”mass J: ....‘ .3 r $5. a.

s ._ $9.574“. a? 3an

.. (44er ..*$” ﬁt...» .
$3 a. 3 .
. A

i. ..
N: “G.“ «Mm. .

3.”.

1w.

.. 3.. .

.wmum. W3”...

;
e I}.

.mﬂwﬁ. ..

w , , ‘ . 3%,}
a a . , 1.2.2.31
, . ﬁi

: . s
.. .an

I

I w
in.

l’

31‘ 1‘
L Itixx.

. .4 3
CL."
- K

\33;§1

 

$3,)...
.1:

.:;ry..

 

 

 

5a....
2!.
If, i (7

t
I:

1‘»

 

. .
: , .. :5 tr.
. January}.

:‘
v

 

 

. 7 n
3%,“
3%. «my ‘

a. J. a. .
.Ha...mh.

£59.:

,2...

15+. ,1
.x .32... .u. 3.131
A; . at .\.
h. J3, ~

L ,
am,
3..
«(143%

was.

3

.
t

Y.

«.3,

 

This is to certify that the
dissertation entitled
Protective Familial Factors Leading to Low Levels

of Substance Use Among Children of Alcoholics

presented by

John Paul Walker

has been accepted towards fulﬁllment
of the requirements for

Ph.D. degreein Marriage & Family Therapy

Major professor

Date/3 (gm/Z 2m /

MS U i: an Affirmative Action/Equal Opportunity Institution 0-12771

 

 

 

LIBRARY

Michigan State

University

 

 

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

 

DATE DUE

DATE DUE

 

DATé lD-UE 3

APR 0 7 2002

Z

 

MA

\-<

f.
A

l. 3,2002

 

a».

a.

 

g’ iii 0531 jug

 

 

 

 

 

 

 

 

 

 

6/01 C'JClFlC/DateDuep65—p. 1 5

 

 

PROTECTIVE FAMILIAL FACTORS LEADING TO LOW LEVELS OF
SUBSTANCE USE AMONG CHILDREN OF ALCOHOLICS

By

John Paul Walker

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Family and Child Ecology

200 l

ABSTRACT

PROTECTIVE FAMILIAL FACTORS LEADING TO LOW LEVELS OF
SUBSTANCE USE AMONG CHILDREN OF ALCOHOLICS

By

John Paul Walker

This study examined if conventionality, emotional autonomy, and parental support
would mediate the relationship between parentiﬁcation and low levels of substance use
among children of alcoholics (COAs). It was hypothesized that the relationship between
high levels of parentiﬁcation and low levels of COA substance use would be mediated by
high levels of conventionality, high levels of emotional autonomy with the alcoholic
parent, low levels of support from the alcoholic parent, low levels of emotional autonomy
with the non-alcoholic parent, and high levels of support hour the non-alcoholic parent.
Subjects were 151 COAs and 151 non-COAs who were primarily female and recruited
ﬁ'orn undergraduate psychology students at Michigan State University. The results of
multiple regressions indicated that none of the potential protective factors mediated the
relationship between parentiﬁcation and low levels of substance use. However,
deidealization of an alcoholic father and high levels of conventionality had a moderating
effect on substance use and this only existed within the collapsed COA group (i.e. COAs
with one and two alcoholic parents) that were not at any more risk for using drugs or
alcohol than the non-COA group. Contrary to predictions, parentiﬁcation alone actually

ﬁmctioned as a risk factor that was related to higher levels of drug use within the collapsed

COA group. Conventionality, the most signiﬁcant moderating variable, was strongly
correlated with low levels of drug and alcohol use across all subgroups within the sample
and appeared to reverse or neutralize the relationship between parentiﬁcation and
heightened drug use within the collapsed COA group.

There was support for a probabilistic view of development (Fitzgerald, et al., 1994)
that has multiple developmental trajectories with multiple outcomes because a family
history of alcoholism did not adequately predict that offspring would perpetuate high levels
of alcohol consumption. Female COAs with one alcoholic parent and non-COAs have
similar substance use levels, but differ in the levels of parentiﬁcation, conventionality,
parental support, and emotional autonomy. COAs with two alcoholic parents were at
signiﬁcantly higher risk for drug use and received the least amount of support ﬁom their
mothers. Hence, parentiﬁed females between the ages of 18 and 24 from families with one
alcoholic who have a seemingly distant relationship with an alcoholic father (i.e. highly
deidealized with low levels of support) appear to be on a protective developmental
trajectory if high levels of conventionality exist and the outcome leads to low levels of drug

and alcohol use.

THIS WORK IS DEDICATED TO MY “PROTECTIVE FACTORS”: WILLIAM J.
SMITH, LAWRENCE POLLMAN, JOSEPHINE KOZLOWSKI, DELLA AND ART
CARIAN, PHYLLIS SHOENHERR, MARY LEA AND MICHAEL HERMANN,
THOMAS GRINDEM, LEE PATRICK SULLIVAN, AND SUSAN JACKSON-

WALKER.

EACH OF THESE GAURDIAN ANGELS HAVE GRACIOUSLY GIVEN ME THE GIFT

OF UNCONDITIONAL LOVE TO ENDURE THIS CHALLENGE

iv

ACKNOWLEDGEMENTS

I am greatly indebted to my wife, Susan J ackson-Walker, for the gentle nudging and
support she has provided me throughout this emotional roller coaster called “graduate
training”. She always believed that I possessed the personal resources to obtain a Ph.D.
even when I did not. Thank you, Susan, for never giving up on me. I want to thank my
good ﬁiends, Gloria Gonzales-Kruger and Ruben Viramontes Anguiano, for their informal
and powerful education about familialism and resilience. They have led me on this journey
through example. I also want to thank Michael Henmann, Richard Jackson, Judith
Jackson, and my mother, Alice Walker, for their continual encouragement to ﬁnish during
times when it would have been so easy for me to quit. I will always be indebted to Knute
Rotto for making it possible to close a chapter in life without sacriﬁcing my job or the
quality of this research. I am extremely fortunate to have had the caliber of mentorship and
support that he has given me. I also need to thank the Rough River Gang for their
“intervention” during our ﬁrst annual retreat. Their masterful use of the “stroke and kick”
technique helped me identify my priorities and see that the stressors of the family life cycle
could no longer be used as excuses to not work on my dissertation.

Finally, I wish to express my deepest appreciation for my major professor, Dr.
Robert E. Lee and my research director, Dr. Hiram Fitzgerald. Their involvement, support,
and interest never wavered. Dr. Lee’s brilliant insights made doing therapy an exhilarating,
powerful growth experience when he was my clinical supervisor. His clinical supervision
through the years has had a profound inﬂuence on my career. His contributions as a co-

chair have deepened my understanding of how resilience and family systems theory

intersect and has made this dissertation a rich journey. I cannot thank him enough for the
gentle, supportive nature of his delivery whenever he provided me with feedback. Dr.
Fitzgerald’s expertise in the alcoholism literature continually challenged my naivety. His
extraordinary commitment to methodologically sound research has been contagious and
taught me that research can be exciting. His enthusiasm for my hypotheses and belief that I
was “on to something” was incredibly validating and gave me the extra fuel that was
desperately needed to ﬁnish. I cannot thank him enough for encouraging me to excel. I
have been a direct beneﬁciary of his expertise. Together, their help has made me be proud

of this work.

vi

TABLE OF CONTENTS

LIST OF TABLES ................................................................................................................. ix
LIST OF FIGURES ................................................................................................................. x
INTRODUCTION ................................................................................................................... 1
Statement of the Problem ........................................................................................... 1
CHAPTER ONE
LITERATURE REVIEW ....................................................................................................... 6
The Risk Status of COAs ........................................................................................... 9
Intemalizing & Extemalizing Disorders ...................................................... 10
Family Dysﬁmction ...................................................................................... 11
Substance Abuse .......................................................................................... 11
Models of Resilience ................................................................................................ 14
Individual Models of Resilience .................................................................. l4
Ecosystemic Models of Resilience .............................................................. 16
Family Systems Models of Resilience ......................................................... 18
Protective Factors ..................................................................................................... 22
Parentiﬁcation .............................................................................................. 22
Conventionality ............................................................................................ 27
Emotional autonomy .................................................................................... 31
Family and Parental Support ........................................................................ 37
Summary ................................................................................................................... 47
Hypotheses ................................................................................................................ 49
Hypotheses for Risk Factors ........................................................................ 49
Hypotheses for Protective Factors ............................................................... 49
CHAPTER TWO
METHOD .............................................................................................................................. 51
Subjects ..................................................................................................................... 52
Procedures ................................................................................................................. 55
Measures ........................................................................................................... 56
COA status ............................................................................................ 56
Parentiﬁcation ....................................................................................... 58
Conventionality ..................................................................................... 60
Parental Support .................................................................................... 62
Emotional Autonomy ............................................................................ 63
Alcohol and Drug Use ........................................................................... 64
Plan of Analysis ................................................................................................ 67

vii

Table of contents (cont’d)

CHAPTER THREE
RESULTS .............................................................................................................................. 70
Preliminary Analyses ................................................................................................ 70
Test of the Hypotheses ............................................................................................. 76
Hypotheses for Risk Factors ..................................................................................... 76
Hypotheses for Protective Factors ............................................................................ 77
Parentiﬁcation and Substance Use ............................................................... 78
Conventionality as a Mediator Variable ................................................ 78
Idealization of a Non-alcoholic Parent as a Mediator ........................... 78
Deidealization of an Alcoholic Parent as a Mediator ............................ 79
High Support From a Non-Alcoholic Parent as a Mediator .................. 79
Low Support From an Alcoholic Parent as a Mediator Variable .......... 79
Deidealization of Father and Conventionality as
Moderating Variables ............................................................................ 80
CHAPTER FOUR
DISCUSSION ................................................................................................................ 83
Limitations ......................................................................................................... 88
Directions for Future Studies. ............................................................................ 91
LIST OF REFERENCES ............................................................................................... 96

viii

Table 1
Table 2

Table 3

Table 4

Table 5

Table 6

Table 7

Table 8

Table 9

LIST OF TABLES

Ethnicity, College Class, & Gender From the Total Sample (N=978) ...... 53
Ethnicity, College Class, & Gender Within Three Groups From the

Final Sample (n=302) ........................................................................... 54
Distribution of Religious Faiths Within Three Groups

From a Sample of College Students ..................................................... 70
The Number of Males and Females Within Three Groups

From a Sample of College Students ..................................................... 71
Father’s Education Level Within Three Groups From a

Sample of College Students ................................................................. 71
Means, Standard Deviations, and F scores for all Regression

Variables in a Sample of College Students .......................................... 72
Pearson Correlations Among All Continuous Regression

Variables in a Sample of College Students .......................................... 74

Regressions Predicting Drug use From Parentiﬁcation,
Conventionality, Among COAs With One Alcoholic Parent and Two
Alcoholic Parents ................................................................................. 80
Regressions Predicting Drug use From Parentifrcation, Father
Deidealization, Among COAs With One Alcoholic Parent and Two
Alcoholic Parents ................................................................................. 81

ix

Figure 1

Figure 2

LIST OF FIGURES

Mediational Model: Inﬂuence of Mediating Variables on the

Relationship Between Parentiﬁcation and COA Substance Use ................. 4
Signiﬁcant Relations Between Parentiﬁcation and Drug Use

as Moderated by Conventionality and Deidealization of Alcoholic

Father Among COAs With One Alcoholic Parent and Two Alcoholic
Parents ............................................................................. 82

Introduction

 

Approximately 15.3 million adults in the United States struggle with alcohol
abuse or dependence (US. Department of Health and Human Services, 1993). An
estimated 28 million people in the United States grow up with at least one parent who
struggles with alcohol addiction (Russel, Henderson, & Blume, 1985). Collective
ﬁndings from the clinical and empirical literatures provide clear evidence that the
offspring of alcoholic parents are at risk for developing a myriad of psychosocial
problems. Although one of the most empirically substantiated risks for children of
alcoholic parents (COAs) is a drug and/or alcohol problem, a signiﬁcant number of
COAs do not develop problems with substances. Several researchers (e. g. Havey &
Dodd, 1993; Burk & Sher, 1988; Werner, 1986) have pointed out the need to focus
more empirical and clinical attention on the substantial sub-population of offspring who
are functioning well to identify protective factors that lead to low levels of substance

use among COAs.

Statement of the Problem

 

This study is the ﬁrst to identify COA status, and measure offspring and parent
substance use while attempting to identify protective familial processes leading to low
levels of substance use during late adolescence. More speciﬁcally, this study explored
if a developmental pathway characterized by parentiﬁcation (independent variable),
conventionality, emotional autonomy, and parental support (potential mediating

variables) led to low levels of substance use (dependent variable) among late adolescent

COAs. A developmental systems perspective that examines variations of functioning in
substance abusing families (Fitzgerald, Davies, Zucker, & Klinger, 1994) guided the
conceptual and methodological framework. The investigator’s clinical experience of
more than 1000 direct client contact hours with substance abusing adolescents and their
families and a subsequent literature review of COA protective factors (Walker & Lee,
1998) provided the foundation for theoretically based hypotheses. This investigator was
struck by how inconsistent his clinical experience with alcoholic families was with the
clinical COA literature. The literature tends to describe a psychological death sentence
for COAs, while the predominate theme in this investigator’s clinical work suggested
that an adaptive form of parentiﬁcation clearly helped a sub-population of COAs thrive.
A review of the clinical and empirical literature suggested that conventionality, parental
support, and emotional autonomy might shape or mediate the impact that parentiﬁcation
would have on family relationships and substance use.

A mediational model was used to explore if conventionality, parental support,
and emotional autonomy were potential protective factors that accounted for a
signiﬁcant relationship between parentiﬁcation and low levels of COA drug and alcohol
consumption. Parentiﬁcation is a concept that has been used in the family therapy
literature to describe family interactions that place children and adolescents in a role
that is typically performed by parents in a given culture (Mika, Bergner, & Baum, 1987;
Boszormenyi-Nagy, Grunebaurn, & Ulrich, 1991). The deﬁnition of parentiﬁcation
used here has a positive connotation. It can be manifested in an adaptive role within
families that have few resources (Minuchin & F ishman, 1981) and lead to greater levels

of self esteem (Goldenthal, 1993) and self-efﬁcacy (Boyd-Franklin, 1989). Adaptive

parentiﬁcation is hypothesized to operate when it is mediated by the potential protective
factors.

Adolescents who embrace a conventional belief system are more likely to have
internalized traditional societal values, develop an internal locus of control, and not
value rebellious behavior (Brook, Brook, Gordon, Whiteman, & Cohen, 1990).
Emotional autonomy (Steinberg & Silverberg, 1986) refers to the process where
children gradually relinquish childhood dependency on parents, perceptions of parental
omnipotence, and begin viewing their parents more and more as fallible people (i.e.
deidealization). Although emotional autonomy has a cognitive and emotional
dimension, deidealization (i.e. cognitive dimension) was the only component of the
concept used in this study. An emotionally supportive parent-adolescent relationship
was deﬁned as one that the adolescent perceives to be an accessible, satisfying source of
support (Sarason, Levine, Bashon, & Sarason, 1983).

The mediating model in Figure 1 outlines a developmental trajectory that was
hypothesized to protect against the development of problems with drugs and alcohol. It
illustrates how parentiﬁcation, conventionality, parental support, and emotional autonomy
may organize in an interconnected causal chain with an outcome that transcends any one
protective factor. The arrowheads indicate the hypothesized direction of effects. Several
potential protective factors were hypothesized to mediate the relationship between high
levels of parentiﬁcation and low levels of substance use. Parentiﬁed COAs were expected
to deidealize their alcoholic parent and score high on measures of conventionality. 'lheir
problematic substance using counterparts with low levels of parentiﬁcation were expected

to have low scores on conventionality and idealize their alcoholic parent. Conventionality

was hypothesized to mediate parentiﬁcation and substance use because greater care-taking
responsibility. towards others (i.e. parentiﬁcation) is likely to encourage low levels of
substance use and other pro-social behaviors that conform to traditional societal norms. In
other words, COAs who deidealize their alcoholic parent and idealize their more
conventional, non-alcoholic parent were expected to emulate the more conventional
parent's value system about drug and alcohol consumption.

Figure l

Mediational Model: Inﬂuence of Mediating Variables on the Relationship Between
Parentiﬁcation and COA Substance Use

 

Mediating Var.
High Levels of
Conventionality

 

 

 

 

Mediating Var.

High’Levels of
Deidealization with
Alcoholic Parent

 

 

 

 

 

Independent Low Levels of

L332 210g: Mediating Var. C2111 grills) and
\> M co 0 se
Parentiﬁcation LOW Levels Of

 

 

 

' Support From

 

 

 

Alcoholic Parent

 

 

 

 

Mediating Var.
High Levels of Support
From
Non-Alcoholic Parent

 

 

 

Mediating Var.
High Levels of
Idealization with
Non-Alcoholic
Parent

 

 

In addition to exploring the inﬂuence of parentiﬁcation, conventionality, and
emotional autonomy, the study also examined how different levels of parental support
impacted substance use. An emotionally disengaged relationship with an alcoholic
parent and an emotionally supportive relationship with a non-alcoholic parent were
hypothesized to be familial dynamics that strengthen the relationship between
parentiﬁcation and low levels of substance use. Prior research has found that the
context of emotional support (Andrews, Hops, & Duncan, 1997) and emotional
autonomy (Fuhrman and Holmbeck, 1995) can inﬂuence whether or not close parent-
adolescent relationships serve detrimental or protective functions. A parental
relationship that is perceived to be a credible source of support was conceptualized as a
protective factor for the parentiﬁed COA who has high levels of emotional autonomy

with their alcoholic parent.

CHART ER ONE

Literature Review

The primary purpose of this study was to utilize a developmental systems
perspective (e.g. Fitzgerald et al., 1994; Gottlieb, 1991; Ford & Lerner, 1992) to
identify mediators that account for the relationship between COA parentiﬁcation and
low levels of substance use during a period in life when the risk for abusing substances
is particularly high. Parentiﬁed COAs with low levels of substance use were expected
to have adaptive relationships with both alcoholic and non-alcoholic parents. COA
conventionality, high levels of idealization, and highly supportive relationships were
hypothesized to characterize protective parent-child relationships with non-alcoholic
parents. On the other hand, high levels of conventionality, high levels of deidealization,
and low levels of support were expected to characterize adaptive relationships with
problematic substance-using parents.

Jurkovic (1997) points out that a comprehensive theoretical approach is needed
to capture the individual, familial, and sociocultural implications of adaptive
parentiﬁcation. Developmental systems theory provides a meta-perspective to depict
“the complexity of the intra-and extra-organism relations that, together, assure the
human diversity” (Lerner, 1991, p. 29) found in COA functioning. This comprehensive
framework is useful to begin unraveling the complexities of transactions between
macrolevel cultural inﬂuences to micro level family and individual cognitive processes
that lead to adaptive outcomes among COAs. The relative plasticity of human

development plays an integral role in producing several developmental trajectories that

is captured in the varying degrees of adaptive behaviors and psychopathology among
COAs

Fitzgerald’s et al. (1994) approach to developmental systems theory speciﬁcally
guided the conceptual and methodological ﬁamework of the mediational model used in
this study. From their perspective, there are many pathways leading to different forms
of substance abuse that extend beyond alcoholism and alcohol abuse. Even siblings
living with the same alcoholic, biological father are not destined to be on a
developmental trajectory with an outcome of alcoholism, because their developmental
and experiential histories are unique on several different levels (e. g. sex, temperament,
perception, etc.). Hence, research models that investigate the causal determinants of
substance abuse must consider intraindividual (e. g. conventionality), interindividual
(e. g. parentiﬁcation, support), contextual (e. g. late adolescence), and organism-
environmental transaction (e. g. bi-directional effects of parentiﬁcation on the adolescent
and parent) inﬂuences leading to low levels of substance use among COAs.

Contextual family therapy theory (Boszormenyi-Nagy, et al., 1991) and the
concept of relational resilience (Walsh, 1996; 1998) were used to generate theoretically
driven hypotheses about protective adolescent cognitions and parent-adolescent
relationships on a micro level. Each of these perspectives was used to quantify how
aspects of family relationships (i.e. parentiﬁcation, emotional autonomy, parent-
adolescent support) and COA cognitions (e. g. conventionality) are expected to transact
and result in low levels of substance use. They provide a detailed lens from which to
conceptualize theoretically driven hypotheses that examine the potentially protective

functions of parentiﬁcation, conventionality, emotional autonomy, and emotional

support. Contextual family therapy theory (Boszormenyi-Nagy et al., 1991) offers a
competency-based lens through which to view the ethical-relational dimension of
parentiﬁcation within alcoholic families, while relational resilience (Walsh, 1996; 1998)
assists in the identiﬁcation of microlevel parent-adolescent interactions that characterize
adaptive parentiﬁcation.

The following literature review summarizes risks that COAs face, models of
resilience, and protective factors (i.e. parentiﬁcation, conventionality, emotional
autonomy, and parental emotional support) that may lead to low levels of substance use.

Cicchetti and Garmezy (1993) point out that an understanding of protective inﬂuences
is intimately tied to an understanding of risk, maladaptation, and psychopathology in
general (and vice versa). Protective mechanisms are tied to risk variables because an
element of risk (i.e. parental alcoholism) is needed to assess the protective utility of
parentiﬁcation, conventionality, emotional autonomy, and parental support (Rutter,
1990). Therefore, a brief review of the literature on the maladaptive psychosocial
outcomes of COAs will establish the parameters of risk (e. g. incidence of substance
abuse, psychosocial maladjustment) that confront COAs and help identify which
variables are most likely to serve a protective function. A review of the empirical
literature about risks across several levels of analysis also provides justiﬁcation to

utilize developmental systems theory to explore protective factors.

The Risk Status of COAs

 

The impact of alcoholism on every level of analysis is so pervasive that it has
been identiﬁed as causing more problems than the combined effects of all other
substances (Royce & Scratchley, 1997). A wealth of research indicates that children
who grow up in homes with alcohol problems are more likely to experience a wide
range of psychological, emotional, and social problems (West & Prinz, 1987), which
may be exacerbated by other clinical diagnoses co-existing with parents' alcoholism
(Sher, 1991). Contextual factors at a macro-level of analysis, such as cultural and ethnic
variation in norms towards alcohol, accessibility, economic deprivation, gender, and
overcrowding offer partial explanations for the prevalence of alcohol and drug use
within families and mediate the development of alcoholism subtypes within individuals
(Zucker, Fitzgerald, & Moses, 1995).

Theoretical and clinical writings have played a large role in shaping the direction
of empirical research on COAs. The clinical literature generally paints a bleak picture
for COAs characterized by general maladaptation and psychosocial impairment. During
the 1970’s and 805 clinicians and the media began popularizing the notion that the
offspring of alcoholics exhibit problems in multiple domains of psychosocial
ﬁmctioning throughout the life span (Sher, 1991). Several authors have written
extensively on how the detrimental impact of alcoholism affects the global functioning
of COAs with one of the most commonly cited problems being an inability to form
intimate, trustworthy relationships (Black, 1982; Brown, 1988; Cermak, 1986). Brown
(198 8) points out that most offspring adopt maladaptive defense mechanisms including

denial, self-blame, the inability to express individual needs and emotions, and the

tendency to be over-controlling in relationships with others.

Irrternalizing and Extemalizing Disorders

 

Early trends to pathologize COAs in the clinical literature and popular media set
a precedent for empirical research to identify dysfunction in their families. For
example, Herjanic, Herjanic, Penick, and Armbruster (1977) found that 54% of children
in a sample with drug and/or alcohol abusing parents could be diagnosed with
adjustment and internalizing disorders. Internalizing symptoms, such as low self-esteem
(Hughes, 1977; Potter & Williams, 1991), anxiety, and depression have been found with
some consistency among children of alcoholics when they are compared to controls
(Anderson & Quast, 1983; Tarter, Hegedus, Goldstein, Shelly, & Alterman, 1984).
Moos and Billings (1982) found that children are more than twice as likely to have
emotional problems when their parent had a period of sobriety that ended in relapse.

Cantwell's (1975) early review of hyperactive children and a more recent review
by Hawkins, Catalano, and Miller (1992) provides evidence that fathers of these
children are more likely to be alcoholic and exhibit antisocial personality traits.
Extemalizing behaviors such as hyperactivity, conduct disorder, oppositional behavior,
and delinquency appear to be some of the most common problems prior to early
adulthood (Sher, Walitzer, Wood, & Brent, 1991; Chassin, Rogosch, & Barrera, 1991)
and have been linked to the development of alcohol problems and antisocial behavior
later in life (e. g. Zucker, Fitzgerald, & Moses, 1995). Although poor academic
performance could be attributed to multiple risk factors, offspring from alcoholic
families in a longitudinal study were less likely to graduate ﬁ'om high school. They

were also more likely to receive school counseling services for discipline problems, and

10

three times more likely to be expelled from school (Miller & J ang, 1977; Deutsch,

Dicicco, & Mills, 1982).

F arnily Dysftmction

 

Addiction has a profound effect on family relationships and overall ﬁrnctioning.
One of the most consistent ﬁndings is that many families with substance abusing
parents are characterized by low cohesion, emotional constriction, and high levels of
conﬂict (Moos & Billings, 1982; F ilstead, McElﬁesh, & Anderson, 1981; Clair &
Genest, 1986; Moos & Moos, 1984). These studies also indicate that family members
are less likely to engage in recreational activities together and hinder the individuation
process during adolescence. Research using small, clinical samples (e. g. Jacob &
Leonard, 1988) has found that alcoholic families are less likely to adapt to stresses due
to impaired problem solving skills and interactions characterized by hostility and

conﬂict.

Substance Abuse

 

Perhaps one of the most commonly cited risks for children of alcoholics is the
development of problems with alcohol and drugs (for a review see Sher, 1991). Results
ﬁ'om twin, adoption and animal studies underscore how the complex relationship
between environment, development, and biology transact to place adolescents and
young adults with alcoholic parents among the highest risk groups for developing
substance abuse problems (Zucker, et al., 1994). One review (Cotton, 1979) found that

children who grow up with alcoholic parents have a six to ten times greater risk to

11

develop substance abuse problems by adulthood. Many studies have found that
offspring risk for addiction goes beyond alcohol to include drugs, such as marijuana
(Wright & Heppner, 1993; Chassin, Rogosch, & Barrera, 1991; Sher et al., 1991),
amphetamines, and cocaine (Johnson, Leonard & Jacob, 1989). COAs are more likely
than non-COAs to begin drinking at earlier ages and escalate in their alcohol use during
adolescence (Colder, Chassin, Stice, & Curran, 1997).

Although the literature on factors leading to alcoholism among females appears
inconclusive, evidence for familial transmission appears to be much stronger for males
than females. Research on female offspring had been relatively sparse until the 1980's,
and most studies have been conducted with predominantly clinical samples ignoring
important within group differences (Mitz, Kashubeck, & Tracy, 1995; Sher, 1991) that
may be indicative of protective familial processes. Several studies have found that
males are proportionately more at risk for substance abuse problems than females
(Schissel, 1993; Berkowitz & Perkins, 1987). Cotton's (197 9) highly cited review
supports a genetic transmission hypothesis and indicates that only 5% of females and
25% of males develop substance abuse problems when paternal alcoholism exists. In
fact, substance abuse among adolescents with alcoholic fathers escalates over a shorter
period of time for boys (Chassin, Curran, Hussong, & Colder, 1996). Adoption studies
that broaden their scope beyond paternal alcoholism transmission rates have found that
maternal alcoholism places females at a much higher, but still lower risk than males
(Bohman, Sigvardsson, & Cloninger, 1981; Harburg, et al., 1990). One non-clinical,
longitudinal study (Webster, Harburg, Gleiberman, Schork, & DiFranceisco, 1989)

measured the volume and frequency of parental and offspring drinking and found that

12

mother's drinking patterns were a better predictor of daughter's alcohol use seventeen
years later. Adult daughters were two times more likely to be heavy drinkers when they
had heavy drinking fathers and three times more likely when their mothers were heavy
drinkers. Other studies have even found that familial transmission among daughters
growmg up with paternal drinking problems are not at any more risk for substance abuse
than daughters of normal fathers (Mitz, et al., 1995; Benson & Heller, 1987; Harburg,
Davis, Caplan, 1982). Male COA's substance use escalates far beyond female use
during adolescence (Chassin, et al., 1996). Those who appear to be most at risk for
substance abuse problems are males who exhibit conduct disordered behaviors and have
fathers with a diagnosis of anti-social personality disorder and some type of chemical

dependency (Zucker, Ellis, Bingham, & Fitzgerald, 1996).

13

Models of Resilience

 

Despite the risk status associated with being a child of an alcoholic (COAs), any
one variable (including a positive family history for alcoholism) appears to be an
imprecise marker of risk due to multiple risk and protective factors that lead to diverse
developmental trajectories on an individual level (Zucker et al., 1995). Studies have
found that despite long held notions of pathology in COAs, many are not seriously
impaired in psychosocial functioning (e. g. Wright & Heppner, 1993; Jacob & Leonard,
1986). The fact that psychological disorders are not pervasive among COAs (West &
Prinz, 1987) and many do not develop substance use problems as adults (Fingarette,
1988; Cotton, 1979) underscores the heterogeneity found within chemically dependent

families and the different developmental pathways that exist among their offspring.

Individual Models of Resilience

 

The concept of resilience grew out of the ﬁeld of developmental
psychopathology (Cicchetti & Garmezy, 1993) and has historically used the individual
as the principal unit of analysis. Developmental psychopathology is interested in how
individuals develop maladaptive and adaptive behaviors in response to stress throughout
the course of development. Although deﬁnitions have varied over time, there is a
general consensus that resilient individuals have the ability to successfully adapt or
maintain competent psychosocial ﬁmctioning despite trauma, risk status, or stress
(Egeland, Carlson, & Sroufe, 1993; Masten, Best, & Garmezy, 1990). Individuals who
develop psychopathology have vulnerability factors that increase susceptibility to

stressors, while resilient individuals have dispositional attributes and environmental

l4

factors that serve protective functions against stress and lead to adjustment (Rutter
1985). Unlike the concept of "invulnerability", resilience is relative and varies over
time because of changing developmental, biological, and environmental contexts. The
context of any given variable dictates its function as a risk or protective variable (Rutter,
1985 ). For example, a variable, such as social support, can potentially have a protective
inﬂuence if risk or stress is present.

Constitutional factors such as gender, temperament, and intelligence are often
conceptualized to moderate stress and are given a great deal of attention in research
utilizing an individual model. The ability to elicit positive regard from others (Werner,
1990), self-efﬁcacy, self—esteem (Rutter, 1985), internal locus of control (Luthar, 1991),
and ego resilience (Block & Block, 1990) are some of the most widely accepted and
researched personality traits of resilient individuals. Several writers on resilience
(Rutter, 1985; Kobasa, 1985; Walsh, 1996) acknowledge that resilient individuals do
not simply avoid stress, they actively interface with it in a way that perpetuates self-
efﬁcacy and strengthens domains of competency. Kobasa (1985) hypotheses that
resilient individuals have "hardy" personalities and possess cognitive structures that
enable them to withstand stressors. These individuals are able to cognitively reframe
difﬁcult events as challenges with opportunities for growth, have a sense of loyalty and
commitment to the activities in their lives, and perceive that they have an inﬂuence on
the events that impact the course of their development.

Rutter and Garmezy are two of the most proliﬁc writers on resilience and have
been instrumental in expanding models beyond individual attributes to include family

factors. Individual attributes, external sources of support, and family cohesion are three

15

domains that are hypothesized by Garmezy (1985) as having the most powerful
inﬂuence on a child's ability to cope. Some of the most important protective qualities of
parent-child relationships are the stability of attachment over time (Rutter, 1985, Farber
& Egeland, 1987), emotional accessibility of the parent (Egeland et al., 1993), and a
sense of warmth and acceptance (Garmezy, 1985). Although family inﬂuences have
been given a considerable amount of attention in resiliency research (e. g. Garmezy,
1985; Rutter, 1990; Masten, Best, & Garmezy, 1990), many individual models neglect
the protective inﬂuence of sibling and parent-child relationships and view resilient

individuals as survivors of destructive familial forces (e. g. Wolin & Wolin, 1993).

Ecosystemic Models of Resilience

 

Walsh (1996) and Egeland et a1. (1993) point out that protective factors need to
be understood in their ecological and developmental context. Developmental systems
approaches (e.g. Fitzgerald et al., 1994; Staudinger, Marsiske & Baltes, 1995; Ford &
Lerner, 1992) posit that the study of resilience needs to extend beyond individual
processes to span multiple levels of analysis of protective factors in an attempt to
unravel a life course leading to low levels of substance use. The theoretical constructs
such as plasticity, multiple developmental trajectories, fusion, and transacting contexts
have been instrumental in building and reﬁning models of resilience. Each of these
concepts assumes that human beings are in constant change with their environments.
While the environment can potentially offer multiple developmental trajectories, the
relative plasticity of human development plays an integral role in producing several

developmental trajectories that can be characterized by varying degrees of resilience or

16

psychopathology.

The identiﬁcation of protective processes within families that have a history of
alcoholism is so complex that a developmental systems orientation is necessary. Fitzgerald
et a1. (1994) have utilized developmental systems theory in the study of alcoholism to
conceptualize a methodological ﬁamework that analyzes patterns of variation families.
Developmental systems offers an excellent theoretical ﬁt to the study of protective factors
because it acknowledges that there is a range of developmental pathways within the
conﬁnes of contextual variables. Research models guided by developmental systems
theory explore domain-speciﬁc sources of variance and bi-directional affects of two or
more domains that “may induce or maintain vulnerability or may provide buffers that
induce or maintain resilience (Fitzgerald et al., 1994, p. 352). Although boundaries of
plasticity vary within developmental trajectories for COAs due to a genetic predisposition
to alcoholism and sociocultural norms that make addictive substances readily available,
development is still viewed as open ended. This probabilistic, rather than deterministic,
view of development accounts for the multiple outcomes that COAs can potentially exhibit
when exposed to similar risks. A probabilistic view of development that has multiple
developmental trajectories with multiple outcomes implies any one-risk variable (e. g.
family dynamics, genetic predisposition to alcoholism, etc.) cannot adequately predict that
offspring will perpetuate intergenerational alcoholism.

Several possible protective factors must be explored to induce or maintain a
developmental pathway leading to the existence of substance use problems (e. g. Gottlieb,
1991; Fitzgerald, et al., 1994). Research models that investigate the causal determinants of

substance abuse need to take into account intraindividual (within the individual),

17

interindividual (between individuals), contextual (social-historical-temporal), and
organism-environment transactional (ecological, bi-directional) sources of variance. The
current study investigates two of these ﬁve levels of analysis: conventionality (i.e.
intraindividual); and emotional autonomy, parentiﬁcation, and supportive parent-adolescent
relationships (i.e. interindividual). Parentiﬁcation, emotional autonomy, parental support
(i.e. interindividual), and conventional belief systems (i.e. intraindividual) are only four
factors that may interact and fuse with genetic predisposition (i.e. intraindividual), and
gender socialization (i.e. contextual). Numerous other factors are likely to guide
individuals down many different developmental pathways because members from alcoholic
families also intentionally and unintentionally alter each other and their contexts (Ford &
Lerner, 1992). For example, it is reasonable to expect that female COAs (i.e. contextual)
are more likely to be in parentiﬁed relationships and hold more conventional belief systems
than males because of the tendency for society, parents, and siblings to socialize girls to be

caregivers (Gilligan, 1982; Weisner, & Gallimore, 197 7).

Family Systems Models of Resilience

 

On a microlevel of analysis, contextual family therapy theory and the notion of
relational resilience ﬁts within the metatheoretical approach the developmental systems
perspective to give a more detailed account of how intraindividual and interindividual
variables transact to serve a protective function. The concept of relational resilience,
contextual family therapy theory (Boszormenyi-Nagy et al., 1991) and other models of

family therapy theory (e.g. Minuchin & F ishman, 1981) view resilience as a systemic

18

quality that is embedded in the process of interdependent relationships. Individual models
of resilience tend to view the family as a static protective or risk variable.

Walsh’s (1996; 1998) concept of relational resilience helps to view the family
interaction domain of a developmental systems perspective in greater detail. It makes a
conceptual shift from the individual to include the family as a developing system that is
continually changing and being changed by multiple contexts (Ford & Lerner, 1992).
The concept of relational resilience views the family as the source of resilience, rather
than an inborn trait or a product of individual initiative (Walsh, 1996; Hawley &
DeHaan, 1996). Thus, resilient families adapt to stressors by utilizing existing
resources, developing new strengths, and reorganizing family relationships to minimize
the disruptive impact of crises. Walsh (1996; 1998) points out that relational resilience
is a phenomenon that lies within the quality of family relationships and can be forged
through adversity. There is a reciprocal exchange of beneﬁts that can lead to adaptive
outcomes for more than one family member. Families with relational resilience are
characterized by afﬁrming belief systems, effective communication patterns, and
relationships that are ﬂexible, cohesive, and adaptive. Hence, a resilient family is one
that is able to adapt and, ultimately, even prosper ﬁ'om crises (Hawley & DeHann, 1996;
Walsh, 1996; 1998).

Minuchin (Minuchin & F ishman, 1981) was one of the ﬁrst family therapists to
acknowledge and explore relational resilience manifested in parentiﬁcation. He points
out that large families or families with few parental resources (e. g. single parent
families) “delegate authority” to ensure efﬁcient functioning in much the same way as

large organizations. One or several older children often take over speciﬁc parental

19

functions as representatives of the parents. This type of an arrangement works well for
the entire family if the duties are well deﬁned and ﬁt within the child’s developmental
capabilities.

Strategic family therapy (Haley, 1987, Madanes, 1981) theory views family
dynamics surrounding parentiﬁcation within the context of hierarchy and power. The
roles of parents are to nurture and protect children. A congruent parent-child hierarchy
is one in which the parent has power over a child because they are responsible for
providing love, guidance, and limits. An incongruent hierarchy exists when a child
holds most of the power in the adult child relationship. The child may take care of the
parent and/or be in charge of making rules and running the day-to-day operations of the
home. Haley (1987) points out that an incongruent hierarchy usually exists because one
or both parents have been undermined by a cross-generational coalition between the
child and another adult to undermine the parent’s authority. It is possible for a child to
have a different power conﬁguration with each parent. In other words, a child may be
parentiﬁed with one parent and not with another. A congruent hierarchy with an
idealized, nurturing, non-alcoholic parent may counterbalance the deleterious effects of
an incongruent hierarchy with an alcoholic parent.

Madanes (1981) has is clear in her conceptualization of how positive aspects of
parentiﬁcation can beneﬁt the family and a form of relational resilience. Her
conceptualization of how parentiﬁcation manifests itself is much different from
Minuchin’s (Minuchin & F ishman, 1981) and Haley’s (1987). From her perspective,
children may have problems that actually increase their parent’s functioning. For

example, a child’s problems in school can provide a parent with respite from obsessing

20

about ﬁnancial problems. The child is benevolently motivated to have school problems
that function as a diversion to make the parent feel less helpless and depressed. Hence,
a child in a parentiﬁed role with an unstable, alcoholic parent might help the family
function better.

Contextual family therapy de-emphasizes pathology and enhances the concept of
relational resilience by placing it in a comprehensive theoretical perspective to explore
the ethical dimensions of parentiﬁcation. There is a focus on the inherent relational
ethics of family life such as, equitable give and take, accountability, and loyalty
(Boszormenyi-Nagy et al., 1991). Contextual family therapy theory is a comprehensive
approach that has evolved out of developmental and individual theories of
psychotherapy (Hargrave & Anderson, 1990). It has made a unique contribution to
family therapy by integrating relational ethics into theory and clinical practice.
Relational ethics is a dimension that addresses fairness and trust within parentiﬁed
relationships. All members in a family are instinctually loyal to one another with each
member committed to behave in any way necessary to meet the needs of its members.
The balance of fairness, between what an individual gives to the family and what one is
entitled to take, is directly related to healthy family ftmctioning. Although children can
be exploited through extreme parentiﬁcation, the development of healthy self-esteem
and self-efﬁcacy is tied to the degree that children are provided the opportunity to give

back to their parents (Goldenthal, 1992).

21

Protective Factors

 

Parentiﬁcation

 

Parentiﬁcation is an interaction pattern within a family structure that gives
parental roles and responsibilities to children and adolescents (Mika et al., 1987).
Children in role reversals or parentiﬁed relationships may take care of siblings, an
impaired parent, or play the role of a mediator between conﬂictual parents. The
developmental status of a caregiving child often dictates the level of responsibility he or
she has in the family and parentiﬁcation type (e. g. consoler, peacemaker, or conﬁdant).
For example, the cognitive and emotional capabilities during adolescence would make
teenagers a stronger candidate than a sibling in early childhood to be parentiﬁed in a
spousal role with a single parent.

A handful of studies have consistently found that COAs are more likely than
controls to score higher on parentiﬁcation measures. Widely held beliefs about parent-
adolescent role-reversals in the clinical literature on alcoholism have been supported by
the research of Goglia, Jurkovic, Burt, and Burge-Callaway (1992). They found that a .
college sample of COAs (particularly female COAs) were more likely than controls to
have assumed caregiving responsibilities in their family of origin. Wright (1992)
replicated Goglia et al.'s (1992) ﬁndings and found that female COAs are more likely to
be in role-reversed relationships speciﬁcally with their alcoholic fathers. Although
healthy emotional autonomy was lost at the expense of marital conﬂict and paternal
alcoholism, daughters in caregiving roles with drinking fathers were also extremely self-
reliant.

The tendency to pathologize families, who place children in caregiving roles

22

appear to ﬂourish in Western, industrialized nations. Rutter (1985) points out that
protective characteristics are often not desirable, positive, or pleasant. This is
particularly true of parentiﬁcation, because it has been normally regarded as
dysfunctional (Hecht, 1979) and even exploitative (Boszormenyi-Nagy, et al., 1991).
However, several diverse cultures practice and rely on sibling caregiving in order to
ensure healthy family ﬁmctioning (Weisner & Gallimore, 1977). Cross-cultural data
indicate that both mothers and younger siblings beneﬁt when an older sibling is an
assistant to parents. Ainsworth (1967) found that sibling surrogate parenting has
beneﬁts for the entire family system because it can decrease maternal caregiving
demands for mothers to be more emotionally and instrumentally accessible to the
family. Based on qualitative, anthropological research, Whiting and Whiting (1975)
assert that a caregiving role helps children to develop skills to be more nurturant and
responsible. Furthermore, while older siblings learn adult functions and how to be
nurturant and independent, younger children are less stressful during periods of
separation from their mothers because sibling caregivers function as alternative
attachment ﬁgures.

Family therapy theorists (e. g. Minuchin & F ishman, 1981; Boszormenyi-Nagy,
et al., 1991; Boyd-Franklin, 1989) have also maintained that some degree of
parentiﬁcation can be beneﬁcial and even necessary. Boyd-Franklin (1989) suggests
that an increase in emotional and instrumental responsibility for adolescents is likely to
serve several adaptive purposes within the family domain and have reciprocal beneﬁts
for all family members. From an individual perspective, caregiving during adolescence

may facilitate a protective autonomy process that permits the adolescent to be separate

23

from one's family by being in a "one up" position within the deidealized, parentiﬁed
relationship, while maintaining a sense of connectedness to the family. From a
structural family therapy perspective (e. g. Minuchin & Fishman, 1981), successful
functioning in large families and single parent families is often made possible by
placing older siblings in auxiliary roles of parenting. Children who assume caregiving
roles within such families act as parental ﬁgures under the auspice of the parents who
ensure that the parental child’s responsibilities do not exceed developmental capabilities
or alienate them from their siblings.

Contextual family therapy asserts (Boszormenyi-Nagy et al., 1991; Goldenthal,
1993) that mild forms of parentiﬁcation are crucial in guiding children down a normal
developmental trajectory. Although young children need to receive more instrumental
and emotional support from their parents in general, parents who gradually provide
opportunities to give (i.e. mild forms of parentiﬁcation) throughout the life-span
enhance their children’s development of self-worth and autonomy (Goldenthal, 1993).
A balance between give and take, also called reciprocity, leads to constructive
entitlement on behalf of the child and ultimately beneﬁts future generations. Parent-
child relationships that do not experience a balance of give and take often cultivate
destructive entitlement and exploit future generations through severe and maladaptive
parentiﬁcation. Extreme forms of parentiﬁcation often leads to an imbalance in the
intergenerational ledger, while adaptive forms of parentiﬁcation within parent-child
relationships that are characterized by mutual respect and partiality leads to constructive
entitlement, loyalty and protection.

Some research and clinical literature has explored forms of adaptive

24

parentiﬁcation that can lead to positive outcomes at individual and familial levels of
analyses. Mika et al. (1987) and Jurkovic (1997) reviewed the family therapy theory
literature to determine the context in which parentiﬁcation can be adaptive. They found
that the responsibilities associated with a parentiﬁed role must be within the child's
developmental competencies and the parents or siblings in the parentiﬁed relationship
must assume complementary, child-like roles with the parentiﬁed child. Another
important component of adaptive parentiﬁcation is that others must acknowledge, value,
and legitimize the child's parentiﬁed role in the family. Mika et al. (1987), however,
does not explicitly address how the family meets the parentiﬁed individual’s emotional
needs. A parentiﬁed relationship has a hierarchical arrangement with an unequal
distribution of power that places the child or adolescent in a "one-up" position with his
or her sibling or parent. Hence, it is important for the parentiﬁed offspring to have a
source of emotional support other than the individual with whom they play a parentiﬁed
role (J urkovic, 1997). Emotional support ﬁom another parent in a non-parentiﬁed
relationship can function as a potential resource that can help to protect the parentiﬁed
individual ﬁ'om becoming overburdened with parental, sibling, and/or spousal
responsibilities.

A very small segment of the COA and divorce literature speculates that
parentiﬁed roles can be adaptive for some adolescents. For example, adolescents who
share in the caregiving of siblings, household tasks, and provide developmentally
appropriate forms of emotional support to family members may beneﬁt from an
increased sense of self-efﬁcacy and independence (Weiss, 197 9; Hetherington; 1989;

Nardi, 1981; Wilson & Orford, 1978; Boyd-Franklin, 1989). Empirical studies on

25

parentiﬁcation or "role-reversals" in alcoholic families have also provided some
evidence that caregiving roles are associated with adaptive behavioral functioning
(Beardslee & Podorefsky, 1988) and self-reliance (Wright, 1992). Schissel (1993)
hypothesizes that lower rates of substance abuse within his sample of female COAs can
be explained by Garmezy's (1985) contention that helping a parent leads to the
development of life skills that make an individual more resistant to the impact of
stressful events.

In summary, a small number of studies have found that parentiﬁcation has both
individual and familial beneﬁts. Quantitative studies seem to indicate that
parentiﬁcation may result in increased self-reliance and self-efﬁcacy, while cross-
cultural, qualitative research has found that children in caregiving roles exhibit an
advanced ability to provide nurturance and tenderness to younger children. Findings
from a review of the family therapy theory literature and cross—cultural studies indicates
that parentiﬁcation is a way in which families make a structural shift to adapt to divorce
and the instrumental and emotional needs of large families. However, the role of
parentiﬁcation within alcoholic families is much less understood. The tendency for
females to be placed in role reversals (Wright, 1992; Goglia et al., 1992) might be one
way in which families reorganize and adapt to an alcoholic parent who is not able to
fully function in a parenting capacity. Although the theoretical literature states that this
form of parentiﬁcation has several familial beneﬁts, it would be ﬁ'uitful for future

research to explore if parentiﬁcation has a protective inﬂuence on COA substance use.

26

Conventionaliﬂ

 

Conventionality has been identiﬁed as one of the most consistent predictors of
COA resilience (Werner, 1986), and a characteristic commonly found to prevent later
substance abuse (Hawkins, Catalano, & Miller, 1992; Brook, Brook, Gordon,
Whiteman, & Cohen, 1990). Conventional adolescents incorporate traditional social
beliefs into their value systems, develop a strong internal locus of control, and value
achievement (Brook et al., 1990). More speciﬁcally, conventional adolescents have a
high intolerance of deviance (e. g. substance use), value school achievement,
responsibility, and church involvement and score low on sensation seeking. Brook et a1.
(1990) was one of the ﬁrst research teams to operationalize conventionality and
investigate its protective function. They found that adolescents ﬁ'om non-clinical, non-
alcoholic families were less likely to use drugs and alcohol if they scored high on
religiosity and academic achievement. Another study found that adolescent drug and
alcohol use actually decreased as commitment to church and academics increased
(Johnson, Bryant, Strader, Bucholtz, Berbaum, Collins, & Noe, 1996). Similar ﬁndings
are conﬁrmed in studies based on samples that do not measure parents' substance use.
For example, affiliations with non-using peers in formal settings, such as Boy Scouts
and church, are related to a reduction in drug and alcohol use, while peer afﬁliations in
less structured settings (Selnow & Crano, 1985; Johnson, 1980) are associated with
increased substance use.

The clinical literature has explored conventionality among offspring of
alcoholics who adopt the “hero” role (see Black, 1982; Wegscheider, 1981). The hero,

mascot, scapegoat, and lost child all function to form a system that adapts to stressors

27

related to parental alcoholism that also enables alcohol consumption to continue. The
hero is often the oldest child who serves to take the focus off the alcoholic parent’s
drinking through conventional, pro-social behavior such as excellence in academics and
sports. The mascot diverts attention away from the alcoholic parent’s problems through
humor, while the scapegoat manages to get the family to refocus their attention on anti-
social behavior such as truancy, poor peer affrliation, and intense conﬂict. Finally, the
lost child plays a less dramatic role characterized by introversion in an attempt to
quietly cope with alcohol related stressors in a way that also does not call attention to
the problems associated with addiction.

Despite the pathological connotations that the “hero” role has received, some
clinical writings and a small amount of empirical research has explored the notion that
adolescents who take on the role may be protected ﬁom developing alcohol problems
due to higher levels of conventionality. Deutsch (1982), Black (1982), and
Wegscheider (1981) have observed that one of the hero’s strengths is that they are the
least likely of their siblings to develop problems with drugs and alcohol. These
clinicians point out that the hero is committed to achievement in conventional activities
and are likely to identify with the non-substance abusing parent who “enables” the
alcoholic parent’s drinking. The scapegoat child, on the other hand, has the tendency to
identify with the alcoholic parent and model their drinking behavior accordingly.

At least two studies have provided empirical support for the protective utility of
the hero’s conventional beliefs and many others have explored the increased risk that
the scapegoat’s anti-social behavior has on developing problems with substances.

Werner (1986) and Potter & Williams (1991) both found that conventional

28

characteristics commonly associated with the hero role are often linked to lower rates
substance abuse and dependence among offspring at risk for developing drug and
alcohol abuse. Although Werner's (1986) study was not directed by a conceptualization
of childhood "roles" in alcoholic families popularized by the clinical literature (see
Black, 1982; Wegscheider, 1981), many of the well-adjusted offspring from a sample of
mentally ill and/or substance-abusing parents had conventional attitudes and
characteristics that correspond to clinical descriptions of the socially competent hero.
The well-adjusted, conventional offspring who demonstrated competence in work and
school domains had a heightened sense of responsibility and self-efﬁcacy and lower
rates of substance use that set them apart from their less successful peers.

In the only study of its kind to speciﬁcally assess for family roles of the
alcoholic family and measure substance use, Potter and Williams (1991) found that
adults in their mid-thirties who ﬁt the clinical descriptions of the hero and mascot had
fewer alcohol problems and higher levels of self-esteem. The items on the questionnaire
they used to assess for the hero role was high in face validity with Brook et al.’s (1990)
conceptualization of conventionality. On the other hand, alcohol problems were more
common in subjects who ﬁt the "scapegoa " description, which is associated with anti-
social behaviors. Both COAs and controls that described themselves as “scapegoats”
were also more likely to report lower levels of self-esteem. However, Potter and
Williams’ (1991) results need to be interpreted with caution due to several
methodological ﬂaws that include poor measures of substance use and the fact that
COAs were recruited ﬁom adult self-help groups.

Studies that explore the ﬂip side of conventionality to measure constructs

29

common to the scapegoat role, such as conduct disordered behaviors and antisocial
personality traits, have found that these characteristics are some of the strongest
predictors of substance abuse and dependence in adulthood. This trend is especially
true for males (see Zucker, Fitzgerald, & Moses, 1995; Zucker, Ellis, Bingham,
Fitzgerald, 1996; Drake & Valiant, 1988). In a 33-year longitudinal study of COAs,
Drake and Valiant (1988) found that paternal alcoholism coupled with an unsupportive
maternal relationship predicts personality disorders in sons. Behavior problems and
truancy in adolescence speciﬁcally predicted anti-social behavior and alcoholism in
adulthood (Valiant, 1983).

In summary, there is not enough evidence to suggest that conventionality results
in low levels of COA substance use. However, conventional behavior and beliefs have
been identiﬁed to be one of the strongest predictors of low levels of substance use
among non-COA adolescents. Potter and William’s (1991) study on a clinical sample of
adult children of alcoholics (ACOAs) sheds a great deal of light on the protective
inﬂuence that the conventional hero role might have substance use during adolescence.
Their study is one of many to suggest that parental alcoholism alone does not seem to
predict anti-social and substance abusing behavior among COAs. More research is
needed to explore if conventionality and other contextual variables mediate the
relationship between parentiﬁcation and COA’s heightened risk for developing

problems with drugs and alcohol.

30

Emotional Autonomy

 

Analytic and neo-analytic theorists have described emotional autonomy as a
process of individuation in which children begin to decentralize the focus that parents
play in their lives (Blos, 1979). As children enter adolescence, they engage in the
process of deidealization, which entails relinquishing idealized perceptions of parents,
renegotiating levels of parent-adolescent intimacy, and re-evaluating childish
dependencies. More recent researchers have pointed out that autonomy is a process that
requires both adolescent and parent to renegotiate their relationship based on mutual
respect (Grotevant & Cooper, 1986). Individuation occurs in the context of intimate
parent-adolescent relationships and supports the adolescent's gradual need for greater
independence. From this perspective, the goal of emotional autonomy is not a linear
detachment from parents with rebellious and chaotic interactions.

Frank, Pirsch, and Wright's, (1988) study on late adolescents indicates that there
are three components of emotional autonomy that permit adolescents to function
separately from parents without denigrating them or fully detaching. First, emotionally
autonomous adolescents do not depend on their parents for determining self-worth.
Second, these adolescents have the capacity to be in charge of negative feelings towards
their parents, rather than being passively consumed by negative reactions. In other
words, rebellious and chaotic interactions are indicative of a destructive detachment
process in which behavior is dominated by and based on negative feelings towards
parents. Thirdly, emotionally autonomous adolescents are able to maintain respect for
their mother and/or father as they struggle with deidealizing their parents.

Blos (1979) also suggests that conﬂict and chaos is not a necessary condition for

31

individuation to occur. Instead, children gradually deidealize their parents as they enter
adolescence to begin the process of individuation and start functioning as autonomous
adults. From this perspective, adolescence is a time of cognitive, behavioral, and
emotional individuation that is not accompanied by excessive tension or estrangement
ﬁom parents. Instead, as children enter into adolescence they gradually deidealize their
parents and experience a moderate degree of internal conﬂict. As adolescents mature,
they begin to let go of childhood dependency on parents, perceptions of parental
omnipotence, and begin to incorporate weaknesses, faults, and vulnerabilities into a
more realistic view of their parents as fallible people.

Steinberg and Silverberg (1986) created the Emotional Autonomy Scale to

 

operationalize Blos's perspective on emotional autonomy. They deﬁne emotional
autonomy as the development of mature, realistic perceptions of parents and the
adolescent's ability to take primary responsibility for emotional stability and decision
making (Lambom & Steinberg, 1993). Emotional autonomy is acquired through the
process of individuation that occurs within a supportive parent-adolescent relationship
(Grotevant & Cooper, 1986; Steinberg, 1990) and is achieved through deidealization
(Frank et al., 1988; Steinberg & Silverberg, 1986). Steinberg and Silverberg’s (1986)
scale is divided into cognitive and emotional domains. The ability to deidealize their
parents and perceive them more as people addresses the cognitive aspects, while non-
dependency and individuation addresses the affective domain.

The relatively few empirical studies done on separation-individuation and
emotional autonomy in alcoholic families have not been designed to identify protective

process within alcoholic families or measure substance use as an outcome. However,

32

Davis, Stern, and VanDeusen’s (1978) early study on autonomy sheds some light on
how emotional autonomy, may be conceptualized as a protective factor. They studied
interactions within a sample of alcoholic families and found that the context for
developing optimal levels of autonomy is characterized by enmeshed relationships
between non-substance using members and a disengaged relationship with the alcoholic
parent. They also found that individuation in the alcoholic family is delayed much
longer than families without substance abuse problems. Although adolescents are
disengaged or more emotionally autonomous with their alcoholic father, they often have
a difficult time separating from the non-alcoholic members of the family (e. g. mother).
Although COA substance use was not measured, future studies could explore if an
"enmeshed" relationship with a non-substance using parent and a "disengaged"
relationship with an alcoholic parent protects the adolescent ﬁ'om problematic alcohol
use if the alcohol-abusing parent has also been deidealized.

Parentiﬁed adolescents who have disengaged relationships and high levels of
emotionally autonomy with their alcoholic parent, but also have enmeshed relationships
and low levels of emotional autonomy with family members who use low levels of
substances, may develop levels of emotional autonomy much like adolescents in
divorced families (Sessa & Steinberg, 1991). Like the divorced family in which the
non-custodial parent becomes peripheral and both parents are likely to reveal more
instances of vulnerability, the individuation process during adolescence may be
challenged by alcohol-related family stress that leads to less parental emotional
accessibility. Hence, COAs may be more likely than non-COAs to develop higher

levels of emotional autonomy and function independently with their alcoholic parent at

33

an earlier age. In fact, Wright, Frank, & Pirsch’s (1991) ﬁndings provide support for an
accelerated individuation process within alcoholic families. They found that that "role-
reversals" in alcoholic families were common and led to higher levels of emotional
autonomy. More speciﬁcally, sons reported greater autonomy during adolescence and
did not seek assistance from their alcoholic fathers (Wright, 1992). Independent
functioning, a component of emotional autonomy, was higher for these males than
female COAs and controls.

Although high levels of emotional autonomy appears to be a trend among
offspring who grow up with an alcoholic parent, a review of the literature indicates that
high levels of emotional autonomy alone have not consistently predicted positive or
negative adjustment during adolescence. High scores have been linked to negative
outcomes, such as substance use, conformity to anti-social peers, and school misconduct
(Lamborn & Steinberg, 1993; Turner, Irwin, & Millstein, 1991; Frank & Burke, 1992).
Premature deidealization, an integral part of emotional autonomy, has been described as
"pseudo-autonomy" (Wright et al., 1991) because the adolescent displays a high degree
of self-governance at the cost of emotional connection to their parents. On the other
hand, high levels of emotional autonomy have also been indicative of positive
outcomes, such as less identify diffusion (Frank, et al., 1988), fewer behavioral
problems, greater social competence, and high academic performance (Furman &
Holmbeck, 1995). Mixed ﬁndings are likely to be a result of methodological issues that
do not take into account the context of the familial environment (Fuhrman & Holmbeck,
1995; Lambom & Steinberg, 1993) or control for COA status. For example, studies that

have found high levels of emotional autonomy associated with substance use did not

34

measure alcohol problems with parents (e. g. Lamborn & Steinberg, 1993; Turner et al.,
1991) or idealization of an emotionally supportive non-alcoholic parent (e. g. Frank et,
al., 1988; Lamborn & Steinberg, 1993; Turner et al., 1991).

Lambom and Steinberg ( 1993) suggest that emotional autonomy must be
measured along with adolescent perception of parental support to accurately predict
adaptive behavioral adjustment. They hypothesized that high scores on emotional
autonomy and supportive relationships with parents would predict better adjustment in
their large sample of non-COA, pre-adolescents. Instead, they found that these
variables were associated with mixed outcomes (e. g. higher substance use and anxiety,
and academic and psychosocial competence). Lamborn (1990) has suggested that high
levels of emotional autonomy are more likely to be related to positive adjustment in the
context of a tumultuous and stressful family environment, rather than supportive
parental relationships per se.

Fuhrman and Holmbeck (1995) conducted a study on 10 to 18 years old
adolescents and borrowed Iamborn’s (1990) hypothesis that a high level of emotional
autonomy would function as a protective factor only when adolescents came from
stressful (i.e. low cohesion and maternal warmth characterized with high conﬂict and
control), single-parent families. They expected that emotional and cognitive distance
would help the adolescent rely on their own internal resources to manage stress.
Consistent with their predictions, high scores on emotional autonomy were often
detrimental in supportive parent-adolescent relationships characterized with parental
warmth and low intensity conﬂict, yet adaptive in parent-adolescent relationships

characterized with low parental warmth and high intensity conﬂict. These conclusions

35

mirror Davis et al.’s (1979) ﬁndings of enmeshed relationships with non-alcoholic
parents and disengaged relationships with alcoholic parents. The adaptive function of
emotional autonomy in Furman and Homlbeck’s (1995) study also differed between
mothers and fathers. Emotional and cognitive distance (i.e. deidealization) ﬁ'om
supportive mothers, but not fathers, was associated with behavior problems.

Three studies on COA’s perceptions of parents' substance abuse provide some
evidence that the cognitive dimension of emotional autonomy, deidealization, may play
a protective inﬂuence against substance abuse. These studies indicate that daughters
who do not identify with their using fathers (Brook, Whiteman, Gordon, & Brook,
1986) and young males who do not perceive their alcoholic fathers in a position of
power (McCord, 1988) are less likely to misuse alcohol as adults. Further evidence
comes from Harburg, DiFranceisco, Webster, Gleiberman, & Schork's (1990)
longitudinal study on COAs that measured both adolescent and parent substance use.
They found that offspring of alcoholics who consumed low levels of alcohol believed
their parent's use was problematic. The authors hypothesize that this "fall-off effect" is
likely to occur when the cross-sex parent has the drinking problem and offspring do not
hold the alcohol-using parent in high esteem.

In summary, the literature on emotional autonomy appears to support the
contention that certain levels of emotional autonomy are protective against COA
substance abuse only when speciﬁc contextual factors are present. The collective
ﬁndings of Davis et al. (1978), Gleiberman, & Schork (1990), and Fuhrman &
Holmbeck (1995) identify elements of a protective developmental trajectory for COAs.

Fuhrman & Holmbeck’s (1995) results clearly indicate that the level of parental support

36

(rather than “enmeshment”) from a non—alcoholic parent might be the most important
contextual variable that regulates the protective ﬁmction of emotional autonomy. Davis
et al.’s (1978) and Gleiberman & Schork’s (1990) studies indicate that deidealization of
an alcoholic parent is the most important component of emotional autonomy in
predicting low levels of COA substance use. In other words, an adolescent who has low
levels of emotional autonomy with a supportive non-alcoholic parent might be protected
ﬁom developing problems with substances if the adolescent also has high levels of

emotional autonomy with an alcoholic parent.

Family and Parental Support

 

A review of the literature indicates that there are no published studies that have
explicitly identiﬁed a sample of COAs to explore whether a supportive relationship with
a non-alcoholic parent protects offspring from developing problems with alcohol or
drugs during late adolescence. This is striking in light of evidence that parents are
direct agents for modeling alcohol use before adolescents experiment with substances,
whereas peers are more inﬂuential for marijuana and alcohol after experimentation with
substance use begins (e.g. Kandel & Andrews et. al., 1987). Parent-adolescent
relationships have indirect effects on substance use with peers because they inﬂuence
adolescents' choices for substance abusing versus non-substance abusing peer groups
(Blanton et al., 1997). Moreover, very few studies have been conducted on the impact
of low levels of parental substance use on the relationship between parental support and
adolescent substance use. In fact, extra-familial sources of support and individual

cognitive processes have received considerably more attention than have supportive

37

relationships within alcoholic families. The reason for this might be due to the
pervasive tendency to pathologize families with an alcoholic parent. Hence, children of
alcoholics are more likely to be labeled "resilient " if they have successfully detached
from their families (e. g. Berlin, Davis, & Orenstein, 1988) and ﬁnd support from an
extra-familial source. For example, Ohannessian and Hesselbrock (1993) examined the
social support networks of adult offspring in their thirties and found that friends were a
more signiﬁcant source of support than family for reinforcing abstinence. Based on
retrospective data, Bennett, Wolin, Reiss, and Teitelbaum (1987), found that COAs who
did not have problems with substance use as adults consistently reported a pattern of
"selective disengagement and reengagement" that allowed them to maintain effective
functioning during times that paternal alcoholism escalated.

Because of the limited research in the area of family support and COA substance
use, as noted above, this review has included studies that have investigated the role of
family and parental support on 1) COA psychosocial functioning, and 2) substance use
levels among non-alcoholic parents and their children. A strong parent-child bond with
at least one parent has been identiﬁed as the single most important protective process
operating to produce resilient outcomes across several, diverse populations at risk
(Rutter, 1990; Garmezy, 1985; Radke-Yarrow & Brown, 1993; Masten, Best, &
Garmezy, 1990; Cicchetti & Garmezy, 1993). Although the clinical literature appears to
suggest that alcoholic families lack supportive, intimate relationships and COAs lack
the ability to form these relationships with others (Black, 1982; Brown, 1988; Cermak,
1986), some empirical studies suggest the opposite. Barnard and Spoentgen (1987)

found that a non-clinical, COA cohort actually had a signiﬁcantly greater capacity for

38

intimate relationships than clinical COAs and a control group. The investigators
speculate that the processes behind this outcome were either a manifestation of healthy
familial relationships or an acquired skill that helped resilient COAs cope with familial
stress. Moreover, pre and post-test measures indicated that a treatment group increased
their ability to engage in greater levels of intimacy. Devine & Braithwaite (1992) give
further support against the clinical observation that alcoholic families are devoid of
intimacy and support. They found that a substantial sub-population of alcoholic
families were capable of cohesive, supportive, interactions and were more likely than
conﬂictual alcoholic families to have offspring who are empathetic and caring to family
members in distress. More recent empirical evidence suggests that families with
alcoholic parents who also have strong emotional bonds provide a warm, supportive
environment that inhibits the development of conduct disorders and depression in young
children (Roosa, Dumka, & Tein, 1996).

A more ecologically driven study by Werner (1986) conﬁrms Barnard and
Spoentgen's (1987) speculation that supportive familial relationships exist in alcoholic
families and provides some evidence that they actually prepare offspring to cultivate
such relationships outside of the home. Werner's longitudinal study (Werner, 1986;
Werner & Smith, 1992) emphasizes the importance of investigating external sources of
support in conjunction with other domains of protective inﬂuences identiﬁed by
Garmezy (1985) to gain a more comprehensive conceptualization of protective
processes that lead to resilience. Werner identiﬁed approximately 50 children in a
sample of 698 that had at least one parent who was struggling with alcohol and/or drug

problems. Following several intervals of data collection, she found that 59% exhibited

39

adaptive outcomes at age 18. Most of these individuals cultivated intimate relationships
with both familial and extra-familial sources, had average intelligence scores, and had
temperaments during infancy that fostered a complementary parent-child subsystem.
Familial support within alcoholic families appears to have a signiﬁcant impact
across the life span and some ﬁndings suggest that lacking an intimate relationship with
at least one parent is a more reliable predictor of psychosocial maladjustment than
parental alcoholism per se. Although COA substance use was not measured, several
studies (Braithwaite & Devine, 1993). Many studies (Obuchowska, 1974; Braithwaite
& Devine, 1993; Jacob & Leonard, 1986; Roosa, Tein, Groppenbacher, Michaels, &
Dumka, 1993; Masini, 1996; Werner & Smith, 1992) have indicated that a supportive
relationship with a non-alcoholic parent has been associated with competent
psychosocial functioning and the absence of antisocial behavior within the context of
paternal alcoholism. In one of the earliest studies on the protective role of a non-
alcoholic parent, Obuchowska (1974) found that elementary age children with alcoholic
fathers who had an emotionally satisfying relationship with their mother demonstrated
social and academic competence to compensate for familial problems. Obuchowska and
other researchers (Jacob & Leonard, 1986; Roosa et al., 1993) interpret the absence of
serious psychosocial impairment (e. g. depression, anti-social behavior) to mean that
mothers play a mediating role in the negative effects of paternal alcoholism on children.
Masini (1996) found that self-esteem levels were higher than controls among male
adolescents with alcoholic fathers when mothers were perceived as a source of support.
Werner and Smith (1992) found that ACOAs who exhibited healthy functioning in their

thirties had cultivated some of the most supportive relationships within their family of

40

origin. ACOAs with primarily alcoholic fathers reported that sibling and maternal
relationships were the most utilized and effective forms of support during childhood,
adolescence, and young adulthood (Children of Alcoholics Foundation, 1992).

The combined effects of supportive familial relationships and cognitive
processes that help cope with alcohol related stress have been found to reduce the
impact of parental alcoholism on psychosocial functioning. Braithwaite and Devine
(1993) ﬁndings from a non-clinical sample of COAs found that overall family cohesion
and an intimate, non-drinking parent-adolescent relationship were the best predictors of
psychological adjustment when alcoholism was not perceived as a threat to the
adolescents' well-being. In other words, a family system that denied the problematic
nature of alcoholism appeared to serve a protective function against anxiety and
depression in adolescents. Similar results from another study (Keane, 1983) found that
positive perceptions of family life in the midst of active parental alcoholism might also
lead to healthy adjustment. The protective function of denial about familial alcoholism
runs contrary to observations in the clinical literature (Deutsch, 1982; Black, 1982;
Royce & Scratchley, 1997). However, Braithwaite and Devine (1993) hypothesize that
a buffering effect is likely to be operating because high levels of familial cohesion
served to protect the children from worry and concern associated with the
acknowledgement of severe parental alcoholism. Another study (Reich, Earls, & Powel,
1988) lends empirical support to such a "sheltering" hypothesis with its ﬁndings that
children who are not exposed to the negative ramiﬁcations of parental drinking are less
likely to have a psychiatric diagnosis. It is difﬁcult to conclude whether denial of

parental alcoholism led to better less substance use among oﬁspﬁng in any of these

41

studies, because parental personality disorders and COA substance use was not
measured.

A small body of family process research provides detailed ﬁndings on the
interactional nature of supportive parent-adolescent relationships within alcoholic
families. Resilient families collectively cope with distal risks, such as poverty (Masten,
et al., 1990), and family relationships (e. g. mother—child) cope with more immediate
familial risks (e. g. alcoholism). For example, Seilhamer, Jacob, and Dunn (1993) point
out that mothers who are unaffected by their husbands' drinking buffer their children
from the negative inﬂuences of paternal alcoholism. Alternatively, positive parent-child
relationships can signiﬁcantly mediate maternal stress related to paternal alcoholism
(Dumka & Roosa, 1993). Dumka and Roosa (1993) also found that families
characterized by relational resilience (Walsh, 1996) were able to utilize existing
resources (i.e. parent-child relationships and mother's marital adjustment) to mediate the
deleterious effects of parental drinking problems. Although problem drinking mothers
and fathers were equally represented, only paternal drinking contributed to family stress.

The complex nature of relational resilience is exempliﬁed by the ﬁnding that mothers
who were invested in relationships with their children provided protection ﬁom paternal
alcoholism for their children and themselves. That is, positive mother-chﬂd subsystems
and mothers' marital adjustment led to lower levels of alcohol-related family stress,
supportive parenting styles (Dumka & Roosa, 1993), and positive mental health proﬁles
for both children and mothers (Roosa et al., 1993).

Another body of literature that does not identify COA status has demonstrated

that supportive relationships with parents lead to better psycho-social adjustment and

42

decreased substance use among adolescents (e. g. Luster & Small, in press; Kandel &
Andrews, 1987; Wills, Vaccaro, & McNamara, 1992; Wills, 1990; Rhodes & Jason,
1990). For example, pregnant, Hispanic and Aﬁican American substance abusing
adolescents are more likely to decrease substance use when they reported access to a
mentor or parent who is perceived as a source of support (Rhodes, Gingiss, & Smith,
1994).

Five studies (Brook et al., 1990; Johnson, et al., 1996; Kandel & Andrews, 1987;
Andrews, Hops, & Duncan, 1997; Blanton, Gibbons, Gerrard, Conger, & Smith, 1997)
have measured parental substance use and parental support to explore drug and alcohol
use among adolescents. Although substance use for parents and adolescents was
measured, COA status was not explicitly identiﬁed using measures such as the CAST in
any of the studies. Andrews et a1. (1997) measured the existence of parent and
adolescent substance use and found that the quality of the parent-adolescent relationship
had a moderating effect on adolescent substance use. Adolescents 11 to 15 years old
modeled parental drug (6. g. marijuana) and alcohol use if they had good relationships
with relatively low conﬂict. Hence, ﬁndings suggested that poor parent-adolescent
relationships could actually protect against adolescent substance use when parents are
using marijuana, alcohol, or tobacco. The investigators point out that close parent-child
relationships alone are not adequate protective inﬂuences against adolescent substance
use. The primary limitation of this study is that the onset or severity of substance use
was not measured.

The protective functions of familial support seem to be most powerful in facilitating

adjustment in families with substance using parents when the reciprocal inﬂuences between

43

individual, familial, and community contexts are explored. Johnson, et al. (1996) evaluated
a community based intervention program and found that substance use among high-risk
youth was reduced when protective factors were operating in multiple domains.
Adolescent substance use decreased along with parents' use when family cohesion, school
involvement, and church involvement increased. Family involvement in church-based,
parenting programs were associated with a decrease in parental alcohol consumption and
was followed by increased adolescent-parent and adolescent-sibling bonding. Moreover,
an increase in global family cohesion was accompanied by an increase in conventional
beliefs among targeted delinquent youth and ultimately a reduction in both drug and
alcohol use. Although severity of substance use was not adequately measured, Johnson et
al.'s (1996) study provides support for the synergistic effects of protective factors on
relational resilience among substance using parents and offspring.

Brook et al. (1990) investigated how protective mechanisms from multiple
contexts (e. g. individual, parent-adolescent relationships, peer group) can interact with
risks (e. g. parental and peer substance use) to impact adolescent substance use. They
speciﬁcally investigated the protective function of a non-substance using parent-
adolescent relationship and conventional attitudes with adolescents at risk for substance
abuse. They found that conventional attitudes became the most powerful predictors of
low levels of substance use among adolescents with drug using peers when there is a
strong attachment to a parent who has similar conventional attitudes. Strong matemal-
adolescent attachments were correlated with non-substance use among adolescents who
had fathers that used drugs when mothers with conventional beliefs were emotionally

stable. Interestingly, the same ﬁndings did not exist for patemal-child subsystems.

44

Instead, a secure relationship with a non-using, emotionally stable father had systemic
reverberations that strengthened bonds with their mothers, buffered the impact of drug
using peers, and enhanced adolescents' conventional attitudes.

Brook et al.'s (1990) ﬁndings of strong parental bonds and adolescent
conventionality were replicated in a study with a non-COA, community sample of
urban, Aﬁican American adolescents. Rhodes and Jason ( 1990) found that adolescents
who had pro-social relationships with their parents were more likely to generalize
conventional attitudes about drug use to their interactions with peers. Adolescents in
supportive parental relationships were found to be assertive with their peers and resist
pressures to engage in heavy drug use. Although COA status was not identiﬁed and
parent substance use was not measured, another study moved a step beyond family
support, and investigated the nature of parent-adolescent relationships in more detail
(Kwakman, Zuiker, Schippers, & de Wuffel, 1988). Using Bowlby's (1980) theory of
attachment, results indicated that insecurely attached adolescents were more likely to
drink for the purpose of social interaction and securely attached adolescents reported
more abstinence and light drinking. It is tempting to generalize these ﬁndings to COAs
and assume that offspring who are securely attached to a non-alcoholic parent are less
likely to use substances abusively. However, more studies are needed to explore the
nature of these relationships in greater depth and identify process level variables in
parent-adolescent attachments among COA samples.

In conclusion, the context of supportive parent-adolescent relationships needs to
be taken into account before it can be determined that support, per se plays a protective

role within a developmental trajectory that leads to low levels of COA substance use.

45

Based on the studies reviewed here, it appears that supportive parental relationships are
most effective in reducing the risk of COA substance use when the parent is not abusing
substances, can provide emotional stability, and is perceived to be a credible source of
support. Brooks et al. (1990) study provides strong support that adolescent
conventionality is a strong mediating factor between parental support and adolescent
substance use. It is also conceivable that these processes would operate in a similar

fashion with regard to COA substance use.

46

Summary

Although COAs have a history of being treated as a homogeneous group with
predictable maladjustment problems, there is overwhelming evidence that protective
factors contribute to the varying degrees of offspring adjustment. Research needs to
identify developmental pathways, rather than generic protective variables, that lead to
adaptive outcomes because COAs are “contextually embedded within a broader set of
systems (biological, intraindividual, interindividual, social and cultural)” (Fitzgerald, et
al., 1994). Parentiﬁcation, conventionality, deidealization, and parental support are
hypothesized to characterize a protective developmental trajectory that leads to low
levels of substance use during a time in adolescence that is often associated with
heightened substance use. It is likely that these protective variables contribute to lower
rates of substance use among a sub-population of COAs because the pathway reﬂects
macrolevel inﬂuences that transact with microlevel inﬂuences to dictate genetic
susceptibility at the individual level (Zucker, et al., 1995).

There are no studies to date that specify parameters of risk by identifying COA
status and substance use of offspring in an attempt to identify protective factors within
parent-adolescent relationships that are associated with low levels of substance use
among late adolescents. The current study used continuous variables in a mediating
model (see Figure l) to provide empirical support for theoretically based hypotheses
that identify variables through which parentiﬁcation leads to low levels of substance
use. The relationship between parentiﬁcation and low levels of substance use was
hypothesized to be mediated by higher scores on conventionality, which inhibit the

development of conduct disordered behaviors in childhood and antisocial personality

47

traits that have been associated with substance abuse in adulthood (see Zucker et al.,
1995; Zucker, et a1, 1996; Drake & Valiant, 1988). Parental support was also
conceptualized as a potential mediating variable, but prior research (Andrews, et al.,
1997; Fuhrman and Holmbeck, 1995) has found that supportive parental relationships
are only protective in particular contexts. Hence, the mediational model shows that
highly parentiﬁed COAs will have lower levels of substance use when they have high
levels of conventionality and receive low levels of parental support from a deidealized
alcoholic parent and receive high levels of parental support from an idealized non-

alcoholic parent.

48

Hypotheses

A total of eight theoretically driven hypotheses were used to identify protective
factors that may lead to low levels of substance use among COAs. The ﬁrst set of
hypotheses tested if COA status within the sample is an independent risk factor for
problematic substance use when compared to non-COAs (Sher, 1991). A second set of
hypotheses tested if conventionality, parental support, and deidealization mediated the
relationship between parentiﬁcation and low levels of substance use. The protective
utility of parentiﬁcation, conventionality, parental support, and emotional autonomy can
be best measured if family history of alcoholism places COAs at greater risk than non-

COAs for using more substances (Rutter, 1985; Sher, 1991).

Hypotheses for Risk Factors

 

l. The COA group will have signiﬁcantly higher rates of substance use than the non-
COA group.
2. The COA group will have signiﬁcantly higher rates of parentiﬁcation than the non-

COA group.

Hypotheses for Protective Factors

 

3. COAs with low levels of substance use will have higher levels of parentiﬁcation
than COAs with high levels of substance use.

4. Conventionality will mediate the relationship between parentiﬁcation and low levels
of COA substance use. That is, higher scores on conventionality will help account

for the relationship between parentiﬁcation and low levels of substance use.

49

. High scores on idealization with a non-alcoholic parent will mediate the relationship
between parentiﬁcation and low levels of COA substance use. That is, idealization
of a non-alcoholic parent will help account for the relationship between
parentiﬁcation and low levels of COA substance use.

. High scores on deidealization with an alcoholic parent will mediate the relationship
between parentiﬁcation and low levels of COA substance use. That is, de-
idealization will help account for the relationship between parentiﬁcation and low
levels of COA substance use.

. High levels of support from a non-alcoholic parent will mediate the relationship
between parentiﬁcation and low levels of COA substance use. That is, higher scores
on non-alcoholic parental support will help account for the relationship between
parentiﬁcation and low levels of substance use.

. Low levels of support from an alcoholic parent will mediate the relationship
between parentiﬁcation and low levels of COA substance use. That is, lower levels
of alcoholic parental support will help account for the relationship between

parentiﬁcation and low levels of COA substance use.

50

CHAPTER TWO

Method

A developmental systems perspective posits that research models investigating
causal determinants of substance abuse need to take into account intraindividual (within
the individual), interindividual (between individuals), contextual (social-historical-
temporal), and organism-environment transactional (ecological, bi-directional) sources
of variance (Fitzgerald et al., 1994). The current study utilized multivariate techniques
to investigate if two of these ﬁve levels, conventionality, emotional autonomy (i.e.
intraindividual), and supportive parent-adolescent relationships (i.e. interindividual),
mediate the relationship between parentiﬁcation and COA substance use.

Data were collected from a sample of college students in social science
undergraduate courses from Michigan State University. A college sample of COAs may be
most appropriate for identifying familial protective mechanisms that lead to low levels of
substance use because college attendance may operate as a natural screening device that
selects offspring who have better coping skills than their siblings and peers who share the
same status (i.e., alcoholic parent). Perkins and Berkowitz (1991) point out that non-
clinical, college samples offer researchers the opportunity to study relatively well adjusted
COAs who are at particularly high risk for problematic alcohol use due to developmental
stage (Pandina & Johnson, 1990) and the prevalence of substance abuse on college
campuses. White (1987) suggests that, regardless of COA status, the peak period for
problematic alcohol use is between 15 and 24 years old. The cohort of COAs that enter

college are less likely to experience academic impairment (Deutsch, DiCicco, & Mills,

51

1982) and are generally less distressed than COAs in self-help groups (Kashubeck &
Christensen, 1992).

Sher (1991) recommends that data collection and statistical procedures be
designed to conﬁrm if a given variable functions as a mediator, moderator, or another
risk variable. He speciﬁcally recommends that research models include a control group
of non-COAs. A control group of non-COAs was used here to help establish the
parameters of risk (Rutter, 1985) and explore whether or not a given variable moderates
or mediates substance use (Sher, 1991). Level of vulnerability can be established by
exploring if level of parentiﬁcation poses different levels of risks to COAs compared to
non-COAs.

Eligible subjects were self-identiﬁed COAs between the ages of 18 and 24 who
lived with two parents (biological or step-parent) for at least eight years during their
adolescent years. Although controlling for family status may have omitted adolescents
from more chaotic single parent families with parental problem drinking, it is assumed
that parameters of risk are best deﬁned with a sample of subjects who have lived with an
alcoholic parent for most of their adolescence. A comparison group of non-COAs was
randomly drawn ﬁom the same sample of college students. COAs with two alcoholic
parents constituted a third group that was analyzed separately. Both the COAs and non-
COAs completed all of the same measures to identify if within and between group

differences existed.

Subjects

The initial goal of data collection was to identify COAs ﬁ'om intact families. A

52

statistical power analysis (Cohen, 1988, 1992) was performed to determine the overall
sample size of COAs needed to test the set of four independent variables. A .05
signiﬁcance level with a medium effect size indicated that a sample of 168 COAs and
168 non-COAs was sufficient. Data were initially collected on 1,083 participants in an
effort to obtain 168 COAs in each group. This number was reduced to 978 because 105
of the 1,083 subjects were between the ages of 18 and 24 and/or did not live in an intact
family since the age of 10. Descriptive statistics on the 978 subjects are presented

below in Table 1.

Table 1

Ethnicity, College Class, and Gender From the Total Sample (N=978)

 

 

 

 

Ethnicity
Caucasian 823
Aﬁican American 67
Asian 46
Hispanic or Latino 21
Native American 1
Other 6
College Class
Freshman 380
Sophomore 289
Junior 209
Senior 78
5th year Senior 22
Gender
Male 256
Female 722
N=978

A total of 151 COAs were identiﬁed within the sample of 978 and divided into

subjects who had two alcoholic parents (n=22) and subjects who had one alcoholic

53

parent (n=129). A sample of 151 non-COAs was randomly identiﬁed from the original
827 non-COAs to serve as the comparison group. Hence, separate analyses were
conducted on the ﬁnal sample that had a total of 302 subjects and consisted of three
groups: 1) COAs with one alcoholic parent; 2) COAs with two alcoholic parents; and 3)
non-COAs. Descriptive statistics on the ﬁnal sample are found in Table 2. All

preceding analyses for this study were conducted on the ﬁnal sample below.

Table 2

Ethnicity, College Class, and Gender Within Three Groups From the Final Sample
(n=302)

 

 

One Alcoholic Parent Two Alcoholic Parents Non-COAs

 

 

Ethnicity
Caucasian 100 l 3 121
Aﬁican American 1 1 6 10
Asian 6 1 13
Hispanic or Latino 8 l 3
Native American 1
Other 3 l 4
College Class
Freshman 44 10 58
Sophomore 40 6 40
Junior 29 3 35
Senior 10 3 15
5th year Senior 6 3
Gender
Male 18 l 38
Female 1 l 1 21 l 13

 

One Alcoholic Parent (n=129); Two Alcoholic Parents (n=22); Non-COAs (n=151)

Most of the ﬁnal sample (n=302) was never married (99%), Caucasian (81%),
Christian (75%) females (81%) with a mean age of 19 that had no children (98%). All

of respondents were from intact families and lived with both parents since they were at

54

least 10 years old. Twenty-ﬁve percent of the families had a stepparent and 75 percent
had never been divorced. The vast majority of subjects had at least one sibling (90%),
39 percent grew up as the oldest sibling, 3 lpercent were the youngest, 20 percent
identiﬁed themselves as the middle child, and 10 percent were only children. The mean
grade point average (GPA) during college was a 3.0 and 3.5 while in high school. Most
of their fathers worked full-time (95%) and over half (57%) of their mothers worked

full-time.

Procedures

Subjects were drawn from undergraduate psychology courses at Michigan State
University and were offered research credits that contributed to their ﬁnal grade as an
incentive to participate in the study. Interested individuals logged onto the psychology
web-based subject pool, viewed restrictions (i.e. intact family status and be between the
ages of 18 and 24), and reviewed the informed consent form. Individuals who agreed to
participate in the study completed a 275-item questionnaire ﬁom any location that had
Internet access. The questionnaire took approximately one hour to complete and
students earned one research credit per 30 minutes of participation. All participants’
responses were anonymous. Anonymity was obtained by deleting any identifying
information from the completed questionnaire once it was down loaded. Furthermore,
data were only reported in aggregate form so that participants would not be identiﬁed or
associated with the data. The primary researcher provided subjects with a telephone
number, address and e-mail address to contact him and two Michigan State University

faculty who were co-chairs on the dissertation with any questions they might have about

55

the study. They were also given information to contact the Chair of the University
Committee on Research Involving Human Subjects (UCRIHS) if they had any questions
about their role and rights as a participant of research. The primary researcher imported
all of the anonymous data from the completed, on-line questionnaires to an SPSS
database for analysis.
Measures

COA Status

Rutter (1990) points out that it is critical to measure the extent and severity of
substance use among family members and COAs to identify the parameters of risk (e. g.
Seilhamer & Jacob, 1989) because protective mechanisms have the clearest effect in
high-risk contexts. The Children of Alcoholics Screening Test (CAST; Jones, 1986)
was used to assess adolescents’ subjective experience of how parental drinking has, or
has not, affected several life domains. Hence, COA status is based on the participants’
self report of their parent’s alcohol abuse or addiction and does not reﬂect any
diagnostic categorization of parents. The CAST is a thirty item yes-no measure that
produces an aggregate score to classify offspring of alcoholic parents. The CAST was
modiﬁed to capture adolescents’ perceptions, feelings, experiences, and attitudes of
each parent. Hence, sixty items assessed if one or both parents were described as
alcoholics. A total of six or more “yes” responses indicated a signiﬁcant level of
parental alcohol related stress and categorized the respondent as a COA. Wright and
Heppner (1993) have referred to a score from two to ﬁve as a sub-clinical population of
COAs

It is crucial to identify if problem parental drinking occurred during adolescence

56

in order to investigate family factors that might protect against risk. Subjects were
identiﬁed as COAs if they answered ‘yes” to six or more items that inquire about their
father’s or mother’s drinking that occurred while they were living at home. This
inclusion criterion ensured that parental problem drinking was active during a period of
adolescence when substances become more and more accessible and alcohol
consumption is expected to escalate in the general population (Johnston, O'Malley, &
Bachman, 1988).

The CAST speciﬁcally assesses subjective feelings, attitudes, perceptions, and
experiences related to parental drinking behavior. Items such as, “Have you ever lost
sleep because of a parent’s drinking?” measures emotional distress. Perceptual distress
is measured by items like “Did you ever think your father was an alcoholic?” Questions
such as, “Did you ever stay away from home to avoid the drinking parent or your other
parent’s response to the drinking?” taps into stressful alcohol-induced family
experiences. Other items explore the subject’s desire to seek out support, such as: “Did
you ever wish that you could talk to someone who could understand and help the
alcohol-related problems in your family?”

The CAST has been used extensively in the COA literature since the early
1980’s. A number of validity and reliability tests have validated Schuckit’s (1980)
assertion that offspring’s perceptions of parent’s alcohol-related problems can be used
to reliably predict a family history of alcoholism when compared to parent reports. For
example, Jones (1986) found that 100% of self-identiﬁed COAs, and COAs who were
identiﬁed during face-to-face, clinical interviews scored six or higher with a validity

coefﬁcient of .78. Maxwell (1985) found a Spearman-Brown split-half (odd-even)

57

reliability coefficient to be .98 for adolescent and adult offspring of alcoholics.
Coefﬁcient alphas were .96 for mothers and .95 for fathers suggesting extremely good
internal consistency within this sample.

Although the validity of close-ended survey instruments has been debated in the
literature, Skinner and Allen (1983) have found no signiﬁcant differences between self-
report questionnaires and face-to-face, clinical interviews. In fact, Gﬁoerer (1985)
found that substance use self-report measures have beneﬁts over interviews, because
interviews do not afford the privacy of self-report questionnaires. Furthermore, self-

report questionnaires tend to facilitate candid disclosure, especially among adolescents.

Parentiﬁcation

 

Degree of parentiﬁcation was measured using a scale developed by Mika, et al.
(1987). Parentiﬁcation is deﬁned as a familial interaction pattern that places children
and adolescents in a role that is typically ﬁlled by a parent in a particular culture.
Subjects responded to items that measured the degree to which they fulﬁlled parental
roles. Ages 14 to 16 are considered to be a critical period during which children ﬁrst
begin the transition into adult status (Erikson, 1963). The scale was chosen because of
its value free conceptualization of parentiﬁcation. Items are worded in a neutral manner
to assess the degree to which respondents provide instrumental and emotional support to
parents and siblings. Other parentiﬁcation scales (e. g. Jurkovic, 1997) approach
parentiﬁcation from a relatively pathological view and assume that caregiving
responsibilities have a negative impact on mental health (e. g. “It often seemed that my

feelings weren’t taken into account in my family”)

58

Mika, et al.’s (1987) parentiﬁcation measure is divided into four sub-scales that
assess parentiﬁcation in spousal, parental, and sibling domains. Likert-form questions
inquire about behavioral patterns that are indicative of parentiﬁcation within each sub-
scale. The spousal sub-scale assesses the degree to which subjects performed conﬁdant
roles with each parent (e. g. “My mother shared personal problems or concerns with me
as if I were another adult”). On the other hand, the parental sub-scale inquires about a
respondent’s role as mediator between parents (e. g. “1 restored peace if conﬂicts
developed between my parents”). The sibling parentiﬁcation sub-scale items measure
the extent to which subjects provided instrumental (e. g. “I was the person responsible
for deciding what action to take if one of my siblings(s) misbehaved, even when my
parent(s) were present”) and emotional support to siblings (e. g. “I provided emotional
support and/or comfort for my siblings”).

An overall score was based on item weights of thirty questions to determine the
degree to which adolescents ﬁll the role as a supportive spouse to a parent; a parent to
their mother and/or father; and a parent to their sibling(s). Although validation of the
scale is in its infancy (Mika et al., 1987), there appears to be a wide distribution of
scores that determine the degree of parentiﬁcation. A Spearman correlation of g = .98, p
< .01 was derived from weighted item rankings of the authors and weighted item
rankings of trained graduate students. A coefﬁcient alpha of .91 was obtained from this
sample indicating that there was good internal consistency. No sub-scale alphas are

available.

59

Conventionality

 

A combination of several scales was used to test the hypothesis that highly
conventional COAs would consume fewer substances than their non-conventional
counterparts. The conventionality measure included 47 Likert-type items that were
divided into six sub-scales that assessed intolerance of deviance, rebelliousness,
responsibility, sensation seeking, religiosity, and orientation to school. A coefﬁcient
alpha of .95 was obtained from this sample, indicating that there was good internal
consistency. Brook et al. (1990) used all of the sub-scales that were used in this study
except Jessor, Turbin, and Costa’s (1998) scale on religiosity and orientation to school.
Brook et al.’s (1990) scale has been used to measure conventionality within three
different studies investigating the protective utility of conventionality against adolescent
substance abuse. Each of the scales Brook et al., (1990) used was reduced to sub-scales
consisting of 6 to 8 items and was subjected to reliability tests. Most of the measures
that they selected had been previously used in research that directly or indirectly
investigated adolescent substance use or problem behavior. Items are phrased in a
developmentally appropriate manner for adolescents who are in their ﬁrst year of high
school to their senior year in college.

Each of the original measures to create the 47-item scale in this study bad sound
psychometric properties and was used extensively within the substance abuse and
juvenile justice literatures. J essor, et al.’s (1998) scales on religiosity and orientation to
school have Cronbach alphas of .89 and .87 respectively. The religiosity scale has 5
items that assess the frequency of church or temple attendance and the use of prayer as a

coping skill (e. g. “T o be able to turn to prayer when you’re facing a personal problem”).

60

Orientation to School assesses commitment to academics using 13 items (e. g. “Being in
college helps me to become the person I’d like to be”). The Attitude Towards Deviance
sub-scale (J essor, Graves, Hanson, & J essor’s, 1968) was modiﬁed by Brook et al.
(1990) to consist of 7 items. It had a Cronbach alpha of .75 in the present study.
Questions inquire about subject’s willingness to drive without a license, smoke
marijuana, or copy another student’s work. Rebelliousness was measured using Brook
et al.’s (1990) modiﬁed version of Smith and Fogg’s scale (1979). It has a Cronbach
alpha of .74 in this study and consists of 7 items that tap into an adolescent’s
willingness to follow rules and social norms (e. g. “When rules and regulations get in
my way I sometimes ignore them”). The coefﬁcient alpha for this sample was .83 in
this study, suggesting good internal consistency for all 47 items.

Brook et al.’s (1990) version of Gough’s scale (1957) includes 6 items that
measure the degree to which respondents are responsible for their own behavior and the
well-being of those around them. Items measure if subjects are willing to help others
(e. g. “There’s no use in doing things for people: you only ﬁnd that you get it in the neck
in the long 11.111”) and whether or not they take responsibility to abide by laws (e. g. “It is
all right to get around the law if you don’t actually break it”). Sensation seeking was
measured using Brook et al.’s 7 item version of Zuckerman, Eysenck, & Eysenck’s
(1978) scale. Brook et al. (1990) found that the sub-scale yielded a Cronbach alpha of
.53. The scale assesses thrill and adventure seeking (“I would like to try parachuting”),
experience seeking (e. g. “I would like to take off on a trip with no preplanned or
deﬁnite routes or timetables”), disinhibition (“I like wild, ‘uninhibited’ parties”), and

boredom susceptibility (“I have no patience with dull or boring persons”).

61

Parental Support

 

Perceived source of support was measured using the Social Support

 

Questionnaire Six: Revised (SSQSR) (Sarason, Sarason, Shearin, & Pierce, 1987) to

 

determine if non-substance abusing COAs seek out their problem drinking versus non-
problem drinking parent for support. The SSQSR is the most appropriate measure of
social support to assess if adolescents perceive that they have a "credible" source of
support other than the individual with whom they play a parentiﬁed role. Each item has
two parts. The ﬁrst part identiﬁes ﬁ'om whom they obtain support and the second part
assesses their level of satisfaction with the support on a six-point Likert-type scale.
There are 6 items that measure the dependability of the respondent’s source of support
(e. g. “Whom can you really count on to be dependable when you need help?”) and
acceptance (“Who accepts you totally, including both your worst and your best
points?”).

The SSQSR was modiﬁed to speciﬁcally assess the level of satisfaction
associated with social support from each parent. The authors of the measure
conceptualize support as a continuous variable that is represented by an aggregated
score. Scores are obtained by counting the total number of people identiﬁed as sources
of support and the number that represents the level of satisfaction with supportive
individuals. However, this study modiﬁed the SSQSR into a mother and father version
and obtained a score on each parent by aggregating scores that reﬂect level of
satisfaction associated with mothers and fathers separately.

Four studies have been conducted to reﬁne the SSQSR's psychometric properties

and assess its utility in quantifying the perceived availability and satisfaction of social

62

support. The authors found that the scale is a reliable instrument with high internal
consistency that yields comparable results to an extensive structured interview (Sarason,
et al., 1987). Each item was factor analytically derived from a large body of items
intended to measure the functions served by social networks (Sarason, Levine, Basham,
& Sarason, 1983). The coefﬁcient alpha for this sample was .89, suggesting good
internal consistency. The availability and quality of social support is signiﬁcantly
related to positive (rather than negative) life changes and permits individuals to pursue
difficult goals under ﬁ'ustrating conditions. Results also indicated that people with
satisfying sources of social support score lower on measures of anxiety, depression, and
hostility, appear to have a more optimistic outlook on life, and score higher on measures

of self-esteem.

Emotional Autonomy

 

Emotional autonomy within relationships with a problem drinking and non-

problem drinking parent was measured using the Emotional Autonomy Scale (Steinberg

 

& Silverberg, 1986). This scale distinguishes subjects' perceptions of the degree of
emotional autonomy that exists with each parent. It was used to measure if COAs who
use low levels of substances are more likely to have high levels of emotional autonomy
with their problem-drinking parent and low levels of emotional autonomy with their
non-problem drinking parent.

The Emotional Autonomy Scale is a 20-item measure that has four sub-scales

 

that operationalize Blos's (1979) conceptualization of individuation. According to this

conceptualization of individuation, adolescence is a time of separation that is not

63

characterized by a stormy, chaotic, rebellious parent-adolescent relationship. Instead, as
children enter into adolescence they gradually relinquish childhood dependency on
parents, perceptions of parental omnipotence, and begin viewing them more and more
as fallible people. The "deidealization" and "perceives parents as people" sub-scales
measure cognitive aspects of emotional autonomy while the "non-dependency" and
"individuation" sub-scales measure affective aspects.

Steinberg and Silverberg (1986) found that an overall emotional autonomy score
has a Cronbach's alpha of .75. Each sub—scale has four to six, four-point Likert—format
items with each exceeding a .60 internal reliability coefﬁcient. The current study
utilized only the deidealization sub-scale (5 items) and modiﬁed the measure to produce
two scales to derive a score that reﬂects the degree of emotional autonomy with each
parent. A coefficient alpha of .83 was obtained from this sample indicating good
internal consistency. Emotional autonomy scores have been found to increase between
the ages of 10 and 16 with girls scoring signiﬁcantly higher than boys. However, Ryan
and Lynch (1989) found that boys score higher than girls. Emotional autonomy is
conceptualized as a continuous variable and a score is obtained by adding up the
numbers that represent the four-point Likert responses that range ﬁ'om strongly disagree
to strongly agree. High scores reﬂect a high degree of emotional autonomy and low

scores reﬂect a low degree.

Alcohol & Drug Use

 

The dependent variable was measured using two scales: the Adolescent Alcohol

 

Involvement Scale (AAIS) (Mayer & Filstead, 197 9) and the Drinking and Drug

 

 

64

History-Revised Edition (Version 3) (Zucker, Noll, & Fitzgerald, 1989). Subjects were

 

given the AAIS to identify those who use alcohol problematically and the Drinking and

 

Drug History-Revised Edition (Version 3) to assess level of drug use. The AAIS

 

provides an in-depth assessment of the quantity, frequency, context in which drinking
occurs (e. g. “I drink alone”), precipitants to alcohol use (e.g. I usually start to drink
because I feel nervous, tense, full of worries or problems”), and perceived consequences
of alcohol use and involvement (e. g. “was in a ﬁght or destroyed property”). It also
provides information about how others perceive their drinking behavior (e. g. “my
family or ﬁiends tell me to get help for my drinking”). Subjects were also asked to
assess their reasons for drinking and disclose how old they were when they had their
fnst experience with alcohol.

The AAIS is a l4-item questionnaire that is used to assess for problematic
alcohol use. A subject’s substance use is categorized as problematic when it interferes
with psychological, social, and/or family domains of functioning. Each item has several
weighted responses which correspond with the degree to which drinking interferes with
one or any combination of the domains. A total score is gleaned from the responses and
places subjects on a continuum from abstinence to misuse. The items on the AAIS are a
compilation of previously veriﬁed indicators that discriminate between adolescent
alcohol use and misuse. An expert panel of psychiatrists and staff members, many who
were in recovery, working in an inpatient substance abuse unit for adolescents rated the
weight of each item on the basis of face validity. A coefﬁcient alpha of .89 was
obtained in this study indicating good internal consistency.

Mayer and Filstead (1979) report that the measure has extremely high face

65

validity, high test-retest reliability when they administered it to a control (infrequent
drinkers) and experimental group residing in inpatient psychiatric care, and well
established norms. Test-retest reliabilities were obtained by administering the
questionnaire to each group two times with a two-week lag. The correlation between
the two average scores was .91 for the experimental group and .89 for the control group.
The authors also administered the questionnaire to a non-clinical sample of 3662
adolescents at four Chicago high schools. Responses were factor analyzed and results
indicated that scores from the AAIS appear to reﬂect normal benchmarks demarcating
adolescents’ use or nonuse of alcohol. The control group of normal adolescents had
scores that ranged from 0-19. Adolescents who use low levels of alcohol had scores that
ranged from 20-41, and problematic drinking adolescents had scores from 42-57.
Scores 58-79 were obtained from hospitalized adolescents. Moberg (1983) administered
the measure through a telephone survey to a random sample of 1014 youth aged 13-17.
Their results were strikingly consistent to Mayer and F ilstead’s (1979) norms and cut-
off scores.

Because it has been found that COAs are at higher risk than controls to use
marijuana and other drugs (Johnson, et al., 1989), data on drug use were collected using

the Drinking and Drug History-Revised Edition (Version 3) (Zucker, Noll, &

 

F itzgerald,l989). Zucker et al.'s measure uses a 9-point Likert scale to assess the
frequency and quantity of marijuana, LSD, inhalants, cocaine, amphetamines,
barbiturates, and other substances. A coefﬁcient alpha of .95 was obtained from this

sample indicating extremely good internal consistency.

66

Plan of Analysis

 

There has been a considerable amount of discussion regarding the confusion
between moderator and mediator variables in research models that attempt to identify
protective variables (e. g. Sher, 1991; Baron and Kenny, 1986). Moderators and
mediators are often conceptualized as protective factors that have the ability to buffer
the consequences of risk, or to substantially reduce risk associated with vulnerability
(Rogosh, et al., 1990; Sher, 1991). They also have the potential to exacerbate risk that
leads to greater vulnerability. According to Baron and Kenny (1986), mediator-
moderator distinctions are essential in determining research models and statistical
analysis. Mediating variables underscore mechanisms that explain how effects occur.
They account for the relationship between the independent and dependent variables. On
the other hand, moderators identify the conditions that must be in place for certain
effects to take place. They interact with independent variables to impact the dependent
variable. Moderating variables are expected to affect the direction and/or strength
between the independent and dependent variable.

Sher (1991) recommends that a series of regression analysis be performed to
determine if a given factor has a mediating or moderating ﬁmction. Regression analyses
were conducted separately for the COA and non-COA groups. The relationship
between parentiﬁcation and low levels of substance use needed to be substantially
weakened when COA substance use was regressed on each of the potential protective
factors to be considered mediators. This assumption rests on Baron and Kenny’s (1986)

three criteria for mediating variables: 1) the independent variable (i.e. high

67

parentiﬁcation) needs to predict the potential mediators (i.e. high conventionality, low
alcoholic parent support, high non-alcoholic parent support, deidealization of alcoholic
parent, and idealization of non-alcoholic parent); 2) the independent variable (i.e. high
parentiﬁcation) is correlated with the dependent variable (i.e. low levels of COA drug or
alcohol use); 3) and the presumed mediators (i.e. high conventionality, low alcoholic
parent support, high non-alcoholic parent support, deidealization of alcoholic parent,
and idealization of non-alcoholic parent) are correlated with the dependent variable (low
levels of COA substance use).

A set of regression techniques was used to test for moderator effects when
variables did not have a mediating function. Variables were considered moderators if
the interaction between parentiﬁcation and the potential protective factors add
signiﬁcant, explained variance over the main effects of parentiﬁcation and each
potential moderating variable (Rogosh, et al., 1990). There were essentially two
regressions to test whether the potential protective factors moderate the relationship
between parentiﬁcation and low levels of drug and alcohol use. Parentiﬁcation was
entered ﬁrst to control for its effects on drug and alcohol use separately. Parentiﬁcation
was expected to have the most predictive utility and would ultimately have a main effect
on COA substance use level if the variables function as moderators. The ﬁrst step was
to enter the main effects of parentiﬁcation and the moderator variables. In the second
step, the interaction of parentiﬁcation and the moderators were then entered, depending
on which is most highly associated with the dependent variable, to allow for the
moderators with the most predictive utility to be entered ﬁrst. Rogosch et al. (1990)

points out that a signiﬁcant two-way interaction is indicative of moderation. Simple

68

regression lines for high and low values of each moderator variable were used to

determine statistically signiﬁcant interactions (Aiken & West, 1991).

69

CHAPTER FOUR

Results

Preliminagr Analyses

 

Preliminary analyses were conducted in two stages. The ﬁrst stage was to assess
the representativeness of the sample and the second was to determine if there were
signiﬁcant differences among the potential protective factors across the subgroups. The
sample was divided into three sub-groups for further analyses on ethnicity, religion, sex,
and parental education: COAs with one alcoholic parent (n=129), COAs with two
parents (n=22), and non-COAs (n=151). Ethnicity and religion did not differ

signiﬁcantly across the three groups (see Tables 3 & 4).

Table 3

Distribution of Religious Faiths Within Three Groups From a Sample of College
Students

 

 

 

Faith One Alcoholic Parent Two Alcoholic Parents Non-COAs
Christian 78.3% 63.6% 72.9%
Not Afﬁliated 17.1% 18.3% 15.2%
Other 3.9% 13.6% 9.3%
Jewish .7% 4.5% 2.6%

X2 =22. 0 “3

One Alcoholic Parent (n=129); Two Alcoholic Parents (n=22); Non-COAs (n=151)
Chi Squares indicated that there were signiﬁcant sex differences (p<.05) with

females (see Table 4) and level of father’s education (see Table 5) unevenly distributed

across groups. There were a total of 57 males within a sample of 302: one had two

alcoholic parents, 18 had one alcoholic parent, and 38 were in the non-COA group.

70

Fathers of non-COAs had more education than fathers of COAs with one alcoholic
parent, and fathers of COAs with two alcoholic parents had the least amount of
education. However, the majority of respondents indicated that both of their parents
had completed some college course work. One-way ANOVAS also indicated that the
differences on paternal education were signiﬁcant between non-COAs and COAs with

one alcoholic parent 13(2, 299)=9.114, p< .05.

Table 4

The Number of Males and Females Within Three Groups From a Sample of College
Students

 

 

 

 

Group Male Female
One Alcoholic Parent 14% 86%
Two Alcoholic Parents .5% 95.5%
Non-COAs 25.2% 74.8%
X2=8.9, p<.05

One Alcoholic Parent (n=129); Two Alcoholic Parents (n=22); Non-COAs (n=151)

Table 5

Father’s Education Level in Three Groups of Participants From a Sample of College
Students

 

Father’s Education One Alcoholic Parents Two Alcoholic Parents Non-COAs

 

 

Less than high school 4.7% 2%
High school diploma 27.9% 31.8% 14.6%
Some college 29.5% 31.8% 25.2%
Four years of college 22.5% 18.2% 25.8%
Master’s degree 11.6 13.6% .5%
Post Master’s degree 3.9% 4.5% 9.9%
X2=19.4, p<.05

One Alcoholic Parent (n=129); Two Alcoholic Parents (n=22); Non-COAs (n=151)

 

The second stage of preliminary analyses was to determine if there were
signiﬁcant differences between the subgroups on the potential protective factors.
Means, standard deviations, ranges, and F scores were derived ﬁom one-way ANOVAS
on all continuous variables in the model (i.e. parentiﬁcation, drug use, alcohol use,
conventionality, father support, mother support, non-alcoholic parent idealization, and

alcoholic parent deidealization), which are included in Table 6.

Table 6

MeansgStandard Deviations, and F Scores for All Continuous Variables in a Sample of
College Students

 

 

 

 

 

One Alcoholic Parent Two Alcoholic Parents Non-COAs
M (SD) M. (SD) M. (SD) E

Variable
Parentiﬁcation 42.43 (19.30)1 51.09 (25.55) 33.50 (17.43) 12.94c
Drug Use 28.40 (35.95) 55.82 (101.70)2 22.34 (31.44) 6.13b
Alcohol Use 32.83 (10.58) 30.59 (11.88) 31.79 (11.88) .51
Conventionality 116.63 (10.91) 114.50 (11.72) 116.76 (10.91) .41
Father Deideal. 3.08 (.56)3 2.97 (.81) 2.72 (.61) 11.79°
Father Support 25.14 (8.30)4 26.18 (8.44) 29.81 (6.15) 14.63c
Mother Deideal. 2.61 (.55) 2.84 (.86) 2.60 (.45) 1.88
Mother Support 32.41 (5.47) 28.90 (9.30)5 32.36 (5.05) 3.891’
@f=2,299)
One Alcoholic Parent (n=129); Two Alcoholic Parents (n=22); Non-COAs (n=151)
“p<.05, t’p<.01, cp<.0001

1The mean for Non-COAs signiﬁcantly differs from COAs with one alcoholic parent
and COAs with two alcoholic parents

2 The mean for COAs with two alcoholic parents signiﬁcantly differs from COAs with
one alcoholic parent and non-COAs

3 The mean for COAs with one alcoholic parent signiﬁcantly differs ﬁom non-COAs

" The mean for COAs with one alcoholic parent signiﬁcantly differs from non-COAs

5 The mean for COAs with two alcoholic parents signiﬁcantly differs from COAs with
one alcoholic parent and non-COAs

Father Deideal. = Father Deidealization

Mother Deideal. = Mother Deidealization

72

Post Hoc Analyses (i.e. Scheffe Tests) indicated that there were signiﬁcant
differences on father deidealization between groups, E (2, 299)=1 1.79, p<.05. That is,
COAs with one alcoholic parent deidealized their fathers M=3.08, p<.05) more than
non-COAs (_M=2.72, p<.05). Moreover, signiﬁcant differences existed on paternal
support between groups, F(2, 299)=14.63, p<.05. That is, COAs with one alcoholic
parent (M=25.14, p<.05) perceived that they had less paternal support than non-COAs
(M=29.81, p<.05). There were also signiﬁcant differences on maternal support between
groups F(2, 299)=3.90, p<.05. That is, COAs with two alcoholic parents M=28.90,
p<.05) perceived that they had less maternal support than non-COAs (M=32.36, p<.05)
and COAs with one alcoholic parent (M=32.41, p<.05).

Table 7 shows correlations among all the continuous variables. Correlations
were conducted on each of the three groups (COAs with one alcoholic parent; COAs
with two alcoholic parents, and non-COAs) and a fourth group that consisted of all
COAs regardless of whether there were one or two alcoholic parents. This latter
collapsed COA group was included because there was a signiﬁcant correlation Q = .20,
p<.05) between parentiﬁcation and drug use that did not exist within the other

subgroups.

73

Table 7

Pearson Correlations Among All Continuous Regression Variables in a Sample of College Students

 

 

Groups 1 2 3 4 5 6 7
One Alcoholic Parent (n= 129)

l. Parentiﬁcation .................... . . .

2. Conventionality .................. .017 . ..

3. Mother Support................... .047 .085

4. Father Support .................... -.l65 .051 .243”

5. Mother Deidealization .......... -.081 -.256" -.591" -.015 . .

6. Father Deidealization ............ .155 -.l 13 -.l67 -.653b .177‘

7. Alcohol use ........................ .083 -.39l" .054 .005 -.011 .003 . . .

8. Drug use ........................... .084 -.402" .066 -.039 .066 -.036 .361"
Two Alcoholic Parents (n=22)

l. Parentiﬁcation ..... ‘ ............... . ..

2. Conventionality .................. -.540" . ..

3. Mother Support................... -.348 .222

4. Father Support .................... -.362 .159 .640"

5. Mother Deidealization .......... -.l70 .031 -.57 1" -.228 . ..

6. Father Deidealization ............ -.240 .207 -.251 -.378 .664” . . .

7. Alcohol use ........................ .375 -.780" .063 -.093 -.112 -.048 . . .

8. Drug use ........................... .358 -.651b -.129 -.181 -.254 -.402 .540"
Collapsed COA (n=151)

l. Parentiﬁcation ....................

2. Conventionality .................. -.097 . . .

3. Mother Support................... -.087 .126

4. Father Support .................... -.191' .064 .308"

5. Mother Deidealization .......... -.080 -. 199' -.595" -.051 . . .

6. Father Deidealization ............ .051 -.044 -.174‘ -.595" -.296" . .

7. Alcohol use ........................ .127 -.448" .068 -.014 -.043 -.003 . . .

8. Drug use ........................... .20‘. -.442" -.056 -.O66 -.037 -.18 1' .363"
Non-COA (n=151)

1. Parentiﬁcation .................... . . .

2. Conventionality .................. .055

3. Mother Support ................... -.151 .179‘

4. Father Support .................... -.051 .178' .420" . . .

5. Mother Deidealization .......... -.094 -.266" -.376" -.164‘ .. .

6. Father Deidealization ............ .060 -.l 17 -. 154 -.430" .516" . . .

7. Alcohol use ........................ -.126 -.486" -.045 .017 .090 -.006

8. Drug use ........................... -.120 -.49 l" -.097 -.037 .057 -.037 .434"

 

N=302

‘p<.05 (two tailed), hp<.01 (two tailed)

74

There was a negative relationship between deidealization of father and drug use
Q‘ = -.18, p<.05) among the collapsed COA group. That is, high levels of father
deidealization were correlated with low levels of drug use. Conventionality and alcohol
use had a negative relationship (1; = -.45, p<.05). The same was true of conventionality
and drug use (I = -.44, p<.05). In other words, high levels of conventionality were
strongly associated with low levels of COA alcohol and drug use. There was also a
negative relationship between conventionality and alcohol use (g = -.49, p<.05) and
conventionality and drug use (g = -.49, p <.05) among non-COAs. That is, high levels
of conventionality were strongly correlated with low levels of alcohol and drug use.

Conventionality and alcohol use had a strong, negative relationship (I = -.3 9,
p<.05) and conventionality and drug use (I = -.40, p<.05) had a strong, negative
relationship among COAs with one alcoholic parent. In other words, high levels of
conventionality were strongly associated with low levels of alcohol use and drug use.
Conventionality and alcohol use had even a stronger negative relationship Q‘ = -.78,
p<.05) among COAs with two alcoholic parents. The same was true of drug use (1 = -
.65, p<.05). That is, high levels of conventionality were strongly associated with low

levels of alcohol and drug use among COAs with two alcoholic parents.

75

Test of the Hypotheses

 

Hypotheses for Risk Factors

 

Consistent with predictions, one-way ANOVA results indicated that there were
signiﬁcant differences between COAs and non-COAs on drug use when COAs had two
alcoholic parents E(2, 299)=6.l3. COAs with two alcoholic parents used more drugs
(M=55.82, p<.05) than non-COAs (M=22.34, p<.05), and COAs with one alcoholic
parent (_M=28.40, p<.01). Contrary to predictions, however, there were no signiﬁcant
differences (p<.05) on alcohol use between COAs in any of the subgroups and non-
COAs (see Table 5). Fathers constituted most of the alcoholics within the COA groups. L;

There were 129 alcoholic fathers in the one parent alcoholic group (n=129) and 11 in
the two parent alcoholic group (n=22). Alcoholic mothers were found only in the COA
group that had two alcoholic parents.

Consistent with predictions, there were signiﬁcant differences (see Table 5)
between COAs and non-COAs on parentiﬁcation Ij (2, 299)=12.94 (p<.0001). Post Hoc
Analyses (i.e. Scheffe Tests) indicated that non-COAs had lower levels of
parentiﬁcation (_M=33, p<.05) than both COAs with one parent M=42, p<.05) and two

alcoholic parents (M=51, p<.05). The mean parentiﬁcation score of the non-COA group

 

M=33.5, p<.0001) was comparable to the mean score in Mika, et al.’s (1987) study of

non-clinical, non-COAs (M=32, p<.05).

76

Hypotheses for Protective Factors

 

Rogosch, et al.’s (1990) procedures were used to determine if the potential
protective factor had a mediating or moderating effect on the relationship between
parentiﬁcation and substance use. The procedure to determine if a factor functioned as
a mediator was run on each potential protective factor within each of the three COA
subgroups and the non-COA group. The procedure that determines if a factor has a
moderating effect was also conducted on each potential protective factor if it did not
serve a mediating effect.

Baron and Kenny (1986) state that the independent variable, parentiﬁcation,
must predict both the dependent variable and the potential mediator and the potential
mediator must predict the dependent variable. Rogosch, et al. (1990) point out that a
factor has a mediating effect if “the strength of the relation between the independent and
dependent variable is eliminated (or substantially weakened) when the dependent
variable is regressed on both the independent variable and the potential mediator” (p.
313). Two tasks were performed to test whether the potential protective factors
moderated the relationship between parentiﬁcation. The main effects of parentiﬁcation
and each potential moderating variable were entered in step one and the interaction
between them was entered in step two (see Table 7). The variable was considered a

moderator if there was a signiﬁcant two-way interaction.

77

Parentiﬁcation and Substance Use

 

Regression analyses showed that the collapsed COA group was the only
subgroup where parentiﬁcation signiﬁcantly predicted drug use (beta = .20 p<.05 ).
None of the other regressions within each of the subgroups, including the non-COA
group was signiﬁcant. Hence, the procedures to determine if the potential protective
factors have a mediating effect were only conducted on the collapsed COA group.
Contrary to hypotheses, however, none of the variables mediated the relationship
between parentiﬁcation and low levels of substance use.

Findings from Pearson correlations contradicted predictions that high levels of
parentiﬁcation would be associated with low levels of COA drug and/or alcohol use (see
Table 6). High levels of parentiﬁcation were actually associated with greater levels of
drug use (I = .20, p<.05) among the collapsed COA subgroup. High levels of
parentiﬁcation were not associated with high levels of drug Q“ = -. 12, p>.05) or alcohol
use (1; = -.l3, p>.05) among non-COAs. In fact, there was not a signiﬁcant direct effect

between alcohol use and parentiﬁcation within any of the subgroups.

Conventionality as a Mediator Variable

 

Parentiﬁcation predicted conventionality (beta = .20, p<. 05) within the
collapsed COA group. Hence, the ﬁrst set of Rogosch, et al.’s (1990) conditions for

demonstrating that conventionality is a mediating variable was met.

Idealization of a Non-alcoholic Parent as a Mediator

Parentiﬁcation did not predict idealization of mother (beta = -.080, p>.05)

78

within the collapsed COA group. Hence, the fnst set of Rogosch, et al.’s (1990)
conditions for demonstrating that idealization of a non-alcoholic parent (i.e. mother in
this sample) as a mediating variable was not met. However, the remaining two

conditions were not met.

Deidealization of an Alcoholic Parent as a Mediator

 

Parentiﬁcation did not predict deidealization of father (beta = .051, p>.05)
within the collapsed COA group. Hence, the fust set of Rogosch, et al.’s (1990)
conditions for demonstrating that deidealization of an alcoholic parent (i.e. father in this

sample) as a mediating variable was not met.

High Support From a Non-Alcoholic Parent as a Mediator Variable

 

Parentiﬁcation did not predict high mother support (beta = -.087, p>.05) within
the collapsed COA group. Hence, the ﬁrst set of Rogosch, et al.’s (1990) conditions for
demonstrating that high levels of support from a non- alcoholic parent (i.e. mother in

this sample) as a mediating variable were not met.

Low Support From an Alcoholic Parent as a Mediator Variable

 

Parentiﬁcation predicted low father support (beta = -. 191, p<.05) within the
collapsed COA group. However, low father support did not predict drug use (beta = -
.066, p>.05). Hence, the ﬁrst set of Rogosch, et al.’s (1990) conditions for
demonstrating that low levels of support from an alcoholic parent (i.e. father in this

sample) as a mediating variable were not met.

79

Deidealization of Father and Conventionality as Moderating Variables

Two regressions were performed on each potential protective variable to
determine if any of them moderated parentiﬁcation and substance use. High levels of
conventionality and deidealization of father were the only potential protective variables
that had a moderating effect and this relationship was only signiﬁcant within the
collapsed COA group. These variables did not function as moderating variables among
COAs with one alcoholic parent, COAs with two alcoholic parents, or non-COAs.
Beta coefﬁcients derived from regression analyses on parentiﬁcation, conventionality
and drug use are shown in Table 8. Table 9 shows coefﬁcients derived from regression

analysis on parentiﬁcation, father deidealization and drug use.

Table 8

Regressions Predicting Drug use From Parentiﬁcation, Conventionality, and Their
Interaction Among COAs With One Alcoholic Parent and Two Alcoholic Parents

 

Dependent Measure

 

 

 

D_rug_U_se

Predictors R2 change B
Step 1

Main effects .30c

Conventionality -.385°

Parentiﬁcation . 1 74"
Step 2

2-way interaction .06c

parentiﬁcation x

Conventionality -.266c

n=1 5 1
“p<.05, cp<.0001

The tabled values for betas were computed at the step that the variable was entered into
the equation.

80

Table 9

Regressions Predicting Drug use From Parentiﬁcation, Father Deidealization, and Their
Interaction Among COAs With One Alcoholic Parent and Two Alcoholic Parents

 

 

 

Dependent Measure

 

 

 

Drug Use

Predictors R2 change [3
Step 1

Main effects .04"

Father deidealization .249b

Parentiﬁcation -. 102
Step 2

2-way interaction .068

Parentiﬁcation x

father deidealization -.216a

ap<.05, bp<.01
N=151

The tabled values for betas were computed at the step that the variable was entered into
the equation.

Simple regression lines were plotted on high and low values (one standard
deviation above and below the mean) of conventionality (Aiken & West, 1991) and
were used to explore the nature of the interaction. Figure 2 shows that there is a strong,
positive relationship between parentiﬁcation and drug use among those who report low
levels of conventionality. That is, higher levels of parentiﬁcation are associated with
higher levels of drug use when COAs report low levels of conventionality. However,
there is a slight negative relationship between parentiﬁcation and drug use when COAs
report high levels of conventionality. In other words, higher levels of parentiﬁcation are
associated with lower levels of drug use when COAs report high levels of

conventionality.

81

Figure 2

Signiﬁcant Relations Between Parentiﬁcation and Drug Use as Moderated by
Conventionality and Deidealization of Alcoholic Father Among COAs With One
Alcoholic Parent and Two Alcoholic Parents

 

 

 

 

 

 

 

 

 

 

Father Deidealization Parent and Conventionality
1.0 1.0
0.8 _ 0.8 _
Low Deidealization
0-5 — \ 0-6 — Low Convauionality
8 03 — 3 0.2 -—
D 0.0 _. :3 0.0 __
on '03 -— 1 no -02 ——
S -0-4 — E -04 — W
-0.6 ._ High Deidealization _06 __ I
08 — 0'8 _ HighConventiomlity
-1.0 l I l -10
-1 0 1 -1 0 1
Parentrﬁcatron Parentiﬁcation
N=151

A similar ﬁnding emerged in the analyses involving COA deidealization of
father and parentiﬁcation (see Figure 2). There is a strong, positive relationship
between parentiﬁcation and drug use among COAs who report low levels of father
deidealization (i.e. father is idealized). That is, higher levels of parentiﬁcation are
associated with higher levels of drug use when COAs report low levels of father
deidealization. However, there is essentially no relationship between parentiﬁcation
and drug use when COAs report high levels of deidealization. In other words, higher
levels of parentiﬁcation are associated with higher levels of drug use when COAs report
low levels of father deidealization, but there is essentially no relationship between

parentiﬁcation and drug use when COAs report low levels of father deidealization.

82

CHAPTER FOUR

Discussion

A series of regression analyses were completed to investigate whether high
levels of parentiﬁcation and low levels of drug and/or alcohol use would be mediated by
high levels of conventionality, alcoholic parent deidealization and low support, and
non-alcoholic parent idealization and high support. None of these potential protective
factors had a mediating effect on the relationship between parentiﬁcation and drug or
alcohol use within a primarily female sample. Alcoholic parent deidealization (i.e.
father deidealization) and high levels of conventionality were the only potential
protective factors that had a moderating effect on substance use and this only existed
within the collapsed COA group that were not at any more risk for using drugs or
alcohol than the non-COA group.

There was some limited support for a probabilistic view of development
(Fitzgerald, et al. 1994) that has multiple developmental trajectories with multiple
outcomes because a family history of alcoholism did not adequately predict that
offspring would perpetuate high levels of alcohol consumption as deﬁned by Mayer and
Filstead (1979). Consistent with some of the literature on female COAs (Mitz, et al.,
1995; Benson & Heller, 1987; Harburg, et al., 1982), all the female COAs in this sample
with one alcoholic parent (i.e. father) were not at signiﬁcantly more risk for using
alcohol than non-COAs. However, COAs with two alcoholic parents were at
signiﬁcantly higher risk for drug use. According to Sher (1991), not all COAs possess a

level of vulnerability that predisposes them to develop problems with substance use.

83

Hence, it cannot be assumed that protective processes are responsible for the
insigniﬁcant results on substance use levels between COAs with one alcoholic parent
and non-COAs.

Despite the fact that there were no data to support the mediating hypotheses,
other ﬁndings from this study are helpful in starting to identify some of the processes
underlying a developmental pathway leading to low levels of COA substance use
(Fitzgerald et al., 1994). Data ﬁom correlations and one-way ANOVAS indicated that
female COAs with one alcoholic parent and non-COAs have similar substance use
levels, but differ in the levels of parentiﬁcation, conventionality, parental support, and
emotional autonomy. Consistent with Brook et al.’s (1990) ﬁndings, conventionality
had the strongest correlation with low levels of drug and alcohol use across all groups
within the sample. Perhaps the most important ﬁnding of this study was the complex,
moderating role that conventionality appeared to play in low levels of drug use among
highly parentiﬁed COAs. Goglia et al.'s (1992) ﬁndings that females COAs are more
parentiﬁed than males were replicated in this study. In fact, all the females in COA
subgroups within this study scored signiﬁcantly higher on parentiﬁcation than females
that were not COAs. Contrary to predictions, however, parentiﬁcation alone actually
functioned as a risk factor that was related to higher levels of drug use within the
collapsed COA group. Conventionality, the most signiﬁcant moderating variable, was
strongly correlated with low levels of drug and alcohol use across all subgroups within
the sample and appeared to reverse or neutralize the relationship between parentiﬁcation
and heightened drug use within the collapsed COA group (see Figure 2).

The relationship between parentiﬁcation and conventionality indicate that

84

parentiﬁcation, alone, cannot be expected to protect adolescents from using high levels
of substances. Instead the inﬂuence of parentiﬁcation on low levels of COA substance
use appeared to be moderated through caregiving responsibilities that probably
reinforced conventional belief systems (Weiss, 1979; Wilson & Orford, 1978; Nardi;
1979; Whiting & Whiting, 1975). Conventional belief systems may inhibit the .
development of conduct-disordered behaviors and antisocial personality traits that
Zucker, et a1. (1996) found to be associated with substance abuse in adulthood. It is
also possible that conventionality moderated the impact of parentiﬁcation on drug use
because greater care-giving responsibility towards others increased the likelihood
offspring would emulate a myriad of pro-social behavior from their non-alcoholic
mothers, including low levels of substance use. Parentiﬁcation alone appears to play a
protective role only in a context that promotes conventional behavior and belief
systems. Another interpretation is that parentiﬁed COAs who used drugs might have
had care-giving tasks that fell outside the realm of their developmental capabilities.
Increased stress associated with roles as a caregiver and the absence of conventional
belief systems, coupled with a family culture that normalizes high levels of substance
use may have predisposed these COAs to non-conventional means of alleviating stress,
such as drug use.

Deidealization of alcoholic fathers, in addition to parentiﬁcation and
conventionality, may be another piece of the causal chain within a developmental
pathway that leads to low levels of drug use. Although idealization of an alcoholic
father among parentiﬁed COAs appears to predict drug use, the reverse, deidealization

of father, was not strong enough to predict low drug use. In other words, deidealization

85

appears to only be a part of the causal chain leading to decreased drug use. Idealization
of an alcoholic father within the collapsed COA group was associated with higher drug
use, presumably indicating that father’s drinking behavior strongly inﬂuenced
daughter’s choice to use high levels of drugs because of their perceptions of parental
omnipotence, and inability to gradually relinquish childhood dependency and begin
viewing their fathers more and more as fallible people. Interestingly, father
deidealization had an interaction effect with parentiﬁcation to serve a moderating effect
on drug use. However, the effect on drug use was fairly weak (see Figure 2).

The strong, positive correlation between father deidealization and parental
support may provide some limited support for Sarason, et al.’s (1983) perspective that
an individual needs to be perceived as a credible source of support before one can feel
conﬁdent enough to access them as a resource. Findings from correlations and one-way
ANOVAS indicated that COAs with one alcoholic parent were more likely to deidealize
their fathers and receive less support from them. From a contextual family therapy
perspective (Boszormenyi-Nagy, et al., 1991), the “one-up” position associated with
caregiving responsibilities and providing support to others may make it difﬁcult for
parentiﬁed children to be on the receiving end of support from their alcoholic fathers.
The ﬁnding that COAs with one alcoholic parent (i.e. the father) reported low levels of
support from their fathers is not surprising when viewed ﬁom this perspective. That is,
daughters may not have viewed alcoholic fathers as a credible source of support because
they viewed them as more fallible and impotent than non-COAs viewed their fathers.

Interestingly, low levels of support ﬁom alcoholic fathers and insignificant levels

of maternal support (when compared to non-COAs) did not place COAs ﬁ'om the

86

collapsed group at higher risk for using drugs or alcohol. The most parsimonious
interpretation of this ﬁnding is that parental support ﬁ'om a non-alcoholic parent has no
relationship to drug use levels among female COAs. On the other hand, this could be
evidence in favor of prior research, which has found that the context of emotional
support (Andrews, Hops, & Duncan, 1997) and emotional autonomy (Fuhrman and
Holmbeck, 1995) can inﬂuence whether or not close parent-adolescent relationships
serve detrimental or protective functions. That is, the level of maternal support within
non-alcoholic families might have been more adaptive or protective for the COAs in this
sample and extremely high levels of support might have been a detriment. For example,
moderate levels of support from non-alcoholic mothers may have helped daughters
maintain a connection with their mother that was ﬂexible enough to receive support
from non-substance abusing individuals outside the family.

The ﬁndings of this study shed light on the intimate relationship between
protective inﬂuences and risk, maladaptation, and psychopathology (Cicchetti &
Garmezy, 1993). COAs who had two alcoholic parents were at highest risk for using
drugs, received the least amount of support ﬁ'om their mothers, did not differ
signiﬁcantly from the other groups on father support, and had fathers who had the least
amount of formal education. The fact that this group had more risk variables than the
others makes their self-reports of low alcohol use and self-reports of high drug use a
provocative ﬁnding that could have many implications for protective factors that have
not been identiﬁed in this study. Their increased use of drugs can be interpreted in
different ways based on the theoretical perspective being used. It could mean that

spousal parentiﬁcation (Mika, et al., 1987) is more prevalent with both parents because

87

she becomes the “parent” on a pragmatic day-to-day basis. As a result, she may not be
able to form stable attachments with others (Ainsworth, Blehar, Waters, & Wall, 197 8),
acquires anti-social personality traits, and abuses substances to exonerate her substance-
abusing parents and indulge in a path of destructive entitlement (Boszormeny-Nagy &
Spark, 1984). On the other hand, her heightened drug use could be the result of social
learning (Bandura & Walters, 1963).

A developmental systems approach (e. g. Fitzgerald et al., 1994; Staudinger,
Marsiske & Baltes, 1995; Ford & Lerner, 1992) to this study has attempted to extend
beyond individual processes to span multiple levels of analysis of protective factors to
help begin unraveling a developmental trajectory that leads to low levels of substance
use. The ﬁndings presented here have started to uncover a developmental trajectory that
protects daughters from a family history of alcoholism. An individuation process that is
characterized by high levels of parentiﬁcation, deidealization of an alcoholic father, and
high levels of conventionality, may be adaptive for female COAs between the ages of
18 and 24. It is tempting to conclude that a seemingly distant relationship with an
alcoholic father (i.e. highly deidealized with low levels of support) is more adaptive for
parentiﬁed daughters if high levels of conventionality exist and the outcome leads to

low levels of drug and alcohol use.

Limitations
The failure to support a mediational model and the limited support on a
moderating model may be partly due to the methodological limitations of this study.

The ﬁndings of a single, cross-sectional, retrospective study utilizing self-report

88

questionnaires cannot be used to deﬁnitely rule out a mediational or moderating model
for protective factors (Rogosch, et al. (1990). Moreover, COA status is based on the
participants’ self report of their parent’s alcohol abuse or addiction and does not reﬂect
any diagnostic categorization of parents. Although these limitations are common to
most large, survey studies, they strongly suggest caution in interpreting the ﬁndings.
Some of the major methodological limitations of this study include the absence of
measures that tap into the biological and genetic aspects of addiction and the inability to
isolate subtypes of alcoholism and comorbidity of alcoholism and other mental health
diagnoses (Zucker, et al., 1996). Zucker et al., (1996) point out that alcoholism is often
accompanied by depression in women and antisocial personality traits in men. Each of
these disorders would greatly inﬂuence the context of risk and are likely to have an
inﬂuence on offspring substance use levels, the degree to which COAs perceive a parent
to be a credible and reliable source of support, and level of behavioral adjustment.

The ﬁndings of this study are not generalizable to diverse samples, because the
majority of the population was college-age, single, childless, Caucasian females who
had parents with some college education. Hence, the ﬁndings of this study may not
hold up across different individual, familial, and cultural contexts. Furthermore, this
sample was obtained from psychology courses only, which are notorious for having
more females than males. Perhaps, more females who go into psychology have
alcoholic parents, or highly parentiﬁed females are more attracted to psychology. It is
impossible to compare parentiﬁcation levels between males and females in this sample
and the differences that might exist within the four sub-scales on parentiﬁcation. For

example, females might be more parentiﬁed on the spousal sub-scale and males might

89

 

be more parentiﬁed on the sibling sub-scale. More stress may be associated with
spousal parentiﬁcation and inﬂuence levels of drug and alcohol consumption. Although
this was not true of the females in this sample, it may be a trend among male COAs.

Another limitation of the study may be in the measures that were used to assess
parentiﬁcation. Parentiﬁcation is a complex concept with some substantial differences
in its meaning as an adaptive versus pathological construct among theories in the family
therapy literature (J urkovick, 1997). However, there are only two published measures
that existed at the time this study was conducted. The scale used in this study (i.e. Mika
et al., 1987) did not tap into the relational ethics dimension of parentiﬁcation, which is a
primary focus of contextual family therapy. For example, it is impossible to know
where parentiﬁed COAs fall on a continuum of entitlement and how this, in turn, might
impact levels of conventionality and substance use. Items from the Therapist Action
Index (Bemal, Flores-Ortiz, Rodriguez, Sorenson, & Diamond, 1990) and the Relational
Ethics Scale (Hargrave & Bomba, 1993) could be used in future studies to assess the
degree of destructive or constructive entitlement among COAs. Such measures might
help identify the reasons why high levels of parentiﬁcation reached maladaptive levels
and were correlated with increased drug use.

Perhaps the association between high levels of parentiﬁcation and high levels of
drug use among COAs within the collapsed group indicates that the scale used to assess
support in this study did not adequately measure the type of familial support needed
when a child is parentiﬁed. The contextual family therapy literature indicates that the
tasks of a parentiﬁed child need to be within their developmental capabilities and the

role needs validation ﬁom the family to result in constructive entitlement. Although

90

children can be exploited through extreme parentiﬁcation, the development of healthy
self-esteem and self-efﬁcacy is intimately tied to the degree that children are provided
the opportunity to give back to their parents (Goldenthal, 1992). A scale that is
speciﬁcally developed using contextual family therapy’s perspective on parentiﬁcation
and measures that assess behavioral adjustment and psychological stress would help
explore adaptive forms of caregiving and the impact that it might have on substance use.
Finally, more studies are needed to differentiate between parental drug and
alcohol dependency, and COA abstinence, substance use, substance abuse and
dependency. Perhaps drug use was more signiﬁcant than alcohol use because the scale
used to measure drugs was far more comprehensive than the scale that measured
alcohol. It is not possible to know if mothers from families with two alcoholic parents
used more drugs than alcohol and if daughters emulated this behavior because the
CAST did not differentiate between drug and alcohol addicted parents. Hence, there is
no empirical evidence to understand why female COAs of two alcoholic parents scored

higher on drug use, but were comparable to non-COAs on their alcohol use

Directions for Future Studies

 

Future studies that have a signiﬁcant number of COAs who use substances more
than non-COAs are needed to increase conﬁdence that moderating and mediating
functions have protective utility. Studies that have sub-groups of substance abusing
COAs, non-problematic substance using COAs, abstaining COAs and non-COAs would
provide solid data to identify protective factors and between and within group

differences on alcohol versus drug consumption. Sub-groups would also need to

91

represent both male and female COAs with a more diverse racial and cultural
background. Diversity in this sample would help provide data to determine if
parentiﬁcation remains a risk variable when gender, race and culture is controlled.
Culture may strongly inﬂuence whether or not parentiﬁcation, alone, is a protective
variable. For example, parentiﬁcation is associated with less taboo in Aﬁican
American, Latino, and Hispanic cultures (Boyd-Franklin, 1989) and may be associated
with less stress. Hence, parentiﬁcation in such a culture may also be associated with
more conventionality and less drug and alcohol use and abuse.

Future studies that have samples consisting of both males and females could
explore if protective processes within female socialization play a role in insulating
females ﬁ'om the risk of substance use and abuse. In their extensive review, Weisner
and Gallirnore (1977) found that the majority of cultures assign caregiving
responsibilities to girls. Gilligan (1982) suggests that females mature and individuate
differently from males because female socialization and individuation involves
interpersonal caring, autonomy, and self-care. Longitudinal studies that utilize a
developmental systems approach can provide data that helps unravel the bi-directional
inﬂuences that peers, parents, and a patriarchal society have on guiding girls through an
individuation process that is characterized by conventional, parentiﬁed behavior. Data
from future studies may suggest that the individuation process of females is accelerated
beyond that of males and results in substance use levels that are more comparable to
non-COAs only when there is one alcoholic parent and there is access to credible
sources of support. Even when alcohol problems occur among girls, the process appears

to be very different ﬁ'om boys. Preliminary results of Fitzgerald, Zucker, Puttler,

92

Caplan, and Mun’s (2000) study suggest that the socialization process of females and
recent increases in alcohol use among girls may result in more alcoholism subtypes than
for males. The age that these subtypes emerge is also different for females.

More studies, with larger sample sizes, are needed to provide data on the
reciprocal effects that biology and environmental variables have on one another among
females growing up with two alcoholic parents. Future studies could investigate if
female COAs need to be at considerable higher risk than male COAs before they begin
using substances at a level that differs from non-COAs. Fitzgerald, Zucker, Puttler,
Caplan, & Mun (2000) have found that the bi-directional inﬂuences of biology and
environment intensify when females have two biological parents that are alcoholic.
More speciﬁcally, their preliminary data suggests that girls who have an alcoholic
mother during early school-age years have more externalizing behaviors than girls who
only have an alcoholic father. Other studies (e. g. Drake & Vaillant, 1988) indicate that
neglect, poor attachments to parents, and parental conﬂict have been some of the
environmental risk factors that are present in the family histories of adult, female
alcoholics.

Preliminary data ﬁ'om longitudinal studies that utilize a developmental systems
approach strongly suggest that the pathways to alcoholism for women may be more
varied and complex than men and often involve depression as part of the causal
structure (Fitzgerald, et al., 2000). The alcoholic women in Fitzgerald et al.’s study had
the densest family history of alcoholism among three subtypes of alcoholism, exhibited
anti-social personality traits and had histories of childhood conduct problems.

Fitzgerald et a1. hypothesize that girls with alcoholic mothers are at more risk than girls

93

with alcoholic fathers for comorbid depression because the mothers in the former group
have fewer resources to have a protective function. Although it was not measured, the
females in the current study who grew up in households with two alcoholic parents
might have been at increased risk for depression.

Research is needed to replicate the ﬁndings in this study to determine if girls are
at higher risk for using drugs over alcohol. Drug or alcohol sensitive strategies could be
implemented in prevention programs that target girls and boys in early adolescence if
females are more likely to use drugs and males are more likely to use alcohol. Trends in
gender speciﬁc drug and alcohol abuse could also mean that clinicians need to be more
thorough in their assessments for drug use among females.

Future studies are also needed to explore the effects of conventionality to help
determine if other contextual variables moderate or mediate its relationship with low
levels of substance use among COAs. For example, strong relationships with
conventional peers might strengthen, or moderate, the relationship between COA
conventionality and low levels of substance use. It is also important to identify the
antecedents of conventionality. Data from future studies might discover that
conventionality is a necessary part of deidealizing an alcoholic parent.

Finally, longitudinal data are needed to help assess how levels of familial
connectedness with alcoholic families might mirror developmental needs as adolescents
move into the early, middle and adult years. Optimal levels of support might help
COAs be connected to the family without destructively individuating during adolescents
and explore how connection to the family might need to change as developmental needs

for autonomy change throughout the life cycle. Although high levels of parentiﬁcation

94

might result in an increase in self-efﬁcacy and independence through caring for others
during adolescence, (Weiss, 1979; Wilson & Orford, 1978; Nardi; 1979; Whiting &
Whiting, 1975; Boyd-Franklin, 1989), it may be maladaptive as they enter adulthood

and begin to marry and have children.

95

LIST OF REFERENCES

Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting
interactions. Newbury Park, CA: Sage.

 

Ainsworth, M. D. (1967). Infancy in Uganda. Baltimore: Johns Hopkins Press.
Anderson, E., & Quast, W. (1983). Young children in alcoholic families: A mental
health needs-assessment and an intervention/prevention strategy. Journal of Primary
Prevention, 3, 174-187.

 

 

Ainsworth, M. D., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of
Attachment. New York: Basic Books.

yr“
.1

Andrews, J. A., Hops, H., & Duncan, S. C. (1997). Adolescent modeling of
parent substance use: The moderating effect of the relationship with parent. Journal of
Family Psychology, 11(3), 259-270.

11‘
I:1

 

Baer, P. E., Garmezy, L. B., McLaughlin, R. J ., Pokomy, A. D., & Wernick, M.
J. (1987). Stress, coping, family conﬂict, adolescent alcohol use. Journal of Behavioral
Medicine, 10(5), 449-466.

 

Bandura, A. & Walters, R. H. (1963). Social Learning and Personality
Development. New York: Holt, Rinehart, & Wintson.

 

 

Barnard, C. P. & Spoentgen, P. A. (1987). Children of alcoholics:
Characteristics and Treatment. Alcoholism Treatment Quarterly, 3(4), 47-65.

 

Barnard, C. P. (1994). Resiliency: A shift in our perception? The American
Journal of Family Therapy, 22(2), 135-144.

 

Baron, R. A., & Kenny, D. A. (1986). The moderator-mediator variable
distinction in social psychological research: Conceputual, strategic, and statistical
considerations. Journal of Personalig and Social Psychology, 51, 1173-1182.

 

Beardslee, W. R & Podorefsky, D. (1988). Resilient adolescents whose parents
have serious affective and other psychiatric disorders: Importance of self-understanding
and relationships. American Journal of Psychiatry, 145(1), 63-69.

Bennett, L. A., Wolin, S. J., Reiss, D. & Teitelbaum, M. A. (1987). Couples at
risk for transmission of alcoholism: Protective inﬂuences. Family Process, 26, 111-
129.

 

96

Benson, C. S., & Heller, K. (1987). Factors in the current adjustment of young
adult daughters of alcoholic and problem drinking fathers. Journal of Abnormal
Psychology, 96(4), 305-312.

 

Berlin, R., Davis, R. B. & Orenstein, A. (1988). Adaptive and reactive
distancing among adolescents from alcoholic families. Adolescence, Vol. XXII(91),
577-5 84. '

 

Berkowitz, A. D., & Perkins, W., H. (1987). Recent research on gender
differences in collegiate alcohol use. College Health, 36, 123-129.

 

Berna], G., F lores-Ortiz, Y., Rodriguez, C., Sorenson, J ., & Diamond, G. (1990).
Development of a contextual family therapy therapist action index. Journal of Family
Psychology, 3(3), 332-331.

 

Black, C. (1982). I Will Never Happen to Me. Denver Colo.: M.A.C. Printing
and Publications.

Black, C. Bucky, S. F., & Wilder-Padilla, S. (1986). The interpersonal and
emotional consequences of being an adult child of an alcoholic. The International
Journal of Addictions, 21, 213-231.

 

 

Blanton, H., Gibbons, F. X., Gerrard, M., Conger, K. J ., & Smith (1997). Role
of family and peers in the development of prototypes associated with substance use.
The Journal of Family Psychology, 11(3), 271-288.

 

Blos, P. (1979). The Adolescent Passgge: Developmental Issues. New York:
lntemational Universities Press.

 

Bohman, M., Sigvardsson, S., & Cloninger, R. C. (1981). Maternal inheritance
of alcohol abuse. Archives of General Psychiatry, 38, 965-755.

 

Bowlby, J. (1980). Attachment and Loss. Vol III: Loss: Sadness and Depression.
New York: Basic Books.

Boszormenyi-Nagy, l & Spark, G. M. (1984). Invisible Loyalties: Reciprocity in
Intergenerational Family Therapy. New York: Brunner/Mazel.

 

 

Boszormenyi-Nagy, 1., Gnmebaum, J. & Ulrich, D. (1991). Contextual therapy.
In A. Gurman & D. P. Kniskem (Eds.),Handbook of Family Therapy (Vol. H PP. 200-
23 8) New York: Brunner/Mazel.

 

Boyd-Franklin, N. (1989). Black Families in Therapy: A Multisystems
Approach. New York: Guilford Press.

 

97

Boyd-Franklin, N. (1989). Black Families in Therapy: A Multisystems
Approach. New York: Guilford Press.

Braithwaite, V., & Devine, C. (1993). Life satisfaction and adjustment of
children of alcoholics: The effects of parental drinking, family disorganization and
survival roles. British Journal of Clinical Psychology, 32, 417-429.

Brook, J. S., Whiteman, M., Gordon, A. S., & Brenden, C. (1983). Older
brother's inﬂuence on younger sibling's drug use. The Journal of Psychology, 114, 83-
90.

Brook, J. S., Whiteman, M., Gordon, A. S., Brook, D. W. (1986). Father-
daughter identiﬁcation and its impact on her personality and drug use. Developmental
Psychology, 22(6), 743-748.

Brook, J. S. Brook, D. W. Gordon, A. S, Whiteman, M. & Cohen, P. (1990).
The psychosocial etiology of adolescent drug use: A family interactional approach.
Genetic, Social, and General Psychology Monographs, 116(2), 111-267.

I}
1.
L

Brown, S. (1988). Treating Children of Alcoholics: A Develomnental
Perspective. New York: John Wiley & Sons.

Burk, J. P., & Sher, K. J. (1988). The "forgotten children" revisited: Neglected
areas of COA research. Clinical Psychology Review, 8 285-302.

Callan, V. J ., & Jackson, D. (1986). Children of alcoholic fathers and recovered
alcoholic fathers: Personal and family functioning. Journal of Studies on Alcohol,
47(2), 180-182.

Cantwell, D. (1975). F amilial-genetic research with hyperactive
children. In D. Cantwell (Ed.), The Hyperactive Child: Diagnosis, Management, and
Current Research (pp. 93-105). New York: Spectrum.

 

 

Cermak, T. L. (1986). Diagnosingand Treating Co-dependence. Minneapolis:
Johnson Institute Books.

Chassin, L. Rogosch, F., & Barrera, M. (1991). Substance use and
symptomatology among adolescent children of alcoholics. Journal of Abnormal
Psychology, 100(4), 449-463. ‘

Chassin, L., Curran, P. J ., Hussong, A. M., Colder, C. R. (1996). The relation of

parent alcoholism to adolescent substance use: A longitudinal follow-up study. Journal
of Abnormal Psychology, 105(1), 70-80.

98

Children of Alcoholics Foundation Inc. (1992). Report of theForum on
Protective Factors, Resiliency, and Vulnerable Children. New York.

 

 

Cicchetti, D. & Garmezy, N. (1993). Prospects and promises in the study of
resilience. Development and Psychopathology, 5, 497-502.

 

Clair, D., & Genest, M. (1987). Variables associated with the adjustment of
offspring of alcoholic fathers. Journal of Studies on Alcohol, 48, 345-355.

 

Colder, C. R. Chassin, L., Stice, E. M., & Curran, P. J. (1997). Alcohol
expectancies as potential mediators of parent alcoholism effects on the development of
adolescent heavy drinking. Journal of Research on Adolescence, 7(4), 349-374.

 

Cohen, J. (1988). Statistical Power Analysis for the Behavioral Sciences. San
Diego, CA: Acadenric Press.

 

Cohen, J. (1992). A power primer. Psychological Bulletin, 112(1), 155-159.

 

Cohler, B. J ., Stott, F. M., & Musick, J. S. (1995). Adversity, vulnerability, and
resilience: Cultural and developmental perspectives (pp. 753-800). In D. Cicchetti and
D. Cohen (Eds.) Developmental Psychopathology. Volume 2: Risk, Disorder, and
Adaptation. New York: Wiley & Sons.

 

Cotton, N. S. (1979). The familial incidence of alcoholism: A review. Jouma_l
of Studies on Alcohol, 40, 89-116.

 

Davis, R, Stern, D. & Vandeusen, J. (1978). Enmeshment-disengagement in the
alcoholic family. In F. A. Seixas (Ed.) Currents in Alcoholism (Vol. 4, pp. 15-27). New
York: Grune & Stratton.

 

Devine, C., & Braithwaite, V. (1992). The survival roles of children of
alcoholics: Their measurement and validity. British Journal of Addiction, 88, 69-78.

 

Deutsch, C., DiCicco, L., & Mills, D. J. (1982). Services for children of
alcoholic parents. In Alcohol and Health Monograph No. 3: Prevention, Intervention,
and Treatment: Concerns and Models. Rockville, MD: NIAAA

 

 

Deutsch, C. (1982). Broken Bottles, Broken Dreams. Teachers College Press:
New York.

 

Drake, R. E., & Valiant, G. E. (1988). Predicting alcoholism and personality
disorder in a 33-year longitudinal study of children of alcoholics. British Journal of
Addiction, 83, 799-807.

 

Dumka, L. E., & Roosa, M. W. (1993). Factors mediating problem drinking and
mothers' personal adjustment. Journal of Fanrily Psychology, 7(3), 333-343.

 

99

Egeland, B., Carlson, E., & Sroufe, L. A. (1993). Resilience asprocess.
Development and Psychopathology, 5, 517-528.

 

el-Guebaly, N., & Offord, D. (1979). On being the offspring of an alcoholic:
An update. Alcoholism: Clinical and Experimental Research, 3, 148-157.

 

Erickson, E. (1963). Childhood and Society (2nd ed.). New York: Norton.

 

F ergusson, D. M., & Lynskey, M. T. (1996). Adolescent resiliency to family
adversity. Journal of Child Psychiatry, 37(3), 281-292.

 

F ilstead, W., McElﬁ'esh, O. & Anderson, C. (1981). Comparing the family
environments of alcoholic and "normal" families. Journal of Alcohol and Drug
Education, 26, 24-31.

 

Fingarette, H. (1988). Heavy Drinking: They Myth of Alcoholism as a Disease.
Berkeley: University of California Press.

Fitzgerald, H. E., Davies, H. W., Zucker, R. A., & Klinger, M. (1994).
Developmental systems theory and substance abuse: A conceptual and methodological
ﬁamework for analyzing patters of variation in families. In L. L’Abante (Ed.) Handbook
of Developmental Family Psychology and Psychopathology. New York: Wiley & Sons,
Inc.

 

Ford & Lerner (1992). Developmental Systems Theory: An Integrative
Approach (pp. 27-90). Newbury: Sage Publications.

Frank, S. J ., Pirsh, L. A., & Wright, V. C. (1988). Late adolescents' perceptions
of their relationships with their parents: Relationships among deidealization, autonomy,
relatedness, and insecurity and implications for adolescent adjustment and ego identity
status. Joru'nal of Youth and Adolescence, 19(6), 571-588.

Frank, S. J ., & Burke, L. (1992). Deidealization and autonomy in late
adolescence: Replications and extensions of earlier ﬁndings. In S. Frank (Chair),
Adolescent Autonomy: Is It All It's Cracked Up To Be? Symposium conducted at the
meeting of the Society for Research in Adolescence, Washington, D. C.

 

Furman, T., & Holmbeck, G. N. (1995). A contextual-moderator analysis of
emotional autonomy and adjustment in adolescence. Child Development, 66, 793-811.

Garmezy, N. (1985). Stress-resistant children: The search for protective
factors. In J .E. Stevenson (Ed), Recent Research in Developmental Psychopathology.
Oxford: Pergamon Press, pp. 213-233 (a book supplement to the Journal of Child
Psychology and Psychiatry, Number 4).

100

Gﬁoerer, J. (1985). Inﬂuence of privacy on self-reported drug use by youths. In
Rouse, B. A., Kozel, N. J ., & Richards, L. G. (Eds) Self-Report Methods of Estimating
Drug Use: Meeting Current Challenges to Validity. NIDA Research Monograph No. 57,
DHHS Publication No. (ADM) , pp. 22-30, Washington: Government Printing Office.

 

 

Gilligan, C. (1982). In a Different Voice. Cambridge, MA: Harvard University

 

Press.

Gittelman, R. S., Mannuzza, R. S., & Bonagura, N. (1985). Hyperactive boys
almost grown up: Psychiatric status. Archives of General Psychiatry, 42, 937-947.

 

Goglia, L. R., Jurkovic, G. J ., Burt, A. M., & Burge-Callaway, K. G. (1992).
Generational boundary distortions by adult children of alcoholics: Child-as-parent and
child-as-mate. The American Journal of Family Therapy, 20(4), 291-299.

 

Goldenthal, P. (1993). Contextual Family Therapy: Assessment and
Intervention Procedures. Sarasota, FL: Professional Resource Press.

 

Gottlieb, G. (1991). Experiential canalization of behavioral development:
Theory. Developmental Psychology, 27, 4-13.

 

Goodwin, D. W., Schulsinger, F., Knop, J ., Mednick, S., & Guze, S. B. (1977).
Alcoholism and depression in adopted out daughters of alcoholics. Archives of General
Psychiapy, 34, 751-755.

 

Gough, H. G. (1957). The California Psychological Inventory. Palo Alto, CA:
Consulting Psychologists Press.

 

Grotevant, H. & Cooper, C. (1986). Individuation in family relationships: A
perspective on individual differences in the development of identity and role-taking in
adolescence. Human Development, 29, 82-100.

 

Harburg, E., Davis, D. R., & Caplan, R. (1982). Parent and offspring alcohol
use: Imitative and aversive transmission. Journal of Studies on Alcohol, 43, 497-516.

 

Harburg, E., DiFranceisco, W., Webster, D. W., Gleiberman, L., & Schork, A.
(1990). Familial Transmission of alcohol use: 11. Imitation of and aversion to parent
drinking (1960) by adult offspring (1977)-Tecumseh, Michigan. Journal of Studies on
A_l_cohol, 51(3), 245-256.

Hargrave, T. D., & Anderson, W. (1990). Helping older people ﬁnish well: A
contextual family therapy approach. Familylherapy, XVII, 1, 9-19.

 

Hargrave, T. D., & Bomba, A. K. (1993). Further validation of the Relational
Ethics Scale. Journal of Marital and Family Therapy, 19, 292-299.

 

101

31m
575

for
lm

1111

l "b

l—I-t

Havey, M. J ., & Dodd, D. K. (1993). Variables associated with alcohol abuse
among self-identiﬁed collegiate COAs and their peers. Addictive Behaviors, 18, 567-
575.

 

Hawkins, D. J ., Catalano, R. F ., Miller, J. Y. (1992). Risk and protective factors
for alcohol and other drug related problems in adolescence and early adulthood:
Implications for substance abuse prevention. Psychological Bulletin, 112(1), 64-105.

 

Hawley, D. R. & DeHann, L. (1996). Toward a deﬁnition of family resilience:
Integrating life-span and family perspectives. Family Process, 35(3), 283-298.

 

Hetherington, M. E., Stanley-Hagan, M., and Anderson, E. R. (1989). Marital
Transitions: A child's perspective. American Psychologist, 44(2), 302-312.

Hecht, M. (1979). Children of alcoholics are children at risk. American Journal
of Nursing, 73(10), 1764-1767.

 

Herjanic, B., Herjanic, M., Penick, E., Tomelleri, C., & Armbruster, R. (1977).
Children of alcoholics. In F. A.Seixas (Ed), Currents in Alcoholism (Vol. 2, pp. 445-
455). New York: Grune & Stratton.

 

Holden, M. G., Brown, S. A. & Mott, M. A. (1988). Social support network of
adolescents: Relation to family alcohol abuse. American Journal of Drug and Alcohol
Abuse, 14(4), 487-498.

Hoffman, J. A. (1984). Psychological separation of late adolescents from their
parents. Journal of Counseling Psychology, 31, 170-178.

 

Hughes, J. M. (1977). Adolescent children of alcoholic parents and the
relationship of ALATEEN to these children. Journal of Consulting and Clinical
Psychology, 45, 946-947.

Jersild, A. T., Brook, J. S., & Brook, D. W. (1978). The psychology of
adolescence (3rd ed.). New York: Macmillan.

 

Jacob, T., Ritchey, D., Cvitkovic, J. F., & Blane, H. T. (1981). Communication
styles of alcoholic and nonalcoholic families when drinking and not drinking. Journal
of Studies on
Alcohol, 42(5), 466-482.

 

Jacob, T. & Leonard, K. (1988). Alcoholic-spouse interaction as a function of
alcoholism subtype and alcohol consumption interaction. Journal of Abnormal
Psychology, 97, 231-237.

102

Jacob, T. & Leonard, K. (1986). Psychosocial functioning in children of
alcoholic fathers, depressed fathers, and control fathers. Journal of Studies on Alcohol,
47, 373-380.

 

Johnson, D. W. (1980). Constructive peer relationships, social development and
cooperative learning experiences: Implications for prevention of drug abuse. Journal of
DrugEducation, 10, 7-24.

 

Johnson, 8., Leonard, K. E., & Jacob, T. (1989). Drinking, drinking styles, and
drug use in children of alcoholics, depressives, and controls. Journal of Studies on
Alcohol, 50(5), 427-431.

 

Johnson, K., Bryant, D., Strader, T., Bucholtz, G., Berbaum, M., Collins, D., &
Noe, T. (1996). Reducing alcohol and other drug use by strengthening community,
family, and youth resilience. Journal of Adolescent Research, 11(1), 36-67.

 

Johnston, L., O'Malley, P., & Bachman, J. (1988). Illicit drug use, smoking, and
drinking by America's high school students, college students, and young adults 1975-
1987. National Institute on Drug Abuse, US. Government Printing Ofﬁce, Washington,
DC.

Jones, J. W. (1982). A Screening Test to Identify Children of Alcoholic
Families. Paper presented at the annual meeting of the American Association for the
Advancement of Science, Clinical Psychology-Neurology Section, Washington, DC.

 

Jones, J. W. (1986). The Children of Alcoholics Screening Test: Test Manual.
Chicago: Camelot.

 

J essor, R., Graves, T. D., Hanson, R. C., & J essor, S. L. (1968). Society,
Personaliypand Deviant Behavior: A Study of a Tri-Ethnic Communig. New York:
Holt, Rinehart, & Winston.

 

Jessor, R., & J essor, S. L. (1977). Problem Behavior and Psychosocial
Development: A Longitudinal Study of Youth. San Diego, CA: Academic Press.

 

 

J essor, R., Turbin, M. S., Costa, F. M. (1998). Protective factors in adolescent
health behavior. Journal of Personality and Social Psychology, 75(3), 7 88-800.

 

Jurkovic, G. J. (1997). Lost Childhoods: The Plight of the Parentiﬁed Child.
New York: Bnmner/Mazel.

Kashubeck, S. & Christensen, S. A. (1992). Differences in distress among adult
children of alcoholics. Journal of Counseling Psychology, 39(3), 356-362.

Kandel, D., B., & Andrews, K. (1987). Processes of adolescent socialization by
parents and peers. The International Journal of the Addictions, 22(4), 319-342.

 

103

Kaufman, J ., Cook, A., Amy, L., Jones, B., & Pittinsky, T. (1994). Problems
deﬁning resiliency: Illustrations ﬁom the study of maltreated children. Development
and Psychopathology, 6, 215-229.

 

 

Keane, J. (1983). Factors related to the psychological well-being of children of
alcoholics. Paper Presented to the National Alcoholism Forum, Houston, TX.

 

Kobosa, S. (1985). Stressful life events, personality, and health: An inquiry into
hardiness (pp. 174-188). IN A. Monat & R. Lazarus (Eds), Stress and Coping (2nd
ed.). New York: Columbia University Press.

 

Kwakman, A. M., Zuiker, F. A., Schippers, G. M., & de Wuffel, F. J. (1988).
Journal of Youth and Adolescence, 17(3), 247-253.

 

Labouvie, E. W., & McGee, C. R. (1986). Relation of personality to alcohol and
drug use in adolescence. Journal of Consulting and Clinical Psychology, 54(3), 289-
293.

 

Lambom, S. D. (1990). Psychosocial Correlates of Emotional Autonomy
Among Securely, Avoidantly, and Anxiously Attached Adolescents. Paper presented at
the meeting of the Society for Research on Adolescence, Atlanta.

 

 

Lamborn, S. D., & Steinberg, L. (1993). Emotional autonomy redux: Revisiting
Ryan and Lynch. Child Development, 64, 483-499.

 

Luster, T., & Small, S. A. (in press). Sexual abuse history and problems in
adolescence: Exploring the effects of moderating variables. Journal of Marriage and the

Family.

 

Luthar, S. (1991). Vulnerability and Resilience: A study of high-risk
adolescents. Child Development, 62, 600-616.

 

Luthar, S., & Zigler, E. (1991). Vulnerability and competence: A Review of
research on resilience in childhood. American Journal of Orthopsychiatry, 6(1), 6-22.

 

Madanes, C. (1981). Strategic Family Therapy. San Francisco, CA.: Jossey-
Bass.

 

Masini, B. E. (1996). Paternal Alcohol Abuse and Psychological Functioning of
Adolescents: An Investigation of ModeratingVariables. Paper presented at the
conference for The Society on Research on Adolescence, Boston, MA.

 

 

104

r“
!.

 

Masten, A. 8., Best, K. M., & Garmezy, N. (1990). Resilience and
development: Contributions from the study of children who overcame adversity.
Development and Psychopathology, 2, pp. 425-444.

Mayer, J. & F ilstead, W. J. (1979). The adolescent alcohol involvement scale:
An instrument for measuring adolescents’ use and misuse of alcohol. Journal of Studies
on Alcohol, 40 (3), 291-300.

Maxwell, (1985). Review of Children of Alcoholics Screening Test. In J. V.
Mitchell, Jr. (Ed), The Ninth Mental Measurements Yearbook (Vol. 1, pp. 307-308).
Lincoln: University of Nebraska Press.

 

McCord, J. (1988). Identifying developmental paradigms leading to alcoholism.
Journal of Studies on Alcohol, 49, 357-362. g-

 

Mika, P., Bergner, R. M., & Baum, M. C. (1987). The development of a scale 5}
for the assessment of parentiﬁcation. Family Therapy, vol XIV(3), 229-235. g

Miller, D. & Jang, M. (1977). Children of alcoholics: A 20-year longitudinal
study. Social Work Research and Abstracts, 13, 23-29.

Minuchin, S & Fishman C. (1981). Family Therapy Techniques. Cambridge,
MA: Harvard University Press.

Mitz, L. B., Kashubeck, S., & Tracy, L. S. (1995). Relations among parental
alcoholism, eating disorders, and substance abuse in nonclinical college women:
Additional evidence against the uniformity myth. Journal of Counseling Psycholggy,
42(1), 65-70.

Moos, R. & Billings, A. (1982). Children of alcoholics during the recovery
process: Alcoholic and matched control families. Addictive Behaviors, 7, 155-163).

Moos, R. H., & Moos, B. S. (1984). The process of recovery ﬁom alcoholism:
Comparing functioning in families of alcoholics and matched control families. Journal
of Studies on Alcohol, 45, 111-118.

 

 

Nardi, P. M. (1981). Children of alcoholics: A role-theoretical perspective. _T_hg
Journal of Social Psychology, 115, 237-245.

 

Needle, R., McCubbin, H., Wilson, M., Reineck, R., Lazar, A., & Mederer, H.
(1986). Interpersonal inﬂuences in adolescent drug use: The role of older siblings,
parents, and peers. The International Journal of the Addictions, 21(7), 739-766.

105

Obuchowska, I. (1974). Emotional contact with the mother as a social
compensatory factor in children of alcoholics. International Mental Health Research
Newsletter, 16, 2-4.

 

Ohannessian, C. M. & Hesselbrock, V. M. (1993). The inﬂuence of perceived
social support on the relationship between family history of alcoholism and drinking
behaviors. Addiction, 88, 1651-1658.

Pandina, R. J., & Johnson, V. (1990). Serious alcohol and drug problems among
adolescents with a family history of alcoholism. Journal of Studies on Alcohol, 51, 278-
282.

 

Penning, M., & Barnes, GE. (1982). Adolescent marijuana use: A review.
International Journal of Addictions, 17(5), 749-791.

Perkins, H. W., & Berkowitz, A. D. (1991). Collegiate COAs and alcohol abuse:
Problem drinking in relation to assessment of parent and grandparent alcoholism.
Journal of Counseling and Development, 69, 237-240.

Potter, A. E., & Williams, D. E. (1991). Development of a measure examining
children's roles in alcoholic families. Journal of Studies on Alcohol, 52(1), 70-77.

 

Radke-Yarrow, M. & Brown, E. (1993). Resilience and vulnerability in
children of multiple-risk families. Development and Psychopathology, 5, 581-5 92.

 

Reich, W, Earls, F., & Powel, J. (1988). A comparison of the home and social
environments of children of alcoholic and non-alcoholic parents. British Journal of
Addiction, 83, 831-839.

 

Rhodes, J. E., & Jason, L. A. (1990). A social stress model of substance abuse.
Journal of Consulting and Clinical Psychology, 58(4), 395-401.

Rhodes, J. E. Gingiss, P. L., & Smith, P. B. (1994). Risk and protective factors
for alcohol use among pregnant aﬁican-american, hispanic, and white adolescents: The

inﬂuence of peers, sexual partners, family members, and mentors. Addictive Behaviors,
19(5), 555-564.

 

Rogosch, F., Chassin, L., & Sher, K. J. (1990). Personality variables as
mediators and moderators of family history risk for alcoholism: Conceptual and
methodological issues. Journal of Studies on Alcohol, 51(4), 310-318.

 

Roosa, M. W., Tein, J ., Groppenbacher, N., Michaels, M., & Dumka, L. (1993).
Mother's parenting behavior and child mental health in families with a problem
drinking parent. Journal of Marriage and the Family, 55, 107-118.

 

106

' _‘1 ’Ah' n

an
L.
1"-

 

 

Roosa, M. W., Dumka, L., & Tein, J. (1996). Fanrily characteristics as
mediators of the inﬂuence of problem drinking and multiple risk status on child mental
health. American Journal of Community Psychology, 24(5), 607-624.

 

Royce, J. E. & Scratchley, D. (1997). Alcoholism and Other Drug Problems.
New York: The Free Press.

 

Russel, M., Henderson, C., & Blume, S. (1985). Children of Alcoholics: A
Review of the Literature. New York: Children of Alcoholics Foundation.

 

 

Rutter, M. (1985). Resilience in the face of adversity: Protective factors and
resistance to psychiatric disorder. British Journal of Psychiatry, 147, 598-611.

 

Rutter, M. (1990). Psychosocial resilience and protective mechanisms (pp. 181-
214). In J. Rolf, A. Masten, D.Cicchetti, K. Nuechterlein, & S. Weintraub (Eds), I_tis_k
and Protective Factors in the Development of Psychopathology. New York: Cambridge
University Press.

 

Ryan, R. M., & Lynch, J. H. (1989). Emotional autonomy versus detachment:
Revisiting the vicissitudes of adolescence and young adulthood. Child Develmment,
60, 340-356.

 

Sarason, I. G., Levine, H. M., Basham, R. B., & Sarason, B. R. (1983). Assessing
social support: The social support questionnaire. Journal of Personality and Social
Psychology, 44(1), 127-139.

 

Sarason, I. B., Sarason, B. R., Shearin, E. N., & Pierce, G. R. (1987) A brief
measure of social support: Practical and theoretical implications. Journal of Social and
Personal Relationships, 4, 497-510.

 

 

Schissel, B. (1993). Coping with adversity: Testing the origins of resiliency in
mental health. The International Journal of Social Psychiatry, 39(1), 34-46.

 

Seilhamer, R. A., Jacob, T. & Dunn, N. J. (1993). The impact of alcohol
consumption on parent-child relationships in Families of alcoholics. Journal of Studies
on Alcohol, 54, 189-198.

 

Selzer, M. L. (1971). The Michigan Alcoholism Screening Test: The quest for
a new diagnostic instrument. American Journal of Psychiatry, 127, 1653-1658.

 

Selnow, G. W., & Crano, W. D. (1986). Formal vs Informal group affiliations:
Implications for alcohol and drug use among adolescents. Journal of Studies on
Alcohol, 47(1), 48-52.

 

Sessa, F. M. & Steinberg L. (1991). Family structure and the development of
autonomy during adolescence. Journal of EarlyAdolescence, 11(1), 38-55.

 

107

 

 

Sher, K. J. (1991). Children of Alcoholics: A Critical Appraisal of Theory and
Research. Chicago: The University of Chicago Press.

 

Sher, K. J., Walitzer, K.S., Wood, R, & Brent, E. E. (1991). Characteristics of
children of alcoholics: Putative risk factors, substance use and abuse, and
psychopathology. Journal of Abnormal Psychology, 100, 427-448.

 

Skinner, H. A., & Allen, B. A. (1983). Does the computer make a difference?
Computerized versusu face-to-face self-report assessment of alcohol, drug, and tobacco
use? Journal of Consulting and Clinical Psychology, 5, 267-275.

 

Smith, .G. M. & Fogg, C. P. (1979). Psychological antecedents of teenage drug
use. In R. Simmons (Ed.), Research in Community and Mental Health: An Annual
Compilation of Research (Vol. 1, pp. 87-102). Greenwich, CT: JAI.

 

 

Springer, S. A., & Gastﬁiend, D. R. (1994). A pilot study of factors associated
with resilience to substance abuse in adolescent sons of alcoholic fathers. Journal of
Addictive Diseases, 14(2), 53-66.

14‘. n “Lu-13‘ ﬂ

‘3

 

Staudinger, U. M., Marsiske, M. & Baltes, P. B. (1995). Resilience and reserve
capacity in later adulthood: Potentials and limits of development across the life span
(pp. 801-810). In D. Cicchetti and D. Cohen (Eds), Developmental Psychopathology,
Volume 2: Risk, Disorder, and Adaptation. New York: Wiley & Sons.

 

 

Steinberg, L., & Silverberg, S. B. (1986). The vicissitudes of autonomy in early
adolescence. Child Development, 57, 841-851.

 

Steinglass, P. (1981). The alcoholic family at home: Patterns of interaction in
dry, wet, and transitional stages of alcoholism. Archives of General Psychiatry, 38,
578-5 84.

 

 

Steinglass, P., Bennett, L. A., Wolin, S. J ., & Reiss, D. (1987). The Alcoholic
Family. New York: Basic Books.

 

 

Tarter, R., Hegedus, A., Goldstein, G., Shelly, C., & Alterman, A. (1984).
Adolescent sons of alcoholics: Neuropsychological and personality characteristics.
Alcoholism: Clinical and Experimental Research, 8, 216-222.

 

Tarter, R. E., Alterman, A. 1., & Edwards, K., L. (1985). Vulnerability to
alcoholism in men: A behavior-genetic perspective. Journal of Studies on Alcohol, 46,
329-356.

 

108

Turner, A., Irwin, C., & Millstein, S. (1991). Family structure, family process,
and experimenting with substances during adolescence. Journal of Research on
Adolescence, 1, 93-104.

 

 

Ullman, A. D. & Orenstein, A. (1994). Why some children of alcoholics become
alcoholics: Emulation of the drinker. Adolescence, 29(113), 1-11.

 

Walker, 1. P. & Lee, R. E. (1998). Uncovering the strengths of children of
alcoholic parents. Contemporary Family Therapy, 20(4), 521-538.

 

Walsh, F. (1996). The concept of resilience: Crisis and challenge. Family
Process, 35(3), 261-281.

Walsh, F. (1998). Strengthening Family Resilience. New York: Guilford Press,

 

Inc.

Weisner, T. S., & Gallimore, R. (1977). My brother's keeper: Child and sibling
caretaking. Current Anthropology, 18(2) 169-190.

 

Weiss, R. S. (1979). Growing up a little faster: The experience of growing up in
a single-parent household. Journal of Social Issues, 35(4), 97-111.

 

Whiting, B., & Whiting, J. M. (1975). Children of Six Cultures. Cambridge:
Harvard University Press.

 

White, H. R. (1987). Longitudinal stability and dimensional structure of problem
drinking in adolescence. Journal of Studies on Alcohol, 48, 541-550.

 

Wills, T. A. (1990). Multiple networks and substance use. Journal of Social and
Clinical Psychology, 9(1), 78-80.

 

 

Wills, T. A., Vaccaro, D., & McNamara, G. (1992). The role of life events,
family support, and competence in adolescent substance use: A test of vulnerability and
protective factors. American Journal of Community Psychology, 20(3), 349-375.

Wilson, C. & Orford, J. (1978). Children of alcoholics: Report of a preliminary
study and comments on the literature. Journal of Studies on Alcohol, 39(1), 121-142.

 

Wegscheider, S. (1981). Another Chance: Hope and Health for the Alcoholic
Family. Palo Alto: Science and Behavior Books.

 

Webster, D. W., Harburg, E., Gleiberman, L., Schork, A., & DiFranceisco, W.
(1989). Familial transmission of alcohol use: 1. Parent and adult offspring alcohol use
over 17 years-Tecumseh, Michigan. Journal of Studies on Alcohol, 50(6), 557-566.

 

109

Werner, E. (1986). Resilient offspring of alcoholics: A longitudinal study from
birth to age 18. Journal of Studies on Alcohol, 47(1), 34-40.

Werner, E. E., & Smith, R. S. (1992). Overcoming The Odds: High Risk
Children From Birth to Adulthood. London: Cornell University Press.

 

 

West, M. 0., & Prinz, R. J. (1987). Parental alcoholism and childhood
psychopathology. Psychological Bulletin, 102(2), 204-218.

 

Wolin, S, Bennett, L., Noonan, D., & Teitelbaum, M. (1980). Disrupted family
rituals: A factor in the intergenerational transmission of alcoholism. Journal of Studies
on Alcohol, 41(3), 199-214.

 

Wolin, S. J ., & Wolin, S. (1993). The Resilient Self: How survivors of
Troubled Families Rise Above Adversity. New York: Villard Books.

 

 

7+“; in; 31..

Woodside, M. (1988). Research on children of alcoholics: past and future.
British Journal of Addiction, 83, 785-792. 2. '

 

 

Wright, V. C., Frank, S. J. & Pirsch, L. A. (1991). The implications of father's
alcoholism for separation/individuation during late adolescence. Unpublished
Manuscript.

 

Wright, V. C. (1992). Late adolescent children of alcoholics: Autonomy and
relatedness in their relationships with their parents. Unpublished doctoral dissertation,
Michigan State University, East Lansing.

 

 

Wright, D. M., & Heppner, P. P. (1993). Examining the well-being of
nonclinical college students: Is it useful to know about the presence of parental
alcoholism? Journal of Counseling Psychology, 40, 324-334.

 

Zucker, R., Noll, R., & Fitzgerald, H. (1989). Drinking and drug history-
Revised edition (Version 3). Unpublished questionnaire. Department of Psychology,
Michigan State
University.

 

 

Zucker, R. A., Fitzgerald, H.E., & Moses, H. D. (1995). Emergence of alcohol
problems and the several alcoholisms: A developmental perspective on etiologic theory
and life course trajectory (pp. 677-711). In D. Cicchetti and D. Cohen (Eds.)
Developmental Psychopathology. Volume 2: Risk, Disorder, and Adaptation. New
York: Wiley & Sons.

Zucker, R. A., Ellis, D. A., Bingham, R. C. & Fitzgerald, H. E. (1996). The
development of alcoholic subtypes: Risk variation among alcoholic families during the
early childhood years. Alcohol Health and Research World, 20(1), 46-54.

 

 

110

Zuckerman, M., Eysenck, S., & Eysenck, H. J. (1978). Sensation Seeking in

England and America: Cross-cultural, age, and sex comparisons. Journal of Consulting
and Clinical Psychology, 46, 139-149.

 

111

*1

1QO ‘15—- .-
.-. .,

 

M IIIIIIIIIIIIIIIIIIIIIIIIIIIIII

mlwillwill”willyllllwlnl