3 .1 n.
.fiwu... hymns"...
.- .
Law”; z
.553. «van.
$4.94” hm,
, .r 2.. .
zfiytwwsazfi
.u....atmw..mm.ar1 1.1.... . W .m

. 5. w... r
. 23.2"“. a“ .

.2
. .n
’1

5.3.2.... I . In.
. ..:.....:...£| ‘ icflgflav
. thud-19811.3". \1 1.2.! 35:51.“!
:3. . x 53.3.: f1!
fink. 7.. . 39‘.

.uliiflu.fiuwv

.. :5 at
.. twang.
I...) 3.3.

1"...
I.- ~ .‘l-i' 3) —
:vya... {Sipf .38!
.51.. X. . £5311-
. 25.1.3.1... . I62. a.
$23.4» .thl‘;
1.2.5.69. 1,2:
tlfiififitinsr
.. mu... : .1 .
3.: It. Fl 53111:!
:i. 2.155%; F3}. . “a: a:
x. . ..
.z... .5 a... g.13.vt2ii§
:. i... 33:1.-dufzb
3.1.3:... 51..
212......
:51}... ‘

.' 3.456%"!

.10.. til.’I.‘O’.r:‘
80.512. (,1. 5.1.3.55}. 4

inn 2,}... .1? s1):
.- ‘Q1! 18.9:y i it!!!
a '3?

vii-’01,..iv‘l. ‘
yeti?“ 3...; < ‘)

 

 

THESE

1.-

This is to certify that the

dissertation entitled

Children's Aggressive Behavior in a Head Start
Sample: Its Relation to Caregiver Psychological and
Environmental Factors and Children's Attachment

Re re ntati
p se ons presented by

Ann Michele Stacks

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Family and Child Ecology

 

 

 

Date Ma; 2. 2002

MS U is 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 ' DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8/01 c:/ClRC/DateDue.p65—p. 1 5

Children’s Aggressive Behavior in a Head Start Sample: Its Relation to
Caregiver Psychological and Environmental Factors and Children’s Attachment
Representations

By
Ann Michele Stacks

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

2002

ABSTRACT
CHILDREN'S AGGRESSIVE BEHAVIOR IN A
. HEAD START SAMPLE:
ITS RELATION TO CAREGIVER PSYCHOLOGICAL AND
ENVIRONMENTAL FACTORS AND CHILDREN’S
ATTACHMENT REPRESENTATIONS
BY
Ann Michele Stacks

The purpose of this study was to investigate caregiver psychological and
environmental factors that contribute to parenting attitudes. attachment
representations in their children, and subsequent child behavior. Fifty-two
caregiver-child dyads participated in the study. Caregiver environmental factors
included the perceived availability of social support and satisfaction with social
support. measured by the Social Support Questionnaire (SSQ. Sarason, Levine,
Basham, & Sarason, 1983). Additionally, life stress was measured using the
Schedule of Recent Events (SRE, Daly, 1984). Caregiver psychological well
being was assessed using the Brief Symptom Inventory (BSI, Derogatis, 1992).
The Six-Year Attachment Doll Play Attachment Classification System (George &
Solomon, 1990, 1996. 2000) was used to assess children’s attachment
representations. Children's aggressive behavior was reported by teachers using
the Teacher Rating Scales (TRS, REynolds & Kamphaus, 1992) and by
caregivers using the Child Behavior Checklist (CBCL. Achenbach, 1991).

Children’s sex did not account for differences in child behavior. The

caregivers cultural background did not account for differences in parenting

attitudes. Due to the small number of children classified as secure (N=2), this

category was dropped from the analysis. Social support network size.
satisfaction with social support, and life stress were not associated with children’s
attachment representations. Caregiver environmental factors were significant
predictors of empathy and role reversal, but were not significant predictors of
values related to corporal punishment. inappropriate expectations. or power-
independence issues. Parenting attitudes and psychological well being were not
associated with children’s attachment representations. Children's aggression at
home and at school did not vary as a function of attachment representations.
The small sample size and the lack of observational data regarding
parent-child interaction were limitations of this study. Additionally, the small
number of children (N=2) classified as secure made it impossible to understand
which factors foster security and the contribution that a secure attachment makes
to social adjustment It also made the comparison between secure and insecure
children impossible. Clearly more research needs to be done before any of the

results can be considered conclusive.

Copyright by
ANN MICHELE STACKS
2002

ACKNOWLEDGEMENTS

During my doctoral study, I have been fortunate to have the support of
family members, friends, and colleagues. I am grateful to The Administration for
Children and Families who funded this research through the Head Start Graduate
Student Research Scholars Grant. As a part of receian this grant, I had the
opportunity to receive the support and mentoring provided by Michael Lopez and
Carole Kuhns, which enhanced my education and gave me a wonderful
opportunity to see a community of scholars. The experiences that were created
by the granting agency were some of the most memorable and educational
pieces of the process of completing my degree.

I would like to thank Dr. Robert Griffore, my major professor, for his help
and availability throughout my doctoral work. Dr. Griffore provided an academic
“secure base” from which I could explore. He has an amazing ability to know
when to let me struggle through a problem, how much help to give, and when to
intervene. He was able to see potential problems that are inherent in a study and
acted as a sounding board for so many important decisions that were made this
year.

I would also like to express my appreciation to Dr. Marsha Carolan, who
has been supportive of every decision that I have made. I will always be grateful
to her for the way that she challenged me as a student and as a therapist. Dr.
Carolan helped me to stay on the correct paflw by asking future oriented

questions in my masters program and early in my doctoral work. Her questions

led me to make a very difficult and important decision to alter the course of my
doctoral study. Without her help I do not think that I would have been prepared
for the work that is still to come. Dr. Carolan also sewed as the advisor to the
grant that I received from the Administration for Children and Families. She was
always available to meet, problem solve, and listen to my thoughts.

The other members of my committee, Dr. Ann Soderman and Dr. Cy
Stewart provided a great deal of support by listening to my thoughts, asking
questions, and providing feedback throughout my doctoral work.

It would never have been possible to complete this research in such a
short amount of time without the help of Brandon Silverthom, Kersten Kimmerly,
and Jamie Goff. They worked diligently on one of the most irmortant parts of my
dissertation, even though their duties were repetitive and difficult. I always felt
like they cared as much about this research as I did and that they always gave
more than 100 percent. I hope that each of them will be able to experience the
same exceptional assistance from others on a project that is as meaningful to
them as my dissertation was to me.

I feel so fortunate to have had the opportunity to work with the Head Start
community. I received a great deal of support from Lucy McCIintic, the family
support workers, and teachers, who welcomed me into the classrooms and
introduced me to families. I am grateful to every parent who took time out of his
or her busy schedule to talk to me, to fill out measures, and to allow me to spend

time with his or her child.

I would have to add another chapter to the dissertation in order to
appropriately thank all of the friends who supported me through this work. Drs.
Kate Rosenblum and Maria Muzik were essential in helping me formulate this
research and gain the skills necessary to carryout the work. I am grateful to
Mara Stein, Lara Campbell, and Nikki Brown for their friendship. They cooked
me dinner when l was hungry, gave me ideas when I didn’t have them, helped
me find excitement when l was ready to give up, and were understanding when
they gave much more than they received. Mara has read more than her fair
share of grants and dissertation drafts without complaints.

Last, but certainly not least, I would like to thank my fanin for their
support through all of my schooling. My parents and brothers have been
available to help in any way that they could. They have put up with bad moods
and exciting times. They have helped me to celebrate every step along this long

and sometimes bumpy road.

vii

TABLE OF CONTENTS

LIST OF TABLES ................................................................................ x
LIST OF FIGURES .............................................................................. xi
CHAPTER 1
INTRODUCTION
Statement of the Problem ............................................................ 1
Purpose of the Study ................................................................... 5
Research Questions .................................................................... 5
Methodology ............................................................................... 6
Overview of Subsequent Chapters .................................................. 7
CHAPTER 2
LITERATURE REVIEW
Introduction ................................................................................ 8
Theories Guiding the Research ...................................................... 8
Attachment Theory .............................................................. 8
Human Ecological Theory ................................................... 13
Conceptual Model ...................................................................... 15
Behavior Problems ..................................................................... 17
Prevalence of Behavior Problems in Preschool Children ............ 18
Sex Differences in Behavior Problems ................................... 19
lntemalizing Behavior Problems ........................................... 21
Extemalizing Behavior Problems... . . ..23
Factors Related to the Onset and Persistence of Extemalrzrng
Problems ........................................................................ 27
Attachment ................................................................................ 35
Attachment in Infancy ......................................................... 36
Attachment in the Preschool Years ........................................ 42
Parental Factors and Attachment Representations in Children.....47
Conclusion ................................................................................. 53
CHAPTER 3
METHODOLOGY
Introducfion ................................................................................ 55
Research Questions .................................................................... 55
Research Design ........................................................................ 56
Conceptual and Operational Definitions ........................................... 56
Sampling Procedure .................................................................... 58
Hypotheses ................................................................................ 63
Instrumentation ........................................................................... 65
Data Collection ........................................................................... 68
Sample Characteristics ................................................................. 69
Data Analysis ............................................................................. 71

viii

Conclusion ................................................................................. 73

CHAPTER 4
RESULTS
Introduction ................................................................................ 74
Frequencies for Relevant Variables ................................................ 74
Major Findings ............................................................................ 80
Conclusion ................................................................................. 90
CHAPTER 5
DISCUSSION
Introduction ................................................................................ 91
Frequencies ............................................................................... 91
Caregiver Environmental Factors and Children’s Attachment
Representations .......................................................................... 94
Caregiver Environmental Factors and Parenting
Attitudes .................................................................................... 95
Parenting Attitudes and Children’s Attachment Representations ............ 96
Caregiver Clinical Symptoms and Attachment Representations ............ 98
Attachment and Child Behavior ...................................................... 99
Aggression at Home and at School ................................................ 100
Implications for Future Research ................................................... 100
Implications for Practice .............................................................. 103
Limitations ............................................................................... 107
Conclusions .............................................................................. 108
APPENDIX A
Family Information ..................................................................... 1 11
APPENDIX B
Informed Consent ...................................................................... 114
Informed Consent (Teachers) ...................................................... 116
REFERENCES .................................................................................. 1 18

ix

LIST OF TABLES

Table 3.1 Child Characteristics ............................................................... 70
Table 3.2 Caregiver Characteristics ......................................................... 72
Table 4.1 Percentage of Caregivers in High Risk, General Population, and
Low Risk Groups for each Nurturing Parenting Subscale .............................. 78
Table 4.2 Means for Aggression, Stressful Events, and Parenting Attitudes
as a Function of Classroom Type ............................................................ 82

Table 4.3 Association Between Caregiver Social Support and Children’s
Attachment Representations .................................................................. 84

Table 4.4 Association BeMen Caregiver Life Stress and Children’s
Attachment Representations .................................................................. 85

Table 4.5 Summary of Linear Regression Analyses for Ecological Factors
Predicting Parenting Attitudes ................................................................. 87

LIST OF FIGURES

Figure 1 Contributions of the Caregiving Environment to Attachment
Representations and Child behavior ......................................................... 16

xi

CHAPTER 1
Statement of the Problem
Violent adolescents have received increased attention in the past five
years. As society struggles to understand the roots of this violence, research
points to the importance of understanding factors that contribute to the onset and
persistence of behavior problems in young children. There is increasing
evidence that parenting behaviors (Egeland, Kalkoske, Gottesman, & Erickson,
1990; Campbell, 1995; Bates, Bayles, Benett, Ridge, & Brown, 1 991 ; Campbell &
Ewing, 1990; Aguilar, Sroufe, Egeland, & Carlson, 2000; George 8 Main, 1979;
Herrenkohl & Russo, 2001), family stress ( Egeland et al., 1990; Campbell, 1995;
Aguilar et al., 2000), parental depression (Campbell & Ewing, 1990; Keenan 8
Shaw, 1994), and the quality of a child’s home environment (Egeland et al.,
1990), are predictive of future aggression. Research has also demonstrated that
early behavior problems are stable over time (Olweus, 1979; Huessmann, Eron,
Lefi<owitz, & Walder, 1984; Mesman, Bongers, & Koot, 2001) and that early
aggression is predictive of later and more severe forms of aggressive behavior
(Loeber 8: Dishon, 1983; Loeber & Stouthamer-Loeber, 1987; Huesmann, et al.,
1984; Magnusson, Stattin, & Duner, 1983; Magnusson, 1983). Clearly preventing
aggressive behavior is important; to do this, it is imperative to understand the
prevalence of early problems.
Behavior problems that are clinically significant are present in
approximately 10-15% of preschool children. Behaviors that are problematic are

also present in normally developing children, but they are often less severe and

present to a lesser degree (Campbell, 1995; Kuppersmidt, Bryant, 8 erloughby,
2000). Most of the research on behavior problems in young children has focused
on externalizing problems (Keenan 8 Shaw, 1997; Campbell, 1995; Mesman et
al., 2001) in samples of boys (Campbell, 1995). Girls, however, are not immune
to externalizing problems and it appears that in the early years, the sex
differences in the prevalence of these behaviors is slight (Zahn-Waider, lanotti,
Cummings, 8 Denham, 1990; Achenbach, Edelbrock, 8 Howell, 1987).

F urtherrnore, internalizing problems may cause significant distress to
these youngsters and play an important role in later behavior problems (Shaw,
Keenan, Vondra, Delliquadri, 8 Giovannelli, 1997). As a result, not only should
more studies regarding behavioral problems be conducted with young children,
but they should also investigate the factors in the child’s natural environment that
contribute to lntemalizing and externalizing problems. Additionally, these studies
need to include girls.

The caregiving environment is central to children’s development. Parent-
child attachment is the result of caregiver responsiveness (Ainsworth, Blehar,
Waters, 8 Wall, 1978; Isabella, 1993; Grossman, Grossman, Spangler, Suess, 8
Unzer, 1985; EgIand 8 Farber, 1984) and has been linked to later attachment
security, which has been shown to be stable from 12 months of age to 6 years, in
normative samples (Main 8 Cassidy, 1988; Wartner, Grossman, Fremmer-
Bombick, 8 Suess, 1994). A great deal of research has documented that
maternal representations of childhood and sensitive responsive caregiving are

linked to attachment security in infancy. However, other research has shown that

while maternal sensitivity is an important factor in the development of a secure
attachment relationship, it is not the only factor (De Wolff 8 van lJzendoom,
1997). Researchers have noted that to better understand attachment, studies
must be conducted that investigate factors in the caregivers environment that
promote his/her ability to provide sensitive and responsive care (Vondra,
DowdeII-Hommerding, 8 Shaw, 1999; Belsky, 1 999).

Attachment theory is relevant to lifespan development however,
attachment research has focused on the period of infancy, adult representations
of attachment, and adult romantic attachment. Research has shown that not all
problems in attachment are rooted in infancy (Cummings 8 Cicchetti, 1990).
Insecurity may develop after infancy as a response to a change in the caregiving
environment (Cicchetti, Cummings, Greenberg, 8 Marvin, 1990). Research on
attachment in the preschool years has been limited due to a lack of readily
available, reliable, and valid instruments to measure a preschool child's
attachment representation (see Solomon 8 George, 1999 for an overview of
these instruments). One measure that has been shown to be useful for children
between the ages of 4 and 7 has recently been validated (George 8 Solomon,
1990, 1996, 2000). Research using this instrument with 6 year-old children has
demonstrated that attachment disorganization is related to aggressive behavior
at home and at school (Solomon, George, 8 De Jong, 1995).

The lack of attachment research with preschool children is problematic
because it perpetuates the notion that attachment is only relevant to infancy,

whereas John Bowlby’s original conception of attachment was that it was a

lifespan concept. While it is true that the period from 6 to 12 months is important
to the development of a secure attachment, attachment continues to play a role
in children’s lives. Furthermore, there is less stability in attachment in high-risk
samples. Research on attachment in the preschool years could identify factors
during this time that promote a secure attachment, thereby decreasing risk
factors for children before they enter school.

One study (Call, 1999) addressed the relationship between adult romantic
attachment and behavior problems in a Head Start sample. While she found that
adult romantic attachment was related to behavior problems in Head Start
children, her sample size was small and her measure of attachment was
problematic. Attachment representations in children are the basis of affect
regulation, relational competence, and prosocial behavior, which all impact
children’s ability to learn and relate to other. By understanding the proportion of
different attachment representations of Head Start children, the classroom needs
of these children can be addressed through possibly reduced class size, teacher
training, and increased mental health services.

Given the high incidence of disorganized attachment in high-risk samples
(van IJzendoom, 1999) and the correlation between disorganized attachment
and aggressive behavior (Solomon, et al., 1995; Lyons-Ruth, Alpem, 8
Repacholi, 1993; Lyons-Ruth 8 Jacobvitz, 1999; Hubbs-Tait, Osofsky, Ham, 8
Culp, 1994; Lyons-Ruth, Easterbrooks, 8 Cibelli, 1997), it is imperative that
studies involving attachment and aggressive behavior are conducted with these

samples. In addition, the stability of aggression over time and the negative

impact that this behavior has on the future of the children and the community,
aggressive behavior must be prevented early. Research conducted with Head
Start children in the area of attachment and children's behavior has not been
conducted, with exception to the study cited earlier, which used a measure of
adult romantic attachment that has not been correlated with attachment in
children (Call, 1999). The overarching goal of Head Start is to promote social
competence and school readiness in children. One way that Head Start
accomplishes this is by strengthening families as the primary nurtures of their
children. Utilizing a Head Start sample provides an excellent opportunity to
assess attachment representations in high-risk preschool children.

This research will make a significant contribution to the existing literature
on the development of aggression by studying how attachment in preschool is
involved in the process of the development of aggression as a result of factors in
the caregiver’s environment. Furthermore, this study will contribute to knowledge
regarding the correlation between caregiver ecological factors, parenting
attitudes, and attachment representations in preschool children.

Purpose of the Study

The proposed study will investigate caregiver psychological and
environmental factors that contribute to parenting attitudes, attachment
representations in their children, and subsequent child behavior.

Research Questions
V The following research questions were developed to accomplish the

identified objectives and achieve the project purpose. This study will address the

following questions for a sample of children enrolled in full day, full year Head

Start and children enrolled in center based, part day Head Start:

. Are factors in the caregiver’s environment associated with children’s
attachment representations?

. Are factors in the caregiver's environment predictive of their ath'tude about
nurturing parenting?

. Are caregiver’s nurturing parenting attitudes associated with children’s
attachment representations?

o Is caregiver psychological well being associated with children’s attachment
representations?

. Do children’s mean aggression scores at home and at school vary as a
function of attachment representations?

. Is aggression at home correlated with aggression at school?

Methodology
In order to achieve the objectives of this study, a correlational design was

used. Caregivers whose children attend Head Start in Ingham County,
Michigan, were informed about the study through flyers that were sent home
by the teachers. In addition the doctoral student was available to meet with
parents and sign them up for the study during their child’s regular school day.
Thirty caregivers from full day, full year Head Start and 32 caregivers from
center based, part day Head Start agreed to participate in the study.
Caregivers who participated filled out self-report measures regarding their life

stress, social support, psychological well being, parenting attitudes, and their

child’s behavior. These measures were completed at their child’s school,
during regular school hours. Later that week, their child was interviewed
using the Six-Year Attachment Doll Play Classification System (George 8
Solomon, 1990; 1996; 2000), which is a semi-structured attachment doll play
interview. Teachers were sent self-report measures regarding the child’s
behavior in school. Once collected, the data were analyzed. Questions 1, 3,
and 4 were analyzed using crosstabulation; question 2 was analyzed using
regression analysis; question 5 was analyzed using ANOVA; and question 6
was analyzed using Pearson correlation. A complete description of the
methodology that was used to carry out this research is provided in chapter 3.
Overview of Subsequent Chapters

Chapter 1 has described the need, purpose, relevant research questions
for this study, and a brief overview of the methodology that will be used to
carry out the study. Chapter 2 provides a review of the theories guiding the
research and the relevant literature regarding attachment in infancy and
preschool and child behavior that is correlated with different attachment
classifications. In addition, the literature review discusses aggressive
behavior and the link between parenfing, aggressive behavior, and
attachment in children. Chapter 3 delineates the specific methodology that
was used to achieve the objectives identified in this chapter. Chapter 4
describes the frequencies for relevant variables and the major findings for
each research question. A discussion of the findings, limitations of the

research, and implications for research and practice is presented in chapter 5.

CHAPTER 2
Literature Review
Introduction

The previous chapter presented a brief review of the literature, which
focused on the importance of preventing behavior problems in young children. It
was suggested that to prevent these problems before children enter school, it is
important to understand the relationship between behavioral problems and
environmental influences is important. Chapter 1 also pointed to gaps in the
existing literature, specifically in the areas of attachment in the preschool years
and maternal ecological factors that contribute to parenting and attachment. The
previous chapter also indicated that the research on internalizing behavior
problems as well as behavior problems in female children is lacking. This
chapter will present the theories that guide the present research before
discussing literature relevant to the study.

Theories Guiding the Research
Attachment theory

Attachment theory posits that the behavior of the infant is organized
around the biological desire to maintain proximity to a parent as a function of
protection (Kobak, 1999; Cassidy, 1999; Bowlby, 1988). Parental responsiveness
in early infancy is related to individual differences in later attachment security.
Through repeated interactions with a parent or parent figure, children learn to
anticipate how their caregiver will respond. They adapt their attachment

behaviors to ensure that they will receive protection from their caregiver (Kobak,

1999). Attachment tends to be enduring and accompanied by intense emotions
(Darling, 1993) which are dependent on the relationship between the individual
and the attachment figure (Bowlby, 1988). As the child grows older, he or she
will cany toward the mental representations of his or her caregiver and impose it
upon new relationships. Attachment theory was first formulated by John Bowlby
and later validated and refined by Mary Ainsworth (Davies, 1999).

Bowlby's Theory of Attachment

John Bowlby began to conceptualize his theory of attachment in the early
1940’s, shortly after his graduation from Cambridge University. During this time,
Bowlby worked as a psychoanalyst in a home for maladjusted boys. His
observations of the boys’ family relationships led to his belief that a child’s early
relationship with his or her mother is important for early development and has an
impact on later functioning (Cassidy, 1999).

Bowlby worked closely with colleagues from other disciplines including
evolutionary biology, ethology, developmental psychology, cognitive science, and
control systems theory to further develop his ideas (Bowlby, 1988). Two
researchers who influenced Bowlby’s thinking were Lorenz (1935) and Harlow
(1958). Lorenz found that goslings became attached to parents who did not feed
them. Harlow observed that infant monkeys exposed to stress preferred a cloth
“mother" monkey over a wire monkey that provided food (Bowlby, 1988). His
belief that attachment behaviors were biologically based and functioned to
protect the child from danger (Kobak, 1999; Cassidy, 1999) came out of his

collaborative effort.

Bowlby’s background in systems theory is apparent in his early writing
about the attachment theory. He used the concept of homeostasis to explain
how attachment functions to promote proximity and thus safety for the child.
Children, he believed, desired proximity to their mothers. When the distance
separating parent and child becomes too great, the attachment system in the
child becomes activated (Cassidy, 1999). Children signal their caregivers in a
variety of ways when their attachment system is activated, including crying,
reaching, and crawling toward him or her. When a caregiver responds and the
child and caregiver are in sufficient proximity, the child’s attachment system is
terminated. It is important to note that the child’s goal is not simply his caregiver,
but instead the maintenance of an appropriate distance at which the child feels
secure (Bowlby, 1982). Bowlby also recognized that the child’s environment
played a role in his theory of attachment. While he considered that factors within
the child contributed to the feeling of stress and danger (injury, illness, or pain),
he also saw that factors within the environment can cause stress and danger,
and thus activate the attachment system (Casady, 1999).

Bowlby believed that attachment was a lifespan phenomenon (Cassidy,
1999; Greenberg, Cicchetti, 8 Cummings, 1990), but that the attachment system
itself became less sensitive to change over time (Solomon 8 George, 1999).
This is evident in his definition of attachment behavior, which he describes as:

“Any form of behaviour that results in a person attaining or
maintaining proximity to some other clearly identified individual who is
conceived as better able to cope with the world. . .Whilst attachment
behaviour is at its most obvious in early childhood, it can be observed

throughout the life cycle, especially in emergencies... The biological
function attributed to it is that of protection.” (Bowlby, 1988, p. 27).

10

Mary Ainsworth’s Contribution to Attachment Theory.

Mary Ainsworth, a student of John Bowlby, contributed significantly to
attachment theory. She conducted experimental observations of mother-infant
separations in a laboratory setting. As a way to measure the dyad’s behavior,
Ainsworth created a technique called the strange situation. Prior to conducting
the strange situation, each mother-infant pair was observed for seventy-two
hours in their home, over the course of one year. As a result the research team
was able to make associations between the babies attachment styles and the
mothers styles of parenting (Karen, 1990).

Infants classified as secure, in the strange situation (group B), greeted
their mothers with pleasure after separation and were easily soothed. Securely
attached infants are thought to be confident in their parent’s ability to be available
in a sensitive and responsive manner (Ainsworth, et al., 1978). As a result, these
infants are confident about their own interactions with others (Weinfield, Sroufe,
Egeland , 8 Carleson, 1999; George 8 Solomon, 2000). Infants classified as
insecure have had the experience that their caregivers are inconsistently
available when the environment is threatening. There are two types of insecure
attachments, insecure-avoidant and insecure-ambivalent. Insecure—avoidant
infants (group A) ignored their mothers upon reunion. Insecure-ambivalent
infants (group C) tended to seek contact with their mother when she returned to
the room, but at the same time would arch away from her and resist her attempts

to soothe him or her.

11

As researchers began to use the strange situation to classify children from
high-risk samples, they found that some of the children were not classifiable.
Mary Main, a student of Mary Ainsworth, and Judith Solomon studied the children
who had previously been unclassifiable and found a fourth categorization, which
they called disorganized/disoriented (group D) (Main 8 Solomon, 1990). Infants
classified as disorganized/disoriented (referred to as disorganized) have had the
experience that their caregivers are also the source of fear (Weian et al., 1999;
George 8 Solomon, 2000). These infants approached their mothers in a
disorganized way such as crawling backward toward her or freezing while
approaching her. Many of the infants that fell into this categorization had been
abused (Main 8 Solomon, 1990) or had an attachment figure who had
experienced trauma or a loss that was unresolved (Main 8 Hesse, 1990).
Stability of Attachment

Attachment influences a child’s later adaptation through beliefs about self
and the relationship (Weinfield et al., 1999). Research has shown that
attachment representations are stable over time in normative samples. Wrthout
intervention, approximately 80% of children assessed at 12 months and again at
age 6 maintained the same attachment classification (Main 8 Cassidy, 1988;
Wartner, et al., 1994). In a sample of maltreated infants, Barnett, Ganiban, 8
Cicchetti, (1999) found that 67% of infants classified as disorganized at 12
months were also classified as disorganized at 18 months. Of those classified as
disorganized at 18 months, 81% maintained that classification at 24 months.

Seventy-five percent of infants classified as secure at 12 months maintained a

12

secure attachment at 18 months, 69% were securely attached at 24 months. It is
thought that any instability is reflective of the ecological risk factors that can
cause changes in the caregiving environment (Vondra et al., 1999; Easterbrooks,
8 Goldberg, 1990; Solomon 8 George, 1999). These factors include parental
psychopathology (DeMulder 8 Radke-Yarrow, 1991), child maltreatment (Carlson
et al., 1989), and family stress (Egeland 8 Sroufe, 1981; Shaw 8 Vondra, 1993).
Research investigating the relationship between factors in the mothers
environment and her ability to provide sensitive and responsive care, which
promotes a secure attachment, has not been adequately explored by attachment
researchers.
Human Ecological Theory

Most of the current research on the development of attachment is not
broad enough because it has focused solely on the caregiving environment in
which the infant is raised. To gain a better understanding of how attachment is
formed, remains stable, or changes over time, it is important to look at the factors
within the caregiving environment that facilitate or inhibit a caregiver’s ability to
provide nurturing care. Human ecological theory provides a framework for the
understanding of how factors in the caregiving environment impact a caregiver’s
relationship with his or her child and thus parent-child attachment in the dyad.

Ecology is the study of the relationship between living organisms and their
natural, human constructed, and social environments (Bubolz 8 Sontag, 1993;
Griffore 8 Phenice, 2001). Human ecologists believe that complexity within a

system cannot be understood by analyzing lower levels of organization (Bubolz 8

13

Sontag, 1993). An ecological perspective when applied to research allows the
researcher to examine multi-Ievel interactions within and between systems.
There are several perspectives within the framework of human ecological theory
including general systems theory, of which attachment theory is partly based,
and psychological ecology (Griffore 8 Phenice, 2001).

This research will use Bronfenbrenner’s psychological approach to human
ecology, which provides a framework for understanding maternal factors
associated with parenting, attachment representations, and child behavior at
home and school. Bronfenbrenner's model states that develwment is always
embedded and expressed through behavior in one’s environment
(Bronfenbrenner, 1979). Bronfenbrenner conceptualizes the environment in four
levels that make up a model of “nested interdependent, dynamic structures”
(Muss, 1996, p. 322) that he called the microsystem, the mesosystem, the
exosystem, and the macrosystem.

The microsystem is the “pattern of activities, roles, and interpersonal
relations experienced by the developing person...with a particular physical,
social, and symbolic features that invite, permit, or inhibit, engagement in
sustained, progressively more complex interaction with, and activity in, the
immediate environment" (Bronfenbrenner, 1993, p. 15). The mesosystem is
comprised of the relationships that exist between two or more settings. “The
exosystem comprises the linkages and processes taking place between two or
more settings, at least one of which does not contain the developing person, but

in which events occur that indirectly influence processes within the immediate

14

setting in which the developing person lives” (Bronfenbrenner, 1993, p. 645).
Bronfenbrenner’s macrosystem is made up of belief systems, resources, and
patterns of social interchange that are imbedded within the culture of systems
(Bronfenbrenner, 1993).
Conceptual Model

Research grounded in attachment theory has mostly focused on parenting
behavior, the parent-child dyad, and child outcome. To further the study of
attachment, it is important to understand the context in which the family develops
and its direct and indirect impact on the parent-child dyad, attachment, and child
behavior. Human ecological theory provides a framework for this. To better
understand the development of problematic behavior, one must consider how
factors within the caregiver, such as psychological adjustment, are influenced
and influence, the caregiving environment. Factors influencing and influenced by
caregiver psychological well being include stressful life events, size of one’s
social support network, and one’s satisfaction with this network. These
components of the caregiver’s environment may influence his or her attitudes
about parenting, such as empathy and physical discipline. It is expected that a
caregiver's attitudes about nurturing parenting are reflective of his or her
behaviors when parenting. As a result, attachment and child behavior may be
indirectly affected by family stress, social support, and caregiver psychological

adjustment through parenting attitudes. These factors also may be additive

15

Contributions of the Caregiving Environment to

Attachment Representations and Child Behavior

Psychological
Adjustment

 

 

 

Attachment
Representation

Parenting

 

 

 

 

Figure 1

l6

 

Child Behavior

 

 

when predicting the variance in child behavior, which is effected through its
relationship with attachment (See figure 1).
Behavior Problems

There has been a great deal of research in developmental psychology
focused on predicting antisocial behavior (Loeber 8 Dishon, 1983; Loeber 8
Stouthamer-Loeber, 1986; Patterson, Capaldi, 8 Bank, 1992) and attempting to
understand the stability of antisocial behavior (Huessman, Eron, Lefltowltz, 8
Walder, 1984), externalizing problems (Keenan 8 Shaw, 1994; Campbell, 1994,
1995), and internalizing problems (Shaw, Keenan, Vondra, Delliquadri, 8
Giovannelli, 1997). Much of the research that has been conducted has focused
on attempting to understand the pathways to behavioral difficulties; that is how
early behaviors lead to later behavior problems (Egeland, Kalkoske, Gottesman,
8 Erickson, 1990). In the past ten years, an increasing amount of research has
pointed to the clinical significance of behavior problems in children under the age
of six and the negative implications that early behavior problems have for school
adjustment (Campbell, 1995).

Many behavior problems receive close attention in the preschool and early
school years possibly because this is the first time that a child is expected to
relate to peers, behave in socially appropriate ways, and conform to new adults
and their rules (Egeland et al., 1990). Studying behavior problems in preschool
can be a challenge (Egeland et al., 1990) because behaviors that are considered
problematic and clinically elevated in preschool are also seen to some extent in

normal populations of preschool children (Campbell, 1995). Another reason that

17

studying behavior problems during the preschool years can be difficult is due to
the rapid developmental change that accompany this developmental stage
(Campbell, Breaux, Ewing, 8 Szumowski, 1984). Campbell (1995), suggested
that determining when a developmentally appropriate behavior becomes a
behavior problem is important. This author suggests that children whose
behavior has become problematic will present with a constellation of symptoms
and will have symptoms that remain stable and last longer than what is
considered normal adjustment to change. They will also have symptoms that
are more severe, that are present in a variety of settings, and they will exhibit
behavior that interferes with their ability to deal with normal developmental
challenges.

Prevalence of Behavior Problems in Preschool Children

Behavior problems in preschool children are often present in normally
developing children, but they are often less severe and present to a lesser
degree, than in samples of children with behavior problems. Across studies, it
appears that approximately 10 to 15% of preschool children have mild to
moderate behavior problems as defined by cut-off scores on measures of
children’s behavior (Campbell, 1995).

Researchers have suggested that children in preschool programs
designed for high-risk children, such as Head Start, might be expected to exhibit
more aggressive behaviors due to their level of poverty (Kupersmidt, Bryant, 8
erloughby, 2000). However, these authors found that the prevalence of
aggressive behaviors in the Head Start classroom was not significantly different

from a sample of children from a community child care program. Approximately

18

10% of Head Start children in their sample showed daily or higher rates of
aggressive behavior. The aufl'iors of this study used frequency of behavior,
rather than a clinical cut-off from a behavior checklist, which could account for
their finding. While this study reported that the prevalence of aggression was the
same in both classrooms, they found that the overall levels of aggression were
higher in the Head Start classrooms compared to the community classrooms.

Sex Differences in Behavior Problems

Studies have shown that a relatively small number of preschool children
are exhibiting significant behavior problems. Boys tend to have more difficulty
than girls regulating their externalizing behavior. (Mesman, Bongers, 8 Koot,
2001; Zahn-Waxler, lanotti, Cummings, 8 Denham, 1990; Achenbach,
Edelbrock, 8 Howell, 1987; Prior, Smart, Sanson, 8 Obkerlaid, 1993). Prior to
the preschool period, it appears that there are no sex differences in the
frequency of behavior problems. For example, Kennan and Shaw (1994) found
that there were no sex differences in the frequency or stability of interpersonal or
object related aggression between 18 and 24 months of age. Mesman and her
colleagues (2001) also found that between the ages of 2 and 3, internalizing and
externalizing problems were as prevalent for girls as they were for boys.

Studies of problem behaviors in children have consistently found that
between the ages of 4 and 5, sex differences in the prevalence and stability of
behavioral problems emerge (Zahn-Waxler et al., 1990; Achenbach et al., 1987).
It appears that beginning at age 4 boys have significantly more externalizing
problems that girls, but that there still is no difference in rates of lntemalizing

problems; this trend continues throughout school age (Mesman et al., 2001).

19

For example, at school age, boys are ten times more likely than girls to exhibit
externalizing disorders (Keenan 8 Shaw, 1997). One study, however found that
preschool girls show significantly higher levels of teacher rated relational
aggression than boys (Crick, Casas, 8 Mosher, 1997).

Accounting for Sex Dr'fl'erence r'n Behavioral Problems

There are a variety of empirically based explanations for the increase in
the level of aggressive behavior in boys. For example, Cummings, Iannotti, 8
Zahn-Waxler (1989) found that after age 4, girls levels of behavior problems
decline; this finding was supported by Prior et al., (1993). Other studies have
found that boys tend to show little or no change in behavior problems (Prior et al,
1993), or an increase in behavior problems (Rose, Rose, 8 Feldrnan, 1989).

This trend has also been shown for children in Head Start (Kupersmidt et al.,
2000). It is not until adolescence when the prevalence of internalizing symptoms
of girls exceeds that of boys (Keenan 8 Shaw, 1997).

It has been suggested (Keenan 8 Shaw, 1997) and supported by Mesman
et al. (1997), that the path to psychopathology is more discontinuous for girls
than for boys. Research shows that girls who develop externalizing disorders are
more likely to be slower than average in developing language and social skills
(Keenan 8 Shaw, 1997). These same authors suggest that girls may show lower
levels of externalizing behaviors because they may “grow out” of problem
behavior. At preschool age, girls’ social development is more advanwd than that
of boys. Through this advanced development, it is thought, that girls learn to

regulate their behavior and emotions. Also, girls’ language skills and empathetic

20

responses in preschool are typically more advanced than boys’ language skills
and ability to respond empathetically. It is possible that these abilities have a
positive impact on the parent-child relationship, which then impacts externalizing
behavior problems that girls may be exhibiting (Keenan 8 Shaw, 1997).
The Development of lntemalizing Problems in Girls

Keenan and Shaw (1997) also suggest several reasons why girls may be
more likely than boys to develop lntemalizing behavior problems. First, it is
possible that girls may have a biological predisposition to internalizing problems.
Second, girls exposed to camgiving environments characterized by abuse and/or
neglect may develop lntemalizing problems. Second, girls are socialized
differently than boys. The authors report that protective factors, such as
advanced verbal skills and empathy, for some girls may actually be risk factors
for others. Keenan and Shaw (1997) explain that girls advanced development in
verbal ability and empathy allow them to take another person’s perspective,
which in turn influences their interpersonal relationships. Girls who develop in a
context where they are exposed to a caregivers chronic distress may develop
feelings of responsibility and guilt, which lead to later lntemalizing disorders.

lntemalizing Behavior Problems
The above section began to offer some possible explanations why,

beginning in preschool, girls and boys may begin to develop different rates of
externalizing problems and why in adolescence girls begin to show higher rates
of internalizing problems. However, most research that has been conducted with

preschool children has focused on externalizing problems. Despite this, there

21

are children who have internalizing symptoms such as fearfulness, sadness,
withdrawal, and somatic complaints in preschool (Campbell, 1995).

The lack of research focusing on lntemalizing problems in preschoolers
could be related to theoretical issues, methodological issues, and to the
discontinuity of behavior problems (Shaw et al., 1997). Shaw and his colleagues
(1997) report that it is unclear how to measure internalizing behaviors in children
under 6, because internalizing problems focus on internal states. Another
problem that contributes to the lack of research on internalizing disorders is the
lack of continuity in children’s expression of internalizing behavior. Also, a
theoretical framework to guide the understanding of how different moods are
related to different developmental periods is lacking.

Factors Related to the Development of lntemalizing Behaviors

Research has found several factors that are related to the development of
internalizing behaviors at 24 and 42 months of age. Two child characteristics
were found to be related to internalizing problems, those were matemal report of
a difficult temperament at 12 and 24 months and the presence of a disorganized
attachment (Shaw et al., 1997). Several caregiver factors appear to be related to
the development of internalizing problems, for example maternal reported
symptoms of depression, maternal aggressive personality, and matemal
satisfaction with social support (Shaw et al., 1998). Family factors such as a high
level of stressful life events (Shaw et al., 1997), family conflict (Shaw et al.,
1997), and parental disagreements about child rearing (Shaw et al., 1998) were

also found to be related to the development of internalizing behavior problems.

22

Preschool children who exhibit internalizing symptoms of withdrawal and
depression are more likely to exhin significantly higher levels of lntemalizing
behavior at school entry than children who did not have symptoms of withdrawal
and depression (Mesman et al., 2001). Clearly, more research needs to be
conducted with samples of preschool boys and girls if we are to be able to say
with any certainty which child, parent, family, and environmental factors are
related to the onset, persistence, and outcome of lntemalizing behavior
problems.

The Relationship Between lntemalizing and Extemalizing Problems

Studies that have examined internalizing problems have mixed findings
(Mesman et al., 2001). Some studies have found that early lntemalizing
problems predict later externalizing problems (Egeland, Pianta, 8 Ogawa, 1996),
while others found no relationship between early lntemalizing behaviors and later
externalizing behaviors (Rose et al., 1989). Still other studies have found that
internalizing problems may act as protective factors in the development of
externalizing problems (Kerr, Tremblay, Pagoni, 8 Vrtaro, 1997).

Extemalizr'ng Behavior Problems

The amount of research on externalizing behaviors far outweighs the
amount of research on internalizing behaviors (Mesman et al., 2001).
Extemalizing behaviors include aggression, opposition, hyperactivity, and
delinquency. Children exhibiting these behaviors tend to be much more
disruptive to a parent or a teacher and are much more likely to cause harm to

another person than children exhibiting signs of an internalizing disorder. This

23

could explain why so much of the research has focused on externalizing
behavior, particularly aggression.

Prospective and retrospective research has documented the stability of
externalizing behavior problems from preschool to adolescence (Campbell, 1995;
Mesman et al., 2001). It appears that disruptive and aggressive behaviors are
already in place and relatively stable by school age (Keenan 8 Shaw, 1994) and
that over time aggressive behavior is approximately as stable as intelligence for
males (Olweus, 1979). Researchers from several disciplines have been
interested in the age of onset, as well as predictors and consequences of
externalizing behavior.

Extemalizr'ng Behavior in Early Childhood

Research that has focused on the stability of behavior problems in early
childhood supports the notion that behavior problems can emerge prior to 24
months of age (Keenan 8 Shaw, 1994). Keenan and Shaw (1994) studied
children 18 to 24 months of age and found that object related aggression is much
more prevalent than interpersonal aggression in the toddler period. Their
research also provided evidence that externalizing behavior may have roots in
toddler behavior. Some researchers began their investigations prior to
toddlerhood and found that mothers who described their infants as more difficult
had preschoolers with behavior problems (Campbell, 1995). Other studies,
however have not found a relationship between early temperament and later
behavior problems (Aguilar, Sroufe, Egeland, 8 Carlson, 2000). One reason for
these different findings could be that the level of infant difficulty interacts with the

24

quality of parenting, developmental problems, and sex of the child to determine a
positive or negative outcome (Campbell, 1995).

In a longitudinal study, Egeland and his colleagues (1990) found evidence
supporting Keenan and Shaw’s earlier findings. Egeland and his colleagues
(1990) set out to determine the persistence of behavior problems and
competence from preschool to first, second, and third grades. They were also
interested in understanding which factors accounted for exceptions to predicted
outcomes. These researchers found that of the children who were identified as
having behavior problems in preschool, 47% had clinically significant (T score >
60) behavior problems in second grade. Additionally, 80% exhibited problem
behavior in at least two of the first three years in school.

The findings by Egeland et al. (1990) supported research previously done
by others. For example, Campbell and her colleagues (1984) found that 50% of
preschool children exhibiting behavior problems also have behavior problems at
age six. In another study, Campbell and Ewing (1990) found that 67% of three
year-olds identified as hyperactive and aggressive also had severe behavior
problems at age nine. Together, these findings support the ideas that behavior
problems may begin early (Keenan 8 Shaw, 1994) and that behavior is stable
from at least preschool to the early school period.

Outcomes for Children with Extemalizing Behavior Problems

Studies Conducted with Preschool Children. Children who are exhibiting

externalizing problems are at risk for a number of later social and academic

problems. For example, Head Start children rated as non-compliant and

25

overactive by teachers engaged in more negative interactions with peers and
were described negatively by peers (Campbell, 1995). Additionally, boys who
exhibited externalizing problems in preschool and first grade were rated as less
socially competent by mothers, fathers, and teachers, than a comparison group
of boys (Campbell, 1994). Physical aggression and opposition, but not
hyperactivity in kindergarten placed boys at high risk for juvenile delinquency
(Nagin 8 Tremblay, 1999).

Studies Conducted with School Age Children. At school age children with
externalizing problems continue to show poor adjustment Tremblay, Pihl, Vitaro,
8 Dobkin (1994) found evidence that school age boys who showed the highest
risk for later delinquency have been shown to be highly impulsive and show low
levels of anxiety. In support of this, 10 to 12-year-old boys who were rated as
non-inhibited and disruptive by peers were more likely than chance to rate
themselves as delinquent at 13 to 15 years. By comparison, boys rated as
inhibited and disruptive were not as likely to become delinquent (Kerr et al.

1 997).

Extemalizing Behavior and School Adjustment. Research has
documented the relationship between externalizing behaviors at school age and
academic difficulties (Patterson, Capaldi, 8 Bank, 1991). For example,
Huesmann, Eron, 8 Yarmel (1987) found that peer-rated aggression at age 8 is a
significant predictor of low attainment in reading, spelling, and arithmetic at age
30. In another study, 7 and 10—year-old boys whose peers perceived them as

aggressive, were more likely than other boys to become aggressive adolescents

26

who also exhibited poor school performance. The girls in this study who were
rated as aggressive also showed poor school performance and were more likely
to receive psychiatric services as adolescents (Serbin, Moskowithz,
Schwartzman, 8 Ledinghan, 1991). Antisocial children have been shown to be
on task in the classroom about 20% less than normally developing children
(Walker, Shinn, O’Neill, 8 Ramsey, 1987). Tremblay (1988) found that 38% of
high-risk children rated as aggressive by kindergarten teachers were failing in
school three years later.
Children who change Paths

Studies have consistently shown that behavior problems remain stable
over time. We can learn a great deal however, from aggressive children whose
behavior did not follow the expected outcome. There is evidence that children
who showed significant behavior problems in preschool, but did not show
significant problems in elementary school have psychosocial factors in common.
Children whose early behavior was problematic, but who changed paths by
school entry had significantly fewer stressful events and came from homes that
were organized, predictable, and more stimulating and responsive to the child’s
needs. These children also had mother’s who became less depressed over time
(Egeland et al., 1990).

Factors related to the Onset and Persistence of Extemalizing Problems

The above research demonstrates that children who show early
externalizing behavior problems are at risk for a variety of continued social and
academic problems. This research underscores the need to prevent this type of

behavior before children begin school. Preventing behavior problems requires

27

that one understands factors related to the onset an persistence of such
problems. Research has shown that child factors, caregiver factors, and
environmental factors, in which the child is raised contribute to the development
and continuation of behavior problems. Overall, researchers agree that
psychosocial factors, early, chronic behavior problems, and cumulative risk
(Aguilar et al., 2000) play an important role in the development and maintenance
of behavior problems in children.

The research points to a number of factors that are related to the onset
and persistence of externalizing behaviors. In keeping with Bronfenbrenner’s
Theory of Human Ecology (Bronfenbrenner, 1979), which provides a framework
for the research that is to be presented in chapters four and five, factors
associated with externalizing behaviors will be presented in sections according to
the levels used by Bronfenbrenner (1979). None of the literature to date has
examined mesosystemic factors that contribute to the development or
persistence of externalizing behaviors. Research of this type is desperately
needed if we are to better understand how to prevent behavior problems in
children.

Microsystemic Factors

Research has consistently examined and found support for a number of
factors in the child’s environment that contribute to the development of
externalizing behavior problems. These factors include initial problem severity,

quality of the home environment, temperament, parenting behavior, and maternal

psychopathology.

28

Problem Severity, Temperament, and Home Environment. The severity of
the behavioral problem at an early age has been shown to contribute to the
maintenance and later severity of behavior problems in elementary school
children. (Campbell 8 Ewing, 1990; Campbell, 1995; Campbell, March, Pierce,
Ewing, 8 Szumowski, 1991). Some studies have supported the idea that early
reports of infant difficulty predict later problem behavior (Campbell 8 Ewing,
1990). Others however, believe that the level of infant difficulty interacts with the
quality of parenting, developmental problems and sex of the child to determine
whether or not a child will exhibit behavior problems (Campbell, 1995).

In a prospective study Aguilar and his colleagues (2000) found that
temperament did not distinguish between groups of children who had persistent
behavior problems beginning in preschool, whose problems began in
adolescence, or children who never exhibited clinical levels of behavior problems
(Aguilar et al., 2000). One factor that may account for the difference in the
findings is that Campbell and Ewing (1990) used mother‘s retrospective accounts
of their infant’s temperament, while Aguilar and colleagues (2000) began
measuring temperament at birth and used multiple measures and infomtants until
12 months of age.

The home environment has also been related to the development of
externalizing behavior problems (Egeland et al., 1990). In general, preschoolers
who had clinically significant behavior problems that persisted into adolescence
tended to live in disorganized homes that were not responsive to the children’s

needs. Adolescents who exhibited behavior problems as preschoolers, but were

29

found to be competent as adolescents came from homes that were more
developmentally stimulating. Furthermore, their homes were predictable and
organized (Egeland et al., 1990).

Parenting Behavior. Campbell (1995) found that researchers representing
a variety of theoretical perspectives agreed that parenting plays a role in the
development of disruptive behavior (Bates, et al.,1991; Farrington, 1978;
Patterson, et al., 1991). The level of parental warmth, degree of responsiveness,
and effective and consistent limit setting impact the parent-child relationship,
thereby impacting child behavior. Campbell (1994), Campbell et al., (1991), and
Campbell and Ewing (1990) compared three groups of boys, those who exhibited
significant problems in preschool and first grade, those who had significant
problems in preschool, but were competent in first grade, and children who were
competent in preschool and first grade. These researchers found that boys who
continued to have problems had mothers who used more negative control to gain
compliance from their children.

Another study (Aguilar et al., 2000) compared four groups of children:
those whose negative behavior began in preschool and continued through
adolescence, those whose behavior was negative in prescth but appropriate in
adolescence, those whose negative behavior began in adolescence, and those
who never exhibited clinical levels of negative behavior. Their findings suggest
that children whose behavior problems started early and persisted through
adolescence were the most likely to have parents who were unavailable,

physically abusive, and neglectful in the first three years of life.

30

Several studies have found a relationship between abusive and neglectful
parenting and childhood aggression (Alesandri, 1991; George 8 Main, 1979;
Herrenkohl 8 Herrenkohl; 1981; Herrenkohl 8 Russo, 2001). Patterson et al.
(1991) believe that family interaction patterns provide reinforcement for
aggression and that aggression is generalized from home to other settings.
These researchers have found that there are two paths to delinquent behavior:
The early-starter model and the late-starter model. Boys on the early start path
begin their antisocial training at home. Patterson et al. (1991) believe that:

“...certain specifiable reactions from the parent will
produce a toddler who displays stable patterns of
coercive behavior and noncompliance. Similar con-
tingencies maintain the performance of these behaviors
in older children as well...lt is assumed that coercive
exchanges are the key ingredient in setting the early
starter process in motion. Once it begins, it seems to
move through a sequence of three stages: (a) The child
shows clearly identifiable antisocial behavior;...(b) the
Child is rejected by the non'nal peer group; and (c)

the child fails in school” (pp. 145-146)

Researchers using attachment theory as a homework to guide their study
have consistently found that behavior problems in preschool and the early school
years are related to infant attachment classification. While many of the above
studies report that parent-child interaction, maternal wamrth and involvement,
and parenting style are important factors related to the onset and persistence of
behavior problems, attachment is consistently not being measured by
researchers interested in children’s behavior. Due to the important role that the
early parent-child attachment may play in the development and continuation of

behavior problems it will be discussed in later sections of this review.

31

Maternal Psychopathology. Research has consistently found that high
levels of maternal depression are related to the onset and persistence of
behavior problems in children (Campbell 8 Ewing, 1990; Egeland et al., 1990;
Campbell et al., 1991; Campbell, 1994, 1995; Keenan 8 Shaw, 1994). Research
that has compared preschool children whose problems persisted over a three
year period, whose problems decreased over a three year period, and whose
behavior was never problematic at a clinically significant level found that
matemal depression was related to the onset and persistence of behavior
problems. Children who had behavior problems in preschool, but were found to
be competent in first grade had mothers whose reported depression levels
decreased over time. Children whose behavior remained problematic had
mothers whose reported level of depression increased over time (Egeland et al.,
1990).

The relationship between maternal depression and child behavior
problems is somewhat unclear. The decreased levels of maternal tolerance for
developmentally appropriate, but difficult behavior may explain the relationship
between depression and behavior problems(Campbell et al., 1991). For example
depressed mothers caring for young children tend to be emotionally unavailable,
provide inconsistent discipline, are likely to be hostile toward their child, and
communicate less with their child (Egeland et al., 1990). Other explanations for
the relationship between maternal depression and child behavior problems

include the idea that behavior problems in children contribute to feelings of

32

depression in caregivers and that children's problems co-occur with family
problems (Campbell et al., 1991).

Overall, research has consistently shown that microsystemic factors play a
role in the development of children’s behavior problems. These factors include
child characteristics, such as the age of onset and severity of the problem and
temperament, as well as parenting factors, including warmth, limit setting,
physical punishment, and attachment. Home environment and matemal
psychopathology also have been shown to impact child behavior.

Exosystemic Factors

Research has not focused specifically on how factors in a caregiver’s
environment may impact child behavior. Factors such as low levels of marital
satisfaction (Campbell, 1994) and low levels of satisfaction with one’s support
network (Campbell, 1995) have been related to externalizing problems in
preschool boys. Campbell’s(1994) comparison of boys who had behavior
problems in preschool and first grade, those whose problems decreased over
time, and those whose behavior was not seen as problematic at either time
period points to a relationship between marital satisfaction and behavior
problems. She found that parents who reported a decrease in marital
satisfaction, along with other factors presented earlier, had children whose
behavior problems persisted over time. Renken, Egeland, Marvinney,
Mangelsdorf, and Sroufe (1989) found that children whose mothers reported less
social support and more stress received higher teacher ratings of aggressive

behavior.

33

Macrosystemic Factors

Belief systems and patterns of social interchange that are imbedded within
the culture of the family system correlate with child behavior problems.
Macrosystemic factors, which may contribute to the development and
maintenance of problem behavior, include poverty and a family history of
stressful life events. It is likely that families living in poverty are also more likely
than families who are not living in poverty to experience a greater number of
stressful life events.

Stressful life events appear to be associated with the onset and
persistence of behavior problems in children (Campbell 8 Ewing, 1990; Egeland
et al., 1990; Campbell et al, 1991; Campbell, 1995; Aguilar et al., 2000).
Children who have severe behavior problems in preschool and continue to have
problems into school age (Campbell 8 Ewing, 1990; Egeland et al., 1990) and
adolescence (Aguilar et al., 2000) live in families whose levels of stress stayed
consistent or increased over time. Children who began preschool with negative
behaviors, but were found to be competent in elementary school had families
whose reported levels of stress had decreased (Egeland et al, 1990). Lower
levels of family stress also distinguished between groups of children who never
exhibited significant levels of behavior problems, and those whose behavior
problems began in preschool (Aguilar et al., 2000). Some models of child and
family adversity suggest that stressful life events, such as maternal depression,
marital discord, single parenting, multiple changes in location, money, or

employment, or other stresses indirectly affect children through their impact on

34

parenting (Belsky, 1984; Patterson, DeBaryshe, 8 Ramsey, l 989; Webster-
Stratton, 1990), specifically on the use of physical discipline (Herrenkohl 8
Russo, 2001).

Clearly a number of factors at a variety of levels make additive and unique
contributions to variance in child behavior. One factor that consistently receives
attention is the impact that parents have on their children’s behavior. Research
has also suggested that factors outside of the parent child relationship exert their
influence directly and/or through the parent—child relationship. One important
factor that is related to children’s behavior and the quality of caregiving is parent-
child attachment.

Attachment

Many studies have linked attachment and behavior problems, including
aggressive behavior (Sroufe, 1983; Sroufe et al., 1984; Renken et al., 1989; Troy
8 Sroufe, 1987; Hubbs-Tait, Osofsky, Hann, 8 Culp, 1994; Solomon, George, 8
De Jong, 1995; DeMulder, Denham, Schmidt, 8 Mitchell, 2000; Lyons-Ruth 8
Jacobvitz, 1999). Understanding this link is important if behavior problems are to
be prevented. Preventing behavior problems is important because early
behavioral problems remain stable and predict later delinquency and crime
(Loeber 8 Dishon, 1983; Loeber 8 Stouthamer-Loeber, 1987; Huesmann, et al.,
1984; Magnusson, et al., 1983; Magnusson, 1988). The research above is an
indication of the factors associated with behavior problems and the negative

outcomes for children who exhibit such behavior.

35

Based on the research presented earlier, it appears that one way to
prevent disruptive behavior in children is through the parent-child relationship.
The internalization of parental standards of behavior, the development of self
regulation, and the development of prosocial behavior has been associated with
the quality of the parent-child attachment (Sroufe 8 Fleeson, 1986) and serves to
protect children from later behavior problems. Campbell (1995) notes that:

“A more detailed assessment of concunent childrearing

practices, maternal sensitivity and responsiveness,

and attachment security in toddler age and preschool

children with and without externalizing problems may

help to clarify the role of attachment in the development

of problems.“ (p. 137)
The sections below review the attachment research in infancy and preschool. It
also provides an overview of the relationship between attachment and behavioral
problems in children, attachment and parenting, and attachment research that
has been done with Head Start children.

Attachment in Infancy

Most of the research on attachment to date has been conducted in infancy
or adulthood. Attachment in infancy functions to protect the infant. When the
attachment system is activated, the infant signals the parent (Kobak, 1999;
Cassidy, 1999; Bowlby, 1988). Parent responsiveness in early infancy is related
to individual differences in later attachment security. Through repeated
interacfions with a parent or parent figure, children learn to anticipate how their
caregiver will respond. Infants adapt their attachment behaviors so that they will
receive protection from their caregiver (Kobak, 1999). There are four functions of

attachment: providing a sense of security; regulation of affect and arousal;

36

promoting the expression of feelings and communication; and senring as a base
for exploration (Davies, 1999). Research on attachment in infancy has focused
on the predictive power of attachment classifications and the relationship
between parental sensitivity and attachment classification.

Attachment and Child Behavior

There has been a great deal of research documenting the relationship
between attachment security in infancy and subsequent child behavior. Secure
attachment is linked to mastery, emotional regulation, and interpersonal
closeness. Attachment and later development are connected through
transactional experiences that influence brain development (Weinfeld, et al.,
1999) and serve as a foundation for learning affect regulation (Isabella, 1993;
Cassidy, 1994; Sroufe, 1996). Much of the research to date has focused on
early parent-child attachment and various child outcomes. The literature cited
below is reflective of this research.

Secure Attachment. Much of what is known about the consequence of
attachment security is the result of a longitudinal study by Sroufe (1983, 1996)
and his colleagues. In this research, he found that sensitive, responsive
caregiving teaches infants that they can readily get their needs met and that they
can have an effect on the world (Sroufe,1983, 1996). These infants are
effectively dependent on their caregivers and grow up to be effectively
independent (Sroufe, Fox, 8 Pancake, 1983).

Children’s history of attachment has been shown to play a role in their
early and later academic and social success. Preschool teachers described

children with a history of secure attachment as empathetic (Waters, erpman, 8

37

Sroufe, 1979). School age children classified as secure in infancy were rated as
more socially active, positive, and popular. These children also described
themselves as less socially anxious than children classified as insecure in
infancy (Bohlin, Hagekull, 8 Rydell, 2000). In addition, preschool children with
secure histories were found to seek out teacher attention in a positive manner
(Weinfeld, et al., 1999) and are rated as more socially competent by teachers
than meir insecure counterparts (Cohen, 1990). At age 10, children classified as
secure in infancy were less dependent on adults at summer camp (Weinfeld, et
aL,1999)

In peer relations, children classified as secure in infancy have not been
found to be bullies or victims in elementary school (Troy 8 Sroufe, 1987) and
they also show less hostility and scapegoatlng of other children than insecure
children (Suess, Grossman, 8 Sroufe, 1992). Children classified as secure at 15
. months of age were found to score significantly higher on measures of social
inititative, prosocial orientation, and popularity than those classified as
ambivalent or avoidant. Securely attached 6—year-olds have been found to be
more competent in their play and conflict resolution in preschool than were
insecurely attached children (Main 8 Cassidy, 1988; Bohlin et al., 2000). Cohen
(1990) also found that securely attached 6 year-old boys were more likely to be
accepted by peers. Furthermore, Sroufe (1983) found that preschool children
with a secure histories were better able to respond flexibly to changing

requirements of situations, especially when frustrated.

38

Insecure Attachment: Ambivalent and Avoidant. Caregivers who are
unresponsive or erratically responsive teach their infants that they cannot
influence their surroundings to meet their needs. As a result, these infants do not
learn to be confident about their autonomy (Sroufe, Fox 8 Pancake, 1983).
Children with insecure-avoidant and insecure-ambivalent histories have been
shown to be more dependent on preschool teachers. Insecure-avoidant children
were found to have more interactions with the teachers and sat next to them
more often during circle time (Weinfeld, et al., 1999). Furthermore, in this
longitudinal study, insecure-avoidant children were described as “mean“ by their
preschool teachers (Waters, et al., 1979). Egeland and his colleagues (1990)
found that of the 16 insecurely attached infants in their sample, 14 emibited
behavior problems in preschool, whereas 15 of the 22 infants classified as
secure were not exhibiting behavior problems in preschool.

Aggression is related to a history of insecure attachment. In a longitudinal
study of attachment, Shaw, Owens, Vondra, Keenan, and \NII'ISIOW (1996) found
that insecure attachment at 12-months of age was a significant predictor of
aggressive behavior, above the clinical cut-off, at 3 to 5 years of age. During
preschool and elementary school, children with insecure attachment
classifications in infancy were found to display more negative affect, anger, and
aggression than were securely attached children (Sroufe, 1983; Sroufe, Schork,
Motti, Lawroski, 8 LaFreniere, 1984).

Boys classified as avoidant in infancy were rated by their teachers in first

through third grades as being more aggressive than boys with secure histories

39

(Renken, Egeland, Marvinney, Mangelsdorf, 8 Sroufe, 1989). Also, children with
insecure-avoidant histories have been shown to be significantly more likely than
other children to victimize play partners. These children were likely to be victims
if they were paired with other avoidant children (Troy 8 Sroufe, 1987). Research
by Cohen (1990) demonstrated that insecurely attached 6 year-old boys were
rated by teachers as less competent and having more behavior problems than
securely attached boys. Furthermore, the boys were less well liked and
perceived as more aggressive by other students in their class.
Disorganized/Disoriented Attachment. Infants classified as disorganized
are at the most risk for later behavior problems (Solomon, et al., 1995). A recent
meta-analysis conducted by van IJzendoom (1999) found that the prevalence of
disorganized attachment is relatively low in middle class, non—clinical samples.
She found that only 14% of infants from middle class samples were classified as
disorganized, compared to 24% from low SES groups. Infants from low SES
groups who have also been maltreated are even more likely to be classified as
disorganized. For example, 82% of infants who were from families with low
incomes and maltreated were classified as disorganized (Carlson, Cicchetti,
Barnett, 8 Braunwald, 1989). Other studies have demonstrated that 62% of
infants growing up in low-income families with depressed mothers were classified
as disorganized (Lyons-Ruth, Connell, Grunebaum, 8 Botein, 1990). Clearly, as
risk factors increase so does the likelihood that an infant or child will develop a

disorganized attachment.

40

A number of family factors have been associated with the development of
disorganized attachment. Factors such as parental depression (Lyons-Ruth 8
Jacobvitz, 1999), insensitive and intrusive caregiving (Carlson, 1998), substance
use in the home (Lyons-Ruth 8 Jacobvitz, 1999), and maltreatment (Carlson,
1998; Lyons-Ruth 8 Jacobvitz, 1999) have been correlated with disorganized
attachment. Children classified as disorganized often have caregivers who have
a history of unresolved loss or who are the source of fear for their children (Main
8 Hesse, 1990).

Infants classified as disorganized are at the greatest risk for later behavior
problems. Shaw and his colleagues (1996) examined the relationship between
infant attachment, measured at 12 months, and clinical levels of aggression at
ages 3 and 5. Disorganized attachment at 12 months was a significant predictor
of 5—year—old aggressive behavior above the clinical cut-off. Furthermore, only
17% of secure infants, 31% of aviodant infants, and 38% of ambivalent infants
had aggressive behavior above the clinical cut off at age 5. Clearly, insecure
infants are at much greater risk for aggressive behavior than children who are
classified as secure in infancy. However, infants classified as disorganized were
almost twice as likely as those classified as insecure, to show clinical levels of
aggression at age 5. In Shaw et al’s (1996) sample, 60% of disorganized infants
had clinically elevated aggression scores at age 5. Other studies support the
findings by Shaw and his colleagues (1996). For example, infants classified as

disorganized at 13 months showed externalizing behaviors in preschool (Hubbs-

41

Tait, et al., 1994; Lyons-Ruth et al., 1993) and at age 7 (Lyons-Ruth, et al.,
1997).

The relationship between aggression and disorganized attachment has
received more attention by attachment researchers than other forms of behavior
disorders. However, while the relationship is significant, only a minority of
disorganized infants become aggressive (Lyons-Ruth 8 Jacobvitz, 1999). Some
research has demonstrated that disorganized attachment status is also related to
other behavior problems. For example, 71% of preschool children and 83% of
school age children from a high risk population, who had been classified as
disorganized at 18 months, showed above normal levels of hostility in the
classroom (Lyons-Ruth, et al., 1993). Disorganized attachment has also been
shown to be significantly related to dissociative symptoms at age 16 and 19
(Ogawa, Egeland, 8 Carleson, 1998, as cited in Weinfield et al., 1999). In a
prospective longitudinal study, infants classified as disorganized at 12 and 18
months scored higher on teacher ratings of dissociative behavior and
externalizing behavior in first, second, third, and sixth grades. In adolescence,
they scored higher on internalizing behavior and overall psychopathology
(Carlson, 1998).

Attachment in the Preschool Years

Although Bowlby considered attachment to be important across the life
span, most of the research conducted thus far has focused on the period of
infancy. Greenberg, et al. (1990) attributed this to several factors. The first is

that most of the attachment researchers were specialists in infancy; second,

42

social development during the preschool years was relatively neglected between
the mid-1960’s and mid-1980’s. Finally, many researchers thought that
attachment was only relevant in infancy, despite what the theory proposes.

Research has shown that not all problems in attachment are rooted in
infancy (Cummings 8 Cicchetti, 1990). Since preschoolers are not yet able to
protect themselves from many sources of danger, attachment continues to play a
large role in their development (Cicchetti, et al., 1990). Insecurity may develop
after infancy as a response to a change in the caregiving environment such as
the birth of a new sibling, traumatic event, or divorce. Also, during this time
children may spend more time away from their parents and as a result, may form
insecure attachments between other adult caregivers or siblings (Ciccheti, et al.,
1 990).

In the late 1980’s, as a response to the lack of attachment research in the
preschool period, attachment scholars fomied a group to discuss how to go
about expanding research in the area of preschool children’s attachment. They
concluded that a theoretical framework for studying attachment in the preschool
years should include the fundamental features of the attachment system. For
example, its function to protect, proximity to a caregiver as a predictable
behavioral outcome, and an internalized felt security (Cicchetti et al., 1990;
Schneider-Rosen, 1990). They believed that, following the theory, attachment
behaviors during the second and third years of life occur with the same frequency
and intensity as they did in infancy (Bowlby, 1982). The circumstances that elicit

attachment behaviors change, however (Schneider-Rosen, 1990).

43

Expanding attachment research to include the preschool years of a child’s
life is important if one is to develop a greater understanding of the behaviors
associated with different attachment styles. Preschool children carry protective
factors, such as a secure attachment, and risk factors, such as stressful events
and family discord fomard from infancy (De Mulder et al., 2000). These factors
may also play an important role in the stability or instability of preschool
children’s attachment representations. Once a representational model is in place
it guides the processing of incoming information, making intemal working models
unlikely to change. However, when the models become inadequate due to
external factors affecting the relationship, internal working models are
restructured (Bretherton, Ridgeway, 8 Cassidy, 1990). Cicchetti et al. (1990)
support this view, they argue that:

“Once attachment develops, it continues to undergo
transformations and reintegrations with subsequent accomplish-

ments such as emerging autonomy and entrance into the peer

world. Thus, children are continually renegotiating the balance

between being connected to others and being independent and

autonomous as they encounter each new developmental phase”

(Cicchetti et al., 1990, p. 3).

Attachment and Children ’8 Behavior
As children enter preschool their attachment relationships become

internalized into a more general representation. Until recently measures of
attachment representations in preschool were not available, and as a result this
is an area that needs to be studied. Research that has been conducted has
generally used an extended version of a laboratory separation reunion developed
by Main and Cassidy (1988). The limited number of studies that have been

conducted using this measure have consistently found that attachment

44

representations in preschool are related to concurrent and later behavioral
functioning. Children classified as disorganized between the ages of 3 and 5
scored lower on maternal ratings of social competence and higher on ratings of
behavioral problems in preschool (Moss, Parent, Gosselin, Rousseau, 8 St-
Laurent, 1996). Two years later, the children from this sample who were
classified as disorganized scored higher on teacher reported behavior problems
and had lower math scores than children classified as secure or resistant. They
also scored lower on teacher reported measures of academic self-esteem (Moss,
Rousseau, Parent, St-Laurent, 8 Saintong, 1998). Preschool children classified
as disorganized have also been shown to exhibit aggression and externalizing
behaviors toward peers (Speltz, Greenberg, 8 DeKlyen, 1990).

Solomon and her colleagues (1995) also used the Main and Cassidy
(1988b) system to classify 6 year-old children along with a new measure of
attachment designed by the first two authors of the study (George 8 Solomon,
1990, 1996, 2000). They found that children’s use of defensive exclusion could
accurately differentiate between the four attachment categories and was
correlated with Main and Cassidy’s (1988b) laboratory reunion. This study
supported the findings of previous research; disorganized children were rated
higher by parents and teachers on both internalizing and externalizing behaviors
(Solomon, et al., 1995).

Attachment and Head Start
The developmental tasks of preschool, to manage affect within social

interaction and to begin to form peer relationships (DeMulder et al., 2000) are

45

closely tied to the development of a secure attachment. A recent report released
by the Child Mental Health Foundations and Agencies Network (2000) indicated
that social and emotional school readiness is an important factor in early school
success and later accomplishments in the workplace. Furthermore, this report
indicated that social and emotional competence is rooted in the attachment
relationships that are developed during the first year of life. The overarching goal
of Head Start is to promote social competence and school readiness in children.
One way that Head Start accomplishes this is by strengthening families as the
primary nurturers of their children.

Clearly, it is important to study attachment representations in children
attending Head Start There are, however, only a few studies that have
investigated attachment using Head Start samples (Call, 1999; Ventura-Cook,
1997; Bost, Vaughn, Washington, Cielinski, 8 Bradbard, 1998). Only one of
these studies (Call, 1999) looked at the relationship between insecure
attachment and behavioral problems. While a significant relationship was found
between insecure adult romantic attachment and behavior problems, the sample
size was very small indicating that more research with this population is
warranted. In addition, the above study used a measure of attachment that has
limited usefulness. Self-report measures of current romantic attachment style
are not correlated with adult measures of attachment that examine the extent to
which one’s parents displayed care and protection (Steele, Steele, 8 Fonagy,
1996), which has been shown to be related to attachment representations in

children (Fogany, Steele, 8 Steele, 1991). More attachment research needs to

46

be conducted with preschool children. Future studies need to improve on
previous research that has attempted to measure attachment using recently
developed assessments. These assessments are often limited in their
usefulness because they fail to differentiate validly among all four main
attachment groups; lack concurrent construct validity with other child attachment
measures; and fail to replicate findings among studies using the same
attachment measure (George 8 Solomon, 1990, 1996, 2000).

Parental Factors Related to Attachment Representations in Children

The above research demonstrates that attachment representations are
stable over time, and that insecure and disorganized attachment have been
correlated with behavior problems in preschool and elementary school children
and adolescents. Because aggression is stable over time and has been
associated with low academic achievement it is important to understand what
factors influence a parent’s ability to form a secure attachment. Research in this
area has been relatively neglected (Belsky, 1999; Vondra et al., 1999).
Social Support

There is evidence that the amount and nature of contact with and support
from significant others affects the way that parents interact with their infants
(Andersen 8 Telleen, 1992). For example, low-income African American
mothers with larger social networks tend to be more responsive in interaction
with their infants (Burchinal, Follmer, 8 Bryand, 1996). Crockenberg (1981) found
that mothers whose infants were assessed as being irritable shortly after birth

and who had low levels of social support were more likely to have insecurely

47

attached infants. When mothers had high levels of social support, infant
irritability did not have an impact on attachment quality. Crittenden (1985) found
that in a group of infants with abusive or neglectful mothers, low social support
predicted insecurity. This may indicate that social support is a mediator of
security for low levels of nurturing parenting during infancy. These studies have
all addressed social support and attachment in infancy, but they do not address
the relationship between parental social support during their child’s preschool
years, parenting attitudes or their child’s attachment representation.
Psychological Well-Being

Parents” psychological health and well being effect the quality of care that
they provide their child (Belsky, 1984). Furthermore, psychologically healthier
parents are more likely than less psychologically healthy parents to have infants
who are securely attached to them (Belsky, 1999). Research has shown that
mothers who describe themselves as self-confident, independent, cheerful,
adaptable, and affectionate were more likely to have secure infants (O’Connor,
1997, as cited in Belsky, 1999). These findings are also true for high-risk and low
SES households (Jacobson 8 Frye, 1991). Depression has been associated
with attachment insecurity (Lyons-Ruth et al.,1990). Cummings 8 Cicchetti,
(1990) note that:

“[Previous] research documents the effects of the physical

unavailability of attachment figures on the development of children.

From the perspective of attachment theory, however, the fact of

physical absence may be less important than the psychological

unavailability of the attachment figure to the child during periods of

separation or loss. Thus, conditions that result in high levels of

psychological unavailability, even when physical absence is not
a factor, may produce similar outmmes in terms of increasing

48

children’s felt insecurity. One instance where this might occur is

when parents have major and extended episodes of unipolar or

bipolar depression” (p.347).

Research by Gelfand 8 Teti (1990) found that children of depressed
mothers were at heightened risk of insecure attachment. Children whose parents
have major depressive disorder were more likely to be classified as insecure than
children whose parents have minor depression or no affect disorder (Radke-
Yarrow, Cummings, Kuczynski, 8 Chapman, 1985). This research, however was
conducted prior to the discovery of the disorganized attachment classification
and as a result can not report the correlation between maternal depression and
disorganized attachment. This research did find, though, that infants whose
parents had the most severe depression also had characteristics of both avoidant
and ambivalent attachment, they labeled these children very insecure.

In a sample of middle income families, DeMulder and Radke-Yarrow
(1991) found that 50 percent of infants and preschoolers whose mothers had
bipolar depression were disorganized. Twenty-five percent of children whose
mothers had unipolar depression were classified as disorganized compared to 18
percent of the children in the control group. Teti, Gelfand, Messinger, and
Isabella (1995) also found that maternal depression was a significant predictor of
disorganization in infancy. In their middle class sample, 40 percent of infants
between 16and 21 months of age, whose mothers were in treatment for

depression, compared to 10 percent of infants in the control group were

classified as disorganized.

49

Some research has not found the expected relation between depression,
disorganized, and insecure attachment (Lyons-Ruth 8 Jacobvitz, 1999). The age
at which attachment security is measured may account for this. In studies where
maternal depression has been found to be related to insecure attachment,
attachment was measured after 15 months of age (Belsky, 1999). Another
possible explanation, is that only depression that results in serious clinical
impainnent is necessary before it has a significant impact on infant
disorganization (Lyons-Ruth 8 Jacobvitz, 1999).

Parenting and Attachment

Sensitive and responsive parenting is of central importance in the
development of a secure attachment. Ratings of matemal sensitivity in the first
year of life are linked to security in the first year of life. This is true for samples
with various SES backgrounds and various cultural backgrounds (Ainsworth,
Blehar, Waters, 8 Wall, 1978; Isabella, 1993; Grossman, Grossman, Spangler,
Suess, 8 Unzer, 1985; Egleland 8 Farber, 1984). Prompt responsiveness to
distress (Crockenberg, 1981), appropriate stimulation (Belsky, Rovine, 8 Taylor,
1984), and warmth, involvement, and responsiveness (Bates, Maslin 8 Frankel,
1985) have also been related to attachment security.

Generally, studies of attachment and parenting have used observational
measures of parent-child interaction. Research examining the link between
parental sensitivity and attachment has utilized many methods including single
home visits, brief laboratory assessments, interviews to assess parental

attitudes, and documentation of the frequency of physical contact. De Wolff 8

50

van IJzendoom (1991) believe that these methods may not capture the concept
of sensitivity as Ainsworth and her colleagues originally operationalized it. A
meta-analysis conducted by De Wolff 8 van IJzendoom, (1997) found that
sensitivity appears to be an important condition for the development of
attachment security, but that it cannot be considered to be the exclusive and
most important factor in the development of attachment. This study
demonstrated that “the transactions between social context or clinical conditions,
on the one hand, and attachment on the other need more careful study to
determine the role of sensitivity, and other aspects of parenting and family life in
the development of attachment security more precisely” (De Wolff 8 van
IJzendoom, 1997).

Research on the relationship between parenting and attachment in
preschool is limited. Parents ability to provide a nurturing environment may be
related to their own history of care and attachment. In one study, George and
Solomon, (1996) found that the caregiving system is organized by a
representational structure that is correlated with the child’s attachment to the
mother and the mother’s representation of her relationships in childhood. Some
of what is known about parenting and attachment in the preschool years is based
on infants who were classified as insecure in infancy, but were functioning well in
preschool. These children had mothers who were more supportive, gave clearer
instructions, and set consistent limits in laboratory play tasks (Erickson, et al.,
1985). It is unclear if the infants’ attachment styles also changed from insecure

to secure. DeMulder and RadkeYarrow ( 1991) found that children classified as

51

disorganized in preschool had mothers who displayed the most negative affective
tone. Another study conducted with preschoolers found that security in

preschool, as measured by the attachment Q-Set, was associated with affectively
appropriate, sensitive, flexible mothering (T eti, Nakagawa, Das, 8 Wirth, 1991).
More research in the area of parenting behavior and attachment in the preschool
years is necessary if one is to better understand this relationship.

Life Stress

Attachment researchers have devoted a great deal of lime studying how
caregiving and maternal sensitivity promote attachment security and how
insecurity and disorganization are related to behavior problems. However,
relatively little research has looked at the factors in the parent’s environment that
contribute to their ability to provide this type of care (Belsky, 1999, Vondra et al.,
1999).

Some studies have indicated that attachment may be influenced by family
stress (DeMulder et al., 2000). Shaw and Vondra (1993) found that cumulative
family stressors, differentiated secure infants from insecure infants, especially for
families with three or more stressors. They also reported that if the stressors
impacted caregiving or maternal functioning, they were more likely to impact
attachment. In a sample of preschool boys, security with mother and family
stress were significantly related to teacher reports of aggression in preschool
(DeMulder et al., 2000). This study also supported the finding that family stress

may influence the quality of the ongoing attachment relationship.

52

Because high-risk, low-income families have a higher proportion of
insecure and disorganized children than do low-risk, middle-income families, it is
important to understand how a secure attachment can be facilitated in high-risk
families. Research has suggested that alleviating the degree of stress impacting
high risk families could help decrease later behavior problems in children and
assist in the development of secure attachments (Call, 1999). This research is
relevant to the current study because it was conducted with a sample of Head
Start families. However, these findings need to be replicated due to the small
sample size of the study and the use of a problematic measure of attachment.

Conclusion

Chapter two has reviewed relevant literature regarding the prevalence and
stability of behavior problems in preschool children as well as factors contributing
to the development and persistence of behavior problems, including internalizing
and externalizing problems. This chapter also explored maternal factors related
to parent-child attachment, the relationship between parenting and attachment
and between parenting and aggressive behavior. Clearly, there has been a great
deal of research regarding attachment in infancy and adult representations of
attachment. Research, however documenting the influence of a parent’s ecology
on the caregiving environment, parental behaviors, and attachment has been
lacking. In addition to this, the research on attachment during the preschool
years has been lacking, despite Bowlby’s lifespan perspective. This study will
address deficits in the attachment research to date and contribute to research on

behavior problems in preschool children by investigating the relationship between

53

concurrent attachment classification and behavior problems. This research will
also make a significant contribution by including girls in the sample. Chapter 3

will delineate the research methodology specific to this study.

54

CHAPTER 3
Methodology
Introduction
The purpose of this study was to investigate caregiver psychological and
environmental factors that contribute to parenting attitudes, attachment
representations in their children and subsequent child behavior at home and at
school. This chapter introduces the methodology used to fulfill the purpose of the
study by including a reiteration of the research questions followed by a
discussion of the research design. In addition, this chapter will discuss the
methodology, sampling, hypotheses, data collection and data analysis
procedures.
Research Questions
. Are factors in the caregiver's environment associated with children’s
attachment representations?
0 Are factors in the caregiver’s environment predictive of their attitude about
nurturing parenting?
a Are caregiver’s nurturing parenting attitudes associated with children’s
attachment representations?
0 Is caregiver psychological well being associated with children’s attachment
representations?
. Do children’s mean aggression scores at home and at school vary as a
function of attachment representations

c Is aggression at home correlated with aggression at school?

55

Research Design
In order to carry out the objectives of this research most effectively, a
correlational design in a controlled setting was used. This study examined
caregiver environmental and psychological factors that contribute to parenting
attitudes and preschool children’s attachment representations. This study also
examined the relationship between caregiver environmental factors,
psychological factors, parenting attitudes, children’s attachment representations
and their aggressive behavior in school and at home. The unit of analysis in this
study was 52 caregivers and their children who attend Head Start in Ingham
County, Michigan.
Conceptual and Operational Definitions

0 Caregiver

Conceptual: The person who provides care for the child.

Operational: For purposes of this study, the caregiver was defined as the

adult who reported that helshe is the person who provided the majority of the

care for the child. This person also reported that they had legal custody of

the child
0 Caregiver environmental factors

Conceptual: The caregiver's report of social support and life stress.

Operational: For the purposes of this study social support was measured by
the caregiver’s score on two scales of the Social Support Questionnaire (SSQ,
Sarason, Levine, Basham, 8 Sarason, 1983). The first scale (SSQ-N) assesses

the caregiver's perception of how available his/her support network is. The

56

second scale (SSQ-S) assesses how satisfied the caregiver is with his/her level
of support. Life stress was measured by the number of items that the caregiver
reported on the Schedule of Recent Events (Daly, 1984).

o Caregiver Psychological Well Being

Conceptual: The caregiver's reported level of psychological symptoms in the
past week.

Operational: For the purpose of this study, caregiver psychological well being
was measured by the caregiver’s score on the general symptom index of the
Brief Symptom Inventory (BSI, Derogatis, 1992). This score is the average of the
nine subscales.

. Nurturing Parenting Attitudes

Conceptual: The caregiver‘s attitudes about nurturing parenting.

Operational: For the purpose of this study, nurturing parenting refers to

attitudes regarding parental expectations, empathy towards children’s needs,

values related to corporal punishment, role reversal, and power and
independence. These attitudes were measured by the caregiver’s standard
score on each of these subscales of the Adult-Adolescent Parenting

Inventory-2 (AAPl-2, Bavolek 8 Keene, 1999).

c Children’s Aggressive Behavior
Conceptual: The child’s aggressive behavior in the home and in the school.
Operational: For the purpose of this study child aggressive behavior in the
home was the caregiver’s report of his/her child’s aggressive behavior on the

Child Behavior Checklist (CBCL, Achenbach, 1991). The raw score on the

57

aggression scale of this instrument was used to measure the child’s aggressive
behavior in the home. Aggressive behavior in school was measured using the
teacher‘s report of the child’s aggressive behavior on the Teacher Rating Scales
(T RS, Reynolds 8 Kamphaus, 1992). Specifically, this was measured using the
raw score on the aggression sub scale of the TRS.

c Children’s attachment representation

Conceptual: The child’s attachment representation.

Operational: For the purpose of this study, children’s attachment
representations were the child’s secure, avoidant, antivalent, or disorganized
classification on the Six-Year Attachment Doll Play Attachment Classification
System (George 8 Solomon, 1990, 1996, 2000).

Sampling Procedure

Data were collected from 62 caregivers whose children were attending
Head Start. Half of the sample was drawn from full day, full year Head Start
classrooms and the other half of the sample was drawn from center based, part
day Head Start classrooms. The sampling procedure was conducted in two
phases.

In phase one, the Associate Director of Head Start provided the doctoral
student with the location of the full day, full year classrooms and the names of
the head teachers for each of those classrooms. There were seven full day
classrooms in Ingham County, all of which were sampled. There were
approximately 17 students in each classroom. The research was explained to

each lead teacher and her consent to participate was obtained. The teacher

58

chose a week during the scheduled data collection time in which no field trips
were planned. Arrangements were made so that parent self-report data could be
collected during times when caregivers drop off and pick up their children.

The week before sampling and data collection was to take place, flyers
describing the research were posted in the school and sent home with children
who were 4 years old. The week that sampling and data collection took place,
flyers that described the research were handed to each parent. The flyers
indicated that caregivers could sign up to participate in the study at the school
that same afternoon. Caregivers were asked to pick a time during that week,
while their child was in school, to fill out self-report measures and information
about their child.

After the caregiver finished filling out self-report information the caregiver
introduced his/her child to the student conducting the research. The child was
told that his/her caregiver had given permission to take the child to a room in the
school building to make a movie about a pretend family. The child was told the
name of the research assistant who would be playing with him or her, the day
that it was going to occur, and that helshe would be introduced to the research
assistant. The child was also told that his/her caregiver and teacher would help
them remember the day that the play time was to occur.

On Friday or Monday following caregiver completion of self-report data,
the caregiver’s child was taken out of class to participate in the doll play
interview. The classroom teacher and the doctoral student introduced each child

to the doll play administrator. Each child was reminded that he or she was going

59

to make a movie about a pretend family and then would return to the classroom.
The interviews were held in an unoccupied room in the building where the child’s
classroom was located. All interviews were video taped so that they could be
transcribed. Prior to the scheduled interview with the child, the doll play
administrator set up the doll house, and began recording. Each video had the
date and the time that the child was interviewed. The video was not turned off
until after the doll play administrator had retumed the child to his/her classroom.
The doll house was set up the same for every interview.

If caregivers were not available during school hours to fill out self-report
measures or they did not want their child taken out of class, an alternate time
was arranged for them to participate. These additional times were in the
evenings and on weekends at the Family and Child Clinic at Michigan State
University.

Thirty-two caregivers signed up to participate in the study. Twelve of
these caregivers failed to follow through with their scheduled appointment, so
they were dropped from the study. Twenty caregivers and their children
participated in phase one of data collection, which took place in August and
September, 2001. Due to the low number of participants, buildings that housed
more than one classroom were sampled again in December, 2001. An additional
twelve caregivers expressed an interest in the study, but two parents failed to
show up for their scheduled appointment to fill out self report measures so they
were dropped from the study, leaving ten caregiver-child dyads who participated.

In phase one, two children refused to leave their classrooms for the doll play

interview. During phase one of data collection three families were dropped from
the sample because they did not fit selection criteria. Two caregivers did not
speak English well enough to accurately fill out the measures; the caregivers.
reported that they were fluent in English, but had a great deal of difficulty
understanding directions and words to describe problem behaviors. One child
was enrolled part time in a pro-primary impaired classroom for language and
social-emotional delays. Another family was dropped from the sanple because
the caregiver had only filled out every other question on only the front pages of
the measurement instruments. When she was asked to complete the surveys
she did so in 10 minutes. Most parents took 40 to 60 minutes to complete the
surveys. This reduced the phase one sanple to 26 parent-child dyads.

Phase two of the sampling procedure was the same as phase one,
however the families that were sampled had children who attended center based,
part day Head Start The Associate Director of Head Start provided the names of
teachers who worked in buildings that housed two or more part day classrooms.
Thirteen classrooms located in Ingham County were sampled. Each classroom
had approximately 17 students who attended the morning session and 17
students who attended the altemoon session. The research was explained to
each lead teacher and her consent to participate was obtained. Arrangements
were made so that data collection could take place when caregivers dropped off
and picked up their children.

The week before data collection was to begin, flyers describing the study

were sent home to caregivers who had 4 year old children. The week that

61

sampling and data collection took place, each caregiver was handed a flyer that
described the research. The flyer indicated that the doctoral student would be at
the school that same afternoon to sign interested caregivers up for the study.
Caregivers were asked to pick a h'me during that week, when their child was in
school, to fill out self-report measures and information about their child.

After the caregiver finished filling out the self-report information, the
caregiver introduced his/her child to the student conducting the research. Each
child was told that his/her caregiver had given permission to take the child to a
room in the school building to make a movie about a pretend family. The child
was told the name of the research assistant who would be playing with him or her
and the day that the doll play was going to occur. The child was also told that
helshe would be introduced to the research assistant, and that his/her caregiver
and teacher would help him/her remember when the playtime would happen.

On Monday following caregiver completion of self-report data, the
caregiver’s child was taken out of class to participate in the doll play interview.
The interviews were held in an unoccupied classroom in the building where the
child’s classroom was located. Data collection took place between September
and December, 2001. Thirty-five caregivers expressed an interest in the study;
two of these caregivers did not show up at their scheduled time to fill out
measures, another parent withdrew from the study after discussing it with her
husband. Data were collected from thirty-two caregiver-child dyads; doll play
interviews were collected from twenty-seven children; Five children chose not to

leave their classrooms for the interview. An additional six dyads were dropped

62

from the sample after data collection was complete because they did not meet
selection criteria. One parent, who the school later reported was unable to read ,
filled out the measures without help, one parent was not fluent in English, and
had been in the country for less than one year. Two of the children did not live
with the parent who filled out the caregiver measures, the teacher reported that
these children saw their parent less than daily. One parent brought her child and
husband to Michigan State, she and her husband each filled out half of the
measures, despite being asked to have only one parent fill out the caregiver
measures. One parent filled out the measures in less than 20 minutes, the
teacher and Family Support Worker reported that they did not think that the
parent could read. This reduced the phase two sample to 26 caregiver—child
dyads.

lf caregivers were not available during school hours to fill out self-report
measures or they did not want their child taken out of class, an alternate time
was arranged for them to participate. These additional times included evenings
and weekend times at the Family and Child Clinic at Michigan State University.

NI of the information that families received and all of the instruments in
this study were submitted to and approved by the University Committee for
Research Involving Human Subjects.

Hypotheses

Ho 1: Caregiver’s perceived social support network size, perceived satisfaction

of social support, and life stress are not associated with children’s

attachment representations.

63

Ha 1: Caregiver’s perceived social support network size, perceived satisfaction of

Ho 2:

social support, and life stress are associated with children’s attachment
representations.

Caregiver’s perceived social support network size, perceived satisfaction
of social support, and life stress are not significant predictors of parental
expectations, empathy towards children’s needs, values related to

corporal punishment, family roles, or power-independence issues.

Ha 2: Caregivers perceived social support network size, perceived satisfaction of

Ho 3:

social support, and life stress are significant predictors of parental
expectations, empathy towards children’s needs, values related to
corporal punishment, family roles, or power-independence issues.
Parental expectations, empathy towards children’s needs, values related
to corporal punishment, family roles, and power-independence issues are

not associated with children’s attachment representations.

- Ha 3: Parental expectations, enpathy towards children’s needs, values related to

Ho 4:

Ha 4:

corporal punishment, family roles, and power-independence issues are
associated with children’s attachment representations.

Caregiver psychological well being is not associated with children’s
attachment representations.

Caregiver psychological well being is associated with children’s

attachment representations.

Ho 5: Children’s mean aggression scores at home and at school do not vary as a

function of attachment representations

Ha 5: Children’s mean aggression scores at home and at school vary as a
function of attachment representations
Ho 6: Children’s aggressive behavior at home is not significantly correlated with
aggressive behavior in school.
Ha 6: Children’s aggressive behavior at home is significantly correlated with
aggressive behavior in school.
Instrumentation

The dependent and independent variable, nurturing parenting, was
assessed by the Adult Adolescent parenting Inventory-2 (AAPl-2, Bavolek 8
Keene, 1999). The AAPI-2 was designed to assess parenting attitudes and child
rearing practices of adolescents and adults. The subscales include: Parental
Expectations, Empathy towards Children’s Needs, Values Related to Corporal
Punishment, Role Reversal, and Power-Independence. The analysis for this
study used the sten scores of the five sub-scales. The internal consistency for
these sub-scales is very good. Alpha levels range from .80 to .92 (Bavolek 8
Keene, 1999).

The independent and dependent variable, life stress was assessed using
the Schedule of Recent Events (Daly, 1984). The Schedule of Recent Events is
a checklist designed to measure the amount of stress experienced from
readjustment to life events. Items include questions about stressful events that
have happened in the past year. Research has shown that the internal

consistency for the scale is very good (alpha=.83).

65

The independent and dependent variable, social support. was assessed
using the Social Support Questionnaire (SSQ, Sarason, et al., 1983). The
strength of the SSQ is that it assesses perceived availability (SSQ-N) of and
satisfaction (SSQ-S) with social support. The SSQ-N instrument yields data on
the number of supports available; the SSQ-S yields data on satisfaction for each
item (Sarason, et al., 1983). Both scales show excellent internal consistency
(SSQ-N, alpha=.97; SSQ-S alpha=.94).

The independent and dependent variable, caregiver psychological well
being, was assessed using the Brief Syn'ptom Inventory (BSI, Derogatis, 1992).
The BSI yields scores in tenns of nine primary symptom dimensions divided by
the total number of responses. The internal consistency is very good with alpha
levels ranging from .71 to .85.

The independent and dependent variable, children’s attachment
representations, was measured using the Six-Year Attachment Doll Play
Classification System (George 8 Solomon, 1990, 1996, 2000). The doll play
interview has inter-rater reliability of 71% and shows concordance between
representation classifications and attachment classifications based on reunion
behavior in the Main and Cassidy (1988) laboratory reunion of 79% (Solomon 8
George, 1999). Furthermore, the Six-Year Attachment Doll Play Classification
System differentiates among all four-attachment groups (Solomon 8 George,
1990, 1996, 2000). The investigator was trained by the creators of the measure
and established inter-rater reliability of 80% using an alternate sample. The

creators of the measure and the doctoral student coded seventeen of the stories

from this sample. Inter-rater reliability for the current sample, between all three
raters was 100%.

The dependent variable, children’s aggressive behavior in the home,
was assessed using the Child Behavior Checklist (CBCL, Achenbach, 1991),
which was filled out by the caregiver. The CBCL is designed to measure
children’s problems as reported by parents. The CBCL has been used widely in
research on social-emotional development with children ages 4-18 (Drotar, Stein,
8 Perrin, 1995). The CBCL reports very good internal consistency with alpha
levels ranging from .88 to .92. For the analysis of this study the aggressive
behavior subscale was used

The dependent variable, children’s aggressive behavior at school, was
assessed using the Behavior Assessment System for Children (BASC, Reynolds
8 Kamphaus, 1992). The BASC is an integrated set of self-reports of
personality, teacher-rating scales (T RS), parent rating scales, a developmental
history, and an observation protocol. For the purposes of this study, the TRS
was used. One strength of the BASC is that in addition to measuring
externalizingbehavior, internalizing behavior, and attention problems, it also
measures the child’s adaptive skills. Intemal consistency for this measure is
quite good with alpha levels ranging from .80 to .90. Furthermore, conelation
between the TRS and Achenbach’s Teacher Report Form are in the .80s and
.90s.

67

Data Collection

All of the parent data were obtained from caregiver self-report. The self-
report measures were completed in 45 to 60 minutes; caregivers filled out these
measures at their child’s school, during regular school hours, or at Michigan
State University. Questionnaires were read to two caregivers who indicated that
they could not read. Children’s behavior was obtained from caregiver and
teacher report. Children’s attachment representations were measured using a
semi-structured doll play interview. Twenty-seven interviews from the full day
sanple and twenty-five interviews from the part day sample took place in an
unoccupied room in the building where the child’s classroom was located. Three
parent-child dyads from the full day sample and two parent-child dyads from the
part day sample came to Michigan State University to participate in the study.
For these families, the caregiver self-report data and the child interview took
place simultaneously.

Prior to the doll play interview, the child was allowed to familiarize
him/herself with a doll house that was constructed for the interview. The
administrator then introduced a story stern about lost pets and the child was
asked to finish the story. Following the introduction, the administrator introduced
four story stems describing a hurt knee, a monster in the bedroom, a separation,
and a reunion story. The children were asked to complete the stories and were
given between three and five minutes to complete the story before the next story

stem was introduced. The interview was videotaped and transcribed. The

68

transcripts of the children’s stories were coded by the doctoral student and each
child was assigned an attachment classification.

Upon completion of the doll play interview and the caregiver self-report
measures, the caregivers were sent a $25 gift certificate to a local grocery
superstore. After the children at each school completed their doll play interviews,
the Teacher Rating Scale was mailed to the child’s teacher. Completed forms
were returned in a self-addressed stamped envelope. The teachers received a
$10 gift certificate to a teacher supply store for every completed measure that
was received.

Sample Characteristics

Sixty-two caregiver-child dyads were recruited from 7 full day and 13 part
day Head Start classrooms in Ingham County, Michigan. It was discovered
during data collection that 10 of the caregiver-child dyads did not meet selection
criteria for the study. As a result of dropping families from the study, the sample
that was analyzed comprised of 52 caregiver-child dyads. Half of the children
were enrolled in full day Head Start (N=26) and half were enrolled in part day
Head Start (N=26).

Child Characteristics

The children were approximately 4 years old (M= 54 months, range 40-68
months). Twenty males and 32 females participated in the study. The majority
of the children were Caucasian (N=15), African American (N=12), or multi-racial
(N= 18). One child was Native American and 6 children were Hispanic. An

overview of child characteristics can be found in Table 3.1.

69

Table 3.1

 

 

 

Child Characteristics
Variable N Percentage
Sex
Male 20 38.5
Female 32 61.5
Classroom
Full day 26 50
Part day 26 50
Race
Caucasian 15 28.8
African American 12 23.1
Hispanic 6 1 1.5
Native American 1 1.9
Multi-racial 18 34.6
Caregiver Characteristics

All of the caregivers filling out measures reported that they had legal
custody of the child who participated in the study, all of the caregivers were the
child’s biological parent. Two fathers and 50 mothers completed self-report
measures. Their ages ranged from 19 to 39 years (M=27.9). The majority of the

parents were either Caucasian (N=24) or African American (N=13). Eight

70

caregivers were multi-racial and 7 were Hispanic. Half of the parents were
single; 18 had never been married, 7 were divorwd and 1 was separated.
Eighteen parents were married and 8 were in a committed relationship that had
lasted longer than 6 months. The caregivers worked an average of 19.4 hours
per week (range 0 to 50 hours) and had an average income of $1,626 per month.
The families had experienced an average of 11.7 stressful events in the past
year (range 1-36). An overview of caregiver characteristics can be found in
Table 3.2.
Data Analysis
Attachment is a categorical variable, while the others are continuous.

Various statistical methods are appropriate for these analyses, depending on
whether or not the attachment will be the dependent or independent variable and
on whether there will be multiple independent variables. The following research
questions were analyzed using crosstabulation.
o Are factors in the caregiver's environment associated with children’s

attachment representations?
0 Is caregiver psychological well being associated with children’s attachment

representations?
e Are nurturing parenting attitudes predictive associate with children’s

attachment representations?
The following research question was analyzed using regression analyses:
0 Are factors in the caregiver’s environment predictive of their attitudes about

nurturing parenting?

71

Table 3.2

 

 

 

Caregiver Characten‘stics
Variable N Percentage
Sex
Male 2 3.8
Female 50 96.2
Race
Caucasian 24 46.2
African American 13 25
Hispanic 7 13.5
Multi-racial 8 15.4
Marital Status
Single 18 34.6
Divorced 7 13.5
Separated 1 1.9
Married 18 34.6
Committed Rel 8 15.4
Mean Range
Age 27.9 19—39
Hours worked/week 19.4 0-50
Income/month $1 .626 0-$4,800

Stressful events/yr 11.7 1-36

 

72

The following research question was be analyzed using Anova:
a Do children’s mean aggression scores at home and at school vary as a
function of attachment representations?
The following research question was analyzed using Pearson correlation:
o Is aggression in the home correlated with aggression in school?
Conclusion
This chapter has delineated the research methodology specific to this
study. This chapter has included the purpose of the research, related research
questions, and hypotheses. From these, the study was designed and the
methodology was determined. The sample was described, data collection
techniques and data analysis procedures necessary for completion of the study

explained.

73

CHAPTER 4
Results
Introduction

The main objective of this study was to investigate caregiver psychological
and environmental factors that contribute to parenting attitudes, attachment
representations in their children and subsequent child behavior at home and at
school. This chapter is divided into two sections. The first section presents
frequencies for relevant variables, the second section presents the major findings
for each hypothesis.

Frequencies for Relevant Variables

Relevant variables in this study include those that pertain to the caregiver
and those that pertain to the child. Frequencies were run for variables that have
established cutoffs to determine clinical levels of symptoms for high risk
parenting practices and for those that are known to exist in a certain percent of a
population. Scores on all caregiver variables were obtained by caregiver self
report. These variables included nurturing parenting and levels of clinical
symptoms. Scores on child variables were obtained by caregiver report, teacher
report, and a semi-structured doll play attachment interview. These variables
included: attachment classification, aggressive behavior at home, and aggressive
behavior at school.
Caregiver Variables

Nurturing Parenting. Nurturing parenting was measured using the Adult

Adolescent Parenting Inventory-2 (AAPI-2, Bavolek 8 Keene, 1999). The AAPl-2

74

is divided into 5 subscales, which measure known patterns of abusive and
neglecting parenting attitudes and child rearing practices. Bavolek and Keene
(1999) suggest that to interpret the results of the AAPl-2, one can compare
sample caregivers to those in the population by comparing standard scores. Raw
scores are converted into standard scores that range from 1 to 10. Low standard
scores (1-4) are characteristic of parents who are practicing abusive parenting,
these parents are labeled high risk and comprise approximately 30 % of the total
population. High standard scores (8-10) are characteristic of a nurturing, non-
abusive, parenting philosophy, these parents are labeled low risk and comprise
approximately 16 % of the total population. Mid-range standard scores (57)
represent parenting attitudes of the general population, approximately 54 % of
parents in the general population fall into this category (Bavolek 8 Keene, 1999).
In order to understand how this sample compares to the population on which this
instrument was normed, the caregivers standard scores were grouped into
categories of low risk attitudes, high risk attitudes, and attitudes of the general
population, according to Bavolek and Keene’s (1999) definitions. Below, each
scale is described and each category within the scale is described. Next, the
percentage of parents, from this sample, who fell into each category, is
presented. This allows the reader to compare this sample with what would be
expected in the population.

The first scale, parental expectations, measures thedegree to which
caregivers understand and accept their child’s developmental limitations. Low

risk caregivers tend to have appropriate expectations of children’s capabilities

75

and tend to encourage self-grth and exploration. Caregivers who fall into the
high risk category generally have a lack of understanding of children’s
developmental capabilities and may expect children to achieve at a higher level
than they are capable (Bavolek 8 Keene, 1999). In the current sample 34.6 % of
caregivers (N=18) were judged to be high risk in this area; 59.6 % (N=31) were
judged to have attitudes of the general population; and 5.8 % (N=3) were
considered low risk.

The second subscale, inability to be empathetically aware of children’s
needs, measures the caregiver’s ability to be aware of his or her child's feelings
and to place the needs of their child as a priority. Caregivers who fall into the low
risk category are sensitive to their child’s needs and see their needs as a priority.
Caregivers who fall into the high-risk category are likely to see the normal
developmental demands of children as bothersome and may have difficulty
assisting their child meet his or her needs (Bavolek 8 Keene, 1999). In the
current sample, 44.2 % of caregivers (N=23) were judged to be high risk; 44.2 %
of caregivers (N=23) had attitudes representative of the general population; and
11.5 % of caregivers (N=6) were judged to be low risk.

The third subscale, belief in the value of corporal punishment, measures a
caregiver’s attitudes about the use of physical punishment to discipline his or her
child. Caregivers who fall into the low risk category tend to use alternative
strategies to physical punishment and value non-violent ways of providing
discipline. Scores in the high-risk category indicate that physical discipline is an

acceptable form of discipline that teaches children appropriate healthy behaviors

76

(Bavolek 8 Keene, 1999). In the current sample, 57.7% of caregivers (N=30)
were judged to be high risk; 40.4% of caregivers (N=21) had attitudes
representative of the general population; and 1.9% of caregivers (N=1) were
judged to be low risk.

The fourth subscale, parent-child role reversal, measures the degree to
which caregivers reverse parent-child roles and the possibility that they look to
their own child for care and comfort. Caregivers who fall into the low risk
category tend to understand and accept the needs of self and child, but do not
approve of meeting their own needs at the cost of the child’s needs. Caregivers
who fall into the high-risk category tend to view children as existing to meet the
needs of their caregivers (Bavolek 8 Keene, 1999). In this sample, 30.8% of
caregivers (N=16) were judged high risk; 51.9% of caregivers (N=27) had
parenting attitudes consistent with the general population; and 17.3% of
caregivers (N=9) had attitudes that are considered to be low risk.

The final subscale, oppressing children’s power and independence,
measures the degree to which caregivers believe that children’s independence
and power needs to be oppressed and the belief that children should comply with
parental requests without question (Bavolek 8 Keene, 1999). Caregivers who fall
into the low risk category generally place value on children feeling empowered
and are likely to give children choices and encourage them to problem solve.
Caregivers who fall into the high-risk category typically place value on obedience
and suppression of children’s feelings (Bavolek 8 Keene, 1999). In this sample,

30.8% of caregivers (N=16) were categorized as high risk; 50% of caregivers

77

(N=26) had attitudes consistent with those of the general population; and 19.2%
of caregivers (N=10) were categorized as low risk. An overview of caregiver
attitudes toward nurturing parenting can be found in table 4.1.

Table 4.1

Percentage of Caregivers in High Risk, General Population, and Low Risk

Groups for each Nurturing Parenting Subscale

 

 

Nurturing Parenting Subscale High Risk General Low Risk
Population

Parental Expectations 34.6 (18) 59.6 (31) 5.8 (3)

Empathy 44.2 (23) 44.2 (23) 11.5 (6)

Corporal Punishment 57.7 (30) 40.4 (21) 1.9 (1)

Parent-Child Role Reversal 30.8 (16) 51.9 (27) 17.3 (9)

Power and Independence 30.8 (16) 50 (26) 19.2 (10)

 

Note. - Cell n’s in parentheses.

Caregiver Psychological Well Being. Caregiver psychological well being
was measured using the Global Severity Index (GSI) of the Brief Symptom
Inventory (BSI, Derogatis, 1993). The Global Severity Index is an indicator of the
respondent’s distress level. It is calculated by combining information about
numbers of symptoms and intensity of distress (Derogatis, 1993). Respondents
whose T-scores are greater than or equal to 63 can be considered to have a
positive diagnosis or case. Approximately 12% of the adult female, non-patient

population have T-scores that are of clinical significance (Derogatis, 1993). In

78

the current sample 21.2% (N=1 1) of caregivers had scores that fell into the
clinical or diagnosable range on the GSI.

Child Variables

Aggressive Behavior at Home. Children’s aggressive behavior at home
was measured using the aggression subscale on the Child Behavior Checklist
(CBCL, Achenbach 8 Edelbrock, 1991). The authors provide clinical cut-off
scores for each of the subscales. T scores from 60 to 63 are considered
borderline clinical and discriminate between deviant and nondeviant groups
(Achenbach, 1991). For the purposes of this study T scores 2 62 were
considered of clinical concern. In this sample 80.8 % (N=42) of children were
categorized as having normal levels of aggressive behavior; 17.3 % (N=9) had a
positive diagnosis or case.

Aggressive behavior in School. Children’s aggressive behavior at home
was measured using the aggression subscale on the Teacher-Rating Scales
(TRS, Reynolds 8 Kamphaus, 1992). The authors report that T scores above 60
are clinically significant. In this sample 23.1% (N=12) of the children had
aggressive behavior that was clinically significant, 76.9% (N=40) of the children
had aggressive behavior that was age appropriate.

Attachment Classification. Children’s attachment representations were
measured using the Six-Year Attachment Doll Play Classification System
(George 8 Solomon, 1990, 1996, 2000). Studies have demonstrated that
approximately 13% of infants from a middle class sample are classified as

disorganized (Main 8 Solomon, 1990). Children in high-risk situations are more

79

likely to be classified as disorganized. For example 28% of infants from multi-
problem families (Spieker 8 Booth, 1988) and 54% of infants from low income
depressed mothers were classified as disorganized (Lyons-Ruth, Connell,
Gruenbaum, 8Botein, 1990). In this sample, 4.7% (N=2) of children were .
classified as secure; 44.2% (N= 19) were classified as avoidant, 20.9% (N=9)
were classified as ambivalent, and 30.2% (N=13) were classified as
disorganized. Due to the small number of children classified as secure, the
expected cell size was less than 5 in all statistical analyses. As a result, secure
children were not included in the analysis.
Major Findings

Group Differences

Preliminary analyses were conducted to examine the possibility of group
differences and to guide future analysis. Previous studies have demonstrated
that there are cultural differences in parenting attitudes, while others have
demonstrated that when socioeconomic status is controlled, these differences
are minimal (Julian, McKenry, 8 McKelvey, 1994). One way ANOVA was used to
determine if caregiver race accounted for significant differences in parenting
attitudes. Results indicate that caregiver race did not account for significant
differences in inappropriate expectations, F(3,48) = 1.185; p = .325, empathy
towards children’s needs, F (3, 48) = 1.145; p = .341, values related to corporal
punishment, F (3, 48) = .477; p = .700, role reversal F (3, 48) = .336; p = .799, or

power-independence issues, F (3, 48) = 2.53; p = .068.

80

Research has indicated that sex differences in child behavior begin to emerge
at preschool age (Zahn-Waxler et al., 1990; Achenbach et al., 1987).T-tests were
run to determine if mean levels of behavior at home and behavior at school
varied as a function of child sex. Results indicated that the mean of children’s
aggressive behavior (boys, M = 9.75; SD = 7.282; girls, M = 8.28; SD = 5.726) at
home did not vary as a function of the child’s sex, t(50)= .810, p = .422. The
child’s sex also did not account for mean differences (boys, M = 9.0; SD = 6.190;
girls, M = 7.0; SD = 5.663) in children’s aggressive behavior in the classroom,
t(48) = 1.179, p = .244.

The Associate Director of Preschool programs indicated that families whose
children attend full day Head Start are different than families whose children
attend part day Head Start. Families whose children attend full day Head Start
are required to be employed full time and qualify for assistance from the State of
Michigan for childcare. T-tests with this sample demonstrated that full day (FD)
and part day (PD) enrollment did not account for differences in aggressive
behavior at home (FD, M = 7.23; SD = 5.332; PD, M = 10.46; SD = 6.941), t(50)
= -1.882, p =.066; aggressive behavior at school (FD, M = 8.08; SD = 5.22; PD,
M = 7.52; SD = 8.869), t(48) = .332, p = .741. Mean levels of family stressful life
events did not vary as a function of class type (FD, M = 11.539; SD = 9.433; PD,
M = 11.885; SD = 10.405), t(50) = -.126, p = .900. Mean scores of inappropriate
expectations (FD, M = 5.04; SD = 1.483; PD, M = 4.73; SD = 1.512), t(50) = .741,
p = .462, empathy (FD, M = 5.23; SD = 2.338; PD, M = 4.69; SD = 2.000), t(50)
=.893, p = .376, corporal punishment (FD, M = 4.15; SD = 1.433; PD, M = 3.81;

81

SD = 1.877), t(50) = .747, p = .458, role reversal (FD, M = 5.96; SD = 2.05; PD,

M = 4.96; SD = 2.200), t(50) = 1.696, p = .096, and power-independence (FD, M

= 5.85; SD = 2.49; PD, M = 5.31; SD = 2.112), t(50) = .840, p = .405 did not vary

by class type. Table 4.2 summarizes means as a function of classroom type.

Table 4.2

Means for Aggression, Stressful Events, and Parenting Attitudes as a Function of

Classroom Type

 

 

Variable Full Day Part Day
M :l: SD M i SD
Child Aggressive Behavior
Home (26) 7.23 :I: 5.33 10.46 :I: 6.94
School (25) 8.08 :I: 5.22 7.52 3: 8.87
Stressful Events (26) 11.54 i 9.43 11.89 :I: 10.41
Parenting Attitudes
Inappropriate Expectations (26) 5.04 i 1.48 4.73 d: 1.51
Empathy (26) 5.23 i 2.34 4.69 i 2.00
Corporal Punishment (26) 4.15 i: 1.43 3.81 :I: 1.88
Role Reversal (26) 5.96 :t 2.05 4.96 :t 2.20
Power Independence (26) 5.85 :I: 2.49 5.31 i 2.11

 

Note. N in parentheses

82

Caregiver Ecological Factors and Children ’3 Attachment Representations

. H01: Caregiver’s perceived social support network size, perceived
satisfaction of social support, and life stress are not associated with children’s
attachment representations.

. Hal: Caregiver’s perceived social support network size, perceived
satisfaction of social support, and life stress are associated with children’s
attachment representations.

Caregiver ecological factors were not associated with children’s attachment
representations. Median scores of social support network size were calculated
and a new dichotomous variable was created. Those whose networks were
smaller had between 0 and 2.52 individuals whereas those with larger networks
had between 2.56 and 6 individuals who offered support. The contingency
coefficient (C) was used to gauge the strength of the relationship. Crosstabs
indicated that social support network size was not associated with children’s
attachment representations and the strength of the relationship was small (C =
.148), x2 (2, N = 40) = .915; p = .633.

Median social support satisfaction scores were calculated to create a
dichotomous variable of satisfaction. Low satisfaction scores were between 0
and 5.6, high satisfaction scores were between 5.74 and 6. The contingency
coefficient (C) was used to gauge the strength of the association. Crosstabs
indicated that satisfaction with social support was not associated with children’s

attachment representations and that the strength of the association was small (C

83

= .229), xZIZ. N = 40) = 2.22, p = .329. Table 4.3 summarizes the association

between caregiver social support and children’s attachment representations.
Table 4.3

Association Between Caregiver Social Support and Children’s Attachment

 

 

Representations
Factors Avoidant Ambivalent Disorganized Total
SSQN
High levels Observed 12 3 6 21
Expected 10.0 4.7 6.3 21.0
Low levels Observed 7 6 6 19
Expected 9.0 4.3 5.7 19.0
Total Observed 19 9 12 40
Expected 19.0 9.0 12.0 40.0
SSQS
High levels Observed 1 1 2 6 19
Expected 9.0 3.8 6.2 19.0
Low levels Observed 8 6 7 21
Expected 10.0 4.2 6.8 21.0
Total Observed 19 8 13 40
Expected 19.0 8.0 13.0 40.0

 

Note. SSQN = social support network size; SSQS = satisfaction with social

suppon

84

A dichotomous variable was created for life stress by splitting the
continuous variable at the median. Low levels of life stress ranged from 0 to 7
and high levels of stress ranged from 8 to 36. The contingency coefficient (C)
was used to gauge the strength of the association. Crosstabs indicated that
stressful life events were not associated with children’s attachment
representations and that the strength of the association was small (C = .241 ),
x2(2, N = 41) = 2.457, p = .293. Table 4.4 summarizes the association between
caregiver life stress and children’s attachment representations.

Table 4.4

Association Between Caregiver Life Stress and Children’s Attachment

 

 

Representations
Factor Avoidant Ambivalent Disorganized Total
Life Stress
High levels Observed 10 3 6 19
Expected 8.8 4.2 6.0 19.0
Low levels Observed 9 6 7 22
- Expected 10.2 4.8 7.0 22.0
Total Observed 19 9 13 41

Expected 19.0 9.0 13.0 41.0

 

85

Caregiver Ecological Factors and Parenting Attitudes

- Ho 2: Caregiver’s perceived social support network size, perceived
satisfaction of social support, and life stress are not significant predictors of
parental expectations, empathy towards children’s needs, values related to
corporal punishment, family roles, or power-independence issues.

. Ha 2: Caregiver’s perceived social support network size, perceived
satisfaction of social support, and life stress are not significant predictors of
parental expectations, empathy towards children’s needs, values related to
corporal punishment, family roles, or power-independence issues.

Table 4.5 reports the relationship between caregiver environmental factors
(social support network size, satisfaction with social support, and stressful life
events) and parenting attitudes. Separate regression analyses were run to
determine if caregiver ecological factors were significant predictors of each
subscale of parenting attitudes. Caregiver ecological factors were not significant
predictors of inappropriate expectations, F (3, 45) = .597, p = .620, values related
towards corporal punishment, F(3, 45) = .979, p = .411, or power-independence
issues, F (3, 45) = 2.391, p = .081. Caregiver ecological factors were significant
predictors of empathy, F (3, 48) = 6.447; p = .001, and role-reversal, F (3, 48) =
3.009, p = .04.

Parenting Attitudes and Children ’8 Attachment Representations

0 Ho 3: Parental expectations, empathy towards children’s needs, values
related to corporal punishment, family roles, and power-independence issues

are not associated with children’ s attachment representations.

86

Table 4.5

Summary of Linear Regression Analyses for Ecological Factors Predicting

 

 

Parenting Attitudes

Parenting Attitudes SSQN SSQS Life Stress Sig. R2
Inappropriate Expectations .030 .176 .079 .620 .038
Empathy -.067 .574” .164 .001“ .301
Corporal Punishment -.040 .012 -.238 .41 1 .061
Role Reversal -. 1 73 .437“ -.005 .040“ .167
Power-Independence .053 .350“ .080 .081 .1 37

 

Note. Standardized betas. Sig = significance of model

*p < .05; “*p < .01; “** p < .001

0 Ha 3: Parental expectations, empathy towards children's needs, values
related to corporal punishment, family roles, and power-independence issues
are not associated with children’ s attachment representations.

Due to the small sample size, it was necessary to make continuous standard
scores on each parenting subscale into ordinal scales. Reducing the number of
categories from 10 to 3 allows for expected counts in each cell to be above 5 or
approach 5 so that cell size was adequate for testing. The categories were made
according to Bavolek and Keene’s (1999) suggestions for interpreting the
parenting profile. The categories that were created were described in an earlier
section of this chapter; the categories are high-risk parenting (standard scores
ranging from 1-4), attitudes consistent with those of the general population

(standard scores ranging from 5-7), and low-risk parenting (standard scores

87

ranging from 8-10). Due to the small number of children classified as secure (N =
2) this category was dropped from the analysis. Crosstabs were run separately
for each parenting attitude subscale, the contingency coefficient (C) was used to
measure the strength of the relationship. Inappropriate expectations were not
associated with children’s attachment representations, x2(4. N = 41) = 4.881, p =
.300, and the strength of the association was moderate (C =.326). Empathy was
not associated with children’s attachment representations, x2(4, N = 41) = 6.976,
p = .137 and the strength of the association was moderate (C = .381). Corporal
punishment was not associated with children’s attachment representations, 12(4,
N = 41) = 2.226, p = .694, and the strength of the association was small (C =
.227). Role reversal was not associated with children’s attachment
representations, X2 (4, N= 41) = 2.807, p = .591, and the strength of the
association was small (C = .253). Power-independence issues were not
associated with children’s attachment representations, x2 (4, N = 41)= 6.860, p =
.143, and the strength of the association was moderate (C = .379).
Caregiver Clinical Symptoms and Attachment Representations
- Ho 4: Caregiver psychological well being is not associated with children's
attachment representations.
. Ha 4: Caregiver psychological well being is associated with children’s
attachment representations.
Psychological well being is defined as the absence of clinical symptoms as
measured by the GSI subscale on the BSI. A dichotomous variable was created

and a crosstab was used to test the association between clinical symptoms (T

88

scores 2 62) and children’s attachment representations. Due to the small
number of children classified as secure (N = 2), this category was dropped from
the analysis. The contingency coefficient (C) was used to test the strength of the
association. There was no association between caregiver psychological well
being and children’s attachment representations, 12 (2, N = 41) = 4.955, p = .084,
and the strength of the association was moderate (C = .328).

Attachment and Child Behavior

- Ho 5: Children’s mean aggression scores at home and school do not vary as

a function of attachment representations.

0 Ha 5: Children’s mean aggression scores at home and at school vary as a
function of attachment representations.

Two separate ANOVA’s were run to determine if aggression at home or at
school varied as a function of children’s attachment representations. Due to the
small number of children classified as secure (N = 2), this category was dropped
from the analysis. Aggression at home, as measured by the raw score on the
aggression subscale of the CBCL, did not vary as a function of children’s
attachment representations, F (2, 38) = 2.63, p = .085. Aggression at school, as
measured by the raw score on the aggression subscale of the TRS, did not vary
as a function of children’s attachment representations, F (2, 38) = 2.43, p = .102.
Aggression at Home and at School
0 Ho 6: Children’s aggressive behavior at home is not significantly correlated

with aggressive behavior at school

89

- Ha 6: Children’s aggressive behavior at home is significantly correlated with
aggressive behavior at school.

Pearson Correlation revealed that children’s aggressive behavior at home, as
measured by the raw score on the aggression subscale of the CBCL, was
significantly correlated with aggressive behavior at school, as measured by the
aggression subscale of the TRS (r = .311; p = .05).

Conclusion

This chapter reviewed frequencies for relevant variables, and presented the

findings for the research hypotheses introduced in chapter 3. The next chapter

will present a discussion of these findings

90

CHAPTER 5
Discussion
Introduch'on
The previous chapter provided the results for this study, which

investigated caregiver psychological and environmental factors that contribute to
parenting attitudes, attachment representations in their children and subsequent
child behavior at home and at school. Included in the previous chapter were
frequencies for relevant variables and the findings for each hypothesis. This
chapter will discuss the findings presented in the previous chapter. First,
frequencies will be discussed, followed by a discussion of each research
question. Finally, limitations, implications and suggestions for future research,
and conclusions will be presented.

Frequencies
Caregiver Variables

There were more caregivers, in this sample, who reported parenting

attitudes that are characteristic of parents who are known to practice abusive
parenting attitudes than in the population on which Bavolek and Keene (1999)
normed the AAPI-2. In general, caregivers from this sample tended to be less
empathetic toward their child’s needs (44.2% compared to 30%), and were more
likely to endorse physical punishment as a form of discipline (57.7% compared to
30%) than would be expected . Only a small number of parents had low risk
attitudes, which are characteristic of a nurturing, non-abusive, parenting

philosophy, as defined by Bavolek and Keene (1999). According to Bavolek and

91

Keene (1999), Approximately 16% of the population have low risk attitudes. In
this sample parents were less likely to have appropriate expectations (5.8%), be
empathetic toward their child’s needs (11.5%), and less likely to use alternatives
to physical punishment for discipline (1 .9%). Caregivers in this sample were
also more likely than the general population to have psychological symptoms that
fell into the clinical or diagnosable range (21.2% compared to 12%, respectively).
Child Variables

Children in this sample had higher levels of aggressive behavior than is
typical. Frequencies from this study indicate that approximately 17.3% had
moderate to severe levels of aggressive behavior at home as designated by cut-
off scores on the CBCL. Furthermore, 23.1% of children were exhibiting
aggressive behavior that was clinically significant, as evidenced by cut-off scores
on the TRS. Other studies have found that 10 to 15 % of preschool children
have moderate (60 < T < 70) to severe (T 2 70) levels of behavior problems.
Children in this sample also were exhibiting higher levels of aggressive behavior
than have previously been found in other samples of Head Start children
(Kupersmidt et al., 2000). Given the low number of children classified as secure
(N = 2) and the relationship between secure attachment and social competence
(Cohen, 1990) and empathy (Waters et al., 1979), the higher than expected
frequency of aggressive behavior is not surprising.

In this sample, girls and boys did not differ significantly in mean levels of
aggressive behavior at home or at school this is contrary to other studies that

found sex differences in externalizing problems to emerge between the ages of 4

92

and 5 (Zahn-Waxler et al., 1990; Achenbach et al, 1987). This finding is
consistent with other research that has been conducted with Head Start children
(Kuppersmidt, et al., 2000).

The percentage of children classified as disorganized is consistent with
other studies of children from multi—problem families (30% and 28%, respectively)
(Spieker 8 Booth, 1988), but less than would be expected in low income samples
of infants who have depressed mothers (30% and 54 %, respectively). This
relatively low number of children classified as disorganized is interesting given
the relatively large number of caregivers in this sample who reported clinical
levels of psychological symptoms. Furthermore, it is interesting and concerning
that only 4.7% (N = 2) children in this study were classified as secure. The low
number of securely attached children could be related to the early child care
histories, not examined by this study, of the children participants, especially when
one considers that 90% of center-based child care is considered to be low quality
(Helbum et al., 1995, as cited in Brazelton 8 Greenspan, 2000). The low number
of securely attached children could also be related to a “crisis of transition”
(Ainsworth, 1990 p. 485). When the study was conducted, some of the child
participants had recently begun preschool (within 3 to 6 months of being
interviemd). However, some of the children in the study had been with the
same teacher for a year before the study began. It the transition to school was
responsible for the small number of children classified as secure, than one would
expect that more children would be classified as secure as the study progressed.

The rate of secure children, however did not increase as the study progressed,

93

actually the children who were classified as secure were interviewed within the
first month of the beginning of the school year.
Caregiver Environmental Factors and Children ’s Attachment Representations

The size of one’s social support network, one’s satisfaction with the
support received, and stressful events in one’s life were not associated with
children's attachment representations. This study examined the association that
social support has on attachment representations, while other studies that found
a relationship between social support and attachment reported that social
support acts as a mediator of security for low levels of nurturing parenting during
infancy (Crittenden, 1985). More research needs to be conducted to support this
earlier finding and to understand if social support network size or quality has a
direct impact on attachment representations.

Unlike other studies which found that family stress influences the quality of
the attachment relationship (DeMulder et al., 2000; Shaw 8 Vondra, 1993),
results from this study did not support the idea that family stress is associated
with different types of insecure attachment. One reason for the lack of
association could be due to the removal of the secure classification from the
analyses. Shaw 8 Vondra (1993) found that cumulative family stressors
differentiated secure infants from insecure infants, especially for families with 3 or
more stressors. The high number of stressors could have contributed to the lack

of secure children in the study.

Caregiver Environmental Factors and Parenting Attitudes

The results of five separate regression analyses indicated that the
independent environmental factor variables (size of social support network,
satisfaction with received support, and life stress) were not significant predictors
of the expectations that caregivers have of their child, their belief in physical
punishment as a form of discipline, or the degree to which parents believe that
children need to be empowered. While the model was not significant, caregivers
who were more satisfied with their support network were more likely to value
empowering children.

Smaller support networks, higher levels of satisfaction with support, and
lower levels of stress were significant pmdictors of empathetic caregiving. The
environmental factors in this model accounted for 30% of the variance in
empathic parenting; satisfaction with social support contributed unique variance
to the model. Additionally, smaller support networks, higher levels of satisfaction
with support, and lower levels of stress were significant predictors of caregiving
that separates the needs of children from the needs of adults (role reversal). The
environmental factors in this model only account for 16.7% of the variance in role
reversal, suggesting that other environmental factors may be more important
than social support and life stress in predicting parenting attitudes. These
findings may be confounded by the relatively few number of parents who
reported low risk parenting attitudes.

While low scores on the AAPI-2 indicate that a parent can be considered
high risk for abusing his or her child and less likely to be considered a nurturing

95

parent, it is difficult to say if parent’s attitudes actually predict if the discipline they
are using can be considered abusive. Furthermore, it is unlikely that the size of
one’s support network, their satisfaction with support, or stressful events would
impact their attitudes about parenting, but it may predict if behavior follows
attitudes. Attitudes may be more of a reflection of a caregivers own history of
being parented. Environmental factors may be more significant in predicting how
parents actually interact with or discipline their children. Future studies .
examining environmental factors and parenting should include attitudes about
parenting and observations of parenting practices.
Parenting Attitudes and Children ’8 Attachment Representations

Crosstabulation indicated that parenting attitudes, thought to be
associated with abusive parenting, were not associated with children’s
attachment representations. Despite the previous finding that disorganization is
more prevalent in samples of children who have been maltreated (Carlson et al.,
1989), measures of parenting attitudes toward discipline may not actually reflect
abusive parenting or account for parent-child interaction that is thought to impact
attachment. The lack of association may also be due the small number of
children classified as secure and the low number of parents reporting an attitude
characteristic of low-risk parenting. However, it is unlikely that these
explanations alone account for the lack of association between parenting
attitudes and children’s attachment representations.

The lack of association makes theoretical sense. Attachment theory

posits that sensitive and responsive caregiving in infancy is related to a secure

96

attachment (Ainsworth et al., 1978; Isabella, 1993; Grossman et al, 1985; EgIand
8 Farber, 1984). Sensitivity in these studies has been defined as the caregiver’s
ability to respond in an appropriate and prompt way to the infant (De Wolff 8 van
IJzendoom, 1997). The measure of parenting attitudes used in the current study
did not measure the degree to which a caregiver responds in an appropriate or
prompt way, observation in the child’s natural environment is a more useful
measure of sensitivity.

Since the function of attachment is protection through the maintenance of
proximity to a caregiver, it makes sense that in infancy, a caregiver’s appropriate
and prompt response is so closely associated with a secure attachment. As
researchers begin to examine the correlates of attachment in preschool,
especially in high-risk samples, where attachment is thought to be less stable,
close observation of children in his or her natural environments is crucial. These
observations will allow researchers to understand how the caregiving
environment in the prescth years, when a child still needs to be protected, but
has gone through significant developmental advancements since infancy,
contributes to attachment security. When combined with measures of
environmental factors, that are thought to impact a caregivers ability to provide
sensitive and responsive care, a clearer picture of the contribution of attachment
theory to development will result.

While measures of parenting attitudes provide interesting and important

information for intervention, they do not measure sensitivity or responsivity, and

97

thus do not provide useful information about the association between parenting
and attachment representations.
Caregiver Clinical Symptoms and Attachment Representations

Crosstabulation indicated that caregiver psychological well-being is not
associated with children’s attachment representations. This finding is contrary to
findings of much of the research that has been conducted (Lyons-Ruth et al.,
1990; Gelfand 8 Teti, 1990). The lack of association in this study could be due to
a variety of factors including the small number of children classified as secure.
Wrthout a larger number of secure children it is impossible to determine if clinical
levels of psychological symptoms differentiate secure from insecure children as it
has in other samples. Furthermore, previous research that has found insecurity
to be related to depression, has found that major depressive disorder, not minor
depression was more likely to differentiate secure from insecure children
(Gelfand 8 Teti, 1990).

It is also plausible that, despite the relatively high number of caregivers
who were exhibiting clinical levels of psychological symptoms, these caregivers
were able to provide sensitive and responsive care to their children. During data
collection a number of parents reported that they knew their scores on the
measure of psychological well-being were high, but they were in therapy to deal
with the issues. These parents, while showing high levels of clinical symptoms,
may be providing sensitive and responsive care because of the services that they
are receiving. Furthermore, other factors may influence this relationship, for

example, social support, life stress, and marital discord. More analyses may

98

provide information about the role that social support plays in moderating the
effects of psychological symptoms.

It is interesting that of the children classified as disorganized in the current
study, none of their parents reported clinical levels of psychological symptoms. It
is possible that this finding is the result of caregivers underreporting symptoms or
that the impact of psychological well-being is greater during the period of infancy
than it is in preschool, possibly because children are exposed to a greater
number of caregivers. It is also possible that environmental factors, not
measumd in this study, account for differences in attachment security in the
preschool period.

Attachment and Child Behavior

Two separate ANOVAs indicated that children’s mean aggression scores
at home and at school did not vary as a function of attachment classification. The
findings in this study are contrary to infant attachment research that has found
that disorganized children are more likely than insecure or secure children to
display hostile behavior in the classroom (Lyons-Ruth et al., 1993). Although
research examining the relationship between attachment in preschool and
concurrent behavior problems is limited, preliminary research has demonstrated
that disorganized children have significantly more behavior problems at home
and at school that insecure or secure children (Solomon et al., 1995).

The lack of association between behavior problems and insecure
attachment has been found in studies that were conducted with infants from low

risk samples, prior to the introduction of the disorganized category (Bates et al.,

1985; Goldberg, Corter, Lojkasek, 8 Minde, 1990). The lack of association in this
study may be related to the small number of secure children or an increasingly
complex relationship between attachment in preschool and concunent behavior
problems. It may be necessary to examine the presence of disorganization in all
stories and begin to look at disorganization as a continuous variable in the
preschool years. An important finding in this study is that there are disorganized
children who are not exhibiting aggression or other behavior problems at home or
at school. This presents an exciting and needed opportunity to understand what
factors mediate behavior problems when children are classified as disorganized.
Aggression at Home and at School

Pearson correlation indicated that children’s aggressive behavior at home
was significantly correlated with behavior at school. This supports another study,
which measured problem behaviors at home and school (Solomon et al., 1995).
These finding make intuitive sense; if children learn at home that aggression gets
results, they are likely to use this behavior at school. It is unclear if this
association is due to communication between teachers and parents, similar
home and school environments, or the persistence of behavior problems once
children enter preschool. These data suggest that for aggressive behavior to
decrease, interventions need to be directed at both environments.

Implications for Future Research

In the past 25 years, researchers have learned a great deal about

attachment in the period of infancy, especially the relationship between sensitive,

responsive caregiving and attachment. Research on attachment in the period of

100

infancy has also made strides in understanding the relationship between
attachment classification and later psychosocial adjustment. Aldiough Bowlby
considered attachment to be relevant throughout the lifespan, there has been
little research on attachment during the preschool years, especially in high-risk
populations. Research with high-risk samples allows scholars to better
understand atypical attachments in the preschool years. Understanding more
about this type of attachment is important because coding of atypical
attachments in the preschool years varies across attachment measures (Barnett
8 Vondra, 1999). By focusing attachment research on high-risk preschool
samples, the precision in which one can code atypical attachments will improve.
One of the inconsistencies in this research was that not all children who
were classified as disorganized had clinical levels of aggressive behavior and
they were not more likely than insecurely attached children to have clinical levels
of aggressive behavior. Barnett and Vondra (1999) suggest that attachment
research has benefited from looking at attachment patterns that were
inconsistent with Ainsworth’s original categorization. In this sample, there was a
great deal of disorganization present in stories of children classified into
organized categories, but because of the way in which the overall attachment
classification is assigned, these children were not classified as disorganized.
Perhaps one of the implications of this is to re-examine disorganization during
the preschool years. It may be that the amount of disorganization present is
more relevant during this time than a child’s status as disorganized. Because

high-risk samples are more likely to include infants and preschoolers with

101

atypical attachments, research with these samples will assist developmentalists
in understanding more about the consequences of atypical attachments. For
example, one can understand if atypical attachment is a risk factor for later
social-emotional problems or if atypical attachments can be used to explain the
processes that are involved in the transformation from risk to psychopathology
(Barnett 8 Vondra, 1999)

In this study, caregiver’s attitudes about discipline were not associated
with children’s attachment representations. This could be the result of using a
measure of self reported attitudes as opposed to direct observation of the parent-
child dyad. It could also be related to the fact that the children in this sample
have been influenced by other relationships with important adults. Future
research in the area of attachment and children’s behavior during the preschool
period should take into account children’s relationships with multiple caregivers.
Many of the children in this study spent at least 20 hours a week with adults in
their classrooms. Some of these children were also in child care before school,
after school, or on weekends. Furthermore, most of the social supports listed by
caregivers were family members, who could have great influence on the child’s
representational model of care, either directly or through contact with the parent.

This study only provided a beginning look at the role of environmental
factors in attachment representations and parenting. Studies in the future need
to be expanded to include observation in the child’s natural environments (home
and school) as well as interview techniques with parents, which will provide

greater depth as to transitions that children are facing. Studies with larger

102

samples, that include multiple measures, by multiple informants of child behavior,
caregiver behavior, and additional environmental factors will be necessary to
inform research about the complex nature of attachment during the preschool
years.

Since comparisons between secure and insecure children were not
possible with this sample, it will be important for future studies to have large
enough samples so that this comparison can be made. Future analyses with this
sample, however will facilitate understanding of the factors that protect insecure
children from behavior problems.

Research on behavior problems in the preschool years should continue to
include assessments of the parent-child relationship, including attachment and
interaction. This research should also focus not just on the correlates of behavior
problems and their stability over time, but instead on factors that reduce the
likelihood of behavior problems at home and at school.

Implications for Practice

This study has many implications for practice. One of the implications is
the support for more funding for Early Head Start. None of the children in this
study was enrolled in Early Head Start, which provides high quality care to high-
risk children under the age of 3. Because this study was conducted at the
beginning of the school year, it allows one to see the many obstacles that
children have faced prior to their entry into Head Start. The frequency data in
this study clearly showed that children and their families have faced many

positive and negative transitions in the past year. They live in families where

103

caregivers are likely to be suffering from psychological symptoms. The children
come to school with more behavior problems than would be expected in a typical
population. The additive impact of multiple risk factors negatively impacts mental
health (Chazan-Cohen, Jerald, 8 Stark, 2001). In addition to the above risk
factors, Head Start children in this sample were less likely to be securely
attached than children in middle class samples, which may impact their ability to
deal with transitions, to behave in socially appropriate ways, and to relate to
peers. When combined, these factors, make it clear that some children are
entering Head Start at a disadvantage, which may contribute to an already
difficult-to-manage classroom.

Attachment security and disorganization tend to be relatively stable even
in high-risk samples (Barnett, et al., 1999). As a result, it is possible that the
children in this sample had experiences in infancy that led to insecure or
disorganized attachment. These relationships impact brain development that is
important in the regulation of emotions and the ability to deal with stress. Recent
research on brain development in infancy has shown that brain maturation in
infants is experience-dependent. The experiences that are critical for brain
development in the first year of life rooted in the transactions between the infant
and his or her primary caregiver (Schore, 2001). The right hemisphere of the
brain is dominant during the first three years of life; this is the same time that
children form attachment relationships. The right hemisphere of the brain is
important throughout the lifespan for processing the physiological and cognitive

components of social-emotional information (Schore, 1994). Developmental

104

psychology and neuroscience are showing that the ability to create and maintain
an internal sense of emotional security during times of stress is established in the
first three years of life through attachment relationships (Schore, 2001).

When one looks at the small number of children in this study who have
secure attachments through a neuroscience lens, the findings are very
troublesome. This perspective suggests that the children in this study may not
have developed the capacity to regulate emotion and cope with stress. Because
a secure attachment relationship is the foundation of mental health and the
context in which healthy social and emotional development continues to grow,
more focus needs to be placed on the opportunities to promote infant mental
health (Chazan-Cohen, et al., 2001). Early Head Start provides an excellent
opportunity to promote healthy child and family development, including secure
attachment relationships. Early Head Start recognizes the importance of social
and emotional needs of infants and the relationship between these needs and
the multiple domains of development (Chazan-Cohen et al., 2001).

Because children in this sample were less likely than children in lower risk
samples to have secure attachment and more likely to exhibit clinically elevated
levels of behavior problems, classroom management and education of the
children may be difficult. One clear implication of this research is that lead
teachers and teacher assistants may need more training to develop skills to help
children in the classroom. Members of the Head Start community, including
teachers and parents have indicated that they would like to understand how to

address the mental health needs of young children and their families (Chazan-

105

Cohen, et al., 2001). Training in social and emotional development, including
attachment, and responding in a sensitive and consistent way may also
challenge children’s current representational models, and foster secure
attachments, Which may moderate future behavior problems and enhance
learning capacity.

One of the goals of Head Start is to support parents as the primary
nurturers of their children. In Ingham County, each classroom is assigned a
family support worker, who assists families with basic needs and connects them
to community resources. This research points to the importance of these
positions to the overall goal of Head Start, which is to promote school readiness.
Family support workers are the important link between home and school. They
have contact primarily with parents and form helpful and supportive relationships
that may influence parent’s attitudes that are associated with abusive parenting.

The higher than expected levels of high-risk parenting, and clinical levels
of child aggressive behavior, and the low number of securely attached children,
despite the lack of association between these variables, underscores the need
for mental health workers in Head Start. These professionals must be trained in
early childhood, be a regular part of the classroom, and have the ability to work
with children both at home and at school. The consistency of behavior problems
at home and at school and previous research that has demonstrated that early
behavior problems, in the context of stressful family and community

environments, remain highly stable suggests that interventions focusing on home

106

and school are a must if children are going to be ready to learn when they leave
Head Start.

One of the most important implications for this study is the need for Early
Head Start in this community. It also points to the need for a curriculum mat
involves parents and is relational in focus. Evaluations of Early Head Start
should compare children and parents who have received Early Head Start
services to those who have not to determine if early environmental factors are
impacting children’s social readiness for Head Start.

Limitations

The relatively short time period in which this study was conducted
contributed to many limitations in the research. The first limitation is the lack of
observational data of child behavior at home and at school and parent—child
interaction. While the information on parenting attitudes provided important
information to Head Start, it did not provide the complex information that has
been so valuable in attachment studies. This type of information would also have
added a great deal to what is known about behavior problems. This study is also
limited in that only one caregiver was included in data collection, when in reality
each child has contact with many important adults who influence their
representational models and behavior.

The second limitation is the small sample size. Although the sample size
is similar to other studies rooted in attachment, the small number of subjects is
problematic when making conclusions about the role of caregiver ecological and

psychological factors in attachment representations in preschool children. The

107

small number of subjects also makes it difficult to determine the relationship
between attachment representations and child behavior, especially because of
the categorical nature of the attachment variable.

The small number of secure children was possibly the largest limitation of
the study because it made it impossible to understand which factors foster
security and the contribution that a secure attachment makes to social
adjustment. It also makes the comparison between secure and insecure children
impossible because the secure category had to be dropped from the analysis.

Conclusions

This chapter has provided a discussion of the findings from this study,
which investigated caregiver psychological and environmental factors that
contribute to parenting attitudes, attachment representation in their children and
subsequent child behavior at home and at school. Overall, this study has
significant implications for Head Start practice and demonstrates the importance
for Early Head Start in high risk communities. It demonstrates that children
entering Head Start have faced a variety of challenges at different levels, which
may impact their ability to learn.

The limited sample size and small number of secure children made it
impossible to understand the correlates of a secure attachment in preschool and
resulting behavior. It did demonstrate that satisfaction with social support is
associated with nurturing parenting attitudes, but that these attitudes are not
associated with insecure or disorganized attachment. Furthermore, it

demonstrated that parenting attitudes do not differentiate between insecure

108

attachment classifications. Life stress, social support, and caregiver
psychological well being were also not associated with insecure or disorganized
attachment classification. The variance in children’s aggression at home and
school could not be accounted for by attachment classification. Clearly more
research needs to be done before any of these results can be considered
conclusive. This study was too narrow to account for the complexities in the
relationship between environmental factors that contribute to caregiving,
children’s attachment representations, and behavior problems at home and at
school. Despite this, it is an important contribution to the small body of literature
on attachment in the preschool years because it provides information on
measures that are potentially not useful in the study of attachment and begins to

add knowledge about contextual factors that are associated with attachment.

109

APPENDIX A

110

Family lnfonnation

Please circle the appropriate answer(s) or fill in the blank for the questions below for
your child who is attending Head Start.

1) My child is
a. Male b. Female

2) His/her date of birth is (please include month, day, year)

 

3) His/her ethnic background is (circle all that apply)
a. White b. African American c. Hispanic d. American Indian
e. Asian 9. Arab h. other

4) The following people live in the home where my child spends most of his/her time
(circle all that apply)

a. biological mother b. biological father
c. biological parent’s significant other d. sibling(s)

e. grandmother f. grandfather

g. foster mother h. foster father

i. parent’s friend(s) j. other relatives
k. other

 

5) How many hours a week, not including the time your child spends in Head Start,
does helshe spend in child care? (if your child does not spend time in
child care outside of Head Start, please go to question 8)

6) If your child is in child care outside of Head Start, is helshe always taken care of by
the same person?
a. Yes b. No

7) If your child is in child care outside of Head Start, what type of care is helshe in
(please circle all that apply)

A relative cares for my child

A friend cares for my child

My child goes to a child care center

My child goes to a child care providers home

Other (please describe)

9.9-9.3!”

 

8) Sometimes children spend a period of time living away from their biological parent(s).
Has your child spent more than 10 days in a row being cared for by someone other
than a biological parent?

a. Yes (please go to question 9) b. No (please skip to question 11)

9) How many times has your child spent more than 10 days in a row being cared for by
someone other than a biological parent?

 

111

10) How long was each separation?

 

 

 

 

The questions below are about the child’s primary caregiver. Please circle the
appropriate answer(s) or fill in the blanksfor the following questions.

11) I am the Head Start child’s

 

a. biological parent b. grandparent c. foster parent
d. other
12) lam
a. Male b. Female‘1
13)I am
a. Single b. Divorced: c. Married

b. Separated c. In a committedrrelationship (more than 6 mo.)

14) Do you have legal custody of thechild described in question 1-10?
a. Yes b. No;

15) How old are you?

 

16) What is your ethnic background? (circle all that apply)
a. White b. African American c. Hispanic d. American Indian
8. Asian 9. Arab h. other

17) Are you employed?
a. Yes b. No (pleaseskipto question 20)

18) How many hours per week do youwork?

 

19) Do you work the same hours every week?

 

20) Are you currently trying to find atjoh?
a. Yes b. No

21) What is your family’s approximate monthly income?

 

112

APPENDIX B

113

Informed Consent
Why are we doing this research?

The purpose of this research is to better understand the needs and suengths of children who attend Head
Start and their caregivers. Specifically we are interested in finding out more about your support network,
recent life changes, and your ideas about parenting. We also are intaested in how your child plays and his
or her behavior both at home and in the classroom. We are interested in positive behaviors, like

coopa'ative play, and negative behaviors, like aggression.
What we are asking you and your child to do?

Weareaskingyoutofill outsomesurveysabout yourselfandyour childwhoattends Head Start. You can
fill these surveys out while your child is still in his or her classroom during regular school hours. If you
have questions about the surveys or would prefer to have someone read them to you, someone will be
available to do this. The ways will take approximately 45 to 60 minutes to complete. The thy after you
complete the stu'veys, we will ask your child to play with us in a room in the school. The researcher who
will be with your child what he or she is playing will begin a story and then ask your child to finish that
story. The stories involve moderately stressful scenarios that are common struggles for many preschool
children. The play is kept at the pretend level, which helps childrar distance themselves fi‘om any stress
they may feel. Also, your child will be able to end the story in any way that he or she wishes, allowing him
or her to resolve any uncomfortable feelings. Your child’s play will be videotaped so that we can make
notes about it late: The play session with your child will take one-half hour. We will also ask your child’s
teacher to tell us about his or her behavior in the classroom. We want to know what he or she does well
and if the teacher has any concerns about his or her behavior.

Participation in this research project is voltmtary. You can decide at any tints that you do not want to
participate in this research. If you child does not want to leave his or her classroom, he or she will be
allowed to stay in the classroom.

It is unlikely that your child will expaience any negative reaction to the play session. In the event that you
have any concerns about your child’s behavior at the day that they participated in the play session you may
contact Dr. Marsha Carolan at (517) 432-3327 or Ann Stacks at (517) 432-2271 ext. 5

Are there any benefits for families who participate?

We Imow that your time is valuable and we are gratefiil that you might be willing to help us understand
more about chilth'en who attend Head Start and their families. After your child finishes his or her playtime
with us we will send you a $25 gift certificate to Meijer. We will ask you to fill your address out on an
envelope so that we can send these to you as soon as possible. You should receive your gift certificate
within 5 business days of your child’s playtime. If your child is in school on the day that the interviews are
conducted, but does not want to leave his/her classroom, you will still receive the $25 gift certificate.

How will we protect your privacy?

Your privacy will be protected to the maximum extent allowable by law. You will be instructed not to put
your name or your child’s name on any survey that you fill out. We will put a number on your surveys.
This number will also be used on the survey that the teacher fills out, on the video of your child’s play, and
on any notes made fi'om that video. If you sign the bottom of this page, indicating that you will voluntarily
participate in this study, this form will be kept separate from the surveys that you fill out. All of the
information we receive will be kept in a locked file cabinet at Midiigan State University. We will not
share information about your child’s play or your answers to our surveys with your child’s teacher or other
Head Start staff. Only the researchers who have been hired to conduct the play session or to make notes
about the video will be allowed to view your child’s play session video. These people have signed a
consent form promising that they will not discuss information about you or your child with anyone who is
not working on the research project. The law requires that if a child tells us that he or she is being hurt by

114

someone, we have to report it to the proper authorities. This is the only time that we would have to break
our promise to keep information about you and your child private. Any information that we share will be
grouped together, no names or other idartifying informatim will be used in these reports.

If at any point you have any questions or concerns about your participation in this program, you may call
the research supervisor, Dr. Marsha Carolan (517) 432-3327. If you have any questions about your rights
as a research participant, you should call Dr. Ashir Kurnar, Chair, University Committee on Research
Involving Human Subjects (517) 355-2180.

Statement

1 voluntarily agree to participate in this study and will allow my child to participate in this study. I am
legally able to give consent for this drild to participate.

 

 

 

Caregiver: Date:
Phone nrnnber:
Witness: Date:

 

 

I give permissiar to my child’s teacher to provide information regarding his or her behavior in the
classroom.

 

 

 

 

Caregiver: Date:
Witness: Date:
Child’s Name:

 

Teacher’s Name:

 

115

Informed Consent
(Teachers)

Purpose of the research

The purpose of this research is to better understand the needs and strengths of childrar who attend Head
Start and their caregivers. Specifically we are interested in finding out more about the support network of
caregivers, the recent changes in their lives, and their ideas about pararting. We are also interested in how
their children play and their behavior at home and at school.

What we are asking teachers to do?

Children will be interviewed during the school day, using a doll play procedure that will take place in an
empty room in the school building. Research staff will make arrangements with you about how this
process will work best in your classroom. After caregivers agree to participate, you will be asked to fill out
a silvey of child behavior. Eadr survey that your are asked to fill out will take between 30 and 40 minutes.

Your participation in the researdr is voluntary. You can decide at any time that you do not want to
participate in this research.

Your time will be reimbursed

It is important that we understand children’s behavior in many settings; you are the best resource for
helping us tmderstand how children behave in your classroom. We rmdastand that your day doesn’t begin
when studarts arrive at school and end whar they leave. We know that you are responsible for a great deal
of paperwork and outside activities. For each survey that you return within 3-4 weeks, you will receive a
$10 gilt certificate to a teacher supply store. You may earn a maximum of $60 in gift certificates. This
incentive is for the time that you spend filling out information about child behavior. It is important to us
that teachers do not urge fimilies to participate in the study.

Your privacy will be protected

Your privacy will be protected to the maximum extent allowable by law. You will be instructed not to put
your name or the child’s name on any sru'vey that you fill out. The information that you give us will be
stored in a locked file cabinet at Michigan State University. We will not share information that you give us
with other Head Start staff or parents. Any information that we share will be grouped together, no names
will be used.

If at any point you have any questions or concerns about your participation in this program, you may call
the research supervisor, Dr. Marsha Carolan (517) 432-3327. If you have any questions about your rights
as a research participant, you should call I}. Ashir Kumar, Chair, University Committee on Research
Involving Human Subjects (517) 355-2180.

Statement

1 voluntarily agree to participate in this study.

Teacher: Date:

 

Witness: Date:

 

116

REFERENCES

117

REFERENCES

Achenbach, T. M. (1991). Manual for the Child Behavior Checklr'st/4-18 and
1991 Profile. Burlington, VT: University of Vermont Department of
Psychiatry.

Achenbach, T. M., Edelbrock, C., 8. Howell, C. (1987). Empirically-based
assessment of the behavioral/emotional problems of 2-3 year old
children. Journal of Abnormal Child Psychology, 15, 629-650.

Adams, C., Hillman, N., & Gaydos, G. (1994). Behavioral difficulties in toddlers:
Impact of sociocultural and biological risk factors. Journal of Clinical Child
Psychology, 23, 373-381.

Aguilar, B., Sroufe, L. A., Egeland, 8., & Carlson, E. (2000). Distinguishing the
early-onsetlpersistent and adolescence-onset antisocial behavior types
from birth to 16 years. Development and Psychopathology 12, 109-132.

Ainsworth, M. D., Blehar, M. C., Waters, E. 8. Wall, S. (1978). Patterns of
Attachment: A Psychological Study of the Strange Situation. Hillsdale.
NJ: Erlbaum.

Ainsworth, M. D. (1990). Epilogue: Some considerations regarding theory and
assessment relevant to attachments beyond infancy. In M. T. Greenberg,
D. Cicchetti, E. M. Cummings (Eds), Attachment In the Preschool Years:
Theory, Research, and Intervention. (pp. 463-488). Chicago: University of

Chicago Press.

118

Alesandn', S. M. (1991). Play and social behavior in maltreated preschoolers.
Development and Psychopathology, 3, 191-205.

Andersen, P., & Telleen, S. (1992). The relationship between social support and

maternal behavior and attitudes: A meta-analytic review. American
Joumel of Community Psychology, 20, 752-774.

Barnett, D., Ganiban, J., 8. Cicchetti, D. (1999). Maltreatment, emotional
reactivity, and the development of type D attachments from 12 to 24
months of age. In J. Vondra 8. D. Barnett (Eds), Atypical patterns of
infant attachment: Theory, research and current directions. Monographs of
the Society for Research in Child Development, 50(1-2, serial No. 209),
167-193.

Barnett, D., & Vondra, J. (1999). Atypical patterns of eady attachment: Theory,
research, and current directions. In J. Vondra 8: D. Barnett (Eds),
Atypical patterns of infant attachment: Theory, research and current
directions. Monographs of the Society for Research in Child

Development, 50(1-2, serial No. 209), 1-24.

Bates, J. E., Bayles, K., Bennett, D. 8., Ridge, 8, 8. Brown, M. M. (1991). Origins
of externalizing behavior problems at eight years of age. In D. J. Pepler &
K. H. Rubin (Eds), The Development and Treatment of Childhood
Aggression (pp. 93—120). Hillsdale, NJ: Edbaum.

119

Bates, J. E., Maslin, C. A., & Frankel, K A (1985). Attachment security, mother-
child interaction, and temperament as predictors of behavior-problem
ratings at age three years. In I. Bretherton & E. Waters (Eds), Growing
points of attachment theory and research. Monographs of the Society for
Research in Child Development, 50(1-2 Serial No. 209), 167-193.

Bavolek, S. & Keene, R. (1999). The Adult Adolescent Parenting Inventory — 2
Handbook. Park City, UT: Family Development Resources.

Belsky, J. (1984). The determinants of parenting: A process model. Child

Development, 55, 83-96.

Belsky, J. (1999). Interactional and contextual determinants of attachment
security. In J. Cassidy & P. R. Shaver (Eds), The Handbook of
Attachment: Theory, Research, and Clinical Applications. (pp. 249-264).
New York: Guilford.

Belsky, J., Rovine, M., & Taylor, D. G. (1984). The Pennsylvania infant and
family development project“ III. The origins of individual differences in
infant-mother attachment: Maternal and infant contributions. Child
Development, 55, 71 8-728.

Bohlin, G., Hagekull, B., & Rydell, A. M. (2000). Attachment and social
functioning: A longitudinal study from infancy to middle childhood. Social
Development, 9(1), 24-39.

120

Best, K. K., Vaughn, B., Washington, W. N., Cielinski, K L., & Bradbard, M. R.
(1998). Social competence, social support, and attachment: Demarcation
of construct domains, measurement, and paths of influence for preschool
children attending Head Start. Child Development, 69, 192-218.

Bowlby, J. (1982). Attachment and Loss: Volume 1. Attachment. New York:
Basic Books.

Bowlby, J. (1988). A Secure Base: Parent-Child Attachment and Healthy
Human Development. London: Basic Books.

Brazelton, T. 8., 8 Greenspan, S. l. (2000). The Ineducible Needs of Children:
What Every Child Must Have to Grow, Learn, and Flourish. Cambridge,
MA: Perseus.

Bretherton, I., Ridgeway, D., 8 Cassidy, J. (1990). Assessing internal working
models of the attachment relationship: An attachment story connletion
task for 3—year-olds. In M. T Greenberg, D. Cicchetti, & E. M. Cummings
(Eds), Attachment in the Preschool Years: Theory, Research, and
Intervention. (pp. 273308). Chicago: University of Chicago Press.

Bronfenbrenner, U. (1979). The Ecology of Human Development. Cambridge,
MA: Harvard University Press.

Bronfenbrenner, U. (1993). The ecology of cognitive development: Research
models and fugitive findings. In R. H. Wozinak & K. Fischer (Eds),
Scientific Environments. Hillsdale, NJ: Erlbaum.

121

Bubolz, M. 8 Sontag, M. S. (1993). Human ecology theory. In P. Boss, W.
Doherty, R. LaRossa, W. Schumm, 8 S. Steinmetz (Eds), Sourcebook of
Family Theories and Methods: A Contextual Approach. (pp. 419-448).
New York: Plenum.

Burchinal, M., Follmer, A., 8 Bryand, D. (1996). The relations of matemal social
support and family structure with matemal responsiveness and child
outcomes among African American families. Developmental Psychology,
6, 1073-1083.

Call, M. J. (1999). Transgenerational attachment, life stress, and the
development of disruptive behavior in preschool children [abstract].
(Doctoral Dissertation, Old Dominion University, 1999). Dissertation
Abstracts Intemational, 60(4-B) AEH9928695.

Campbell, S. B. (1994). Hard-to-manage preached boys: Extemalizing behavior,
social competence, and family context at two-year followup. Joumel of
Abnormal Child Psychology, 22(2), 147-166

Campbell, S. B. (1995). Behavior problems in preschool children: A review of
recent research. Joumel of Child Psychology and Psychiatry, 36(1), 113-
149.

Campbell, 8. B., 8 Ewing, L. J. (1990). Follow-up of hard-to-manage
preschoolers: Adjustment at age 9 and predictors of continuing symptoms.

Joumel of Child Psychology and Psychiatry, 31(6), 871 -889.

122

Campbell, S. B., Breaux, A. M., Ewing, L. J., 8 Szumowski, E. K. (1984). A one-
year follow-up study of parent-referred “hyperactive” preschoolers.
Joumel of the American Academy of Child Psychiatry, 23, 243-249.

Campbell, S. 8., March, C. L., Pierce, E. W., Ewing, L. J. (1991). Hard-to-
manage preschool boys: Family context and the stability of externalizing
behavior. Journal of Abnormal Child Psychology, 19(3), 301-318.

Carlson, V., Cicchetti, D., Barnett, D., 8 Braunwald, K. (1989).
Disorganized/disoriented attachment relationships in maltreated infants.
Developmental Psychology, 25, 525-531.

Cassidy, J. (1994). Emotion regulation: Influences of attachment relationships.
In N. Fox (Ed), the development of emotion regulation. Monographs of the
Society for Research in Child Development, 59(2-3, Serial No. 240).

Cassidy, J. (1999). The nature of the child’s ties. In infancy and childhood. In P.
Shaver 8 J. Cassidy (Eds) Handbook of Attachment (pp. 3-20). New
York: Guilford.

Chazan-Cohen, R., Jerald, J., 8 Stark, D. (2001). A commitment to supporting
the mental health of our youngest children. Zero to Three, 22(1), 4-12.

Child Mental Health Foundations and Agendas Network (2000). A Good
Beginning: Sending America ’s Children to School with the Social and
Emotional Competence they Need to Succeed. Bethesda, MD: The

National Institute of Mental Health.

123

Cicchetti, D., Cummings, E. M., Greenberg, M. T., 8 Marvin, R. S. (1990). An
organizational perspective on attachment beyond infancy: Inrplications for
theory, measurement, and research. In M. T. Greenberg, D. Cicchetti, 8
E. M. Cummings (Eds), Attachment in the Preschool Years: Theory,
Research, and Intervention (pp. 349). Chicago: University of Chicago
Press.

Cohen, D. A (1990). Child-mother attachment of six-year-olds and social
competence at school. Child Development, 61, 152-162.

Crick, N. R., Casas, J. F., 8 Mosher, M. (1997). Relational and overt aggression
in preschool. Developmental Psychology, 33(4), 579-588.

Crittenden, P. M. (1985). Social networks, quality of child rearing, and child
development. Child Development 56, 1299-1313.

Crockenberg, S. B. (1981). Infant irritability, mother responsiveness, and social
support influences on the security of infant-mother attachment. Child
Development, 52, 857-869.

Cummings, E. M., Iannotti, R. J., 8 Zahn-Waxler, C. (1989). Aggression between
peers in early childhood: Individual continuity and developmental change.
Child Development, 69, 14-27.

Daly, E. B. (1984). Relationship of stress and ego energy to field-dependent
perception in older adults. Perceptual and Motor Skills, 59(3) 919-926.

Darling, L. M. (1993). Exploring Adult Attachment in Non-Clinical Samples:
Validating a Four Category Model. Unpublished doctoral dissertation,
Michigan State University.

124

Davies, D. (1999). Attachment as a context of development. In D. Davies (Ed),
Child Development: A Practitioner's Guide (pp. 13-43). New York:
Guilford.

DeMulder, E. K., 8 Rakde-Yarrow, M. (1991). Attachment with affectively ill and
well mothers: Concurrent behavioral correlates. Development and
Psychopthology, 3, 227-242.

DeMulder, E. K., Denham, S. Schmidt, M., 8 Mitchell, J. (2000). Q-Sort
assessment of attachment security during the preschool years: Links from
home to school. Developmental Psychology, 36(2), 274-282.

Derogatis, L. R. (1993). Brief Symptom Inventory: Administration, Scoring, and
Procedures Manual. (3" Ed. ). Minneapolis: National Computer Systems,
Inc.

De Wolff, M. S. 8 van IJzendoom, M. H. (1997). Sensitivity and attachment: A
meta-analysis on parental antecedents of infant attachment. Child
Development, 68(4), 571-591.

Drotar, D., Stein, R., 8 Perrin, E. (1995). Methodological issues in using the
Child Behavior Checklist and its related instruments in clinical child
psychology research. Journal of Clinical Child Psychology, 24(2), 184-
192.

125

Easterbrooks, M. A., 8 Goldberg, W. A (1990). Security of toddler-parent
attachment: Relation to children’s sociopersonality functioning during
kindergarten. In M. T. Greenberg, D. Cicchetti, 8 E. M. Comings (Eds),
Attachment in the Preschool Years: Theory, Research, and Intervention.
(pp. 221-244). Chicago: University of Chicago Press.

Egleand, B., 8 Sroufe, L. A. (1981). Developmental sequale of maltreatment in
infancy. New Directions for Child Development,11, 77-92.

Egeland, B., 8 Farber, E. A. (1984). lnfant-rnother attachment: Factors related
to its development and changes over time. Child Development, 55, 753-
771.

Egeland, 8., Kalkoske, M., Gottesman, N., 8 Erickson, M. F. (1990). Preschool
behavior problems: Stability and factors accounting for change. Joumel of
Child Psychology and Psychiatry, 31(6) 891 -909.

Egeland, B., Pianta, R... 8 Ogawa, J. (1996). Early behavior problems:
Pathways to mental disorders in adolescence. Development and
Psychopathology, 8, 735-749.

Farrington, D.- P. (1978). The family backgrounds of aggressive youths. In L.
Hersov, M. Berger, 8 D. Shaffer (Eds), Aggression and Antisocial
Behavior in Childhood and Adolescence (pp. 73-93). Oxford: Pergamon.

Farrington, D. P. (1983). Offending from 10 to 25 years of age. In K. T.
VanDuesen 8 S. A. Mednick (Eds), Prospective Studies of Crime and
Delinquency (pp. 17-37). Boston: KIuwer-Nijhoff.

126

Farrington, D. P. (1987). Early precursors of frequent offending In J. Q. Wilson,
8 G. C. Loury (Eds), From Children to Citizens, Vol 3: Families, Schools
and Delinquency Prevention (pp. 27-50). New York: Springer-Vedag.

Fogany, P., Steele, H., 8 Steele, M. (1991). Maternal representations of
attachment during pregnancy predict the organization of infant mother
attachment at one year of age. Child Development, 62, 891-905.

Gelfand, D., 8 Teti, D. (1990). The effects of nratemal depression on children.
Clinical Psychology Review, 10, 329-353.

George, C. 8 Main, M. (1979). Social interactions of young abused children:
Approach, avoidance, and aggression. Child Development, 50, 306-318.

George, C., 8 Solomon, J. (1990, 1996, 2000). Six-Year Attachment Doll Play
Classification System. Unpublished classification manual. Mills College,
Oakland, CA.

George, C. 8 Solomon, J. (1996). Representational Models of Relationships:
Links Between Caregiving and Attachment. Infant Mental Health Joumel,
17(3), 198-216.

George, C. 8 Solomon, J. (1999). The measurement of attachment security in

infancy and childhood. In P. Shaver 8 J. Cassid (Eds) Handbook of
Attachment (pp. 287-316). New York: Guilford.

Goldberg, 8., Carter, C., Lojkasek, M., 8 Minde, K. (1990). Prediction of
behavior problems in 4-year-olds born prematurely. Development and
Psychopathology, 2, 1 5-30.

127

Greenberg, M. T., Cicchetti, D., Cummings, E. M. (Eds). (1990). Attachment in
the Preschool Years: Theory, Research, and Intervention. Chicago:
University of Chicago Press:

Grifl‘ore, R. J. 8 Phenice, L A. (2001). The Language of Human Ecology: A
General Systems Perspective. Dubuque, IA: Kendall/Hunt.

Grossmann, K., Grossmann, K. E., Spangler, G., Suess, G. 8 Unzner, L. (1985).
Maternal sensitivity and newborns’ orientation responses as related to
quality of attachment in northern Germany. In I. Bretherton 8 E. Waters
(Eds), Growing points otattachment theory and research. Monographs of
the Society for Research in Child Development, 50(1-2, Serial No. 209),
233-257.

Herrenkohl, R. C., 8 Herrenkohl, E. C. (1981). Some antecedents and
developmental concequences of child maltreatment. In R. RizIey 8 D.
Cicchetti (Eds), Developmental Perspectives on Child Maltreatment (pp.
57-76). San Francisco: Jossey-Bass.

Herrenkohl, R. C., 8 Russo, M."J. (2001). Abusive early child rearing and early
childhood aggression. Child Maltreatment, 6(1), 3-16.

Hubbs-Tait, L., Osofsky, J., Ham, 0., 8 Culp, A. (1994). Predicting behavior
problems and social competence in children of adolescent mothers.
Family Relations, 43, 439:446.

Huesmann, L. R., Eron, L. D., Lefltowitz, M. M. 8 Walder, L. O. (1984). Stability
of aggression over time and generations. Developmental Psychology, 20,

1120-1134.

128

Huessman, L. R., Eron, L. D., 8 Yarmel, P. W. (1987). Intellectual functioning
and aggression. Joumel of Personality and Social Psychology, 52, 232-
240.

Isabella, R. (1993). Origins of attachment: Maternal interactive behavior across
the first year. Child Development, 64, 605-621.

Jacobson, S. W., 8 Frye, K. F. (1991). Effect of maternal social support on
attachment: Experimental evidence. Child Development, 62, 572-582.

Julian, T. W., McKenry, P. C., 8 McKelvey, M. W. (1994). Cultural variations in
parenting: Perceptions of Caucasian, African American, Hispanic, and
Asian-American Parents. Family Relations, 43(1) 30-37.

Keenan, K., 8 Shaw, D. (1994). The development of aggression in toddlers: A
study of low income families. Joumel of Abnormal Child Psychology,
22( 1), 53-78.

Keenan, K., 8 Shaw, D. (1997). Developmental and social influences on young
girls’ early problem behavior. Psychological Bulletin, 121(1), 95-113.

Kerr, M., Tremblay, R. E., Pagani, L., 8 Vitaro, F. (1997). Boys’ behavioral
inhibition and the risk of later delinquency. Archives of General
Psychiatry, 54, 809-816.

Kobak, R. (1999). The emotional dynamics of disruptions in attachment
relationships: Implications for theory, research, and clinical intervention.
In J. Cassidy 8 P. R. Shaver (Eds), The Handbook of Attachment. (pp.
21-43) New York: Guilford.

129

Kupersmidt, J. 8., Bryant, 0., 8 erloughby, M. T. (2000). Prevalence of
aggressive behavior among preschoolers in Head Start and community
child care programs. Behavior Disorders, 26(1), 42-52.

Liaw, F., 8 Brooks-Gunn, J. (1994). Cumulative familial risks and Iow-birthweight
children’s cognitive and behavioral development. Journal of Clinical Child
Psychology, 23,360-372.

Loeber, R. 8 Dishon, T. (1983). Early pmdictors of male delinquency: A review.
Psychological Bulletin, 94, 68-99.

Loeber, R., 8 Stouthamer-Loeber, M. (1986). Family factors as correlates and

predictors of juvenile conduct problems and delinquency. In M. Tonry 8
N. Morris (Eds), Crime and Justice. Chicago: University of Chicago
Press.

Loeber, R., 8 Stouthanrer-Loeber, M. (1987). Prediction. In H. C. Quay (Ed),
Handbook of Juvenile Delinquency_(pp. 325-382). New York: erey.

Lyons-Ruth, K., Alpem, L., 8 Repacholi, B. (1993). Disorganized infant
attachment classification and maternal psychosocial problems as
predictors of hostile-aggressive behavior in the preschool classroom.

Child Development, 64, 572-585.

Lyons-Ruth, K., Connell, D., Gruenbaum, H., 8 Botein, S. (1990). Infants at
social risk: Maternal depression and family support services as mediators
of infant development and security of attachment. Child Development, 61,
85-98.

130

Lyons-Ruth, K., Easterbrooks, A., 8 Cibelli, C. (1997). Infant attachment
strategies, infant mental lag, and maternal depressive symptoms:
Predictors of internalizing and externalizing problems at age 7.
Development and Psychopathology, 3, 377-396.

Lyons-Ruth, K., 8 Jacobvitz, D. (1999). Attachment disorganization: Unresolved
loss, relational violence, and lapses in behavioral and attentional
strategies. In P. Shaver 8 J. Cassidy (Eds), The Handbook of
Attachment. (pp. 520-554). New York: Guilford.

Magnusson, D. (1988). Individual Development fiom an lnteractional
Perspective. Hillsdale, NJ: Lawrence Erlbaum Associates.

Magnusson, D., Stattin, H., 8 Duner, A. (1983). Aggression and criminality in a
longitudinal perspective. In K. T. Van Dusen, 8 S. A. Mednick (Eds),
Prospective Studies in Crime and Delinquency. (pp. 277-301 ). Boston:
KIuwer-Nijhoff.

Main, M., 8 Cassidy, J. (19883). Security in infancy, childhood, and adulthood:
A move to the level of representation. In I. Bretherton 8 E. Waters (Eds),
Growing points in attachment theory and research. Monographs of the
Society for Research in Child Development, 50(1-2, Serial No. 209), 66-
104.

Main, M. 8 Cassidy, J. ( 1988b). Categories of response to reunion with the
parent at age 6: Predicted from infant attachment classifications and

stable over a 1-month period. Developmental Psychology, 24, 415-426.

131

Main, M. 8 Hesse, E. (1990). Parents’ unresolved traumatic experiences are
related to infant disorganized attachment status: Is frightened and/or
frightening parental behavior the linking mechanism. In M. T. Greenberg,
D. Cicchetti, 8 E. M. Cummings (Eds), Attachment in the Preschool
Years: Theory, Research, and Intervention. (pp. 161 -1 82). Chicago:
University of Chicago Press.

Main, M. 8 Solomon, J. (1990). Procedures for identifying infants as
disorganized/disoriented during the Ainsworth Strange Situation. In M. T.
Greenberg, D. Cicchetti, 8 E. M. Cummings (Eds), Attachment in the

Preschool Years: Theory, Research, and Intervention. (pp. 121 -1 60). Chicago:

University of Chicago Press.

Mesman, J., Bangers, I. L., 8 Koot, H. M. (2001). Preschool developmental
pathways to preadolescent internalizing and externalizing problems.
Joumel of Child Psychology and Psychiatry, 42(5), 679-689.

Moss, E., Parent, 8., Gosselin, C., Rousseau, D., 8 St-Laurent, D. (1996).
Attachment and teacher-reported behavior problems during the preschool
and early school-age period. Development and Psychopathology, 8, 511-
525.

Moss, E., Rousseau, D., Parent, 8., St-Laurent, D., 8 Saintong, J. (1998).
Correlates of attachment at school age: Maternal reported stress, rnother-
child interaction, and behavior problems. Child Development, 69, 1390-
1405.

132

Muss, R. E. (1996). Urie Bronfenbrenner’s ecological perspective of human
development. Theories of Adolescence. (pp. 312-336). New York:
McGraw-Hill.

Nagin, D., 8 Tremblay, R. E. (1999). Trajectories of boys’ physical aggression,
opposition, and hyperactivity on the path to physically violent and
nonviolent juvenile delinquency. Child Development, 70(5), 1181-1196.

Olweus, D. (1979). Stability of aggressive reaction patterns in males: A review.
Psychological Bulletin, 86, 852-857.

Pagani, L., Boulerice, B., 8 Tremblay, R. (1997). The influence of poverty on
children’s classroom placement and behavior problems. In G. Duncan 8
J. Brooks-Gunn (Eds), Consequences of Growing up Poor (pp. 311-339).
New York: Russell Sage Foundation.

Patterson, G. R., Reid, J. B., Jones, R. R., 8 Conger, R. E. (1975). A Social
Learning Approach: Vol. 1. Families with Aggressive children. Eugene,
OR: Castalia. '

Patterson, G. R., De Baryshe, B. D., 8 Ramsey, E. (1989). A developmental
perspective on antisocial behavior. American Psychologist, 44, 329-335.

Patterson, G. R., Capaldi, D., 8 Bank, L. (1992). An early starter model for
predicting delinquency. In D. J. Pepler 8 K. H. Rubin (Eds), The
Development and Treatment of Childhood Aggression (pp. 139-188).
Hillsdale, NJ: Erlbaum.

133

Prior, M., Smart, M. A., Sanson, A., 8 Oberklaid, F. (1993). Sex differences in
psychological adjustment from infancy to 8 years. Journal of the American
Academy of Child and Adolescent Psychiatry, 32(2), 291 -304.

Radke-Yarrow, M., Cummings, M., Kuczynski, L., 8 Chapman, M. (1985).
Patterns of attachment in two and three year olds in normal families and
families with parental depression. Child Development, 56, 884-893.

Renken, B., Egeland, 8., Marvinney, D., Mangelsdorf, S., 8 Sroufe, L. A. (1989).
Early childhood antecedents of aggression and passive withdrawal in early
elementary school. Journal of Personality, 57, 257-281.

Reynolds, C. R., 8 Kamphaus, R. W. (1998). Behavioral Assessment System for
Children Manual. Circle Pines, MN: American Guidance Service.

Rose, S. L., Rose, S. A., 8 Feldman, J. F. (1989). Stability of behavior problems
in very young children. Development and Psychopathology, 1, 5-19.

Sarason, I. G., Levine, H. M., Basham, R. B., 8 Sarason, B. R. (1983).
Assessing social support: The social support questionnaire. Joumel of
Personality and Social Psychology, 44(1) 127-139.

Schneider-Rosen, K. (1990). The developmental reorganization of attachment
relationships: Guidelines for classification beyond infancy. In M. T.
Greenberg, D. Cicchetti, 8 E. M. Cummings (Eds), Attachment in the
Preschool Years: Theory, Research, and Intervention. (pp. 185-220).
Chicago: University of Chicago Press.

Schore, A. N. (1994). Afiect Regulation and the Origin of the Self: The
Neurobiology of Emotional Development. Mahwah, NJ: Erlbaum.

134

Schore, A. N. (2001). Contributions from the decade of the brain to infant mental
health: An overview. Infant Mental Health Joumel, 22(1-2), 1-6.

Serbin, L. A., Moskowitz, D. S., Schwartzman, A. E. 8 Ledingham, J. E. (1991).
Aggressive, withdrawn, and aggressive/withdrawn children in
adolescence: into the next generation. In D. J. Pepler 8 K. H. Rubin
(Eds), The Develogment and Treatment of Childhood Aggression (pp. 55-
70). Hillsdale, NJ: Erlbaum.

Shaw, D. S., 8 Vondra, J. I. (1993). Chronic family adversity and infant
attachment security. Journal of Child Psychology and Psychiatry and

Allied Disciplines, 34, 1205-1215.

Shaw, D. S., Owens, E. B., Vondra, J. l., Keenan, K., 8 \Mnslow, E. (1996).
Early risk factors and pathways in the development of early disruptive
behavior problems. Development and Psychopathology, 8, 679-699.

Shaw, D. S., Keenan, K., Vondra, J. I., Delliquadri, E., 8 Giovannelli, J. (1997).
Antecedents of preschool children’s internalizing problems: A longitudinal
study of low-income families. Journal of the American Academy of Child
and Adolescent Psychiatry, 36(12), 1760-1767.

Shaw, D. S., Wil‘lSlOW, E. B., Owens, E. B., 8 Hood, N. (1998). Young children’s
adjustment to chronic family adversity: A longitudinal study of low-income
families. Journal of the American Academy of Child and Adolescent
Psychiatry, 37(5), 545-553.

135

Smith, J., Brooks-Gunn, J., 8 Klebanov, P. (1997). Consequences of living in
poverty for young children’s cognitive and verbgal ability and early school
achievement. In G. Duncan 8 J. Brooks-Gunn (Eds), Consequences of
Growing up Poor (pp. 132-189). New York: Russell Sage Foundation.

Solomon, J., George, C. 8 De Jong, A. (1995). Children classified as controlling
at age six: Evidence of disorganized representational strategies and
aggression at home and at school. Development and Psychology, 7, 447-
463.

Speltz, M. L., Greenberg, M. T., 8 DeKlyen, M. (1990). Attachment in
preschoolers with disruptive behavion A comparison of clinic-referred and
nonproblem children. Development and Psychopathology, 2, 31-46.

Spieker, S. J., 8 Booth, C. (1988). Maternal antecedents of attachment quality.
In J. Belsky 8 T. Newworski (Eds), Clinical Implications of Attachment
(pp. 300-323). Hillsdale, NJ: Erlbaum.

Sroufe, L. A. (1983). Infant-caregiver attachment and patterns of adaptation in
preschool: The roots of rnaladaptation and competence. In M. Perlmutter
(Ed), The Minnesota Symposia on Child Psychology: Vol 16.
Development and Policy Concemr'rrg Children with Special Needs (pp. 41 -
83). Hillsdale, NJ: Eribaum.

Sroufe, L. A. (1996). Emotional development. Attachment: the Dyadic
Regulation of Emotion (pp. 172-191). New York: Cambridge University

Press.

136

Sroufe, L. A. (1996). Emotional Development: The Organization of Emotional
Life in the Eariy Years. New York: Cambridge University Press.

Sroufe, L. A., Fox, N., 8 Pancake, V. (1983). Attachment and dependency in
developmental perspective. Child Development, 54, 1615-1627.

Sroufe, L. A., Schork, E., Motti, E., Lawroski, N., 8 LaFreniere, P. (1984). The
role of affect in social competence. In C. Izard, J. Kagan, 8 R. Zajonc
(Eds), Emotions, Cognition and Behavior (pp. 289-319). New York:
Cambridge University Press.

Steele, H., Steele, M., 8 Fogany, P. (1996). Associations among attachment
classifications of mothers, fathers, and their infants. Child Development,
67, 541-555.

Stevenson, J., Richman, N., 8 Graham, P. (1985). Behaviour problems and
language abilities at three years and behavioral deviance at eight years.
Journal of Child Psychology and Psychiatry, 26 215-230.

Suess, G. J., Grossmann, K. E., 8 Sroufe, L. A. (1992). Effects of infant
attachment to mother and father on quality of adaptation in preschool:
From dyadic to individual organisation of self. International Journal of
Behavioral Development, 15, 43-65.

Teti, D. M. Bakagawa, M., Das, R., 8 Wrrth, O. (1991). Security of attachment
between preschoolers and their mothers: Relations among social
interaction, parenting stress, and mothers’ sorts of the attachment Q-Set.
Developmental Psychology, 27(3), 440-447.

137

Teti, D. M., Gelfand, D. M., Messinger, D. S., 8 Isabella, R. (1995). Maternal
depression and the quality of early attachment: An examination of infants,
preschoolers, and their mothers. Developmental Psychology, 34, 361-
376.

Troy, M., 8 Sroufe, L. A. (1987). Vrctimization among preschoolers: The role of
attachment relationship theory. Journal of the American Academy of Child
and Adolescent Psychiatry, 26, 166-172.

Tremblay, R. E., Pihl, R. O., V’rtaro, F., 8 Dobkin, P. L. (1994). Predicting early
onset of male antisocial behavior from preschool behavior. Archives of
General Psychiatry. 15, 732-739.

Walker, H. M., Shinn, M. R., O’Neill, R. E., 8 Ramsey, E. (1987). A longitudinal
assessment of the development of antisocial behavior in boys: Rationale,
methodology, and frist-year results. Remedial and Special Education,
8(4), 7-16.

van IJzendoom, M. H., Schuengel, C., 8 Bakerrnans-Kraneburg, M. J. (1999).
Disorganized attachment in early childhood: Meta-analysis of precursors,
concomitants, and Sequale. Developmental Psychopathology, 11(2), 225-
249.

Ventura-Cook, E. E. (1997). Attachment representations and socioemotional
characteristics of lower-Income solitary-passive, solitary-nonpassive, and
nonwithdrawn preached children [abstract]. (Doctoral Dissertation,
University of New Orleans, 1997). Dissertation Abstracts International,
57(8-B) AAM9701562.

138

Vondra, J. I., DowdeII-Homrnerding, K. 8 Shaw, D. S. (1999). Stability and
change in infant attachment in a low income sample. In J. l. Vondra 8 D.
Barnett (Eds), Atypical Attachment in Infancy and Early Childhood Among
Children at Developmental Risk. Monographs of the Society for Research
in Child Development, 64(3, Serial No. 258), 119-144.

Wartner, U., Grossmann, K, Fremmer-Bombik, E., 8 Suess, G. (1994).
Attachment patterns at age six in south Germany: Predictability from
infancy and inrplications for preschool behavior. Child Development, 65,
1 014-1027.

Waters, E. erpman, J., 8 Sroufe, L. A. (1979). Attachment, positive affect, and
conrpetence in the peer group: Two studies in construct validation. Child
Development, 50, 821-829.

Webster-Shanon, C. (1990). Stress: A potential disruptor of parent perceptions
and family interactions. Journal of Clinical Child Psychology, 4, 302-312.

Weinfleld, N. S., Sroufe, L. A., Egeland, B., 8 Carleson, E. A. (1999). The nature
of individual differences in infant-caregiver attachment. In J. Cassidy 8 P.
R. Shaver (Eds), The Handbook of Attachment (pp. 68-88). New York:
Guilford.

Zahn-Waxler, C., lanotti, R. J., Cummings, E. M., 8 Denham, S. (1990).
Antecedents of problem behaviors in chidlren of depressed mothers.
Development and Psychopathology, 2, 271 -291.

139