children properly could very well lead to the youths’ delinquency.

Wells and Rankin (1988) found similar results when studying direct
parental controls (monitoring their children’s behavior). Their ﬁndings indicated
that the relationship between direct parental controls and delinquency is not linear
and direct, but complex and nonlinear. In other words, family process variables
are important in the analysis ofjuvenile delinquency and need to be considered
more seriously. The problem, however, is that most researchers simply have not
done a good enough job of operationalizing and analyzing those variables. Most
researchers simply have not taken the time to choose and analyze signiﬁcant
family process variables.

in an important recent study, Canter (1982) used family process variables
to examine gender differences in family bonding as a causal factor of

delinquency. She hypothesized that boys and girls have different socialization

22

experiences that result in differential bonding to conventional social order. More
important than Canter’s hypothesis or her ﬁndings were the variables She used
and how she analyzed them. Canter used variables such as family involvement
(the amount of time spent with family per week), parental inﬂuence (how much the
parents inﬂuenced what the child thought and did), family importance, family
aspirations, family norrnlessness, and family social isolation.

Finally, a researcher was looking at how family members interacted with
one another. An attempt was being made to examine not only the relationships
between and among family members, but the feelings underlying those
relationships as well. Unfortunately, Canter did not have many variables that
were clearly deﬁned as dealing with speciﬁc family issues. She did not delve
deeply enough into family functions and issues to gain a full understanding of
family functions and their effect on delinquency (Cernkovich 8 Giordano, 1987).
Although it only scratched the surface as far as family process variables go,
Canter’s study is important because it laid the groundwork for using variables that
deﬁne and stress the importance of family processes, as opposed to focusing
merely on the structural aspects of the family.

Picking up where Canter left off, Cemkovich and Giordano (1987) more
closely examined and deﬁned the family process variables. In their study they
pointed out that family interaction and attachment have a prominent position in
social control theory but have been inadequately measured in research. They

also stated that previous researchers showed that the theory does not conceive

23

the family as being as important as school, peers. and other social factors. As for
the researchers who did look at family issues, Cemkovich and Giordano pointed
out that the methods they used were either inadequate or outdated.

In forming the basis for their research, Cemkovich and Giordano stated
that, in the foundation for social control theory, Hirschi (1969) deﬁned attachment
as communication between parent and child. Cemkovich and Giordano took that
concept one step further, stating that attachment to parents is not just one
variable (the ability to communicate); rather, it is multidimensional and is
composed of many other variables. They referred to these variables as family
interaction dimensions; they present a signiﬁcant step forward in the study of
family functionalism and delinquency.

Speciﬁcally, Cemkovich and Giordano identiﬁed seven family interaction
dimensions: control and supervision, identity support, caring and trust, intimate
communication, instrumental communication, parental disapproval of peers, and
conﬂict. Another important factor is the way the researchers measured these
dimensions. Instead of asking only one or two questions about each dimension,
they posed many and complex questions, such as “How often do you purposely
not talk to your parents because you are mad at them?” ”How often do you talk
to your parents about your problems and your friends?” and “How often do you
talk to your parents about things that you’ve done that make you feel guilty?” and
had them respond to such statements as ”One of the worst things that could

happen to me is ﬁnding out i let my parents down,” ”My parents seem to wish I

...

24

was a different type of person,“ and ”In my free time away from home, my parents
know who I’m with and where I am.”

The preceding are the types of variables and measurements that need to
be used in order to enhance the understanding of juvenile delinquency. Each of
these measures takes into account the complex and interrelated factors that exist
within the family unit and gives the researcher a look at the underlying factors that
contribute to delinquency. These variables and measures must be closely
examined and improved on if any advances in studying delinquency are to be

made.

CHAPTER III

DESIGN AND METHODOLOGY

intmsluﬂim

In this study, the researcher examined family process variables and their
inﬂuence on delinquent behavior. This study was premised on the notion that the
family is a fundamental element in the socialization of children; it was undertaken
to measure qualitative and interactive aspects of family life relative to delinquent
behavior. Many researchers have focused on family structure variables such as
broken homes and Single-parent families and their effect on delinquent behavior
(Chilton 8 Markle, 1972; Glueck 8 Glueck, 1950; Mastueda 8 Heimer, 1987;
Rankin, 1983). However, those studies did not yield signiﬁcant or conclusive
ﬁndings regarding the cause of delinquent behavior (Cernkovich 8 Giordano,
1987)

Previous researchers have dealt primarily with stnictural aspects of the
family. What is puzzling about this is that there have been indications that it is the
complex interactive relationships within the family unit, and not Structural aspects
of the family, that affect delinquency (Nye, 1958).

Although some researchers examined the qualitative functions of the

family, they failed to include in their studies a number of family process variables,

25

26
such as discipline, activities, communication. and family conﬂict (Canter, 1982;
Johnson, 1985; Noriand et al., 1979). Therefore, the researcher’s objective in this
study was to examine and analyze family process variables (i.e., family
communication, family conﬂict, family discipline, and family activities), and their
relationship to youths’ delinquency.

The research process for this study took place in three phases. The ﬁrst
phase involved identifying an appropriate sample and compiling the data. The
second phase involved identifying the independent and dependent variables for
the study. The ﬁnal phase involved choosing and applying statistical procedures

suitable for the data analyses.

Researchﬂesign

To achieve the objectives of this research, the researcher used data
contained in the National Survey of Children, a longitudinal study of youths that
was conducted in 1976, 1981 , and 1987. Because of the lower delinquency rates
in 1976 and 1987, this researcher examined the data for 1981. That was an
advantageous year because at that time most of the children in the survey were
in their middle teenage years, which is a prime time for delinquent behavior.
Moreover, the focus of the 1981 questionnaire was marital conﬂict and its
disruption of children, which may be related to the individual child’s level of

delinquent behavior.

27

It is noteworthy that not only children, but also parents and/or legal
quardians, were questioned and examined in the National Survey of Children.
This is relevant because past studies focused almost exclusively on children,
thereby limiting the research to the child’s perception or interpretation of reality.
By questioning both parents/guardians and children, a more comprehensive
analysis is possible. For example, a researcher can examine how a parent’s
attitudes, communication, supervision, and disciplinary skills inﬂuence the
behavior of the child.

Another advantage to using data from the National Survey of Children was
the size of the sample. More than 1,400 children and their parents/guardians
were interviewed. This large sample Size, combined with the greater number of
questions designed to measure dynamic aspects of family Interaction, made the

survey that much more appealing for the purposes of this study.

Emcedntes
The present study involved a secondary analysis of the National Survey
of Children for the years 1976, 1981, and 1987. The survey was administered in
three waves. Wave one of the survey was administered in 1976 and was
sponsored by the Foundation for Child Development The purposes of that wave
of the study were to assess the physical, social, and psychological well-being of
different groups of American children; to develop a proﬁle of the children’s lives

and the quality of care they received; to analyze relationships between the

 

28
conditions of the children’s lives and measures of child development and well-
being; and to replicate items from previous national studies of children and their
parents in orderto analyze trends overtime (Zill, Furstenburg, Peterson, 8 Moore,
1990).

The Institute for Survey Research at Temple University conducted ﬁeld
work, and interviews were concluded with 2,031 children and the parent who had
the most knowledge about the child (in most cases, it was the mother). In spring
1977, follow-up surveys were sent to schools that the children in the survey
attended. Of the teachers who were questioned, 80% responded to the survey
(Moore 8 Peterson, 1990).

The children who participated in the study were between the ages of 7 and
11 at the time of the ﬁrst survey; speciﬁcally, they were born between September
1, 1964, and December 31, 1969. The original sample was a multi-staged,
stratiﬁed sample of households in the continental United States that contained at
least one child. There were 2,193 households in the ﬁrst sample group. For that
sample, data were obtained for 2,031 children from 1,747 households, for a
response rate of 80%. In households with two eligible children, both of them were
interviewed; if there were three or more children, two were selected at random
(Moore 8 Peterson, 1989).

Black households were oversampled to yield interviews with approximately
500 black children. Within each racial group, several measures were employed

to ensure that each eligible child had an equal probability of being selected.

29
Weights were developed forthis oversampling and to correct for minor differences
between census and sample ﬁgures for age, gender, and race of the children, as
well as residential location.

The focus of the second wave of the survey, which was administered in
1981, was to determine the effects of marital conﬂict and disruption on the
children sampled. Second interviews were sought with all of the children who had
been found in 1976 to be living in high—conﬂict or disrupted families. Ninety
percent of those children had been relocated and interviews were conducted with
more than 90% of them, giving a response rate of 82%. A total of 1,423 children
completed the second interview: 1,036 whites and 337 blacks (Moore 8
Peterson, 1989).

The goals of the follow-up survey were to develop a proﬁle of the
behavioral and mental health of the children at various stages in the marital-
disruption process and to examine the child, parent, and family factors that were
thought to inﬂuence the risk of childhood problems that could be attributed to
marital disruption (Moore 8 Peterson, 1989).

In addition to the questions that were asked in the original survey, new
data were obtained with regard to parent-child interaction and topics that are more
relevant for teenagers, such as drinking, smoking, sexual activity, drug use, and
delinquency. In households that had experienced a divorce, separation, or some
other type of marital disruption, questions were asked about the child’s

relationship with the parent who was living outside the home.

30

Wave three ofthe National Survey of Children was conducted in 1987, with
primary funding coming from the Center for Population Research and the National
Institute of Child Health and Human Development. The purpose of the third wave
was to examine the social, psychological, and economic well-being of the sample
group as they became adults; primary interest was in the sexual activity and
fertility of the group. Questions that dealt with these issues, concerning sexual
activity, contraception, pregnancy, and childbearing, were added to the survey. ’

In the present study, the researcher conducted a secondary analysis of
data from the National Survey of Children. Speciﬁcally, the writer focused on data
that were collected during the second wave of the survey, in 1981, because
delinquency was more likely to occur during that stage and the survey contained
questions that speciﬁcally addressed delinquency issues. The purpose was to
examine the effect of family process variables (i.e., family communication, family

conﬂict, family discipline, and family activities) on youths’ level of delinquency

D . I' [II S l

Of the children in the survey, 72.7% were of a nonminority (or white),
23.6% were black, and 2.7% were Spanish American; only .2% were Oriental,
and .6% were described as 'other.‘ The sample was almost evenly split with
regard to gender; 49.9% were male and 50.1% were female. As for the parents

in the survey, 73.4% were of a nonminority (or white), 23.6% were black, and

31
2.5% were Spanish American; only .1% were Oriental, and .2% were described
as ”other.”

With regard to mothers’ educational background, 45.1% had graduated
from high school, and of that number, 20.6% had gone on to attend college. Of
those who went to college, 7.9% received a bachelor’s degree, 1.7% received an
associate’s degree, 1.1% earned a master’s degree, and .1% received a
doctorate degree. As for fathers’ educational background, 36.2% had graduated
from high school, and of that number, 26.9% had attended college. Of those who
went to college. 8.8% received a bachelor’s degree, 2.9% went on to earn a
master’s degree, 1.3% received an associate’s degree, and .4% worked toward
a doctorate degree.

The composition of the family unit of children in the survey was as follows:
64% of the children said that they lived with both of their biological parents, 16.2%
lived with their mother, 9% lived with their mother and stepfather, 5.1% lived with
their mother and another adult, 2% lived with their father, and 1.5% reported that

they did not live with either of their biological parents.

Variables
Dependenmariable
Level of delinquency was the dependent variable. This consisted of status,
property, and personal offenses. Status offenses pertained to how old the child

was the last time he or She ran away from home. Property offenses concerned

32

how often the child had stolen something and whether he or She had ever
damaged any property (i.e., vandalism). Personal offenses indicated the number

of times the child had hurt someone badly (i.e., assault).

independenLMatiablﬁs

Because previous studies did not include qualitative variables, this
research focused exclusively on some of the qualitative dynamics of the family.
These family process variables included the family’s level of communication,
family social leisure activities, family conﬂict, religiosity, parental supervision, and
parental discipline. All of these variables are complex and are intertwined.
Almost all of them relate to or inﬂuence another variable, either directly or
indirectly. Therefore, it was important to measure each family process variable
thoroughly. To that end, a conﬁrmatory factor analysis was concluded to identify
multiple indicators of a singular family dynamic and to reduce the quantity of data.
Each variable in this study had a corresponding question or questions on the
survey that were designed to measure all aspects of the relationship between and
among members of the family.

A factor analysis of the family process variables did not conﬁrm multiple
indicators of any one particular family process variable. Thus, only single
measures were used to describe a family’s level of communication, social leisure

activities, conﬂict, religiosity, parental supervision, and parental discipline.

33

Two questions from the survey concerned levels of family communication.
The level of communication was measured in terms of the frequency with which
parents discussed decisions with their children, and how much say the child had
in establishing the family rules.

Family social leisure activities focused primarily on the number of times the
family dined out together, whereas family conﬂict described how often the family
argued about-rules. The level of religiosity was Indicative of the number of times
the family attended religious services together, and parental supervision mirrored
the amount of unsupervised time the child was afforded per day. Parental
discipline was assessed in terms of the frequency with which the parents
administered discipline.

For measurement ofthe independent variables, a factor analysis was used
to help organize the variables and eliminate any insigniﬁcant variables. A
correlation analysis was then conducted to determine the strength of the
relationship between the dependent and independent variables. Finally, with the
independent variables that were shown to have signiﬁcance under the correlation
analysis, a multiple regression analysis was used to determine further the

relevance of those relationships.

CHAPTER IV

RESULTS

introduction

The researcher’s purpose in this study was to examine the inﬂuence of family
process variables (i.e., family communication, family conﬂict, family discipline, and
family activities) on youths’ level of delinquency. A correlation analysis of selected
family process variables and delinquency was performed ﬁrst. This analysis Showed
the amount of association between the aforementioned family process variables and
Status, property, personal, and drug offenses. This analysis was further speciﬁed
through the inclusion of race, gender, and type of family unit (Single versus multiple-
parent households) as control variables. A discussion of the ﬁndings regarding
these family process variables and their association with delinquency follows.

Another objective of the study was to assess the predictability of delinquency
given a juvenile’s family situation. Hence, regression analyses were used to
determine the association between the family process variables and status, property,
personal, and drug offenses. Results of these analyses also are discussed in the

following sections.

 

35

and_Status_foenses

Status offenses was measured by the variable ”number of times that you
have run away from home in the past 12 months.“ Thirteen percent of the sample
group had run away from home at least once. Of that number, 81% were white and
18% were minority; 40% were male and 59% were female.

As Shown in Table 1, the variables number of times a week the family eats
dinner together and frequency ofdiscipline showed negative relationships (-.10 and
-.05, respectively) with running away from home. This suggests that activities such
as sitting down and sharing a meal together and frequency of discipline deter a child
from running away from home“.

The ﬁndings regarding unsupervised time seemed surprising at ﬁrst.
Although the correlation of this variable with running away was fairly weak (1: = -.04),
the negative relationship was the opposite of what one would expect. This ﬁnding
suggests that the more unsupervised time the youth has, the lower the chances of
his or her running away. If one takes into consideration, however, that a child would
have little desire to run away if he or She had all the free time desired, and that a
child who is under constant supervision may tend to feel suffocated and want his or
her ”own space, the negative relationship makes sense.

The variable with the strongest correlation to running away, frequency of
arguments about rules (r = -.11), is an indication of family conﬂict. This negative

association suggests that the more the parent and child argue, the less apt the child

36

Table 1: Association between family process variables and Status offenses:
frequency of running away from home.

 

 

 

 

 

 

TOTAL SAMPLE
Correlation N Corr. l
Argue about rules 1,353 -.11“ I
# of times per week eat dinner together 1,389 -.10"
Frequency of discipline 1,121 -.05

 

 

 

 

   

" = one-tailed ” = two-tailed

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Multiple Regression . Beta SE Beta 1 Sig. oft

# of times per week eat -.09 .04 -2.6 .01
dinner together

Argue about niles -.08 .03 -2.5 .01
Frequencyofdiscipline -.08 04 -2.4 .02 I
Sex dummy -.02 04 -5.6 .58

Race dummy -.06 .04 -1.7 .09 1
Family dummy -.10 .04 -2.8 .01 I

—_—=—

 

 

 

MultB 3’ Adj.Bz I E(Eqn) ISig.ofEI B’CH I ECl-l I Sig. ofCH I

 

Table 1: Continued.

37

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

WHITE MALES
Correlation N Corr.
# of times per week eat dinner together 510 -.13"’"
Frequency of discipline 430 -.12*
4i
Argue about rulcs 500 -.1 1*
How much say in family nJleS 524 .10'
Took family out to dine 526 .09
Parents discuss decisions with child 511 .07
* = one-tailed *" = two-tailed
Multiple Regression Beta SE Beta 1 Sig. oft
# of times per week eat -.13 .06 -2.3 .02
dinner together
Frequency of discipline -.10 .06 0.8 .44
Argue about mles -.12 .06 -2.2 .03
How much say in family .07 .06 1.2 .22
rules I
Took family out to dine .06 .06 1.1 .28
Parents discuss decisions 04 .06 0.7 .49
with child
Family dummy -.14 .05 -2.6 .00 AI
1 —r 7

 

 

 

 

 

 

 

 

I MultB I 82 AdI.B2 E(Eqn) ISig.ofE

 

B’CH I ECH I Sig.ofCH

 

 

 

L30 I .09 .06

 

2.8 I .oo

.09 I 2.8

.00

 

38

Table 1: Continued.

 

 

 

 

 

WHITE FEMALES
Correlation N Corr.
# of timcs per week eat dinner together 506 -.11"
Argue about mics 498 -.10'

 

 

 

 

* = one-tailed " = two-tailed

 

 

 

 

 

 

Multiple Regression “I Beta SE Beta 1 Sig. oft
# of times per week eat -.09 .05 -1.7 .09
dinner together ‘
Argue about rules -.05 .05 -1.0 .33 I
Family dummy -.19 .05 «3.7 .00 I

 

 

 

 

 

 

 

 

[MultBI 32 I Adj.B’ I E(Eqn) ISig.ofEI BZCH I ECH I Sig.orch=I

.29 .09 .06 2.9 .00 .09 2.9 .00

39

Table 1: Continued.

 

 

 

 

 

 

MINORITY MALES
__==
Correlation N Corr.
Unsupervised time 184 -.10
How much say in family rules 180 .10

 

 

 

 

* = one-tailed " = two-tailed

 

 

 

 

 

 

 

 

 

 

 

 

 

E _
Multiple Regression , Beta SE Beta 1

Unsupervised time -.14 .13 -1.1
How much say in family .12 .13 0.7
rules

#oftimes perweekeat .17 .13 1.3
dinner together

Family dummy -.12 .14 —0.9

l “——

 

Sig. ofE 132 CH
.40 .16 .00 1.0 .45 .16 1.0 .45

 

 

Table 1: Continued.

 

 

40

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MINORITY FEMALES
F a
Correlation N
If of timcs per week eat dinner together 192
Unsupervised time 203
Frequency of discipline 141
Parents discuss decisions with child 192
Took family out to dine 202
* = onetailed “ = two-tailed
Multiple Regression Beta SE Beta 1 Sig. oft
# of times per week eat -.05 .10 -0.5 .63
dinner together
Unsupervised time -.07 .10 -0.7 .50
Frequency of discipline -.20 .10 -2.0 .05
Parents discuss decisions -.11 .10 -1.1 .29
with child
Took family out to dine -.16 .10 -1.5 .13
Family dummy

 

 

 

 

41
is to run away. Although arguing shows conﬂict, arguing is a form of communication
and interaction; it reﬂects the desire for stronger parental supervision and parent-
child interaction and thwarts truanting from home.

Then again, the negative relationship of how often parent and child argue
(-.01) and the positive relationship of whether parents discuss decisions with the
child (.01), relative to truanting from home (although weak), seem to suggest a lack
of open communication. The former correlation Indicates a lack of communication,
whereas the latter could be Interpreted as one-way communication and not indicative
of parent and child openly discussing issues, thereby leaving truanting as an option.

Interestingly, how often a family attends religious services (.03), frequency of
taking the family to the movies (.02), and frequency of taking the family out to dine
(.02) seemed to increase the chances of the child’s running away. This Is surprising
because, on the surface, these variables all seem to describe an open, cohesive
family that communicates and does activities together. However, the research
ﬁndings suggest that the aforementioned dynamics are positively correlated with
truanting from home.

Further analysis Showed that, under controlled conditions, the frequency of
dining together, discipline, and arguments over rules were inversely correlated with
truanting from home for white males. On the other hand, the amount of verbal input
into determining family rules was positively associated with running away from home

for white males. Moreover, for white females, only the frequency of dining together

42
and arguments over rules were signiﬁcantly (and inversely) correlated with truanting
from home.

However, when race was included in the correlation analysis, unsupervised
time was signiﬁcantly correlated with truanting from home for black males and
females. In addition, only the frequency of the family dining together (a < .05) was
inversely correlated with running away from home for black females.

The multiple regression analysis (Table 1, total sample) gave no suggestion
that any of the included family dynamics were strong predictors of running away.
However, the data indicated that frequency of arguments over rules (beta = 0.08),
dining together (beta = -.09), and discipline (beta = -.08) were signiﬁcant factors that
were inversely associated with truanting from home.

Moreover, the data suggested that there were familial differences in frequency
of truanting (beta = .10). In particular, family structure was a predictor of truanting.
Youths from Single-parent homes were more likely than adolescents from homes
with more than one parent to run away. On the other hand, the data did not suggest

that race or gender was a signiﬁcant predictor of truanting from home.

BII' I'BI E'IE It'll
andﬁmneﬂxﬂftenses

Two types of property offenses were examined in this study: damage to
property and theft. Twenty-three percent of the sample group said that they had
damaged property at least once. Of that number, 80% were white and 19% were

minority; 74% were male and 26% were female. A correlation analysis of the

43

offense damage to property (see Table 2) Showed that frequency of dining together
at home (1: = -.09), dining out ([ = -.06), discipline (1: = -.1 1), and arguments over rules
(r = -.12) were negatively associated with frequency of damaging property.

However, the data also showed that youths’ Involvement in decision making
(r = .07) and amount of unsupervised time ([ = .06) were signiﬁcantly associated with
the frequency of damaging property, in a positive direction. Further analysis showed
that, under controlled conditions, only the frequency of dining out had a Signiﬁcant
inverse association with frequency of damage to property for minority male
adolescents (see Table 2).

For minority females, the data showed that input into family rules and
arguments over family rules were signiﬁcantly associated with the frequency of
Involvement in property damage. In particular, input into family rules (1: = .19) was
positively correlated with property damage, whereas arguments overfamily rules (I =
-.22) was inversely associated with damage to property. Thus, for minority females,
communication regarding family rules inﬂuenced the frequency of involvement in
damage to property.

Unlike the situation with status offenses, a multiple regression analysis of the
data showed a number of signiﬁcant relationships (see Table 2). As with the
previous correlation analysis, the regression beta weights showed the importance
of communication in that frequency of arguments about rules (-. 12), dining out (-. 10),
and eating dinner together (-.08) were predictors of frequency of involvement in

property damage. In Short, the regression analysis showed that the aforementioned

44

Table 2: Association between family process variables and property offenses:

 

 

 

 

 

 

 

damage to property.
TOTAL SAMPLE

Correlation N Corr. I
Argue about rules 1,353 -.12**
Frequency of discipline 1,121 -.11“
if of times per week eat dinner together 1,389 -.09“
Parents discuss decisions with child - 1,391 .07*
Unsupervised time 1,422 .06“ I
Took family out to dine 1,421 '- .06" I

 

 

 

 

” = one-tailed " = two-tailed

lwm

 

 

 

 

 

 

 

 

 

Multiple Regression Beta SE Beta I Sig. oft
Argue about mics ~ -.12 .03 -3.5 .00
Frequency of discipline -.04 .03 -1.1 .27

# of times per week eat -.08 .03 -2.4 .02 I
dinner together

Parents discuss decisions .03 .03 1.0 .34

with child

Unsupervised time .05 .03 1.6 .1 1 I
Took family out to dine -.10 .03 -2.9 00 I
Sex dummy -.29 .03 -8.5 .00 I
Family dummy 4.20 ‘ .03 -0.0 99 I

 

 

 

 

 

 

 

 

45

Table 2: Continued.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

WHITE MALES
-I =l===—I=—=— .
Correlation N Corr.
Argue about mics 500 -.13“
Unsupervised time 526 .12: I
if of timcs per week eat dinner together 570 -.09 I
How often attend religious services 500 -.07
‘ = one-tailed ** = two-tailed
Multiple Regression Beta SE Beta 1 Sig. oft
Argue about mles -.10 .06 1.8 .08 l
Unsupervised time .08 06 1.4 .18
# of timcs per week eat -.08 06 -1.4 .17
dinner together
How often attend religious -.10 06 -1 8 .08
services
Family dummy 06

 

 

 

 

 

 

 

.26 .07 .03 I 2.0 .02 .07 2.0 I .02

 

 

46

Table 2: Continued.

 

 

 

 

 

 

 

 

 

 

 

WHITE FEMALES
T-ﬁ Correlation N Corr.
Argue about mics 498 " .11"
# of times per week eat dinner together 510 -.1 1"
Unsupervised time 509 .10”
Took family out to dine 510 -.09
———=.l——=_=_=.=

 

 

 

* = one-tailed *" = two-tailed

Multiple Regression

 

“ Argue about mles

 

 

# of times per week eat
dinner together

Unsupervised time .14 .05 2.6 .01
Took family out to dine -.08 .05 -1.6 .12

Family dummy -.01 .05 -0.1 .91 I

 

 

 

 

 

 

 

 

 

 

 

   

  

MultB 32 Adj. 3’ E(Eqn) Sig. ofEI BZCH I ECH I Sig. ofCH I
.26 2.2 .01 .07 2.2 .01 "

      

47

Table 2: Continued.

 

 

 

 

 

 

 

MINORITY MALES
Correlation N Corr.
Took family out to dine 183 -.19‘
Argue about mics 173 .07
Parents discuss decisions with child 182 .16
Frequency of discipline 116 .14
Unsupervised time 184 -.13
" = one-tailed

" = two-tailed

 

 

 

 

 

 

 

 

 

Multiple Regression Beta SE Beta 1 Sig. oft I
Took family out to dine -.19 .13 -1.5 .15
Argue about mics .17 .12 -1.4 .15
Parents discuss decisions .12 .13 0.7 .35
with child
H Frequency of discipline .21 .13 1.6 .12
Unsupervised time -.16 .12 -1.4 .17

 

 

 

 

T Mum ”—7 m2. " seq. 7 sees son“

 

 

 

Sig. of CH ;

 

.54 .29 .15

2.1

.03

2.1

Table 2: Continued.

48

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MINORITY FEMALES
Correlation N Corr.
Argue about mles 182 -.22*
How much say in family mles 198 .19 I
How often attend religious services 200 09 l
Parents discuss decisions with child 192 -.08 I
Frequency of discipline 141 -.07 I
Took family out to dine 202 .06 I
" = one-tailed ” = two—tailed -
Multiple Regression Beta SE Beta 1 Sig. oft
Argue about mleS -.29 .12 -2.5 .01
How much say in family -.07 .11 -0.6 .55
mics
How often attend religious .04 .11 0.4 .73
services
Parents discuss decisions -.05 .11 -0.5 .62
with child
Frequency of discipline -.11 .11 -1.0 .33
Took family out to dine .05 .11 0.5 .62
Family dummy .

 

 

 

 

 

 

 

 

 

 

IIMultB 3’ Adj. 3’ E(Eqn) Sig. ofEI BZCH ECH Sig. ofCH II
II .42 .18 .06 1.5 .15 I .18 1.5 .15 II

 

 

 

 

 

 

 

49
family dynamics were inversely associated with frequency of involvement in property
damage. This was particularly true for male adolescents. No signiﬁcant racial or
family stmctural effects were observed. Moreover, the family variables included in
the regression analysis explained only 12.5% of the variance relative to the
frequency of involvement in property damage.

With regard to theft, 19% of the sample group admitted to stealing something
at least once. Of that number, 80% were white and 20% were minority; 60% were
male and 40% were female. The correlation analysis (T able 3) showed a positive
relationship between frequency of discipline ([ = .17) and arguments over mles ([ =
.06) and involvement in theft. Table 3 also depicts a negative association between
the amount of unsupervised time (r = -.12) and frequency of involvement in theft.

The correlations for involvement in theft were the opposite of what they were
for the other type of property offense—damage to property. This ﬁnding might be
attributed to the fact that stealing is a different type of offense and may be inﬂuenced
more by the youth’s peers than by his or her family. Stealing could also be a form
of rebellion. A high rate of discipline and frequent arguments over mles may make
a child feel dominated by his or her parents and thus compel the youngster to rebel.

Further analysis Showed that, for minority adolescents, with the exception of
number of times eat dinner together for minority males, none of the family dynamics
was signiﬁcantly correlated with theﬂ. This was true for both minority males and

females (se Table 3).

50

Table 3: Association between family process variables and property offenses:

 

 

 

 

theft.
TOTAL SAMPLE
I Correlation N E Corr. 1
Frequency of discipline 1,121 .17“ I
Unsupervised time 1.422 _.1 2.. l
Argue about mles 1,353 .06‘

 

 

 

 

 

" = one-tailed " = two-tailed

 

 

 

 

 

 

 

 

Multiple Regression Beta SE Bite-t;- t Sig. oft
Frequency of discipline . .14 .03 4.5 .00
Unsupervised time -.12 .03 -3.8 .00
Argue about mles .05 .03 1.8 .07
Sex dummy .06 .03 1.8 .07
Race dummy .10 .03 2.8 .01
Family dummy .10 .04 3.0 .00

 

 

 

 

 

 

 

 

Mult.B 32 Adj.3’ E(Eqn) Sig. ONE 3’ CH ECH Sig.ofCH

 

 

 

 

 

51

Table 3: Continued.

 

 

 

 

 

 

 

 

WHITE MALES
Unsupervised time 526 -.19“ II
Frequency of discipline 430 .17" I
Parents discuss decisions with child . 511 -.07 I
Argue about mics 500 .07
How often attend religious services 500 -.06

 

 

* = one-tailed ** = two-tailed

 

 

 

 

 

 

 

 

 

 

 

 

 

Multiple Regression Beta SE Beta 1 Sig. oft II
Unsupervised time i -.16 .05 -3.2 .00
Frequency of discipline .16 .05 3.1 .00
Parents discuss decisions -.06 .05 -1.1 .26
with child
Argue about mles .03 .05 0.6 .53

I How often attend religious -.04 .05 -0.9 .39
servrces
I Family dummy -.00 .05 -.08 .94 II

 

 

 

 

Mult.B 3’ Adj.Bz E(Eqn) Sig.ofE B’CH ECH Sig. ofCH
* .27 .07 .05 2.7 .00 .07 2.7 .00 '

52

Table 3: Continued.

 

 

 

 

WHITE FEMALES
Correlation N Corr.
Frequency of discipline 434 .24“
Unsupervised time 509 -.10*
Argue about rules

 

 

 

 

* = one-tailed " = two-tailed

       
   

 

Multiple Regression

 

Frequency of discipline .19 .05 3.8
Unsupervised time -.09 .05 -1.8
Argue about mles .04 .05 0.9

 

 

 

 

 

 

 

 

 

 

Table 3: Continued.

 

 

 

 

 

 

MINORITY MALES
— ——
Correlation N Corr.
# of times per week eat dinner together 181 .13
ll Frequency of discipline 116 .08
How much say in family rules 180 .05

* = one-tailed ** = two-tailed

 

 

 

 

 

 

 

 

Multiple Regression - Beta SE Beta 1
# of times per week eat .21 .10 2.2
dinner together
Frequency of discipline .07 .11 0.7
How much say in family .05 .10 0.5
rules
Family dummy .30 .10 3.0

 

.40 .16 .06

1.6

 

.11

 

.16

1.6

 

 

*E '7

MultB 32 Adj. 32 E(Eqn) Sig. ofE 32 CH ECH Sig. of CH .

.11

Table 3: Continued.

 

 

 

 

 

 

 

 

 

 

 

 

 

MINORITY FEMALES
Correlation N Corr.
How often attend religious services 200 .08
Took family out to dine 202 .06
Unsupervised time 203 -.06
Frequency of discipline 141 -.05
“ = one-tailed “ = two-tailed
Multiple Regression Beta SE Beta 1 Sig. of: I
How often attend religious .10 .10 1.0 .31

services

 

Took family out to dine

 

 

Unsupervised time

   

 

Frequency of discipline

  

 

 

 

Family dummy

 

 

 

 

 

 

55
In the regression analysis (see Table 3, total sample), the beta weights for
frequency of discipline (.14), unsupervised time (-.12), and argue about rules (.05)
all followed the same trend as they did in the correlation analysis. Again, this
suggests that stealing may be more a result of the youth’s rebelling against his or
her family (or perhaps even being more inﬂuenced by peers) than of the family
process variables.

BII' I'EI EiIE 1!.“
andEersonaLQﬁenses

With regard to personal offenses. 13% of the sample group admitted to
having hurt someone at least once. Of that number, 74% were white and 26% were
minority; 67% were male and 33% were female. Results of the correlation analysis
for personal offenses. in this case the number of times a juvenile had physically hurt
someone, are shown in Table 4. Although frequencies of dining together, going out
to dine, discipline, and arguments over rules were inversely associated with personal
assaults. only the frequency of discipline ([ = -.18) had a significant negative
association. The frequency with which parents discussed decisions with the child
had a signiﬁcant positive association with frequency of personal offenses ([ = .06).

Variables such as frequency of dining together ([ = -.03), level of input into
family rules (i: = .01), amount of unsupervised time (t = .02), and frequency of
arguments about mles ([ = -.01) had weak correlations with frequency of personal

offenses. Discipline and parental discussions with children regarding decisions

56

Table 4: Association between family process variables and personal offenses (how
often the juvenile physically hurt someone).

 

 

 

 

 

 

 

 

 

 

 

 

 

   
  
   
  
  
 

 

 

 

 

 

TOTAL SAMPLE
Correlation N
Frequency of discipline 1.121
Parents discuss decisions with children 1,391
* = one-tailed " = two-tailed
Multiple Regression Beta SE Beta I
Frequency of discipline -.14 .03 -4.4 .00
Parents discuss decisions ' .04 .03 1.3 .20
with children
Sex dummy -.11 .03 -3.5 .00
Race dummy .04 .03 1.1 .27
Family dummy

 

 

 

 

 

  

 

 

 

 

 

Table 4: Continued.

57

WHITE MALES

 

Correlation

   

 

 

 

  

 

Frequency of discipline 430 -.17"
How often attend religious services 500 .07
Took family out to dine 526 -.06

   

 

 

 

* = one-tailed "" = two-tailed

Parents discuss decisions with child

 

l “

 

   

 
 

 

  

 

 
 

 

  

 

 

 

 

 

 

MultB B’ Adj.32
i .21 .05 .01

1.6

 

.10

 

.04

Multiple Regression Beta SE Beta
Frequency of discipline -.15 .05 -2.9 .00
How often attend religious .05 .05 1.0 .31
services
Took family out to dine -.04 .05 -0.7 .50
Parents discuss decisions - .06 .05
with child

E(Eqn) Sig. ofE B’CH ECH Sig. ofCH l
1.6

 

 
 

.10

_ ___ ___J

58

Table 4: Continued.

 

 

 

 

 

 

 

 

 

WHITE FEMALES
I ‘ ' ‘ ' T T _ _T _ T 1
Correlation N Corr. g

Frequency of discipline 434 -.12"

How often attend religious services 484 .06 I
' = one-tailed “ = two-tailed
I r

Multiple Regression Beta SE Beta

Frequency of discipline . -.10 .05

How often attend religious .03 .05

services

Family dummy -.05 .05

 

 

 

 

 

 

 

MultB 3’ Attila2 E(Eqn) Sig. ofE B’CH ECH Sig. ofCH '
.18 .02 -.oo .88 .56 | .02 I .88 | .56

59

Table 4: Continued.

 

 

 

 

 

 

MINORITY MALES
Correlation N Corr.
Took family out to dine 183 -.16"
Parents discuss decisions with child 182 -.16
Frequency of discipline 1 16 -.12
Argue about rules 173 -.10
How often attend religious services

 

 

 

 

‘ = one-tailed “ = two-tailed

 

 

 

 

 

 

 

 

Multiple Reegrssron

Took family out to dine -.30 .12 -2.6
Parents discuss decisions -.21 .11 2.0
with child

Frequency of discipline -.09 .11 -0.9
Argue about rules -. 14 .10 -1.4
How often attend religious -.14 .10 -1.5
servrces

Family dummy -.15 .10 -1.5

 

 

 

 

 

 

MulLB 13’ Adj. 3’ E(Eqn) Sig. ofE BZCH
.40 .16 .06 1.6 .11 .16

 

 

 

 

60

Table 4: Continued.

 

 

 

 

 

 

 

MINORITY FEMALES
Correlation N
# of times per week eat dinner together 192
Unsupervised time 203
Frequency of discipline 141
Parents discuss decisions with child 192
Took family out to dine 202 .05

 

 

 

 

‘ = one-tailed " = two-tailed

   
   
  
  
   
 
 

 

 

 

 

 

 

 

 

 

 

 

Multiple Regression 8? SE Beta 1

# of times per week eat -.20 .10 -2.1 .04
dinner together

Unsupervised time .19 .10 2.0 .04
Frequency of discipline -.17 .09 -1 .8 .08
Parents discuss decisions -.13 .09 -1.4 .16
with child

Took family out to dine .06 .09 0.6 .56
Family dummy

 

 

 

(9% ts _ -s--.---s-_--_-,stﬁs__-m_-_-_- _t H.
l Mult.B 32 Adj. 3’ E(Eqn) Sig. ofE 32 CH ECH Sig. ofCH l
l .34 .12 .03 1.4 .20 .12 1.4 .20 ‘

 

61
were the only factors that were significantly associated with frequency of
involvement in personal offenses.

Similar results were observed in terms of race. in particular. Table 4 shows
that, for white juveniles, frequency of discipline was significantly correlated with
frequency of involvement in personal offenses. This was true for both males ([ =
-.17) and females (1: = -.12). However, for minority juvenile males, taking the family
out to dine and frequency of discipline had strong negative associations with
frequency of involvement in personal offenses; parents discussing decisions with
their children also had a significant correlation with personal offenses. Frequency
of discipline and parents discussing decisions with their children also had significant
associations for minority females, but number of times per week the family eats
dinner together and unsupervised time. respectively, had significant negative and
positive relationships.

The regression analysis (Table 4, total sample) showed a negative beta for
frequency of discipline (-.14) and a positive beta for parental discussions of decision
making (.04) relative to the frequency of involvement in personal offenses. This
finding corresponds with the correlation analysis, in that frequency of discipline and
involvement in discussions of decisions made by parents predicted youths’
frequency of involvement in personal offenses. These data do not suggest that race.
sex, and family composition (number of adults in the household) are predictors ofthe

frequency of involvement in personal offenses. However, it should be noted that the

62

aforementioned variables explained just a small percentage of the variance (32 =
.06) in frequency of involvement in personal offenses.

BII' IiEI EiIE 1!.”
andAlmhoLandegQﬂenses

Eighty-ﬁve percent of the sample group said that they had drunk alcohol at
least once. Of that number, 81% were white and 19% were minority; 50% were
males and 50% were females. The only variable that signiﬁcantly correlated with
frequency of drinking alcohol was frequency of attendance at religious services (r =
.09) (see Table 5).

However, when the data were examined according to race, attendance at
religious-services was not signiﬁcantly associated with frequency of drinking for
white or black adolescents. Moreover. none of the family dynamics included in the
analysis was signiﬁcantly associated with drinking for either white or black males.

For white females, frequency of attendance at religious services was
positively correlated with drinking ([ = .12), whereas for black females, arguing about
rules was positively associated with drinking (1: = .17), and dining out together ([ = -
.21) and frequency of discipline (1: = -.13) were inversely correlated with frequency
of consumption of alcohol.

The multiple regression analysis showed that frequency of attendance at
religious services (beta = .08) was a predictor of frequency of alcohol consumption

for adolescents (see Table 5. total sample). In addition, the regression analysis

63

Table 5: Association between family process variables and alcohol and drug

offenses (how often the juvenile drank alcohol).

 

 

 

 

 

 

 

 

 

TOTAL SAMPLE
T Correlation N Corr. 1
How often attend religious services 1.362 .09“
' = one-tailed “ = two-tailed
Multiple Regression Beta SE Beta 1 Sig. oft l
i
How often attend religious .08 .03 2.2 .03 I
services 7
Sex dummy .14 .04 3.8 .00 1
Race dummy -.11 .04 -3.0 .00 I
Family dummy .01 .04 .154 .88

 

 

 

32 Mi 3’ E (Eon)

 

 

   

 

.21 .04 .03 3.4 .00

 

.04

3.4

 

 

 

Table 5: Continued.

 

 

 

 

 

 

WHITE MALES
Correlation N Corr. I
Took family out to dine 526 .07 I

 

 

 

 

 

 

* = one-tailed '* = two-tailed

 

 

 

 

 

 

 

 

 

MultB 3’ Adj.3’ E(Eqn) Sig. ofE 3’CH

Multiple Regression Beta SE Beta 1 Sig. oft I
Took family out to dine .03 .06 0.6 .58
Family dummy -.04 .06 0.8 .45

 

ECH Sig. ofCH

 

.21 .04 .01 1.3 .25 .04

1.3

.25

Table 5: Continued.

65

 

 

 

   

 

WHITE FEMALES
Correlation N Corr
How often attend religious services 484 .12‘
Frequency of discipline 434 -,03

 

 

‘ = one-tailed ” = two-tailed

# of times per week eat dinner together

Multiple Regression - Beta

 

 

 

 

     

 
  

 

  

 

dinner together

SE Beta
How often attend religious .11 .05
services ‘
Frequency of discipline -.06 .05
# of times per week eat -.08 .05

  

 

 

 

Family dummy

 

MultB 3’ A33 E(Eqn) Sig.ofE 3’ CH

 

 

 

ECH

Sig. of CH

 

.23 .05 .02

1.8

 

.05 .05

1.8

.05

 

_ l

l
l

l

66

Table 5: Continued.

 

 

 

 

MINORITY MALES
l—
Correlation N Corr.
Took family out to dine 183 .07
# of times per week eat dinner together 181 -.07 1
Parents discuss decisions with child 182 .07

 

 

 

‘ = one-tailed " = two-tailed

 

 

 

 

 

 

 

Multiple Regression Beta SE Beta 1 Sig. oft
Took family out to dine .00 .14 .03 .98
# of times per week eat -.07 .13 -.51 .61
dinner together
Parents discuss decisions -.06 .14 -.42 .68
with child
Family dummy

 

 

 

 

 

MultB 3’ Adj. 3’ E(Eqn) Sig. off
.39 .15 -.01 .93 .52

 

 

 

 

 

67

Table 5: Continued.

MINORITY FEMALES

    

 

 

 

 

Correlation
Took family out to dine 202 -.21' II
Argue about mles 182 .17
Frequency of discipline 141 -.13
Parents discuss decisions with child 192 -.06

 

 

 

 

 

" = one-tailed “ = two-tailed

 

 

 

 

 

 

 

 

 

 

Muipie segment * see If I Be I H I I ' sig. i I ‘;
Took family out to dine -.33 .11 -3.0 .00 I
Argue about rules .06 .11 0.5 .62 I
Frequency of discipline -.12 .11 -1.1 .25 I
Parents discuss decisions -.21 .11 -2.0 .05 I

l with child l
How much say in family .01 .11 0.1 .89 I
mles I
Family dummy .02 .11 0.2 .88

 

 

 

68
showed that sex and race were predictors of frequency of alcohol consumption, with
male and white juveniles consuming alcohol more frequently than others.

The fact that a positive relationship was found between attendance at
religious services and frequency ofalcohol consumption was surprising. One would
think that the opposite would be true. given the position taken against consumption
of alcohol by religions in general. Perhaps alcohol use by minors could be a form
of rebellion against the strict moral codes of most religions. or the pressures
imposed by religion might contribute to alcohol use.

For drug offenses, use of two drugs—marijuana and cocaine/crack-was
examined. Fifty-three percent of the sample group admitted to having experimented
with marijuana at least once. Of that number, 80% were white and 20% were
minority; 51% were male and 49% were female. The correlation analysis for
marijuana use showed a strong negative relationship between that offense and
frequency of arguments about household rules (-.11) and a strong positive
relationship with frequency of attendance at religious services (.10) (see Table 6).
Also showing signiﬁcant relationships to marijuana use were the number of times per
week the family eats dinner together (-.09), frequency of discipline (-.07), and
unsupervised time (.07). Once again, frequency of attendance at religious services
was found to be signiﬁcant, whereas the other variables followed the trends that
were displayed in previous analyses.

The multiple regression analysis for marijuana use showed fairly high beta

weights for the signiﬁcant variables (see Table 6. total sample). The beta weights

69

Table 6: Association between family process variables and alcohol and drug
offenses (how often the juvenile used marijuana).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

TOTAL SAMPLE
Correlation N Corr.
Argue about rules 1,353 -.11"
How often attend religious services 1.362 .10”
# of times per week eat dinner together 1,389 -.09"
Frequency of discipline 1,121 -.07"
Unsupervised time 1.422 .07' I
* = one-tailed " = two-tailed
IT- Multiple RegLression Beta SE Beta 1 Sig. oft
Argue about rules -.09 .03 -2.5 .01
How often attend religious .08 .03 2.4 .02
servrces
# of times per week eat -.10 .03 -2.8 .01
dinner together
Frequency of discipline -.06 .04 -1.8 .07
Unsupervised time .10 .03 2.9 .00
Sex dummy -.01 .04 -o.2 .87 I
Race dummy -.05 .03 -1.3 .18 I
Family dummy -.01 .04 -0.4 .71 I

 

 

 

 

 

 

 

Muit3 3’ Ad]. 32 E(Eqn) Sig.ofE 3’CH ECH Sig.ofCH
' 3.8

 

Table 6: Continued.

70

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

WHITE MALES
Correlation N Corr.
Argue about rules 500 -.17“
Unsupervised time 526 .15
Frequency of discipline 430 -.09
Parents discuss decisions with child 511 .09
# of times per week eat dinner together 510 -.08
How often attend religious services 500 .06
How much say in family mles 524 .05
" = one-tailed “ = two-tailed
Multiple Regression Beta SE Beta t Sig. oft ]
Argue about rules -.11 .06 -2.0 .05
Unsupervised time .07 .06 1.3 .19
Frequency of discipline -.10 .06 1.8 .09
Parents discuss decisions .08 .06 1.4 .17
with child
# of times per week eat -.06 .06 -1.0 .31
dinner together
How often attend religious .07 .06 1.2 .22 I
servrces
How much say in family .03 .06 -o.5 .60 II
rules
Family dummy .02 .06 0.3 .76J
I Mult 3 I 32 I Adj. 3’ I EtEqn) I Sig. ofE I 3’CH I ECH I Sig. ON
I .27 I .08 I .04 I 2.2 02 I . 7 I 2.2 I

 

9:1

Table 6: Continued.

71

 

 

 

 

 

 

WHITE FEMALES
III: Ctrrelation —- N Corr. 1
Unsupervised time 509 .16“
Argue about rules 498 -.12
How often attend religious services 484 .12‘
Frequency of discipline 434 -.09 I

 

 

 

" = one-tailed ‘* = two-tailed

 

 

 

# of times per week eat dinner together 506 -.07 I

 

 

 

 

 

 

 

Multiple Regression Beta SE Beta 3
Unsupervised time .17 .05 3.2
Argue about mles -.06 .05 -1.1
How often attend religious .09 .03 1.8
services
Frequency of discipline -.09 .05 -1.7 .10
# of times per week eat .04 .05 0.7 .47
dinner together

 

 

 

 

 

 

 

72

Table 6: Continued.

 

 

 

 

 

 

MINORITY MALES
Correlation N Corr. II
How often attend religious services 178 .18' II
Argue about rules 173 -. 12 I
How much say in family mles ' 180 .09 I
Took family out to dine
Unsupervised time

 

 

 

   

* = one-tailed “ = two-tailed

 

 

 

 

 

 

 

 

 

 

 

 

 

Multiple Regression Beta SE Beta 1 Sig. oft I
How often attend religious .17 .13 1.3 .18
seerces
Argue about rules -.02 .13 -0.1 .90
How much say in family .04 .13 0.3 .77
mles I
Took family out to dine -.03 .15 -0.2 .83 I
Unsupervised time -.01 .13 -0.2 .83 I
Family dummy - l

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

73
Table 6: Continued.
MINORITY FEMALES

Correlation N Corr. I
How much say in family rules 198 -.19* l
Unsupervised time 203 .18‘
# of times per week eat dinner together 192 -.17
Argue about rules 182 .08
Frequency of discipline 141 .08 I
Took family out to dine 202 -.06 II

_7 I

 

 

* = one-tailed “' = two-tailed .

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Multiple Regression Beta SE Beta =- t Sig. oft I
How much say in family -.10 .1 ‘l -0.9 .37 I
rules
Unsupervised time .01 .12 0.1 .95 II
# of times per week eat -.33 .11 -3.0 .00
dinner together
Argue about rules .09 .12 0.8 .46
Frequency of discipline .08 .1 1 0.8 .45
Took family out to dine -.16 .11 -0.1 .15
Family dummy -.09 .11 _-0.8 .44 II

 

 

II Mult3 I 3’ I Adj.3’ I E(Eqn) ISig.ofEI 3’CH I ECH I Sig.ofCH I
l

I .42 I .18 .06 I 1.5 I .14 I .18 I 1.5 I .14 I

 

74
for frequency of arguments about rules. frequency of attendance at religious
services, number of times per week the family eats dinner together, frequency of
discipline, and unsupervised time were -.09, .08, -.10, -.06, and .10, respectively.

The major difference between the findings regarding alcohol and marijuana
use was that family factors played a more important role in the latter (in fact, no
family factors showed any strong significance with regard to alcohol use). This
finding may be attributed to the fact that. in American society. alcohol is considered
to be a more acceptable drug than marijuana. and children will experiment with it
regardless of what the family does.

With regard to crack/cocaine use. 17% of the youths sampled said that they
had tried it at least once. Of that number, 88% were white and 12% were minority;
55% were male and 45% were female. Table 7 shows that religion did not have any
significant relationship to cocaine/crack use (.01). Instead, family factors played a
much more prominent role. Frequency of arguments about rules had a strong
negative relationship with crack/cocaine use (-.16), as did frequency of discipline
(-.09). Unsupervised time had a positive correlation (.07) with crack/cocaine use,
and number of times per week the family eats dinner together had a negative
correlation (-.06).

Table 7 (total sample) shows fairly high beta weights for some of the family
process variables. Frequency of arguments about rules had the highest beta weight
(-.15), and unsupervised time had a positive beta weight of .11. This suggests that

communication and lack of supervision contribute heavily to cocaine/crack use. The

75

Table 7: Association between family process variables and alcohol and drug
offenses (how often the juvenile used crack/cocaine).

 

 

 

 

 

 

 

 

TOTAL SAMPLE
Correlation N Corr.
Argue about rules 1,353 -.16"
Frequency of discipline 1,121 -.09**
Unsupervised time 1.422 -.O7**
# of times per week eat dinner together 1,389 -.06"

 

     

' = one-tailed ‘" = two-tailed

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Multiple Regression . Beta SE Beta 1 Sig. oft
Argue about rules -.15 .03 -4.3 .00
Frequency of discipline -.07 .04 -2.0 .05
Unsupervised time - .1 1 .03 3.3 .00 h
#.of times per week eat -.06 .03 -1.7 .09 1
dinner together
Sex dummy -.03 .04 -0.7 .47
Race dummy -.14 .04 -3.7 .00
Family dummy -.04 , .04 -1.1 .29

 

 

 

 

 

 

MultB B’ I AdiB2 I E(Eqn) Sig. ofE Bch ECH ISig.ofCH1I

I .26 I .07 .05 I 5.3 .00 .07 5.3 I .oo

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

76
Table 7: Continued.
WHITE MALES
Correlation N Corr. I
Argue about rules 500 .19“
Unsupervised time 526 -.16"
# of times per week eat dinner together 510 .10
Parents discuss decisions with child 511 -.09
Frequency of discipline 430 .07

 

 

' = one-tailed ” = two-tailed

 

 

 

 

 

 

 

 

 

Multiple Rressonegi 5
Argue about rules .14 .06 -2.2 .03 J
Unsupervised time -.24 .06 4.3 .00
# of times per week eat .12 .05 2.2 .03
dinner together
Parents discuss decisions -.00 .06 -0.0 .97
with child
Frequency of discipline .06 .06 -1.1 .29
Family dummy -.O3 .05 -O.6 .52

 

 

 

 

 

 

 

 

 

.32 .10 .07 3.2 .00 | .10 I 3.2 I .oo

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

77
Table 7: Continued.
WHITE FEMALES
Correlation N Corr.
Unsupervised time 509 .16"
Frequency of discipline 434 -.08 I
How much say in family rules 509 .06
Parents discuss decisions with child 506 -.06
" = one-tailed *‘ = two-tailed
Multiple Regression Beta 55 Beta 1 Sig. oft I
Unsupervised time .10 .05 1.8 .07
Frequency of discipline -.O6 .05 -1.2 .25
How much say in family .08 .06 1.4 .15
mles
Parents discuss decisions -.04 .05 ~0.8 .43
with child
Family dummy -.01 .05 -o.3 .73 I

 

 

 

 

 

 

Sig. ofE B’CH Sig. ofCH i
.25 .06 .03 2.1 .02 .06 2.1 .02

 

 

78

Table 7: Continued.

 

 

 

 

 

 

MINORITY MALES
m... II
Frequency of discipline 116 .36“
Unsupervised time 184 .21'
How much say in family rules 180 -.20"
Argue about rules 173 .18
How often attend religious services 178 -.16

 

 

 

 

 

‘ = one-tailed *‘ = two-tailed

  
    

 

 

 

 

 

 

 

 

 

 

 

Multiple Regression
Frequency of discipline -.30 .13 -2.3 .02
Unsupervised time -.27 .1 1 -2.3 .02 I
How much say in family -.11 .12 -1.0 .33 1
rules
Argue about rules .01 .13 0.1 .93
How often attend religious .32 .11 2.8 .01
servuces
Took family out to dine

 

MultB 3’ Adj. 32 E(Eqn) Sig. ofE B’CH I ECH Sig.ofCH

 

.56 .31 .18 2.4 .02 .31 2.4 .02

 

 

 

Table 7: Continued.

79

MINORITY FEMALES

 

Correlation

ﬂ

 

How often attend religious services

200

 

 

Unsupervised time

" = one-tailed ” = two-tailed

 

 

 

 

 

 

 

 

 

 

 

MultB 3’ Adj-82 EiEqn) SieofE 3ch ECH SigoO'CH I

 

Multiple Regression Beta SE Beta 1 Sig. oft
How often attend religious - .10 .11 0.9 .38 I
services
Unsupervised time ‘ -.11 .12 —0.9 .36 7
Family dummy

 

 

.35 .12 u -.00 .97

 

.48

.12

!
.48 i

,j

80

negative beta weights of frequency of discipline (-.07) and number of times per week
the family eats dinner together (-.06) seemed to coincide with this ﬁnding.

Of the three drug and alcohol offenses, crack/cocaine use was the only one
with which religion was not signiﬁcantly associated. This ﬁnding might be attributed
to the fact that cocaine is considered a harder drug than alcohol and marijuana, and
therefore it has different characteristics. Or it could be that most families who
frequently attend religious services are somewhat removed from the inﬂuences that
lead to crack/cocaine use. In any case, this ﬁnding, as well as other signiﬁcant

results from the study, is discussed in greater depth in the next chapter.

CHAPTER V

SUMMARY, CONCLUSIONS, LIMITATIONS, AND RECOMMENDATIONS

Summary

The researchers main objective in this study was to examine the inﬂuence
of family process variables (i.e., family communication, family conﬂict, family
discipline. and family activities) on yoUths’ level of delinquency. The researcher
conducted a secondary analysis of data from the National Survey of Children,
which was conducted in 1976: 1981, and 1987.

Factor analysis was used ﬁrst, with the intention of ﬁnding multiple
indicators of family process variables. None was found, so single measures were
used instead. A correlational analysis was then performed to show the level of
association between speciﬁc family process variables and the level of
delinquency. Race. gender, and family structure (single versus two-parent) were
used as control variables. A regression analysis was then used to determine the
inﬂuence of the family process variables on status, property, personal, and drug
offenses.

For status offenses (measured as the number of times the youth had run
away from home in the last year), the correlation analysis showed that, for white
males, the frequency of dining together, frequency of discipline, and arguing over

mles were all inversely correlated with twanting from home. In contrast, verbal

81

82

input was positively associated with running away from home. For white females,
however, only the frequency of dining together and arguing over rules were
signiﬁcantly inversely correlated with running away from home.

Interestingly, none of the family dynamics was signiﬁcantly correlated with
running away for minority males. For minority females. only the frequency of
dining together was inversely correlated with running away.

A multiple regression analysis indicated that the family process variables
were not strong predictors of running away. The only signiﬁcant ﬁnding from this
analysis was that family structure was a predictor of truancy. That is, children
from single-parent families were more aptto run away from home than were those
from homes with more than one parent.

Property offenses were divided into two types: damage to property and
theft. The correlation analysis for damage to property indicated that. for white
males. frequency of dining together. discipline, and arguing over rules all were
inversely related to property damage. The analysis also showed that white males'
involvement in decision making and unsupervised time were both positively
related to property damage. However, for these youths. only the frequency of
dining out had a signiﬁcant inverse association with property damage.

With regard to theft. the correlation analysis indicated positive relationships
with frequency of discipline and arguing over mles, and a negative relationship
with unsupervised time, for white males. The correlation analysis revealed that,

for both male and female minorities, none of the family process variables was

83
signiﬁcantly associated with theft. A regression analysis had the same results as
the correlation analysis.

Personal offenses were deﬁned as the number of times the child had
physically hurt someone in the last year. Only the frequency of discipline was
found to have a signiﬁcant negative effect among white males. and parents
discussing decisions with the child had a signiﬁcant positive effect.

For minority males. family communication skills were found to have a
strong negative inﬂuence on delinquent behavior. Taking the family out to dine
and frequency of discipline were the two strongest variables that had a negative
effect. I

The same scenario existed for minority females; the same two variables
had strong negative inﬂuences on delinquency. Unlike the case with minority
males. however. unsupervised time had a solid positive association with personal
offenses for minority females.

The regression analysis afﬁrrned what was found in the correlation
analysis. That is. frequency of discipline and discussing decision making had a
signiﬁcant negative and positive effect. respectively, on personal offenses.

For alcohol offenses, the only signiﬁcant correlation was a positive one
with frequency of attendance at religious services. When race was included,
however, this association was not signiﬁcant for minority females. Arguing over
rules and having a say in decision making had an inverse relationship with

drinking alcohol for white females. and frequency of dining at home together was

 

84
positively associated with drinking, whereas dining out with the family had an
inverse association with the consumption of alcohol. A regression analysis
showed that the frequency of attending religious services was a strong predictor
of drinking alcohol and that while males consumed alcohol more than did the
other groups.

Drug offenses were divided into two categories: marijuana and crack!
cocaine use. Whereas attending religious services followed the same trend as
it did in relation to alcohol use. other variables such as arguing about rules, eating
dinner together. discipline, and supervision all had signiﬁcant inverse
relationships with smoking marijuana. The regression analysis afﬁrmed what was
found in the correlation analysis: the same variables were found to be strong
predictors of marijuana use.

Attending religious services, however. did not have a signiﬁcant
relationship with the use of crack/cocaine. A correlation analysis revealed that
arguing about mles. discipline. and dining together all had inverse associations
with crack/cocaine use. whereas unsupervised time had a positive association
with crack/cocaine use. Again, the regression analysis conﬁrmed what was found

in the correlation analysis.

Conclusions
When examining the ﬁndings of this study, it became clear that
communication (even if it was in the form of arguing) was the factor most related

to predicting delinquency. Arguing can be considered a helpful way of relieving

85
tension between parent and child, which may eliminate the youth’s need to act out
in delinquent behavior. Having an open relationship with the parent may 'open
the door' to effective communication because both the child and the adult would
feel comfortable discussing serious topics such as drug use and sex with each
other.

Another family process variable that was effective in predicting delinquent
behavior was dining together (both at home and going out to eat). Although it is
doubtful that the act of eating together is in itself a deterrent to delinquency, it is
possible that other dynamics are involved in dining together that may help curb
delinquent behavior. Families that dine together may feel more comfortable and
relaxed with each other, which leads to stronger relationships within the family.
As a result, they communicate more effectively and have stronger bonds than
families who do not dine together.

As expected, frequency of discipline and supervision predicted
delinquency, for all the reasons one would expect. If children are disciplined fairly
and consistently, they will realize the consequences of their actions and stop
acting out in order to avoid punishment. Similarly, ifthey are supervised well,
they may not have the chance to engage in delinquent behavior. It should be
noted that discipline was almost always associated with communication of some
sort, which would lead one to conclude that discipline alone is not enough to act
as a deterrent. In fact, too much discipline or disproportionate discipline may

cause the youth to rebel and act out in even more delinquency.

 

86

The same can be said for supervision. Too much supervision may stiﬂe
the child and cause him or her to rebel. Instead, parents should use supervision
along with communication to instill the norms and values that they want the child
to abide by. Ideally, the child will have a sense of the parents being
“psychologically present“ even when they are not there, and thus the child will
think twice before engaging in delinquent actions and eventually monitor his or
her own behavior. I

In terms of alcohol and drug use, it seems that the child’s views tend to
reﬂect those of the parent or guardian. All other factors being equal, ifthe parents
are more tolerant of alcohol (or use it themselves) than other drugs, their children
will be more apt to try drinking. With the exception of alcohol use (in which
frequency of attending religious services was the only signiﬁcant factor), other
family process variables such as communication and discipline had the same
relationship with marijuana and crack/cocaine use as they did with the other
offenses. The fact that American society is more accepting of alcohol than other
types of drugs may offer some explanation as to why the other family process
variables did not have any signiﬁcant relationship with alcohol consumption.

l'ill' IB II' I
EunbeLBesearob

Although this study provided useful insights into the relationship between
family process variables and delinquency, it did have some limitations. To begin

with, the National Survey of Children (from which the data for this study were

87

taken) did not focus entirely on family process variables or delinquency. The
primary focus of that survey was to examine the overall well-being of American
children—to provide a proﬁle of the children’s lives and to give an indication of the
type of care they received.

Ideally, a longitudinal study should be conducted that would focus entirely
on family process variables and their relationship to delinquency. in this way the
family process variables can be examined in greater depth. Communication
could be broken down and examined more closely. For example, different types
of communication could be examined, as could frequency of communication and,
most important, the quality of the communication. The same can be said for all
of the types of family dynamics that were discussed in this study. More qualitative
measures could be used with each one, which would give a better indication of
what is really occurring in the family unit.

The subjects in this survey were predominantly white. in future studies,
minorities should be better represented. Conducting research with a wider
minority representation would provide a better understanding of all American
children and the similarities and/or differences that exist among them.

Another aspect that could have been examined in greater depth was
offense types. Although a fairly good representation of offense types was
included in this study, other types of offenses could be looked at as well. Truancy

from school, shoplifting, and possession of weapons, all of which are common

 

88

offenses, could be examined to include more of the types of delinquency that
exist.

All of the above-mentioned issues are important and worth examining more
closely. Such future research may lead to other types of family process variables

that were not included in this study that could have a signiﬁcant effect on

delinquency.

 

' REFERENCES

REFERENCES

Arnold. w. R. (1983). .luyanilomisoondmmgolinguonoy Boston:

Houghton-Mifflin.

Bahr, S. (1979). Family determinants and effects of deviance. In W. Burr, R.

Hill, F. I. Nye, &l. Reiss (Eds) ContemooranLtboonesamnotamny

New York. Free Press.

Browning, C. (1960). Differential impact of family disorganization on male
adolescents. Socialﬁrobioms. 8. 37-64.

Canter, R. J. (1982). Family correlates of male and female delinquency.

Criminology. 29.149-167.

Cemkovich, S. A., 8 Giordano, P. C. (1987). Family relationships and
delinquency. Criminology. 25. 295-321.

Chilton, R. J, &Markle, G. E. (1972). Family disruption, delinquent conduct

and the effect of subclassiﬁcation. Wm, 31,

93-99.

Cloward R. A., & Ohlin, L. E. (1960). Doﬁnoueimnooomnunny. Glencoe.

IL. Free Press.

Glueck, S., &Glueck, E. (1950). Wm New York:

Commonwealth Fund.

Gold, M. (1963) WWW. Ann Arbor: University of

Michigan Press.

Haskell, M. R, 8. Yablonsky, L. (1982). Jumnﬂedolinouonoy Boston:

Houghton-Mifﬂin.
Hirschi, T. 1969). W Berkeley: University of California

Press.

89

 

90

Jackson, T. (1957). The differential impact of family disorganization.
AmonoanﬁooiologioaLBowou. 22(5). 505 512.

Johnson, R. E. (1986). Family structure and delinquency: General patterns
and gender differences. Criminology, 25, 65-84.

Laub, J. H., 8 Sampson, R. J. (1988). Unraveling families and delinquency:
A reanalysis of the Gluecks' data. Criminology, 25, 355-379.

Mastueda, R. L, 8Heimer, K. (1987). Race, family structure, and
delinquency. A test of differential association and social control

theories. AmericanﬁooologioaLBoyioiy. 52. 826-840

McCord, J. (1991). Family relationships, juvenile delinquency, and adult
criminality. Criminology. 29. 397-417.

Monahan, T. (1957, March). Family status and the delinquent child: A
reappraisal and some new ﬁndings. W, 95, 251-258.

Morash, M., 8 Chesney-Lind, M. (1991). Areformulation and partial test of
the power control theory of delinquency. Justigeﬂuanedy, 9.

Norland, s., Shover N., Thornton, w. E., 8James, J. (1979). Intra-family '
conﬂict and delinquency. W, 22, 223-240.

Nye. I. F. (1958). Eamily_relationsnios_and_dolinouont_bonayior. New York:
lMIey.

Parsons, T., 8Bales, R. F. (1955). Eamily._socialization._and_imeraction
oroooss. Glencoe, IL: Free Press.

Patterson, G. R., 8 Dishion, T. J. (1985). Contributions of families and peers
to delinquency. Criminology, 29, 63-79.

Rankin, J. 8 Wells, E. (1990). The effect of parental attachments and direct
controls on delinquency .LoumaLoLBesoarcLinCrimoand
Dolinouonoy. 21.140-165

Rosen, L. (1985. Family and delinquency: Structure or function?
Criminology. 29. 553-573.

Shoemaker, D. J. (1984). IbooriosoLdolinouonoy. New York: Oxford

University Press.

91

Slocum, W., 8 Stone, C. L. (1963). Family cultural patterns and delinquency
type behavior. Marriagoanooenayior 25. 202-208.

Sterne. R. (1964). Qalinouontoonduotonobrokoniiomo. New Haven. CT:

University Press.

Trojanowicz, R. C., 8 Morash, M. (1987). WWW
control. Englewood Cliffs, NJ: Prentice-Hall.

Van Voorhis, P., Cullen, F. T., Mathers, R. A., 8 Garner, C. C. The impact of
family structure and quality on delinquency A comparative assessment
of structural and functional factors. Criminology, 26, 235-261.

Wells, E. L., 8 Rankin, J. (1986). The broken home model of delinquency.

Analytic issues. JoumaLoLBosoaroanCnmoaimoelinouonoy. 29. 68-
93.

Wilkenson, K. (1974). The broken family and juvenile delinquency: Scientiﬁc
explanation or ideology? Sodalﬁmoloms. 21. 726-739.

Zill, N., Furstenburg, F. Jr. Peterson, J., 8 Moore, K. (1990). NationaLSonLoy
MW. Ann Arbor,

MI: Inter-University Consortium for Political and Social Research.

nICHIGeN sran UNIV. LIBRARIES
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIHI
31293015654373