. _ 4 1 .34.
V iaafif....amfi? ,4.
. a424,...» :. «L. . . 4
. r. J. ‘2 .—
auiygz. x. , . ,

. a

31". v”.

._‘d..

1

. 2... s23

usfluwfian

.. ,2. .

I”? W
1...:

‘ 2. .uflunwna.’ r .
.x... 33.. a; .3
..

 

.
. J?
.u.:... .9
Lit. ...
.3: .1. R .

s4 «IQ.

. flaw? «an
“awn...“umn .r

3L

 

 

;
. .
“Jami“..fi
:5”! (Lu! ‘1 1:“

€
1

 

 

 

 

Winn“ h: v ..
a: t. 2?! Pt.
3‘... 1h... :tisl;

13031.23
. 1..

.himrut. 1513;?
A .. .L . ‘
15...". ‘

 

 

 

.15.... .c. . ..

THESIS
5 l LIBRARIES
J o 0 b MICHIGAN STATE UNIVERSITY

EAST LANSING, MICH 48824-1048.

This is to certify that the
dissertation entitled

Postdivorce Parental Relationships and Well-being
among African American, Hispanic, and Caucasian
College Students

presented by

Janet M. Kinney

has been accepted towards fulfillment
of the requirements for the

PhD. degree in Counseling Psychology

 

 

A Queue ALMA, PA ( b .

(Major Professor’s Signature
08/zs/Q5

Date

MSU is an Affirmative Action/Equal Opportunity Institution

’ .7 o —- -‘v —- ‘ ~— g, 7‘ - v ‘7 '— 9 - v 4’ fl *' v '- ' Fyrgv—H-‘ -“W QM“ fiv—K‘

 

 

 

PLACE IN REfURN 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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 chlFlC/DateOuoindd—pds

POSTDIVORCE PARENTAL RELATIONSHIPS AND WELL-BEING AMONG
AFRICAN AMERICAN, HISPANIC, AND CAUCASIAN COLLEGE STUDENTS

By

Janet M. Kinney

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Counseling, Educational Psychology and Special Education

2005

ABSTRACT

POSTDIVORCE PARENTAL RELATIONSHIPS AND WELL-BEING AMONG
AFRICAN AMERICAN, HISPANIC, AND CAUCASIAN COLLEGE STUDENTS

By

Janet M. Kinney

Adolescents' relationships with parents have long been acknowledged as
important contributors to psychosocial development. The purpose of this study
was to examine parental attachment and interparental conflict in an effort to
explain why some adolescents fare well after the divorce of their parents while
others do not. Of particular interest were the separate and unique contributions
of attachment style (e.g., secure, insecure) with mother (MlAtt) and maternal
initiated conflict with father (MICON) to attachment style with father (PIAtt) and
adolescent well-being (WB). Attachment to parents was examined by comparing
MIAtt and P/Att with regard to three domains: affective quality of relationships,
facilitator of independence, and source of support. Participants were 58 African
American (AA), 28 Hispanic (H), and 93 Caucasian (C) 18-23 year old young
adults from a Midwestern university who came from divorced families in which
mothers were the custodial parent. Across the racial/ethnic groups, participants
viewed their relationships with mother and father as both independent and
interdependent. Path analysis of the two groups indimted a statistically

significant relationship between the entire model and the outcome variable. The

prsr

AR"!

m

957‘

proposed model accounted for a greater portion of variance in WB within the
AAIH group (R2 = .79) than within the C group (R2 = .35). The findings supported
the hypotheses that 1) secure MIAtt and low levels of MICON correspond with
secure P/Att, 2) secure MIAtt and PIAtt resulted in higher levels of general,
spiritual, and psychological WB, and 3) H and AA participants demonstrate more
adaptive strategies to MICON than C participants. Significant differences
between the groups are illustrated and discussed. The findings suggest that it is
important to evaluate the relationship between parental attachment within the
context of family process variables that may explain the parental divorce-child

psychosocial adjustment relationship.

Dedication
This dissertation is dedicated to the Lord my God, my family and my
friends, without which my dream could never have come true. Thank you for
believing in me and supporting me in your own unique way. I especially
recognize and appreciate my husband, Jack, and my daughters, Samantha and

Sarah. I love you very much and recognize the sacrifices you have made as l

have completed this journey.

iv

.I
Tarra E
Ph D
unwarj
I
pray?
SUDDC‘
Com."
expene
interns

and pr:

90W C»:

We 3M

 

ACKNOWLEDGMENTS

First, I would like to thank the members of my Dissertation Committee, J.
Tawa Sina, Ph.D., Gloria Smith, Ph.D., Robert Caldwell, Ph.D., and Lee June,
Ph.D., for lending their special expertise. Special thanks to Dr. Smith for her
unwavering support and the stability she provided to the Counseling Psychology
program. My sincere appreciation goes out to Dr. Sina for her many years of
support and guidance in her roles as chairperson of my Doctoral Dissertation
Committee, advisor of my doctoral program, supervisor of my practicum
experience at the Counseling Center, and Training Director during my doctoral
internship. Irrespective of her role, Dr. Sina has modeled integrity, genuineness,
and professionalism. She encouraged me to perseverance and motivated me to
grow personally and professionally. For this and much more, I thank her greatly.

Next, I would like to specially thank Robert Fabiano, Ph.D. for the major
role he played in my doctoral process. His support, training, and direction, in
addition to his friendship, were invaluable to me. I would also like to recognize
and thank Ellie Bossi, Barb Walkington, Bets Caldwell, Mary Ball, Rosalyn Davis,
and Karen Brown for their support and assistance.

Finally, I would like to acknowledge the years of professional and personal
support provided by my doctoral classmates, Pei-chen Hsu, Anne Kubal, Sarah
Raymond, Steve Pence, and Ariel Agosto. The support system and friendships

we developed often served as my encourager and saving grace.

UST
UST'

CHAC

CHAF‘
UTER

CHAP‘
MET-H,

 

CHAP:
RESUL

TABLE OF CONTENTS

LIST OF TABLES
LIST OF FIGURES

CHAPTER 1
Introduction
Attachment Theory
Influence of Attachment on Separated/Divorced Families
Relationship Between Maternal and Paternal Attachment
Maternal Conflict with Father
Race, Separation/Divorce, and Attachment
Age and Attachment
Parental Attachment and Adolescent Well-being
Research Questions
Summary and Hypotheses

CHAPTER 2
LITERATURE REVIEW
Postdivorce Father Involvement and Father-Child Relationships
Mother-child, Father-child, and Mother-Father Relationships
Parental Conflict
Attachment Theory
Age and Attachment
Race/Ethnicity and Attachment
Summary

CHAPTER 3
METHOD
Participants
Measures
Design
Data-Gathering Procedures
Data Analysis

CHAPTER 4

RESULTS
Preliminary Analyses
Correlational Analyses
Path Analyses

viii

A—l—i—l
comahommoihw-s

20
20
22
24
26
29
33
38

4O
4O
4O
50
51
52

59
59
59
75

CHAPTER 5

DISCUSSION 94
Parental Attachment, Parental Conflict, and Adolescent Well-being 94
Limitations 105
Relevance to Practice 107
Future Research 109

APPENDICES
APPENDIX A Forms and Demographic lnforrnation 113
APPENDIX B African American & Hispanic Group Path Analysis 1 120
APPENDIX C Caucasian Group Path Analysis 1 134
APPENDIX D Caucasian Group Path Analysis 2 142

APPENDIX E African American 8. Hispanic Group Path Analysis 2 157
REFERENCES 186

vii

Tatse

Table

Table

Tat-e

Table

Table

Table

T3216

Table

Tatie

Table

 

LIST OF TABLES

Table 1 Correlation Coefficients of Adolescent Well-being Measures

Table 2 African American and Hispanic Group Correlation Coefficients
of Well-being Scores and Independent Variables

Table 3 Caucasian Group Correlation Coefficients of Well-being
Scores and Independent Variables

Table 4 African American and Hispanic Group Correlations of Main
Variables

Table 5 Caucasian Group Correlations of Main Variables

Table 6 African American and Hispanic Group Pearson Correlation
Coefficients of Endogenous Variables and Possible Covariates

Table 7 Caucasian Group Pearson Correlation Coefficients of
Endogenous Variables and Possible Covariates

Table 8 Pearson Correlation Coefficients African American and Hispanic
Group Attachment Subscales, Well-being, and Maternal conflict
with Father

Table 9 Pearson Correlation Coefficients Caucasian Group Attachment
Subscales, Well-being, and Maternal conflict with Father

Table 10 African American and Hispanic Group Pearson Correlation
Coefficients Endogenous and Exogenous Variables Used in
Path 1

Table 11 African American and Hispanic Group Path 1 Observed and
Reproduced Correlations

Table 12 Summery for Causal Effects for African American and
Hispanic Group Path 1

Table 13 Caucasian Group Pearson Correlation Coefficients Endogenous
and Exogenous Variables Used in Path 1

Table 14 African American and Hispanic Group Path 1 Observed
Correlations and Standardized Path Coefficients

viii

59

6O

6O

67

67

72

72

75

78

78

81

 

 

Table

Table
Table

Tatie

Tame

Tatle

Table

IBCIE

Table

Table

Table

Table

Table

 

Table 15 Pearson Correlation Coefficients Caucasian Group Endogenous 84
and Exogenous Variables Used in Path 2

Table 16 Caucasian Group Path 2 Observed and Reproduced Correlations 85
Table 17 Summery for Causal Effects for Caucasian Group Path 2 Outcome 86

Table 18 Pearson Correlation Coefficients African American and Hispanic 90
Group Endogenous and Exogenous Variables Used in Path 2

Table 19 African American and Hispanic Group Path 2 Observed and 90
Reproduced Correlations

Table 20 Summery for Causal Effects for African American and Hispanic 93
Group Path 2

Table A1 Caucasian Group Descriptive Statistics for Demographic 117
lnforrnation

Table A2 Caucasian Group Descriptive Statistics for Subscales and 117

Full-scale Parental Attachment Measures.

Table A3 African American and Hispanic Group Descriptive Statistics 118
for Demographic Information

Table A4 African American and Hispanic Group Descriptive Statistics 119
for Subscales and F uII-scale Parental Attachment Measures.

Table A5 Cronbach’s Alpha 1 19

Table B African American and Hispanic Group Path 1 Decompositions 133

and Calculations of Reproduced Correlations

Table D Caucasian Group Path 2 Decompositions and Calculations of 154
Reproduced Correlations
Table E African American and Hispanic Group Path 2 Decompositions 180

and Calculations of Reproduced Correlations

ix

 

 

A F
Fiver:
U

Flat/E
U

5.3.. ,5

45“

LIST OF FIGURES

Figure 1 Input Path Diagram 1: African American! and Hispanic
Group Path Analysis 1

Figure 2 Input Path Diagram 2: Caucasian Group Path Analysis 1
Figure 3 Input Path Diagram 3: Caucasian Group Path Analysis 2

Figure 4 Input Path Diagram 4: African American and Hispanic
Group Path Analysis 2

Figure B1 African American and Hispanic Group Stepwise Multiple
Regression for Path Analysis 1

Figure 82 African American and Hispanic group Normal Probability
Plot of the Regression Standardized Residuals for Path
Analysis 1

Figure B3 African American and Hispanic Residuals Scatter plot for
Path Analysis 1

Figure C1 Caucasian Group Stepwise Multiple Regression for Path
Analysis 1

Figure C2 Caucasian group Normal Probability Plot of the Regression
Standardized Residuals for Path Analysis 1

Figure C3 African American and Hispanic Residuals Scatter plot for
Path Analysis 1

Figure D1 Caucasian Group Stepwise Multiple Regression for Path
Analysis 2

Figure DZ Caucasian Group Normal Probability Plot of the
Regression Standardized Residuals of Path Analysis 2

Figure D3 Caucasian Group Residuals Scatter plot for Path Analysis 2
Figure D4 Caucasian Group Path 2 Paternal Promotion of Independence
Figure 05 Caucasian Group Path 2 Paternal Feelings of Affect

Figure D6 Caucasian Group Path 2 Father Feelings of Support

75

81

91

121

131

132

135

140

141

143

153

153
156
156
156

 

 

th'e

Fi§§Jl

 

Figure E1 African American and Hispanic Group Stepwise Multiple
Regression for Path Analysis 2

Figure E2 African American and Hispanic group Normal Probability
Plot of the Regression Standardized Residuals for Path
Analysis 1

Figure E3 African American and Hispanic Residuals Scatter plot for
Path Analysis 1

Figure E4 African American and Hispanic Group Path 2 Paternal
Promotion of Independence

Figure E5 African American and Hispanic Group Path 2 Paternal
Feelings of Affect

Figure E6 African American and Hispanic Group Path 2 Paternal
Feelings of Support

xi

157

178

179

184

184

185

 

Int.

Postdivorce Parental Relationships and Well-being among African American,
Hispanic, and Caucasian College Students

Introduction

Adolescents are highly valued in our society and therefore, there is
concern for their adjustment and well-being. Parental attachment and conflict
merit attention in the context of adolescents' well-being post-divorce and
therefore, will be the foci of this investigation. The number of adolescents with
divorced parents has continued to grow over the last several decades and since
1972, one million American children every year have seen their parents divorce
(US. Bureau of the Census, 2005). Some adolescents display negative
behavioral and/or psychosocial effects (e.g., decrease in school performance,
delinquency, depression, and difficulties with interpersonal relationships) that can
be attributed to the divorce of their parents, while others do not. The factors that
lead to negative effects for some adolescents are not clear, but there is a
continuing effort to identify and understand this area. One plausible explanation
for why some adolescents exhibit psychosocial and behavioral difficulties post-
divorce may be attachment. Specifically, the types of relationship children have
with their non-residential fathers. The father-child relationship in turn might be
affected by the type of relationship adolescents have with their custodial mothers.
Another factor may be the degree of conflict exhibited by the custodial mother
toward the noncustodial father following a divorce. Parental conflict has been
shown to have negative impact on psychosocial adjustment across a wide array
of family structures and settings (Davies & Cummings, 1994: Erel & Burman,

1995), but has yet to be studied in the context of adolescents' perceived well-

 

‘7’.“
and

C35:
fa {'11

W96

”55:;

362x

being and adjustment after divorce among different racial groups. The
percentage of those who are divorced has increased across all racial groups, yet
the vast majority of research on divorce focuses on the experience of
Caucasians, leaving a considerable gap in the literature.

Contemporary changes in family structures and in social attitudes toward
divorced families underscore the importance of studying the diverse roles fathers
play in their children’s lives (Harris & Salt, 1999). Given that almost half of the
children born in the last 10 years will experience the divorce of their parents
(Gallagher, 1996; Hetherington, 1992), there is growing concern about the
consequences of father separation on the child. Currently, 23% of American
children live primarily with their mothers compared with less than 3% living
primarily with their fathers (US. Census Bureau, 2001 ). Consequently, divorce
research has continued to focus primarily on families in which mother has
custody and father is the visiting parent. Of those children not residing with their
fathers, 35% see their father at least once per week and some several times per
week (Selzer, 1998).

Divorce an be a prolonged process and many practitioners and
researchers have focused on the consequences of this process on children’s
well-being. The literature indicates that parental divorce is significantly
associated with child psychosocial adjustment (for reviews, refer to Amato, 1993;
Amato & Keith, 1991); however, there is substantial variability in children’s
response to divorce (Amato, 1993; Jekielek, 1998; Seltzer, 1994; Whiteside &

Becker, 2000). This variability has led to the investigation of family process

 

5533'
have
benNi
fixhe

«m4
553'. m

Toma
exar
ates
Iowa

Afiat

SEWI
res;
teem

D's-r?

variables that may explain the parental divorce-child psychosocial adjustment
relationship. Parent-child relationships and parental conflict are among the
family process variables that have been associated with the adjustment of
children after parental divorce (Whiteside & Becker, 2000). Although few studies
have examined family process variables to clarify the long-term association
between parental divorce and psychosocial adjustment of older adolescents,
father-adolescent and father-young adult relationships have been found to be
significant predictors of young adult psychosocial adjustment (Summers,
Forehand, An'nistead, & Tannenbaum, 1998). Missing in the research is the
examination of the father-adolescent relationship in relation to 1) the mother-
adolescent relationship and 2) the degree of conflict exhibited by the mother
toward the father following a divorce.
Attachment Theory

Over the last several decades, attachment theory (Bowlby, 1969) has
served as an important framework for conceptualizing the parent-child
relationship. Parental attachment refers to the enduring emotional bond and
feelings of security children perceive in their relationship with parents. A basic
premise of this theory is that the infant’s initial experiences of attachment
become cognitively represented in the form of “internal working models” of the
self and others. These working models incorporate expectations about how
worthy one is and how accessible and responsive others will be to providing
support and protection. In regards to fathers, few studies have examined the

benefits of paternal attachment in postdivorce families, and little research has

 

diVOhf

specifically explored the effects of divorce on the father-child relationship. Amato
and Booth (1996) found that some young adults who experienced divorce as
children show less affection for their noncustodial father than for their custodial
mother. Yet, why this is the case for some young adults and not others has not
been identified. We do know however that children who perceive their fathers as
accepting, supportive, and trustworthy are more likely to report feeling closer to
their fathers (Stevenson & Black 1995).
Influences of Attachment on Separated/Divorced Families

Secure attachment emerges from the supportiveness and safety of the
child’s environment (Thompson, Lamb, & Estes, 1982), which is affected by the
type of relationship between the parents (Das Eiden, Teti, & Corns, 1995). When
the caregiving environment changes because of parental separation or divorce,
the style of attachment (secure versus insecure) may change as well (Asendorpf
8: Wilpers, 2000; Belsky, Campbell, Cohn, & Moore, 1996). Within families
where mothers are the custodial parent, father-child attachment style has been
shown to be vulnerable to changes in the caregiving environment (Cummings &
Watson-O’Reilly, 1997) and to the relationship children have with their mothers
(Doyle, Markiewicz, Brendgen, Lieberman, & Voss, 2000; Hojat, 1998). The
influence of mothers on children’s well-being is prevalent in the research. Within
divorced families, children’s attachment classifications have been found to
change in relation to changes in maternal behaviors (Vaughn, Egeland, Sroufe, &
Waters, 1997) and children’s perceptions of maternal attachment, rather than

paternal attachment, have been linked to psychosocial adjustment in adulthood

 

«$.-
A...

at
RSI;

590:1:
93's.»

”55a:

(Hojat, 1998). In addition, mothers appear to have the ability to facilitate or
impede the noncustodial father's involvement with his children (Braver, Wolchik,
Sandler, Fogas, & Svetina, 1991), thereby encouraging or dissuading the father-
child relationship. A next step in this line of inquiry was to explore how children’s
attachment style with custodial mothers influenced children’s attachment style
with noncustodial fathers.
Relationship between Maternal and Paternal Attachment

Attachment style is often viewed as stable and consistent across different
relationships due to the strong influence of the internal working model
(Bartholomew & Horowitz, 1991; Hazan & Shaver, 1987), however most
researchers report independence in the attachment classifications for mother-
child and father-child dyads (Grossman, Grossman, Huber, & Warner, 1981;
Kenny, Lomax, Brabeck, & Fife, 1998; Main & Weston, 1981). In their meta-
analyses of 14 studies, Van lJzendoorn and DeWolff (1997) stated, “By and
large, infant attachment security does not appear to generalize substantially
across relationships within the family system” (p. 607). They concluded that
infant attachment security is more relationship-specific than infant-specific.
However a meta-analyses of 11 studies involving 672 participants found that
attachment to mother and attachment to father were interdependent (Fox,
Kimmerly, & Schafer, 1991). That analysis indicated that children classified as
secure to one parent are highly unlikely to be classified as insecure to the other
parent and vice versa. Still other studies have found that attachment in parental

relationships cannot be sufficiently captured without examining consistency

 

acres
Ase:
stre

iv!"
U'hbt

mid.
12 s:
deve:
pare:
(first:

WW3"
" V \I

pater

across relationships as well as relationship-specific qualities. For example,
Asendorpf and Wilpers (2000) found that changes in young adult attachment
style reflected changes in their parental relationship style, and they also found
support for cross-relationship consistency of attachment between the two
parental relationships. The inconsistency of findings lent merit to further
examination of the relationship between the three family dyads, i.e., the mother-
child, the father-child, and the mother-father relationship. In a meta-analysis of
12 studies, Whiteside and Becker (2000) concluded that relationships children
develop with their fathers cannot be viewed independently of the quality of the
parental alliance. Custodial mothers serve as primary caregivers and have more
direct influences on children’s everyday activities, so it was important to explore
whether children’s style of maternal attachment influenced children’s style of
paternal attachment.
Maternal Conflict Wrth Fathers

Adolescent psychosocial adjustment has been associated with supportive
co-parenting, low conflict between parents, and better personal adjustment of
parents, while adolescent difficulties have been associated with continued anger
and disagreement between parents (Hetherington, Cox, & Cox, 1982; Wallerstein
8. Kelly, 1980). According to Kurkowski, Gordon, and Arbuthnot (1993), children
from divorced families are caught in the middle of parental conflicts significantly
more often and experience more stress than children from intact homes. Ahrons
(1994) found that fifty percent of white, middle-class divorced couples engage in

bitter, open conflict and only twelve percent were able to create pleasant, Iow-

 

Cox.
sac;

Tall:

disc:

invc:

conflict relationships after the divorce. Interparental conflict has been identified
consistently as a major correlate of behavior problems in children across a wide
array of family structures and settings (for reviews Davies 8. Cummings, 1994;
Erel & Burman, 1995). A meta-analysis of 57 studies focusing on the style of
paternal involvement indicated that enduring emotional bonds are more strongly
associated with positive child outcomes than with frequency of contact (Amato &
Gilbreth, 1999). However, without spending time with their children, it is less
likely that fathers will have the opportunity to develop and/or maintain a quality
relationship with their children. The level of father involvement after divorce can
be substantially affected by the co-parental relationship (Hetherington, Cox, &
Cox, 1982; McKenry, Price, F ine, & Servovich, 1992). Conflict with the former
spouse is one of the major barriers to maintaining father involvement (lhinger-
Tallman, Pasley, 8. Buehler, 1993).

Interestingly, Whiteside and Becker (2000) in their meta-analysis
discovered the impact of cooperation between parents and the level of father’s
involvement with children were different for the mother-child relationship than for
the father-child relationship. Whereas cooperative parenting facilitated the
father-child relationship through greater frequency of father visits with their
children and higher-quality relationships with children, cooperative parenting,
hostility between parents, and greater frequency of father visits were negatively
related to the mother-child relationship style. Fathers that were less collaborative
were associated with mothers that exhibited more warmth and support of children

(Whiteside & Becker, 2000). Therefore, even though high conflict between

 

pa'e

A"
,.
CC'Vb

mct.‘
355-:

stjte

Over 1

736:

If”.
'a

parents interferes with the mother-child and father-child relationships,
cooperative parenting and high paternal involvement may correspond with
mothers having less time with their children and feeling less validated in their role
as a mother. Thus, in addition to investigating the interdependence of the
mother-child and father-child attachment style, it is important to evaluate the
association between maternal conflict with fathers and father-child attachment
style. If mothers benefit from less father involvement and poor father-child
relationships, mothers may be consciously or subconsciously sabotaging their
children’s-relationship with their fathers. In this study, the construct of maternal
conflict with father will be measured by examining adolescent perceptions of
maternal hostility toward father, maternal facilitation] interference of paternal
involvement, and cooperative parenting behaviors of mothers.
Race, Separation/Divorce, and Attachment

A significant limitation of the literature on parental separation/divorce and
adolescent adjustment is its exclusive focus on the experience of Caucasian,
middle class children and adolescents (Hines, 1997; Ogbu, 1988). The
percentage of those who are separated or divorced has increased across all
racial groups. Census Bureau statistics show that in 2004, 15.5% of African
Americans (4.3 and 11.2% respectively), 11.3% of Caucasians (1.7 and 9.6%
respectively), and 10.6% of Hispanics (3.5 and 7.1% respectively) age 18 and
over were separated or divorced (US. Bureau of the Census, 2005). African
American and Hispanic children are more likely than Caucasian children to be

living with one parent; 57% for African American children, 32% for Hispanic

 

chi}:
T“
(L! a

df‘iOl

eTICC
Aft:

. a...
that

"Bar.
Set:

Date

31kg,

605:6

children, and 22% for White children in 1996 (US. Bureau of the Census, 1996).
Thirty-seven percent of these one-parent households are the result of parental
divorce (US. Bureau of the Census, 1996).

It is imperative that researchers broaden their perspective in this area to
encompass race and ethnicity. Separation and divorce rates are higher for
African American families, but we know little about the effects of divorce and the
influence of the parent-adolescent relationship on African American youth and
the literature contains virtually no information on the effects of divorce on
Hispanic youth (Hines, 1997). Although few studies have examined
race/ethnicity to clarify the long-term association between parental divorce and
psychosocial adjustment of older adolescents, those that have run analyses
differentiating between African American and Caucasian youth report
inconsistent results. Some researchers have reported no significant differences
between African American and Caucasian adolescents’ attachment style
(Bakermans-Kranenburg, Van lJendoom, & Kroonenberg, 2004; Mason, 2005;
Schlack, 2003; Van IJzenboorn & Krooneneberg, 1988) or their response to
parental divorce (Rice, Cunningham, & Young, 1997: Veneziano & Rohner, 1998;
Zimmerman, Salem, & Maton, 1995), while others have reported significant
differences (Dombusch, Carlsmith, Bushwall, Ritter, Leiderman, Hastorf, &
Gross, 1985; Famworth, 1984; Hanson, 1999; Lopez, Melendez, & Rice, 2000).

A culturally infused understanding of attachment leads to the
acknowledgement that parenting practices are embedded within the cultural

context of a parent-child relationship and greater attachment-related autonomy or

proximity is encouraged as an element of the working model and attachment
style of that culture (Van lJzendoorn & Sagi, 1999). For example, the pattern of
establishing extended families appears to be an adaptive strategy common to
African American and Hispanic individuals (Harrison, Wilson, Pine, Chan, 8:
Buriel, 1990). Three times as many African American children under the age of
18 live with their grandparents, as do Caucasian Americans (US. Bureau of the
Census, 1996). Kenny and Perez (1996) found that 27% of minority students
identified their primary attachment figure as someone other than a parent. A
caregiving environment that consists of a network of more or less stable
relationships between the individual and several caregivers may be optimal
(Donley, 1993) and may shelter the adolescent from some of the negative
consequences associated with parental divorce, including insecure attachment
style and parental conflict. In this study, racial/ethnic differences in postdivorce
parental relationships and well-being among African American, Hispanic, and
Caucasian adolescents were explored.
Age and Attachment

The role of attachment relationships in promoting well-being during late
adolescence (18-23 year olds), has been receiving increased attention
(Woodward, Pergusson, 8. Belsky, 2000). In an attempt to identify predictors of
college adjustment and well-being among late-adolescent students, researchers
have studied the influence of secure parental attachment (Arrnsden &
Greenberg, 1987; Brack, Gay, & Matheny, 1993; Bradford & Lyddon, 1993;
Kenny, 1987; Kenny & Donaldson, 1991; Larose and Boivin, 1998; Mattanah,

10

Hancock, & Brand, 2004; Rice, Fitzgerald, Whaley, 8: Gibbs, 1995; Vrvona,
2000). Parental figures serve as a secure base from which adolescents explore
and as a source of comfort in times of stress (Kems, Klepac, & Cole, 1996;
Larson, Richards, Moneta, Holmbeck, & Duckett, 1996). From an adolescent
developmental perspective, researchers found that adolescent development
tends to proceed best when adolescents develop some autonomy from parents,
while simultaneously maintaining an ongoing supportive and close parent-
adolescent relationship (Grotevant, 1989; Grotevant & Cooper, 1986). The
experience of entering college has been described as a naturally occurring
separation experience that activates attachment-related processing because the
adolescent must navigate, explore, and master a novel environment (Kenny,
1987). A number of studies have shown that students securely attached to their
parents report better social, academic and emotional well-being in college
(Bradford & Lyddon, 1993; Holmbreck & Wandrei, 1993; Kenny & Perez, 1996;
Larose & Boivin, 1998; Rice, et al, 1995; Schultheiss & Blustein, 1994), greater
social connectedness with friends and less loneliness (Blain, Thompson &
Whiffen, 1993; Brack, Gay, & Matheny, 1993), and greater psychological well-
being (Arrnsden 8: Greenberg, 1987; Cavell, Jones, Runyan, Constantin-Page, &
Velasquez, 1993; Vrvona, 2000). The symbolic or physical presence of parental
attachment figures can create conditions of safety, security, and guidance that
enable students to explore the novel environment with feelings of confidence
(Holmbreck & Wandrei, 1993; Papini & Roggman, 1992) and obtain mastery so

that students are less likely to respond to challenges with disappointment and

11

feelings of helplessness (Annsden & Greenberg, 1987). The degree to which
students exhibit well-being is strongly related to their style of parental attachment
(Klasner 8. Pistole, 2003).

Although studies have included students from a variety of ethnic
backgrounds, the influence of attachment theory on the well—being of college
students has been studied mostly among white middle-class college students
(Holmbreck & Wandrei, 1993; Kenny, 1987; Kenny & Donaldson, 1991; Rice 8.
Cummins, 1996). As a result, little is known about attachment and well-being
among nonmajority racial/ethnic groups (Kenny 8. Rice, 1995). For African
American students at predominantly White universities, some researchers
(Dewitt-Parker, 2000; Duesterhaus, 2004; Hinderiie & Kenny, 2002) found that
parental attachment type was predictive of college well-being, the development
of life goals, and the development of extrinsic aspirations. Other researchers
have found secure parental attachment positively associated with psychological
well-being among racially and ethnically diverse college students, including
Hispanic and African American students (Kenny & Perez, 1996; Taub, 1995).
These findings are consistent with the relationships found between parental
attachment style and late-adolescent well-being among Caucasian samples.
Additional evaluation of the influence of parental attachment style on late-
adolescent well-being is warranted, particularly among majority and nonmajority
students from divorced families. In this study, attachment to parents was

examined by comparing mother-attachment and father-attachment with regard to

12

three subscales: feelings of affects, parents’ promotions of independence, and
parents as sources of support.
Other Background Variables

In addition to age, several other factors are frequently examined and/or
controlled for in studies involving parental attachment and divorce. Potential
covariates noted in the literature included postdivorce marital status (Lopez et al.,
2000), frequency of father contact (Hetherington, et al., 1982; Wallerstein & Kelly,
1980), socio—economic status factors (Whiteside & Becker, 2000), current living
arrangements (McCormick & Kennedy, 2000), time since divorce (Lewis, Feiring,
& Rosenthal, 2000), number of children in household, household income,
parental educations, and age (Hanson, 1999). All participants provided the
following information: age, gender, race, socio-economic status (SES),
educational level, SES and educational level of both parents, age at time of
parent's separation, age at time of parent’s divorce, length of the marriage, years
resided in one-adult home, current living situation (e.g., parent, self, other),
number of siblings, birth order, parents current marital status, age at time of
remarriage, frequency of father contact currently and at time of divorce, and a
description of contact with father. In addition, as measures of well-being may be
confounded by social desirability (i.e., tendencies to answer questions in terms of
social appropriateness rather than true well—being) and ethnic identity (i.e.,
attitudes and knowledge of who one is as a member of an ethnic group), a social
desirability scale and ethnic identity measure was included in the data collection.

After data was collected, it was analyzed to determine covariates. Covariates,

13

namely, unwanted variance, were controlled in the final analysis and are
discussed in the results section.
Parental Attachment and Adolescent Well-being

The concept of well-being has been defined and measured in many
different ways, by many different researchers (Adams, Bezner, & Drabbs, 2000;
Adams, Bezner, & Steinhardt, 1997; Ardell, 1986; Depken, 1994; Greenburg,
1985; Travis 8. Callander, 1990). Within adolescent research, strong emphasis is
placed on psychological and spiritual dimensions of well-being. Therefore, the
dimensions of psychologiml and spiritual well—being will be measured in this
study, in addition to the physical, social, emotional, and intellectual dimensions of
well-being.

Noticeable emphasis on spirituality has been evident in the literature and
research over the past fifty years (refer to Adams et al., 2000 for a review).
Dunn stated in 1966 that “we can no longer ignore the spirit... as a factor in our
medical and health disciplines... which are designed as though the sum total of
our concern is for the body and the mind... leaving to metaphysics and religion
the affairs of the spirit" (p. 216). Over the past couple decades, several
researchers have included spirituality as a dimension of well-being (Chapman,
1987; Depken, 1994; Myers, 1992; Travis & Callander, 1990; Whitmer 8.
Sweeney, 1992). Although the conceptualizations of the spiritual dimension vary
among researchers, 3 common thread includes a sense of meaning and purpose
in life (Adams et al., 2000; Chapman, 1987; Hettler, 1984; Seaward, 1995;

Whitmer & Sweeney, 1992). Krippner (2005) defined spiritual as “an awareness

14

of a broader life meaning that transcends the immediacy of everyday physical
expediency” (p. 225). Likewise, Paulson (2005) defined spirituality as an
awareness of a broader life meaning, extending beyond everyday matters and
natural concerns. The conceptualization of spiritual well-being in this manner
allowed students to respond to the spiritual questionnaire from either a secular or
religious context. Based on their own belief systems, the students decided
individually whether their sense of meaning and purpose in life was fulfilled via
mystical peak experiences, connection to other people, linkage to a greater
power than the self, or some other source. In this study, the spiritual dimension
of well-being was defined as a positive sense of meaning and purpose in life.

Self-esteem has been empirically shown to be a good indicator of general
psychological health and will be used in this study to measure adolescents’
psychological well-being. In studies of adolescent’s well-being, the adolescents
self-esteem is frequently assessed (Kenny, Lomax, Brabeck, 8: Fife, 1998;
Leonardi 8. Kiosseoglou, 2000; McCormick, & Kennedy, 2000; Meyers, 1998).
Self-esteem refers to an attitude of self-acceptance, self-approval, self-respect,
and self-worth. Whereas parental divorce has been shown to have long-term
detrimental effects on self-esteem (Billingham & Abrahams, 1998), several
studies have shown that the level of adolescents’ self-esteem is directly related
to the style of attachment to parents (Carranza & Kilmann, 2000; Hojat, 1998;
McCormick, & Kennedy, 2000; Meyers, 1998). Carranza and Kilmann (2000)
found that young women in intact families who perceive their fathers as

unapproachable and unsupportive demonstrate insecure attachment patterns

15

 

 

arc

 

 

3:3

wl‘.’

 

pate

sci),

and lower levels of self-esteem than those that perceive their fathers as
approachable and supportive. In addition, Hojat (1998) reported that satisfaction
with mother was significantly associated with higher self-esteem in their children.
McCurdy and Scherrnan (1996) found that adolescents who have

experienmd parental divorce and subsequent mother-custody arrangements are
at risk for high conflict and poor quality of emotional attachment with fathers,
which could limit their overall sense of well—being. Thus, the assertion that a
positive relationship with mother is a protective factor against the adverse effects
of divorce (McCombs 8 Forehand, 1989) has been contradicted by the finding
that a conflictual relationship with even one parent is associated with poorer well-
being (Lapsley, Rice, & Shadid, 1989; McCurdy & Scherrnan, 1996). Adolescent
well-being appears to be associated with attachment to fathers, especially in the
subscales of the affective quality of the relationship and fathers as sources of
support (McCurdy & Scherrnan, 1996). The relationship between adolescents’
well-being and style of attachment with custodial mothers and with noncustodial
fathers has supplied valuable information regarding the adolescents’ overall
adjustment to the divorce environment and has provided insight and direction for
practitioners working with families, divorcing parents, and adolescents.
Research Questions

Numerous attempts have been made to account for the differences in
children’s reactions to parental divorce. Variables such as maternal stress,
parenting style, conflict between parents, absent parents, custody arrangements,

schedules, economic resources, parental behavior, attachment quality and

16

environmental factors (e.g., remarriage, father adjustment, mother adjustment)
have all been researched (Whiteside & Becker, 2000). However, few, if any,
studies have simultaneously examined parent—adolescent attachment style,
maternal conflict with father, gender, well-being and the interactional effects
between these variables among African American, Hispanic, and Caucasian
adolescents. The goal in this study was first to examine the main effects and
interactive effects of style of maternal attachment and maternal conflict with
father on the style of paternal attachment within each racial/ethnic group. This
shed light on whether postdivorce style of matemal—adolescent attachment and
maternal conflict with father each contribute important and unique variance to
predictions of the style of the father-adolescent attachment and whether the
interaction of these factors also contributes significant variance to these
predictors. Secondly, an evaluation of the adolescents’ general, psychological,
and spiritual well-being provided some indication of the adolescents’
internalization of the main effects and suggested possible impacts of these
effects. Thirdly, other background variables (e.g., social desirability and ethnic
identity) that were significantly related to focal variables were controlled for in
order to enhance the overall predictive power of the results. Given these
objectives, the following questions were addressed within and between each of
the racial/ethnic groups:
1. Among adolescents from divorced families, is there a relationship between
a) the style of maternal attachment and the style of paternal attachment,

b) the degree of maternal conflict with father and the style of paternal

l7

 

F.)

3

Sun:

"We:

la'nl

 

atar

afid

 

attachment, c) the style of maternal attachment and adolescents well-
being, and d) the style of paternal attachment and adolescents well-being?
2. Among adolescents from divorced families, controlling for the quality of
maternal attachment, does the degree of maternal conflict with father
uniquely predict the style of paternal attachment?
3. Does the style of maternal attachment moderate the relationship between
the degree of maternal conflict with father and the style of paternal
attachment?
4. Among adolescents from divorced families, holding background variables
constant, does maternal attachment style and maternal conflict with father
uniquely predict the relationship between the style of paternal attachment
and adolescent well-being?
Summary and Hypotheses

Attachment theory provides a strong conceptual framework from which to
investigate the relational bond between parents and adolescents from divorced
families. Yet, little research has been done applying the basic assumptions of
attachment theory to the explanation of why some adolescents experience long-
term effects of divorce while other adolescents do not. Even less research has
been done in this area with postdivorce African American and Hispanic youth. In
this study, the impact custodial mothers have on father-adolescent relationships
and adolescent well-being was evaluated for African American, Hispanic and

Caucasian adolescents.

18

 

 

and

ate .

 

The purpose of this study was to explore the associations of both 1)

maternal attachment and 2) maternal conflict with father with paternal attachment

and well-being among adolescents from divorced families. Drawing from

attachment theory and the literature on divorce, it was hypothesized that:

1.

3.

Among adolescents from divorced families, secure attachment to mother
and low levels of maternal conflict with father would correspond with
secure attachment to father. Conversely, insecure attachment to mother
and high levels of maternal conflict with father would correspond with

insecure attachment to father.

. Secure attachments to parents would be associated with higher levels of

general, spiritual, and psychological well-being, while insecure
attachments to parents would be associated with lower levels of general,
spiritual, and psychological well-being.

In regards to race/ethnicity, it was hypothesized that:

a. No significant differences would be found differentiating Hispanic
and African American adolescents’ from Caucasian adolescents’
parental attachment patterns.

b. Hispanic and African American adolescents’ would demonstrate
significantly more adaptive coping strategies to insecure parental
attachment style and maternal conflict with father as evidenced by
greater overall well-being than the Caucasian adolescents’ with

insecure parental attachment style and higher levels of conflict.

19

 

 

Literature Review

While the majority of adolescents from divorced families fall into the
normal range on measures of both psychological and cognitive functioning (Kelly,
2000), many adolescents show prolonged negative reactions to the divorce
process, and great variability exists amongst both of these groups (Jekielek,
1998; Seltzer, 1994; Whiteside & Becker, 2000). Parental divorce during
childhood may adversely impact the style of the parent-child relationship, leading
to long-term psychosocial adjustment difficulties (Emery & Forehand, 1994). The
literature indicates that parental divorce presents unique obstacles to the
development and maintenance of positive parental relationships, especially with
noncustodial fathers. This section will review six lines of research relevant to the
proposed study: 1) postdivorce father involvement and father-child relationships,
2) mother-child, father-child, and mother-father relationships, 3) parental conflict
and child well-being, 4) attachment theory, 5) age and attachment, and 6)
race/ethnicity.
Postdivorce Father Involvement and Father-Child Relationships

The population of divorced Americans continues to be substantially high,
estimated at nearly 20 million in 1998, up from 11 million in 1980 (Raymond,
2001). Since 1960, the divorce rate in the United States has more than doubled
and the last-reported U.S. divorce rate as of May, 2005, is 38% (National Center
for Health Statistics, 2005). A meta-analysis of 37 studies examining the long-
terrn relationship of parental divorce on children and adolescents found they

were more likely to have psychological and social difficulties and to score lower

20

on measures of the quality of mother-child and father-child relationships (Amato
& Keith, 1991). One proposed explanatory idea for these findings posits that the
negative relationship between parental divorce and the adjustment of
adolescents results from disrupted parent-child relationships (Emery & Forehand,
1994). Although the potentially negative effects of divorce can be mitigated by
children maintaining good relationships with both parents (Barnes, 1999), several
studies indimte a precipitous decline in father contact after divorce, with 23% to
30% of sampled children and sampled fathers reporting no contact with each
other during the first year postdivorce (Furstenberg, Morgan, 8. Allison, 1987;
Mitchell, 1985; Seltzer, Schaeffer, & Chamg, 1989). Attempts to elucidate the
reasons for the decrease in noncustodial father-child contact after divorce have
been minimal (Dudley, 1991), with most efforts identifying loosely connected
predictors of father involvement (Stone & McKenry, 1998). However, since
fathers with high frequency of visitation are more likely to have high-quality
relationships with their children (Whiteside & Becker, 2000), and high-quality
relationships are positively related to children’s well-being (Hetherington, Cox, &
Cox, 1982), lack of father involvement following divorce needs to be understood.
To do this, the father-child relationship needs to be examined in the context of
both the level of father’s involvement and the relationship between the parents.
The research conducted by Hetherington, Cox, and Cox (1982) and by
Wallerstein and Kelly (1980) concluded that in the two years following parental
separation, positive adjustment for children is associated with supportive

coparenting and low conflict between parents, whereas child difficulties are

21

 

 

associated with disagreement and inconsistency between parents and continued
anger between parents. While supportive coparenting and low conflict between
postdivorce parents is ideal, this request may present a tall order for parents to
fill. No matter how smoothly parents manage the divorce process, when they
part, they face the reality that the time children spend with one parent inevitably
mandates separation from the other parent. The impact of this appears to be
particularly relevant in regard to maternal role identity. Families with high
paternal involvement and cooperation correspondingly have less residential time
with mother and mothers may oppose the impact this has on their relationship
with their children. For example, Little (1992) found that mothers were more
satisfied with parenting when they had primary custody than when they have joint
custody, indicating a preference for more maternal control of the mother-child
and the father-child relationships. Kruk (1993) identified ways in which the
custodial mother discouraged father contact and Lehr and MacMillan (2001)
found that poor father relationships with custodial mother resulted in differences
in access to the children as well as loss of day-to-day contact with them. These
findings highlight the importance of evaluating the father-child relationship in
relation to the mother-father dyad and the mother-child relationship.
Mother-Child, Father-Child, and Mother-Father Relationships

While most studies have focused on either the parent-child or the mother-
father relationship, theorists and investigators have increasingly recognized that
the different dyads within the family (mother—child, father-child, mother-father) are

mutually interdependent (Cook, 1994; Lytton, 1990). A meta-analysis of 11

22

35

c?

EX
TE;

let

ha:

tor:

L."

We

studies involving 672 participants found that attachment to mother and
attachment to father were significantly related (Fox, Kimmerly, & Schafer, 1991 ).
In another meta-analysis of 12 studies, researchers report consistent positive
correlations between mother-child attachment style and father-child attachment
style (Whiteside & Becker, 2000). This means that when one parent has a good
relationship with the child, it is likely that the other parent does also. Conversely,
children classified as secure to one parent are highly unlikely to be classified as
insecure to the other parent. Interdependence is found between the mother-child
relationships, the father-child relationships, and the mother-father relationship’s
impact on child behaviors, yet each has unique distinctive influences. For
example, Mathijssen, Koot, Berhulst, DeBruyn, and Oud assessed the relative
association of the different family dyads and discovered that both the mother-
child and the mother-father relationship were positively related to child behavior.
However, whereas the mother-child relationship was consistently related to
externalizing behavior of children, the mother-father relationship was particularly
related to internalizing behavior of children. Offer, Ostrov, & Howard (1998)
found that both mother-child and father-child relationship style significantly
correlated with adolescents’ self-image; however, the mother-child relationship
had a greater influence. Miller and Stubblefleld (1993) concluded that disclosure
to mothers was more common than disclosure to fathers, but the strongest
predictor of disclosure to one parent was disclosure to the other parent. These
studies illustrate that the influence of family relations on the child cannot fully be

understood by studying one dyad in isolation. Whiteside (1998) found that

23

althoug
Wow:
the co;
reéatc:
dyfiam
Comm

Parent

m
Pores?
“WI
330295
bene-
Cox. 8
and A:

 

£2"

:3

“-u
1

 

although a substantial number of studies examined the mother-child dyad
following divorce, fewer have included information about the father-child dyad or
the coparent alliance. The parent-child relationships and mother-father
relationship must each be included when examining the relationships of family
dynamics and between family dynamics and child well-being (Davies 8.
Cummings, 1994; Rothbaum & Weisz, 1994).
Parental Conflict

Interparental conflict and the parent-child relationship have each been
associated with child adjustment after parental divorce (Amato, 1993; Emery 8.
Forehand, 1994; Summers et al., 1998). Supportive co-parenting, low conflict
between parents, and better personal adjustment of parents are associated with
adolescent psychosocial adjustment, while continued anger and disagreement
between parents have been associated with adolescent difficulties (Hetherington,
Cox, 8. Cox, 1982; Wallerstein & Kelly, 1980). According to Kurkowski, Gordon,
and Arbuthnot (1993), children from divorced families are caught in the middle of
parental conflicts significantly more often and experience more stress than
children from intact homes. Ahrons (1994) found that fifty percent of white,
middle—class divorced couples engage in bitter, open conflict and only twelve
percent were able to create pleasant, low-conflict relationships after the divorce.
Interparental conflict has been identified consistently as a major correlate of
behavior problems in children across a wide array of family structures and
settings (for reviews Davies & Cummings, 1994; Erel & Burman, 1995). In a

recent meta-analysis, Amato and Keith (1991) found that parental conflict

24

dive!
(He:

Cont

mfoh

 

 

accounted for more of the negative consequences of divorce than both parental
absence and economic disadvantage.

Interparental conflict plays a central role in the frequency of father
visitation and consequently with the father-child relationship style (Whiteside &
Becker, 2000). A meta-analysis of 57 studies focusing on the style of paternal
involvement indicated that enduring emotional bonds are more strongly
associated with positive child outcomes than with frequency of paternal contact
(Amato & Gilbreth, 1999). However, without spending time with their children, it
is less likely that fathers will have the opportunity to develop and/or maintain a
quality relationship with their children. The level of father involvement after
divorce can be substantially affected by the co—parental relationship
(Hetherington, Cox, & Cox, 1982; McKenry, Price, Fine, & Servovich, 1992).
Conflict with the former spouse is one of the major barriers to maintaining father
involvement (lhinger-Tallman, Pasley, & Buehler, 1993).

Continuing conflict with the ex-wife has been shown to impact both how
fathers feel about their children and what they do in relation to them (Goldberg,
1990). Therefore, it is not surprising that parental conflict is related to insecure
attachments (Lewis, Fairing, & Rosenthal, 2000), but it is interesting that
conflictual mother-father relations have more negative effects on father-child
relations than on mother-child relations (Cummings 8: Watson-O’Reilly, 1997).
Similarly, a meta-analysis conducted by Whiteside and Becker (2000)
demonstrated that the impact of cooperation between parents and the level of

faflwer’s involvement with children were different for the mother-child relationship

25

than for the father-child relationship. Cooperative and supportive coparenting
skills were significantly associated with greater frequency of father visits and
father-child relationship style, whereas cooperation and greater frequency of
father visits negatively related to the mother-child relationship style (Whiteside &
Becker, 2000). Fathers that were less collaborative were associated with
mothers that exhibited more warmth and support of children (Whiteside &
Becker, 2000). Orbuch, Thornton, and Cancio (2000) found that the mother-
daughter relationship improved after divorce. Therefore, even though high
conflict between parents interferes with the mother-child and father-child
relationships, cooperative parenting and high paternal involvement may
correspond with mothers having less time with their children and feeling less
validated in their role as a mother. Thus, a more complex relationship between
divorce, parental conflict, and child well-being is implied. If mothers benefit from
less father involvement and poor father-child relationships, mothers may be
consciously or subconsciously sabotaging their children’s relationship with their
fathers. Hanson’s (1999) results suggested that children’s exposure to conflict
accounted for about 11% of the effects of divorce on child well-being, suggesting
that conflict only partially explains why parental divorce and child well-being are
related. In this study, the construct of maternal conflict with father was measured
by examining adolescent perceptions of maternal hostility toward father, maternal
facilitation! interference of paternal involvement, and cooperative parenting
behaviors of mothers.

Attachment Theory

26

Over the last several decades, attachment theory (Bowlby, 1969) has
served as an important framework for conceptualizing the parent-child
relationship. Attachment theory is based on the notion that there are individual
differences in the way infants become emotionally attached to their primary
caregivers and that these differences influence a child’s perceptions of self,
others, and resources for emotional self-regulation in times of crisis (Bowlby,
1969). An infant’s relationship with its caretakers evolves into a working model
that organizes behavior (Ainsworth & Bowlby, 1991). Initially, attachment
behavior manifests in crying, smiling, and vocalizing which serve to keep the
infant in close proximity to the caregiver for purposes of protection (Bowlby,
1969). During the early stages of development, the infant forms internal
representations of self, attachment figure, and the environment (Bowlby, 1969).
The attachment system thereby becomes a component of the cognitive structure
subject to developmental change as well as environmental influence (Ainsworth,
1 989).

One basic principle of attachment theory is that attachment relationships
continue to be important throughout the lifespan (Ainsworth, 1989). If the
caregiver is experienced as warm, responsive, dependable, and encouraging of
autonomy, secure attachment characterizes the relationship (Hazan & Shaver,
1987). However, if the caregiver is experienced as cold, neglectful, inconsistent,
rejecting, or intrusive, an insecure attachment is likely to characterize the
relationship (Hazan & Shaver, 1987). Assessments of attachment relations

during adolescence have indiwted continuity of both mother- and father-child

27

attachment over time (Lopez, 1996; McCormick & Kennedy, 1994), as well as
similar proportions of persons in each attachment category as reported in infancy
(Hazan & Shaver, 1987). Attachment security has been shown repeatedly to be
an important correlate of social and emotional adjustment throughout the life
span (Bartholomew & Horowitz, 1991).

One way to look at the caregiving environment is to examine the effects of
separation/divorce on attachment styles. It is widely recognized that the
caregiving environment, particularly maternal behavior and characteristics, is
related to continuity of the attachment classification (Egeland & Farber, 1984).
However, parental separation/divorce captures many aspects of the caregiving
environment because it has a direct impact on parents, children, and the
emotional and social experiences in the family (Davies, & Cummings, 1994).

The decision to separate sets in motion a significant and potentially stressful
transition that frequently includes intense negative emotion and disruption of
everyday structures and routines. Thus separation/divorce provides the type of
environment that impacts early childhood attachment as well as effecting change
in attachment representation (Lewis, et al., 2000). Subsequently, we know that
the processes associated with the marital separation experience tend to change
the parent-child relationship, but it is difficult to ascertain exactly which variables
are significant in this transformation of parent-child relationships. So the parental
behaviors and the coparenting partnership should be considered jointly to

understand the full set of influences on the child.

28

 

Points in time of parental separation and of parental divorce are seldom
differentiated within the divorce literature and the terms are used interchangably.
Several of the more recent studies use both terms in reference to the breakup of
a parental relationship. This may be in response to the growing number of
parents that do not go through the legal process of marriage, and as a result, do
not go through the legal process of a divorce. A child in a single-parent home in
1996 was as likely to be living with a parent who had never been married (36%)
as with a parent who was divorced (37%) (US. Census Bureau, 2001). In the
current study, participants’ age at time of separation and age at time of divorce
was requested and will be evaluated.

Age and Attachment

In an attempt to identify predictors of college adjustment, researchers
have studied the influence of secure parental attachment (An'nsden &
Greenberg, 1987; Brack, Gay, & Matheny, 1993; Bradford & Lyddon, 1993;
Kenny, 1987; Kenny 8. Donaldson, 1991; Larose and Boivin, 1998; Mattanah, et
al., 2004; Rice, et al., 1995; Vivona, 2000). A number of psychologists have
emphasized the importance of autonomy and individuation as key developmental
tasks facing late-adolescent college-students (Arnstein, 1980; Chickering, 1969).
It has been proposed that students with a stronger and healthier sense of self will
be better equipped to handle the challenge of disengaging from childhood and
learning to function in the college environment as autonomous adults, including
self-organization of coursework and preparation for exams, negotiating new

social environments and relationships, and attention to career and identity issues

29

(Grayson, 1998; Kenny & Rice, 1995). From an adolescent developmental
perspective, researchers found that adolescent development tends to proceed
best when adolescents develop some autonomy from parents, while
simultaneously maintaining an ongoing supportive and close parent-adolescent
relationship (Grotevant, 1989; Grotevant & Cooper, 1985). Conversely, those
adolescents who strive too strongly to separate from their parents appear
isolated and withdrawn and are at increased risk of behavioral problems (Ryan &
Lynch, 1989). Attachment is defined as an enduring emotional bond that forms
between the parent and the child across the life span (Rice, et al., 1995) and
secure parental attachment helps the adolescent develop a positive view of self
and expect positive, supportive interactions with others (Griffin & Bartholomew,
1 994).

The experience of entering college has been described as a naturally
occurring separation experience that activates attachment-related processing
because the adolescent must navigate, explore, and master a novel environment
(Kenny, 1987). The college transition may be viewed correspondingly as a
second “strange situation” (Kenny, 1990), in which adolescents with a history of
secure attachment to their parents experience them as a secure base that
anchors and supports exploratory behavior and provides anxiety-reducing
functions (Baldwin & Fehr, 1995; Elicker, Englund, Sroufe, 1992; Hazan &
Shaver, 1994). A number of studies have shown that students securely attached
to their parents report better social, academic and emotional adjustment in

college (Bradford & Lyddon, 1993; Holmbreck & Wandrei, 1993; Kenny & Perez,

3O

1996; Larose 8 Boivin, 1998; Rice, et al, 1995; Schultheiss 8 Blustein, 1994),
greater social connectedness with friends and less loneliness (Blain, Thompson
8 Whiffen, 1993; Brack, Gay, 8 Matheny, 1993), less psychologically distressing
symptoms such as depression and anxiety, and less alcohol use (Arrnsden 8
Greenberg, 1987; Cavell, Jones, Runyan, Constantin-Page, 8 Velasquez, 1993;
Vivona, 2000). The symbolic or physical presence of parental attachment figures
can create conditions of safety, security, and guidance that enable students to
explore the novel environment with feelings of confidence (Holmbreck 8
Wandrei, 1993; Papini 8 Roggman, 1992). Attachment security provides the
resource for mastery so that students are less likely to respond to challenges
with disappointment and feelings of helplessness (Arrnsden 8 Greenberg, 1987).
In addition, students having the expectation that others will be helpful and
responsive are more likely to approach faculty for assistance and to develop
supportive friendships (Florian, Mikulincer, 8 Bucholtz, 1995; Lopez, 1996).
Thus the degree to which students’ well-being is resilient to the college transition
is strongly related to their style of parental attachment (Klasner 8 Pistole, 2003).
Although studies have included students from a variety of ethnic
backgrounds, the influence of attachment theory on the well-being of college
students has been studied mostly among white middle-class college students
(Holmbreck 8 Wandrei, 1993; Kenny, 1987; Kenny 8 Donaldson, 1991; Rice 8
Cummins, 1996). As a result, little is know about attachment and wellobeing
among nonmajority racial/ethnic groups (Kenny 8 Rice, 1995). In a study of

attachment and psychological well—being among racially and ethnically diverse

31

first-year college students, Kenny and Perez (1996) found the secure parental
attachment was negatively associated with psychological symptoms of distress.
For African American students at predominantly White universities, other
researchers (Dewitt-Parker, 2000; Duesterhaus, 2004; Hinderlie 8 Kenny, 2002)
found that parental attachment type was predictive of college adjustment, the
development of life goals, and the development of extrinsic aspirations. Taub
(1995) found that among Hispanic, African American and Caucasian students,
one of the most significant independent contributors to autonomy among college
students was the parental role of providing emotional support. These findings
are consistent with the associations found between parental attachment style and
late-adolescent well-being among white samples.

Additional evaluation of the influence of parental attachment style on late-
adolescent well-being is warranted, particularly among majority and nonmajority
students from divorced families. College students from intact families may
perceive their parents as a greater source of support, display more career
decidedness, and exhibit more financial connectedness than students whose
parents are divorced (Scott, 1999). However, studies have found that growing up
with secure parent attachment, as compared to a preoccupied or dismissing
attachment style, may increase an adolescents resilience to possible negative
effects of parental divorce (Hannum 8 Dvorak, 2004; Hayashi 8 Strickland,
1998). In this study, attachment to parents was examined by comparing mother-
attachment and father-attachment with regard to three subscales: feelings of

affects, parental promotion of independence, and parents as sources of support.

32

Race/Ethnicity and Attachment
As previously mentioned, the majority of research on divorce has focused

on homogeneous samples drawn from White, middle-class contexts, with little
attention given to African American or Hispanic marriages (Feldman 8 Elliott,
1990; Guidubaldi, 1988; Hines, 1997; Ogbu, 1988; Power, Hauser, 8 Kilner,
1989). According to Billingsley (1990), relatively little attention has been given to
the impact of divorce on African Americans because of researchers’ inability to
dispel the pathological approach to black families which emphasizes single-
parent families and teen pregnancies. Although African Americans are
associated with unique types of family patterns, most studies on Black families
have de-emphasized internal family dynamics and have focused on family
pathologies (Billingsley, 1990, Crawley 1988). Statistics indicate that
approximately 47% of African American women separate from their husbands
within 10-15 years of marriage, compared with 28% of Caucasian women.
Some researchers use this phenomenon to guide their research and explain their
results. For example, Shaw, Winslow, and Flanagan (1999) stated,

“We also expected the effects of pre- and postdivorce

marital status to be more pronounced among European

American rather than African American children, due to

the greater prevalence of two-parent families among

European American families. That is, single—parent

status may be accepted as more normative among

33

African American children regardless of whether or not a
divorce has occurred” (p. 744).

The approach to African American families as a deviant form has
obscured the diversity of African American families based on socialization
experiences, geographical regions, and socioeconomic backgrounds (Lawson 8
Thomson, 1994). According to Ainsworth (1989), the attachment model is
applicable across racial/ethnic settings, but some researchers have questioned
the ethnic specificity and measure of the attachment construct (Le Vine 8 Miller,
1990). There is some evidence that African American adolescents may be
particularly sensitive to interpersonal matters and family relationships (DeCindio,
Floyd, Wilcox, 8 McSevency, 1983; Hughes 8 Demo, 1989) and that they
perceive higher levels of intimacy with parents than do Caucasian adolescents
(Cemkovich 8 Giordano, 1987), even after controlling for socioeconomic status
(Giordano, Cemkovich, 8 DeMaris, 1993). African American adolescents have
been found to value relationships with parents, particularly mothers, more highly
than Caucasian adolescents (D’Augelli 8 Hershberger, 1992). Figures from the
US. Bureau of the Census (2000) indicate that African American and Hispanic
children will spend significantly longer periods in a single-parent, mother-headed
household than Caucasian children.

In contrast to the above mentioned results, a meta-analysis of the long-
term consequences of parental divorce for adolescent well-being found that the
effect sizes tended to be stronger for Caucasians than for African Americans for

separation and divorce, one-parent family status, material quality of life,

34

occupational quality, and educational attainment (Amato 8 Keith, 1991).
Caucasian samples were associated with significantly stronger negative effect
sizes than were African American or mixed-race samples (b = .14 and -.43,
respectively; Amato 8 Keith, 1991). Other studies have found racial/ethnic
similarities on measures of adolescent-parent relationships (Rice, Cunningham,
8 Young, 1997; Veneziano 8 Rohner, 1998; Zimmerman, Salem, 8 Maton,
1995). For example, in a study of 21 African American and 37 Caucasian fathers
and their children, Veneziano and Rohner (1998) found that only perceived
paternal acceptance was significantly related to African American and Caucasian
children’s psychological adjustment and that father involvement by itself was not
related to either group’s psychological adjustment.

Regarding marital conflict, African American women seem to exhibit
greater levels of disrupted parenting before divorce than afterward, whereas
Caucasian mothers tended to show more disrupted parenting following the
marital separation than before (Shaw, Winslow, 8 Flanagan, 1999). These
findings may be significant in explaining why African American females
experienced less depression and adjusted more positively to single-parenthood
than their Caucasian counterparts (Gove 8 Shin, 1989; Menaghan 8 Lieberman,
1986; Weiss, 1975). Lawson and Thomson (1994) found that African American
men reported postmarital friendship relationships with former spouses and that
these relationships satisfied their need for attachment, perhaps through shared
parenting and social networks. This relationship between African Americans

postdivorce mother-father dyads warrants further investigation.

35

Whereas the research on postdivorce parental relationships and
adolescent well-being is limited for African Americans, it is virtually absent for
Hispanic adolescents. In one study, Lindahl and Malik (1999) found more
similarities than differences between Hispanic American and Caucasian families
in a study of intact families, reporting that marital conflict and disengaged family
alliances were associated with child behavior problems and dysfunctional
patterns of family interaction in both ethnic groups. They concluded, “although
some of the patterns of interrelatedness are influenced by culture, many are not,
suggesting that ethnic differences may somehow, but not always, affect how
family subsystems are related to one another” (Lindahl 8 Malik, 1999; p. 22).
One significant difference Lindahl and Malik (1999) found was that a hierarchical
parenting style was associated with boys’ behavioral problems for Caucasian
families, but not for Hispanic families. In Hispanic families, only lax and
inconsistent parenting was associated with externalizing behaviors. This finding
is supported by a study done by Zayas and Solari (1994).

In exploring the development of adaptive strategies, Zayas and Solari
(1994) found Hispanic families place a lot of emphasis on family solidarity and on
individuals’ sense of obligation to the family. A parenting practice that
incorporates these goals involves the insistence on children’s conformity to
parental and extended family authority (Zayas 8 Palleja, 1988). This strategy
promotes the importance of family relatedness and helps in racial identity
development of Hispanic children. Parental divorce may disrupt this adaptive

strategy that helps to not only protect the family’s continuity, but also to

36

preserves its culture. A goal of this study was to gain insight into internal family
dynamics and extend the attachment literature to Hispanic families.

Healthy well-being, which is influenced by parental attachment style, is
culturally defined (Rothman, Weisz, Pott, Mijake, 8 Morelli, 2000) and should be
evaluated within cultural contexts. Parenting practices are embedded within the
cultural context of the parent-child relationship and greater attachment-related
autonomy or proximity is encouraged as an element of the working model and
attachment style of that culture (Van lJzendoorn 8 Sagi, 1999). For example,
the pattern of establishing extended families appears to be an adaptive strategy
common to African American and Hispanic individuals (Harrison, Wilson, Pine,
Chan, 8 Buriel, 1990). Three times as many African American children under the
age of 18 live with their grandparents as do Caucasian Americans (US. Bureau
of the Census, 1996). Kenny and Perez (1996) found that 27% of minority
students identified their primary attachment figure as someone other than a
parent. A caregiving environment that consists of a network of more or less
stable relationships between the individual and several caregivers may be
optimal (Donley, 1993) and may shelter the adolescent from some of the
negative consequences associated with parental divorce, including insecure
attachment style and maternal conflict with father. An understanding of the
network of stable relationships may be important to obtaining a culturally infused
theory of attachment.

In this study, racial/ethnic differences in postdivorce parental relationships

and well-being among African American, Hispanic, and Caucasian adolescents

37

were explored. Though some inconsistencies exist, a synthesis of the literature
appears to support the postulate that no significant differences will be found
between Hispanic and African American adolescents’ and Caucasian
adolescents’ pattern of attachment. According to Bakermans-Kranenburg, Van
lJzendoorn, and Kroonenberg (2004), children of African American, Hispanic,
and Caucasian families in the United States may be exposed to culturally specific
experiences, but these do not alter the relationship between attachment security
and emotional responsiveness and support of parents. However the research
also appears to support the postulate that college students of color may be more
resilient and less affected by the stressors associated with parental divorce.
Summary

Attachment theory should provide a strong conceptual framework for
investigating the relational bond between parents and adolescents from divorced
families. The purpose of this study is to explain why some adolescents fare well
after the divorce of their parents while others do not. One underlying assumption
of this study is that postdivorce relationships with both parents are important for
the well—being of children. Parental divorce may adversely impact the quality of
the parent-child relationship, leading to long-term psychosocial adjustment
difficulties (Emery 8 Forehand, 1994). The literature indicates that parental
divorce presents unique obstacles to the development and maintenance of
positive parental relationships and overall well-being. Few studies have
evaluated the impact of parental divorce and maternal conflict with fathers on

parental attachment and the subsequent influence on adolescents’ well-being.

38

The cunent study helps to fill this gap in the literature. Additionally, this study
extends the attachment literature relevant to racial/ethnic groups and clarifies

some of the inconsistencies reported in previous research.

39

Method

Participants

The participants were 179 (141 female and 38 male) students recruited
from a major research institution located in an urban setting (population
exceeding 250,000) in the Midwest. The racial/ethnic composition was 52%
White (n = 93; 76 female, 17 male), 32.4% African American (n = 58; 47 female,
11 male), and 15.6% Hispanic/LatinolMexican American (n = 28; 18 female, 10
male). Age ranged from 18—24 (mean = 19.97). Education level Included 66
freshman (36.8%), 41 sophomores (22.9%), 42 juniors (23.5%), 27 seniors
(15.0%), and 3 graduate students (1.7%). All participants were from families in
which biological parents divorced prior to age 16 and mothers had primary
physical custody. Age of participant at time of parents’ separation ranged from
age one to age 16 (n = 122, mean = 7.3) and age of participants at time of
parents’ divorce ranged from birth to age 16 (n = 179, mean = 7.2 years). Fifty-
seven students failed to indicate their age at the time of their parents divorce, but
for those that did, no significant difference was found between age at separation
and age at time of divorce (Cohen’s d = .016; effect size = .008). Frequency of
contact with father prior to age 18 included daily (n = 39, 22%), weekly (n = 73,
41%), monthly (n = 37, 21%), and a couple times a year (n = 30, 17%). Student
participants were also single with no children and had at least an 8th grade
reading level.
Measures

All participants completed the following self-report measures.

40

Personal data questionnaire (PDQ).

This questionnaire solicited demographic information from participants as
well as information related to parents’ divorce. The questionnaire included
information on adolescents’ age, gender, race, educational level, educational
level of both parents, age at time of parents’ separation, age at time of parents”
divorce, length of the marriage, years resided in one-adult home, number of
siblings, parents’ current marital status, age at time of remarriage, frequency of
contact with mother and with father currently and at time of divorce, and a brief
description of contact with father.

Parental Attachment Questionnaire (PA Q; Kenny, 1 98 7).

Parental attachment was operationalized along a continuous scale
indicating the degree of maternal attachment and paternal attachment with
regard to three subscales: feelings of affect, parental promotion of
independence, and parents as sources of support. This measure provided full-
scale attachment scores and individual subscale scores.

The Parental Attachment Questionnaire is a 55-item Likert scale designed
to assess the perceptions of young adults about their parents’ attitudes towards
them and their relationships with their parents (Kenny, 1987). F ifty-two of the
fifty-five items were used for this study. The three items that were omitted did not
specifically address maternal and paternal attachment (e.g., “When I have a
serious problem or an important decision to make, I work it out on my own
without help,” “When I have a serious problem or an important decision to make,

I discuss the matter with a friend,” and “When I have a serious problem or an

41

important decision to make, I talk to a professional, such as a psychologist,
member of the clergy, etc”). The PAQ contains three subscales: affective
quality of attachment, parental fostering of autonomy, and parental role in
providing emotional support. Subjects answer each question using a five-point
rating scale ranging from 1 (not at all) to 5 (very much). Scores are calculated for
each of the three subscales, as well as aggregated to get a full-scale score. The
higher the number, the more positive the relationship between the subjects and
their parents, as perceived by the student completing the survey. Examples of
items from the feelings of affects scale include, “During visits or time spent
together, my mother! father was a person I wanted to be with all the time,”
“During visits or time spent together, my mother] father was a person towards
whom I felt cool and distant,” and “When I go to my mother/ father for help, I
continue to feel unsure of myself.” Examples of items from the autonomy scale
include, “In general, my mother/ father respect my privacy,” “In general, my
mother/ father restrict my freedom or independence,” and “In general, my mother/
father take my opinions seriously.” Examples of items from the emotional
support scale include, “Following time spent together, I leave my mother] father
with warm and positive feelings,” “Following time spent together, I leave my
mother] father feeling let down and disappointed" and “When I have a serious
problem or an important decision to make, I know that my mother/ father will
know what to do.” The reliability of the attachment measure was assessed by
Kenny (1987) through test-retest and internal consistency methods. Test-retest

reliability over a 2-week interval was .92 for the measure as a whole and ranged

42

from .82 to .91 for the three scales derived from factor analysis. Cronbach
coefficient alpha was calculated for each of the three scales, yielding coefficients
of .96, .88, and .88. Internal consistency for the entire measure is Cronbach
alpha of .93 for male college students and .95 for female college students.
Evidence of construct validity was obtained by correlating each of the three factor
scales with subscales of the Moos Family Environment Scale (Moos, 1985). The
three factor scales are theoretically consistent with Ainsworth et al.’s (1978)
conceptualization of attachment as an enduring affective bond.

Inter-Parental Conflict Scale (IPC; Schwarz, 1999)

This construct measured the adolescent’s perception of mother’s negative
feelings and behaviors regarding father. The Inter-Parental Conflict Scale (IPC)
was designed to assess the frequency of overt conflict between spouses. There
is a spousal form and a child form, but since this study sample included only
children from divorced families, only the child form was used. It contains a list of
34 topics, grouped in four categories: finance and responsibility (6 topics),
personal characteristics (13 topics), childrearing practices (12 topics), and family
activities (3 topics). College students rated the frequency with which they
perceived the mother initiated conflict over each of the 34 topics. IPC items are
anchored on a 7-point Likert-type scale, ranging from 0 (never) to 6 (at least once
a week). The total IPC score is obtained by taking the mean of numerical rating
for all 24 topics. Some sample topics included “insensitivity,” “methods of

discipline,” “providing financial assistance,” “punctuality” and “father’s degree of

43

involvement with children.” The internal consistency for ratings by college
students is .90 and the one-week test-retest reliability is .86 (Schwarz, 1999).

Adolescent Well-being

The construct of adolescent well-being is defined as a state of being
content and healthy and successful, permitting the experience of consistent,
balanced growth. Several dimensions of well-being are involved in human
existence, including psychological, spiritual, physical, social, emotional, and
intellectual. As stated earlier, within adolescent research, strong emphasis is
placed on psychological and spiritual dimensions of well-being (Adams, Bezner,
8 Drabbs, 2000; Adams, Bezner, 8 Steinhardt, 1997; Ardell, 1986; Depken,
1994; Greenburg, 1985; Travis 8 Callander, 1990) in comparison to the other
four dimensions (e.g., physical, social, emotional, and intellectual). In this study,
thme measures of well-being were utilized in an attempt to capture the full
essence of well-being among this college-aged population, therefore more weight
was given to the psychological and spiritual dimensions of well-being in the total
score of well-being. The standardized scores from the three measures of well-
being were added to formulate an aggregated score of well-being. The
Perceived Wellness Survey provided an estimate of all six dimensions of well-
being and the Current Thoughts Scale provided an estimate of the psychological
dimension of well-being. The spiritual dimension was measured using the
Personal Meaning Index of the Life Attitude Profile — Revised. The spiritual
dimension of well-being has been characterized as “one’s meaning of life” by

many researchers (Gatrad, Sadig, 8 Sheikh, 2003; Krippner, 2005; MacKinlay,

44

2001; Paulson, 2005; Schlesing, 2005). Krippner’s (2005) definition of spiritual
was “an awareness of a broader life meaning that transcends the immediacy of
everyday physical expediency” (p. 225). Likewise, Paulson (2005) defined
spirituality as an awareness of a broader life meaning, extending beyond
everyday matters and natural concerns. In this study, the spiritual dimension of
well-being was defined as a positive sense of meaning and purpose in life.
Based on their own belief systems, the students decided individually whether this
meaning and purpose in life was fulfilled via mystical peak experiences,
connection to other people, linkage to a greater power than the self, or some
other source.

Cunent Thoughts Scale (C TS; Heatherton 8 Polivy, 1991).

This measure of state self-esteem represents the psychological dimension
of well-being at the time the scale is completed. Self-esteem was
operationalized along a continuous scale indicating the worthiness that the
adolescent holds and maintains with regard to him or herself in social, academic,
family and personal areas of experience at that particular time. The CTS is a 20-
item self-administered inventory assessing attitudes toward the self in social,
academic, family, and personal areas of experiences. Scores are indicators of
the sense of worthiness that individuals hold toward themselves at that moment.
CTS presents respondents with generally favorable (e.g., “I feel confident about
my abilities”) or generally unfavorable statements about the self (e.g., “I am
worried about what other people think of me”), which they indicate as: not at all
(1), a little bit (2), somewhat (3), very much (4), or extremely (5). The CTS is

45

scored as a Likert scale. The original sample for which the scale was developed
in 1991 consisted of 428 undergraduates enrolled in Erindale college of the
University of Toronto (mean age = 20.3). According to Heatherton and Polivy
(1991), the measure of sampling adequacy (MSA) values for each item were
found to all be over .80, indicating an acceptable ratio of interitem correlation in
partial correlation coefficients (total MSA = .92). The scale also has a high
degree of internal consistency (a = .92) and a robust factor structure. Test-retest
values ranged from .62 to .71.

Personal Meaning Index of the Life Attitude Profile - Revised (LAP-R;
Reker, 1992).

In agreement with previous researchers, (Adams et al., 2000; Chapman,
1987; Hettler, 1984; Seaward, 1995; Whitmer 8 Sweeney, 1992), the spiritual
dimension was defined as a positive sense of meaning and purpose in life. The
conceptualization of spiritual well-being in this manner allows students to
respond from either a secular or religious belief system. The 48-item LAP-R was
designed to measure the multidimensional nature of attitudes toward life. The
Personal Meaning Index (PMI) of the Life Attitude Profile — Revised (Reker,
1992) provides a focused measure of personal meaning toward having life goals,
a mission or purpose, a sense of direction from past, present, and future, and
having a logically integrated and consistent understanding of self, others, and life
in general. The PMI is derived by summing the Purpose (PU) and Coherence
(CO) dimensions. “Purpose provides thrust and direction to one’s life. Implicit in

purpose is the notion of worthwhileness and what is of central importance in a

46

person’s life” (Reker, 1992; p. 14). The subscale contains 8 items scored from 1
(strongly disagree) to 7 (strongly agree). Total scores range from 8 to 56, with
higher scores indicating a greater life purpose. Sample items include “Basically, I
am living the kind of life I want to live,” and “In achieving life’s goals, I have felt
completely fulfilled." In the initial psychometric study of the Life Attitude Profile-
Revised, the internal consistency for the PU subscale was .86 (Reker, 1992),
which compares favorably to the internal consistency found by Adams, Bezner, 8
Drabbs (2000), a = .87.

The Coherence dimension is defined as a global orientation that
expresses the extent to which one has a pervasive, dynamic feeling of
confidence that one’s internal and external environments are predictable and that
there is a high probability that things will work out as well as can reasonably be
expected. “Implicit in coherence is a sense of order and reason for existence, a
clear sense of personal identity, and greater social consciousness” (Reker, 1992;
p. 15). The subscale contains 8 items scored from 1 (strongly disagree) to 7
(strongly agree. Total scores range from 8 to 56, with higher scores indicating a
greater life purpose. Sample items include “I have a philosophy of life that gives
my existence significance” and “I have a clear understanding of the ultimate
meaning of life.” In the initial psychometric study of the Life Attitude Profile-
Revised, the internal consistency for the CO subscale was .82 (Reker, 1992).

Perceived Wellness Survey (PWS; Adams 8 Bezner, 1997).

Perceived wellness is the sense that one is living in a manner that permits

the experience of consistent, balanced growth in the emotional, intellectual,

47

physical, psychological, social, and spiritual dimensions of human existence
(Adams 8 Bezner, 1997). Sample items from each dimension are, respectively,
"In general, I feel confident about my abilities," "In the past, I have generally
found intellectual challenges to be vital to my overall well-being,” ”I expect to
always be physically healthy," "In the past, I have expected the best," "My friends
will be there for me when I need help," and "I believe there is a real purpose for
my life.” Each dimension is represented by 6 items ranging from 1 (very strongly
disagree) to 6 (very strongly agree).

The dimensional scores are integrated into a wellness composite score by
combining the magnitude or mean of each dimension with the balance or
standard deviation among dimensions. Scores range from 3 to 29, with higher
scores indicating greater wellness. The PWS has been shown to possess very
good estimates of factorial and construct validity, as well as internal reliability (a =
.91; Adams 8 Bezner, 1997), which compares favorably to the internal
consistency found by Adams, Bezner, 8 Drabbs (2000), a = .91.

Marlowe-Crowne social desirability scale — M-C form C (Reynolds, 1982).

Measures of well-being and reports of parental relationship quality may be
confounded by social desirability (i.e., tendencies to answer questions in terms of
social appropriateness rather than true well-being), so a social desirability scale
was included in the data collection. The Marlowe-Crowne social desirability
scale — M-C form C is a 13 item short version of the original 33 item scale. The
M-C form C demonstrates an acceptable level of reliability (Kuder-Richardson

formula 20 = .76) and compares favorable with the reliability of the standard form

48

(KRzo = .82) and M-C form XX (KRzo = .79), although it has one-third less items
than the original form. Concurrent validity was examined via correlations
between the Marlowe-Crowne short form and the standard version and the
Edwards Social Desirability Scale. The product-moment correlation coefficient
with the standard version is .93 and with the Edwards scale it is .41. The
correlation with the Edwards scale is consistent with the correlation of .35 found
by Crowne and Marlowe (1960). Examples of items from the scale are “I am
always courteous, even to people who are disagreeable” and “I have never
deliberately said something that hurt someone’s feelings." The shorter form is
recommended as a viable short form for use in the assessment of social
desirability response tendencies. This form is brief and easy-to-administer.

Multigroup Ethnic Identity Measure (MEIM; Phinney, 1992).

Measures of well-being and reports of parental relationship quality may be
confounded by ethnic identity (i.e., attitudes and knowledge of who one is as a
member of an ethnic group), so an ethnic identity measure was included in the
data collection. Phinney developed this 12-item, 4-point (1 = strongly disagree, 4
= strongly agree) instrument to address, conceptually and methodologically,
ethnic identity as a general phenomenon across groups. The components
measured by the MEIM are ethnic affirmation and belonging (feelings of
attachment to one's group, ethnic pride, attitudes toward one’s group) and ethnic
identity achievement (understanding one’s ethnicity, commitment and secure
knowledge of who one is as a member of an ethnic group). The instrument also

assesses attitudes toward ethnic groups other than one’s own. Reliability for the

49

MEIM was established using two samples (the largest being high school
participants) (Phinney, 1992). Examples of items from the measure are “l have a
clear sense of my ethnic background and what it means for me" and “l have
spent time trying to find out more about my ethnic group, such as its history,
traditions, and customs.” Because the current study involves a university
sample, psychometric information for the college sample is reported. The
reliability coefficient for the overall measure was .90. Reliability coefficients for
the subscales were .86 for Affirrnation and Belonging, .80 for Ethnic Identity
Achievement, and .74 for Other-Group Orientation. The MEIM has subsequently
been used in dozens of studies and has consistently shown good reliability,
typically with alphas above .80 across a wide range of ethnic groups and ages.
Design

Due to limitations in randomly assigning subjects and in manipulating the
predictive variables, a correlational field study was the appropriate design to use
in this research endeavor. Determining the relationships among the variables
(e.g., the extent the variables are related) and causal modeling are of primary
concern in this study. Correlational field designs are known to have good
generalizability due to factors such as real life settings and naturally occurring
variables. However, due to the lack of control in the experiment, there are
potential threats to both internal and external validity. Conclusions drawn from
causal modeling with correlational data are valid and unbiased only if the
assumed model accurately represents the real causal processes (Tate, 1992). In

order to minimize the extraneous error variance, 1) a thorough review of the

50

. iterature was provided, 2) clear operational definitions of each variable was
established in accordance with previous research, 3) reliable and valid measures
were selected to evaluate each variable, 4) purposive sampling criteria were
defined, and 5) the data was analyzed utilizing techniques that best fit the
characteristics of the design and data.

08 (“a-Gathering Procedures

Participants were recruited via oral and written announcements in classes,

organizations, and residential living areas. The recruitment announcement
asked students to voluntarily take part in a research project exploring young

ad I...- lts’ perceptions of themselves and their parental relationships after the
divorce of their parents. Participants completed all questionnaires either

ind ividually or within a group setting. At the onset, the informed consent forms
were administered and discussed. Subjects were informed that participation was
volu ntary and they could withdraw from the study if desired. Volunteers were
assured that their privacy would be protected to the maximum extent permissible
und er the law. Consistent with the informed consent form, courses of action
were discussed to address negative consequences should they present. After
the consent forms were collected, each participant was given a packet of the
questi<>nnaires, which she or he took home to complete. Completed
questi<>nnaires were either returned to this investigator or to the classroom
instru ctor to be picked up by this investigator. Each measure within a packet was

COded so that the researcher knew which surveys were completed by the same

51

person, but no personally identifiable information was on any of the measures.
Packets were returned upon completion.

Each participant completed: Personal Data Questionnaire, Parental
Attachment Questionnaire (PAQ), Inter-Parental Conflict Scale (IPC), Current
Thoughts Scale (CTS), Personal Meaning Index of the Life Attitude Profile —
Revised (PMI), Perceived Wellness Survey (PWS), Marlowe-Crowne Social
desirability measure (M-C 3), and The Multigroup Ethnic Identity Measure
(MEIM). The order of testing for each participant was randomized to control for
order effects. Participants had as much time as needed to complete the forms,
but total time was approximately 30-40 minutes. The investigator was available
to answer questions. Persons who completed all forms had the opportunity to
have their names entered into a drawing for one of three $100 awards.
Participants who opted to enter the drawing (all except two) provided their email
addresses on the informed consent forms. Winners were drawn randomly from
the informed consent forms and notified via email. Checks were sent via USPS.
Copies of the UCHRIS approval, recruitment announcement, and consent form
are provided in Appendix A.

Data Analysis

Simple correlations and path analysis were appropriate to investigate this
study’s specific research questions and hypotheses. Multiple regression was
used in performing path analysis. All variables were either continuous or
dichotomous, so this type of analysis was the most powerful in determining the

causal interrelationships among the variables, (i.e., how each predictor was

52

relating to each other in predicting paternal attachment and well-being). The
analysis was completed separately for the two groups: Caucasians (N = 93) as
one group and African American (N = 58) and Hispanic Americans (N = 28) as
the second group. The data were assessed for significant findings within the
groups and between the groups.

All model variables were screened for missing data outliers and tested for
assumptions. Identification of outliers was done by Conducting a preliminary
regression to calculate Mahalanobis distance. The Explore procedure was
completed to determine if any cases exceeded the chi square criterion of 16.27
(df =3) for the first path analysis and 24.32 (df =7) for the second path analysis.
No outliers were found in either group. Creating scatterplots (see Figures B3, p.
131; C3, p. 140; D3, p. 152; and E3, p. 178) and residuals plots (refer to Figures
82, p. 130; C2, p. 139; DZ, p. 152; and E2, p. 177) assessed test assumptions.
The results suggested normality, linearity, and homoscedasticity of the data.
Correlation matrices were then created for all the model variables (refer to Tables
4 and 5, p. 64; 8 and 9, p.72) and multicollinearity was evaluated. The presence
of minimum and maximum relationship between independent and dependent
variables was checked. At least moderately strong relationships were found
between the full-scale independent variables and the dependent variables (r >
.30). None of the correlations between the full-scale independent variables was
greater than .70, so the assumption regarding multicollinearity was not violated.
All variables were retained. SPSS also performed collinearity diagnostics on the

variables as part of the multiple regression program and eliminated variables

53

 

accordingly. These analyses indicated that the assumptions of multiple
regression have not been violated.

No statistical procedure for evaluating the assumptions of path analysis
exists (for review of assumptions, refer to Mertler 8 Vannatta, 2003) since they
deal specifically with the degree to which the causal model has been correctly
specified. Conclusions drawn from causal modeling with correlational data,
according to Tate (1998), are confined to the following limitation: the results of
causal modeling are valid and unbiased only ifthe assumed model adequately
represents the real causal processes. However, there is no empirical test that
can evaluate the extent to which the correct model has been selected and
described. Tate (1998) suggests that the credibility, reasonableness, and utility
of a proposed model be used to address the assumptions regarding correct
model specification. In the present study, the extensive literature review, formal
theory, personal observations and experiences, and the empirical data all played
a substantial role in bringing about the causal specification of the models, and as
a whole lend credibility to the results.

Research question one was examined using simple correlations for each
group (i.e., African American/Hispanic group and Caucasian group; refer to Table
4 and 5, p. 64; 8 and 9, p. 72). The style of maternal attachment was correlabd
with the style of paternal attachment by looking at the relationship between full-
scale maternal attachment and full-scale paternal attachment, as well as the
relationships between each of the attachment subscales (i.e., feelings of affects,

parents’ promotions of independence, and parents as sources of support). The

54

degree of maternal conflict with father was then conelated with full-scale paternal
attachment and each subscale of paternal attachment. F ull-scale maternal
attachment and each subscale of maternal attachment were correlated with
adolescent well-being and lastly, full-scale paternal attachment and each
subscale of paternal attachment were correlated with adolescent well-being.
Adolescent well-being was analyzed using an aggregated score of well-being
compiled from the three measures of well-being. The well-being scores were
standardized so that the scores were expressed in standard deviations from the
mean. The standardized scores from the three measures of well-being were
then added to formulate an aggregated score of well-being. Using correlation
analysis, ethnic identity and social desirability were found to be covariates within
the African American and Hispanic group and social desirability co-varied with
the dependent variables within the Caucasian group (refer to tables 6 and 7, p.
67).

Regression analyses were conducted to establish the possibility of cause-
and-effect relationships among the set of variables that had been logically
ordered on the basis of literature and previous research. Using regression
analysis in this manner permitted an examination of whether or not the pattern of
intercorrelations among the variables fit my underlying theory of which variables
were affecting other variables. Specifically, stepwise regression analysis was
used to assess how matemal attachment style and maternal conflict with father
uniquely influence paternal attachment style and adolescent well-being.

Research questions 2, 3, and 4 each include multiple predictor variables and

55

were therefore examined using separate hierarchical MR analyses. The
covariates of ethnic identity and social desirability within the African American
and Hispanic group and social desirability within the Caucasian group were
controlled by putting them in the multiple regression first, as a block. Regression
analyses were conducted for the endogenous variables and the full-scale
exogenous variables and then with the subscale exogenous variables.

Path analyses established the causal ordering of the variables for the
African American and Hispanic group and the Caucasian group. Path
coefficients for the full-scale variables can be seen in the first two path diagrams
(Figure 1, p. 75 and Figure 2, p. 81). For the African American and Hispanic
group, coefficients were then used to calculate the reproduced correlations
through the path decompositions, which are displayed in Table B, p. 132.
Reproduced correlations were calculated and were compared to the empirical
correlations (refer to Table 11, p. 78). Only two reproduced correlations
exceeded the criterion of a .07 difference. Thus, it was concluded that the
revised model was consistent with the empirical data. The final step was to
calculate the direct, indirect, and total effects for each endogenous variable (refer
to Table 12, p. 78).

The path analysis for the Caucasian group resulted in only direct causal
effects between the exogenous and endogenous variables and no significant
relationship was found between paternal attachment and well-being (Figure 2
and Table 14, p. 81). Consequently, stepwise regression analyses were

conducted again using the attachment subscales instead of the full-scale

56

attachment variables (refer to Figure 3, p. 84; Figures D4, 05, and 06, p. 155).
Reproduced correlations were calculated through the path decompositions (refer
to Table D, p. 153)_and were compared to the empirical correlations (refer to
Table 16, p. 85). Only two reproduced correlations exceeded the criterion of a
.07 difference. Thus, it was concluded that the revised model was consistent
with the empirical data. The next step in this analysis was to calculate the direct,
indirect, and total effects for each endogenous variable (refer to Table 17, p. 86).
The following step was to construct a path diagram for the African American and
Hispanic group using the attachment subscales. Path coefficients for the
subscale attachment variables superimposed into one path can be seen in path
diagram 4 (p. 91) (refer to E4, E5, E6, pp. 183-184, for the individual paternal
attachment subscale paths). Reproduced correlations through the path
decompositions were calculated (refer to Table E, p. 179). Reproduced
correlations were then compared to the empirical correlations (refer to Table 19,
p. 90). The revised model was consistent with the empirical data. The final step
was to calculate the direct, indirect, and total effects for each endogenous
variable (refer to Table 20, p. 93).
Significance

Utilization of a modeling process by which covariables were first entered in
a stepwise fashion, followed by independent variables allowed for the relative
influence of exogenous and endogenous variables to be illuminated.
F urtherrnore, this process allowed for an examination of how these influences

Changed to examine how these influences changed when other theoretically

57

importar'
wastes“.I
and prov
outcome
variation
contribu-
variance
the othe
explalmr

variables

 

important variables were added to the equation. Significance of relationships
was tested using correlations, stepwise multiple regression, and path analysis
and provided an index of the relationship between the entire model and the
outcome variables. The models convey which variables account for the most
variation in the outcome variables, with the Beta weights indicating the relative
contribution of each variable to the model. Delta R2 indicates the amount of
variance in the outcome variable that the model accounts for when controlling for
the other variables and it indicates whether the interaction of the variables is
explaining any variance in the dependent variable above and beyond the

variables individually.

58

PIBIIITIIF.

I
D
I
”6416'
loTableI
Conata:
|

Table 1

m:
State 8

Spiral;
FLIS .
Psyche
hgflflepJ
" COTTE
' Corr.

study, ‘
ThoUgh.
Caplure
Spiritua;
and Pa
Pelcelw
mdudv

“lemme

 

I. As 0"

Results
Preliminary Analyses
Descriptive statistics are summarized in Tables A1, A2, A3, and A4, pp.
116-118. Cronbach’s coefficient alphas are high, ranging from .844 - .957 (refer

to Table A5, p. 118).

 

 

 

 

 

 

Correlation Analyses
Table 1 Correlation Coefficients of Adolescent Well-being Measures
Life State Perceived Spiritual
Attitude Self- Wellness Dimension
Profile Esteem PWS
Life Attitudes Survey 1.00
State Self-Esteem .625“ 1.00
Perceived Wellness .698" .698“ 1.00
Spiritual Dimension .774" .646“ .889“ 1.00
PWS
Psychological .692” .581 ** .850" .739**
Dimension PWS

 

 

 

 

 

 

 

** Correlation is significant at the 0.01 level (2-tailed)
* Correlation is significant at the 0.05 level (2 tailed)

Adolescent well-being is the endogenous variable of most interest in this
study. Three separate measures of well-being (Life Attitude Profile, Current
Thoughts Scale, and Perceived Wellness Survey) were included in an attempt to
capture a thorough estimate of the variable. Life Attitude Profile measured
spiritual well-being, Current Thoughts Scale measured psychological well-being,
and Perceived Wellness Survey provided a general index of well-being. The
Perceived Wellness Survey (PWS) addressed six dimensions of well-being,
including spiritual wellness and psychological wellness. The correlation
coefficients for the three measures and the two dimensions are provided in Table

1 . As can be seen in the table, internal consistency was found among all the

59

‘—I

Table 2 .

\.

 

 

 

 

 

 

 

 

Table 2 African American and Hispanic Group Correlation Coefficients of Well-

bein Scores and Independent Variables

 

 

 

 

 

 

 

 

 

 

 

 

 

Aggregated Psychological Spiritual General Well-

Well-being Well-being Well-being being Score

Score Score Score
MIAtt .644“ .575“ .611“ .679**
P/Att .581“ .584“ .575" .610“
MICON -.014 -.057 -.105 -.036
MIAFF .612" .612" .639" .702**
M/IND .416” .416" .487" .511**
MISUP .432" .432** .440" .506“
P/AF F .500“ .500“ .525" .567"
P/IND .633" .633" .566" .551"
P/SUP .361” .361 ** .378” .432”
ETH .488“ .549" .546" .542“
SocD -.543** -.515** -.442** -.501**

 

 

 

 

 

Maternal Attachment (MlAtt), Maternal conflict with Father (MICON), Paternal Attachment (P/Att),
Maternal promotion of independence (MIIND), Maternal feelings of affect (MIAF F), Maternal
feelings of support (MSUP), Paternal feelings of affect (P/AF F), Paternal promotion of
independence (P/IND), Paternal feelings of support (P/SUP), Ethnic Identity (ETH), Social

Desirability (SOCD)

*" Correlation is significant at the 0.01 level (2-tailed)
“ Correlation is significant at the 0.05 level (2 tailed)

Table 3 Caucasian Group Correlation Coefficients of Well-being
Scores and Independent Variables

 

 

 

 

 

 

 

 

 

 

 

 

 

Aggregated Psychological Spiritual General Well-
Well-being Well-being Well—being being Score
Score Score Score
MIAtt .399“ .280” .487” .443“
P/Att .350“ .436“ .308“ .505"
MICON -.377** -.421** -.266** -.405**
M/AF F .351“ .206“ .420** .387"
M/l ND .377“ .407** .434” .441“
MISUP .377** .209“ .485“ .404"
P/AFF .347" .421" .292" .483“
P/lND .276“ .351 ** .228* .448"
P/SUP .333“ .410** .343” .462"
ETH .372“ .250” .399” .384"
SocD -.074 -.016 -.121 -.157

 

 

 

 

 

Maternal Attachment (MlAtt), Maternal conflict with Father (MICON), Paternal Attachment (PIAtt),
Maternal promotion of independence (MIIND), Maternal feelings of affect (MIAFF), Maternal
feelings of support (MSUP), Paternal feelings of affect (P/AF F ), Paternal promotion of
independence (P/IND), Paternal feelings of support (P/SUP), Ethnic Identity (ET H), Social

Desirability (SocD)

'* Correlation is significant at the 0.01 level (2-tailad)
" Correlation is significant at the 0.05 level (2 tailed)

6O

 

 

m€BSUI

:58.p

estima‘
PVIS).
psvcbc
Include
Profile
referes
Percei
TOUT e:

and C:

measures, as well as between the two measures of psychological well-being (r
=.58, p< .01) and the two measures of spiritual well-being (r = .77, p < .01).

The well-being scores were standardized and aggregated to form an
estimate of spiritual well-being (Life Attitude Profile + spiritual dimension of
PWS), an estimate of psychological well-being (Current Thoughts Scale +
psychological dimension of PWS), and a total estimate of well-being, which
included all three full-scale measures (Current Thoughts Scale + Life Attitude
Profile + Perceived Wellness Survey). This total estimate of well-being is merely
referenced as “well-being” in the analyses that follow. The scores from the
Perceived Wellness Survey are referred to as general well-being. Each of the
four estimates of well-being were then correlated with the other study variables
and compared (refer to Tables 2 and 3).

Consistency was found between the four estimates of well-being and the
other variables in this study. If a significant relationship was found with one
estimate, it was also found with the other three, and conversely, if an insignificant
relationship was found with one estimate, an insignificant relationship was found
with the other three estimates. In addition, relationships were consistently either
negative or positive across the estimate. Next, the statistical significance of the
differences between the correlation coefficients was tested. The r values were
converted into standard scores (i.e., z scores) and were then plugged into the
following formula to obtain 20b, scores: 20b, = (21 - 22) /the square root of (1/N1 +
1IN2). Within the African American and Hispanic group, no statistically significant

differences were found between any of the well-being estimate correlations for

61

119$an

differer-I h
psyche F

5;; rrl..a

regress

represe

 

the same variable. Within the Caucasian group, one statistically significant
difference was found. This difference was between the correlate for
psychological well-being and mother as a source of support and the correlate for
spiritual well-being and mother feelings of support, thereby indicating that
maternal feelings of support explains significantly more of the variance in spiritual
well-being than in psychological well-being.

Given that only 1 out of the 132 differences between correlations was
significant, the researcher concluded that it would be redundant to run multiple
regression with each of the four measures. The aggregated (total) score of well-
being appears to be an inclusive and meaningful measure of well-being and
represents adolescent well-being in the statistical analyses that follow. These
findings indicate that secure attachments to parents are associated with higher
levels of general, spiritual, and psychological well-being, while insecure
attachments to parents are associated with lower levels of general, spiritual, and
psychological well-being, thus supporting hypothesis two.

Pearson product-moment correlations among full-scale maternal
attachment (MlAtt), full-scale paternal attachment (PIAtt), maternal conflict with
father (MICON) and well-being (WB) are shown in Tables 4 and 5 (p. 64).
Consistent with prior research (e.g., Carranza 8 Kilmann, 2000; Hojat, 1998;
McCormick 8 Kennedy, 2000; Meyers, 1998), maternal and paternal attachment
were found to have a strong positive correlation with well-being within the African
American and Hispanic group (r = 0.64, p < .01 and r = 0.58, p < .05,

respectively) and a moderately strong positive correlation within the Caucasian

62

grour
with :
of se

posit

mall
rest
corr
1686

R05

deg

group (r = .40, p < .01 and r = .35, p < .01, respectively). Therefore, students
with a more secure attachment style with either parent depicted a greater degree
of sense of well-being. As hypothesized, paternal style of attachment was
positively correlated with maternal style of attachment within both groups (r =
0.34, p < .01 and r = .23, p < .05, respectively) and negatively correlated with
maternal conflict with father (r = -0.23, p < .05 and r = -0.31, p < .01,
respectively). However, maternal style of attachment was not significantly
correlated with maternal conflict with father, a finding that differs from prior
research (e.g., Amato, 1993; Emery 8 Forehand, 1994; Lewis, Fairing, 8
Rosenthal, 2000; Summers et.al., 1998; Whiteside 8 Becker, 2000).

In summary, within both groups, there is a significant relationship between
a) the style of maternal attachment and the style of paternal attachment, b) the
degree of maternal conflict with father and the style of paternal attachment, c) the
style of maternal attachment and adolescent well-being, and d) the style of
paternal attachment and adolescent well-being. These findings support
hypothesis one for both groups, that is, secure attachment to mother and low
levels of maternal conflict with father correspond with secure attachment to
father. Conversely, insecure attachment to mother and high levels of maternal

conflict with father correspond with insecure attachment to father.

63

 

 

 

 

Sir:
Wit
bet

Arr

TE};

re:
be

an

Table 4 African American and Hispanic Group Correlations of Main Variables

 

 

 

 

 

MIAtt MICON P.Att WB
MIAtt 1 .00
MICON -.187 1.00
P/Att .336” -.228* 1 .00
WB .644“ -.014 .581* 1.00

 

 

 

 

 

 

 

Maternal Attachment (MlAtt), Maternal conflict with Father (MICON),
Paternal Attachment (PIAtt), Well-being (WB)
** Correlation is significant at the 0.01 level (2-tailed)

* Correlation is significant at the 0.05 level (2 tailed)

Table 5 Caucasian Group Correlations of Main Variables

 

 

 

 

 

MIAtt MICON P.Att WB
MIAtt 1 .00
MICON -.040 1 .00
PlAtt .232" -.314** 1 .00
WB .399“ -.377** .350" 1 .00

 

 

 

 

 

 

 

Maternal Attachment (MlAtt), Maternal conflict with Father (MICON),
Paternal Attachment (PIAtt), Well-being (WB)
““ Correlation is significant at the 0.01 level (2-tailed)

“ Correlation is significant at the 0.05 level (2 tailed)

Hypothesis two was also supported. Secure parental attachments had
strong positive correlations with general, spiritual and psychological well-being.
Within the Caucasian group, a moderately strong positive relation was found
between maternal conflict with father and well-being, but within the African
American and Hispanic group, the data indicated there was not a significant
relationship. Thus, secure attachment to mother corresponds with secure
attachment to father, yet interparental discord does not appear to influence the
well-being of the child. While research supports negative impact of interparental
conflict on adolescent well-being, it is important to remember that the majority of
research has been done using predominantly Caucasian samples. A different
pattern may exist with Hispanic and African American adolescents than is found

among Caucasian students, contrary to hypothesis 3. Given these results and

the fr
relate
being

deta

Inter
for s
lnvo
and
varrr
pate
are

for l

tabl

the finding that maternal attachment and maternal conflict with father were
related to paternal attachment, but not to each other, nor consistently to well-
being (even though paternal attachment was related to well-being), a more
detailed analysis of the data was warranted.

In order to control unwanted variance that may be affecting or biasing the
interpretations of the data, Pearson product-moment correlations were generated
for several factors that have been identified as covariants in previous studies
involving parental attachment and divorce. These factors are listed in Table 6
and 7, p. 67. The correlations between these factors and the dependent
variables of this study (e.g., parental attachment, parental feelings of affect,
paternal promotion of independence, father feelings of support, and well-being),
are found in Table 6 for the African American and Hispanic group and in Table 7
for the Caucasian group. Age at time of separation was not included in the
tables because it was not significantly different than age at time of divorce
(Cohen’s D = .016, effect size = .008) and 57 students failed to provide an age
(37% of the African American and Hispanic group and 27% of the Caucasian
group). In examining Table 6, two variables stand out as significant covariates,
specifically, ethnic identity and social desirability. Both ethnic identity and social
desirability had a strong relationship with well-being (r = 0.49 and —0.54,
respectively, p < .01) and were treated as covariates in further analysis of the
data. Ethnic identity also correlated with full-scale paternal attachment (r = 0.33,
p < .05) and the three paternal attachment subscales (r = 0.33, 0.26, 0.22, p <

.05, respectively), with the strength of the relationships ranging from small to

65

rnedr

<.OE

depe
appe
aux:
cone
they
nfiat
rnot
02D

025

Gee

medium. Social desirability correlated with paternal feelings of affect (r = -.023, p
< .05) and paternal promotion of independence (r = -0.24, p < .05).

As is evident in Table 6, several other factors co-varied with the
dependent variables of this study. While most of these significant relationships
appear commonsensical (i.e., as students grow older they experience more
autonomy-promoting behaviors from their fathers), the majority have small
correlations (e.g., r = .10 - .29, Cohen, 1988), suggesting weak relationships and
they were not included in the regression analysis as covariates. Even though the
relationships were week, the significant positive relationships between 1)
mother’s education level and the paternal attachment subscales (r = 0.25, 0.31,
0.27; p < .05) and 2) adolescent well-being and mother’s education level (r =
0.25, p < .05), father’s education level (r = 0.29, p < .01), and the student’s
education level (r = 0.22, p < .05) are quite interesting and were addressed in the

discussion section of this paper.

66

Table 6 African American and Hispanic Group Pearson Correlation Coefficients
of Endogenous Variables and Possible Covariates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

P/Att P/AFF P/IND PISUP WB
_Age .142 .091 .267* .072 .267*
Education Level .083 .052 .167 .036 .217*
_Age at Divorce 253* 261* .141 .276" .159
Lengm parents Committed Rel. -.025 .062 -.O77 -.058 .007
No. of Siblings with same parents .078 .146 .031 .024 .073
Freq maternal contact currently -.172 -.093 -.258* -.069 -.400**
Frequency maternal contact < 18 .071 .077 .131 .009 .017
Freq paternal contact currently -.461** -.481** -.204 -.454** -.127
Frequency paternal contact < 18 -.281** -.341** .142 -.467** .228*
Mother’s education level 247* 313* .003 275* 250*
Father’s education level .017 .118 -.028 .045 .288“
Ethnic Identity .327" .330“ 256* 216* .488"
Social Desirability -.182 -.233* -.243* .022 -.543**

 

 

 

Paternal Attachment (PIAtt), Paternal feelings of affect (P/AF F ), Paternal promotion

of independence (PlIND), Father feelings of support ((PISUP), Total Wellobeing measure (WB)

*" Correlation is significant at the 0.01 level (2-tailed)
* Correlation is significant at the 0.06 level (2 tailed)

Table 7 Caucasian Group Pearson Correlation Coefficients
of Endgqenous Variables and Possible Covariates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

P/Att P/AF F P/IND PISUP WB
Age .161 .146 .122 .166 .170
Education Level .027 .004 .053 .026 .127
Age at Divorce -.017 -.052 -.149 .135 -.207*
Length parents Committed Rel. -.002 -.043 -.113 .136 -.199
No. of Siblings with same parents -.069 -.078 -.252* .105 -.163
Freq maternal contact currently .005 .001 -.022 .025 -.241*
Frequency maternal contact < 18 .104 .101 .006 .150 -.021
Freq paternal contact currently -.503** -.478** -.171 -.611** -.257*
Frequency paternal contact < 18 -.497** -.409** -.251*. -.663** -.011
Mother’s education level .100 .136 .071 .030 230*
Father’s education level .105 .117 -.057 .176 -.005
Ethnic ldentig .145 .131 .053 .203 .372"
Social Desirability -.304** -.326** -.178 -.297** -.074

 

 

 

 

 

 

 

Paternal Attachment (PIAtt), Paternal feelings of affect (P/AF F), Paternal promotion
of independence (P/IND), Father feelings of support ((PISUP), Total Well-being measure

(WB)

** Correlation is significant at the 0.01 level (2-tailed)
* Correlation is significant at the 0.06 level (2 tailed)

67

frequr
morit
Liker
betwi
there
more
are r
1981
visit
the

mor
atta

evlc

onl
aft;
Of :
thr
wit

ail

ml

It is important to note that the possible choices to questions about
frequency of contact with parent (e.g., mother or father) included daily, weekly,
monthly, couple times a year, and not at all. The responses were scored on a
Likert scale ranging from 1 to 5, with 1 being daily. The negative correlations
between frequency of contact and attachment subscales and well-being are
therefore interpreted as higher levels of contact with parent are associated with
more secure attachment styles and/or more positive well-being. These findings
are consistent with prior research (Hetherington, et al., 1982; Wallerstein 8 Kelly,
1980). Several of these correlations involving frequency of contact with parent
indicate a moderately strong relationship (r = .30 - .49, Cohen, 1988), but given
the nominal nature of the data and the logical structure of the relationships (i.e.,
more contact, more secure attachment style, or vice versa, more secure
attachment style, more contact), these correlations served as supportive
evidence to other findings, but were not included in the regression analysis.

In examining the possible covariates with the Caucasian group (Table 7),
only social desirability stood out as a significant covariate with paternal
attachment (r = -0.30, p < .01) and the paternal attachment subscales of feelings
of affect (r = -0.33, p < .01) and father feelings of support (r = -0.30, p < .01). All
three were moderately strong relationships. Ethnic identity neither correlated
with full-scale paternal attachment, nor any of the subscales of paternal
attachment, yet it did have a moderately strong correlation with well-being and
was thus evaluated for covariance in the regression analysis. Using stepwise

multiple regression, ethnic identity did not account for a significant amount of the

68

vanar

vafial

exha
sgni
weal
cont
anal
(Zau
dep
vwel

lNhl

Io

variance in well-being within the Caucasian group when the other independent
variables were controlled.

The Caucasian group did not significantly correlate with as many
extraneous variables as the African American and Hispanic group. The
significant relationships (refer to Table 7, p. 67) not already mentioned, were
weak (r = .10 -.29) or were between the dependent variables and frequency of
contact with parent. Again, these variables were not included in the regression
analysis for the reasons stated previously. It is interesting to note that within the
Caucasian group, the only significant correlation between education level and the
dependent variables was found between mother’s education and adolescent
well-being (r = 0.23, p < .05), unlike the African American and Hispanic group in
which several significant relationships were related to education level.

To summarize the covariate data, ethnic identity and social desirability
were significant covariates within the African American and Hispanic group and
needed to be entered first when doing regression analysis. Within the Caucasian
group, social desirability significantly correlated with the dependent variables and
was thus entered first when doing multiple regression analysis. In addition,
correlates of frequency of parental contact with the dependent variables
promoted the idea that more secure attachment styles and more contact with
parents go hand in hand.

Table 8 and Table 9 (p. 72) present the Pearson product-moment
correlations among maternal conflict with father, well-being, and the subscales

for maternal and paternal attachment. Within the Caucasian group, internal

69

consistency was found among the subscales for paternal attachment, (r = 0.75,
0.82, 0.53, p < .01), and among the subscales for maternal attachment (r = 0.64,
0.89, 0.52, p < .01). Within the African American and Hispanic group, internal
consistency was found between paternal feelings of affect and both paternal
promotion of independence (r = 0.54, p < .01) and father feelings of support (r =
0.88, p < .01), but only a moderate correlation was found between father feelings
of support and paternal promotion of independence (r = 0.38, p < .01). A similar
pattern was found among the maternal subscales in this group (r = 0.60, 0.84,
and 0.30, respectively, p < .01). This pattern was evaluated later in light of
additional analysis.

Examining the correspondence between secure attachment to mother with
secure attachment to father within the African American and Hispanic group, six
of the nine correlations among the subscales were significant. Paternal
promotion of independence positively correlated with maternal feelings of affect,
maternal promotion of independence, and maternal feelings of support (r = .43, p
< .01; r = 0.40, p < .01; and r = 0.23, p < .05 respectively). Paternal feelings of
affect positively correlated with maternal feelings of affect and maternal
promotion of independence (r = 0.25 and 0.27, p < .01, respectively). Paternal
feelings of support only correlated with maternal promotion of independence (r =
.30, p < .01). Note that the strength of the correlation is similar to that found
between paternal feelings of support and paternal promotion of independence (r

= .38, p < .01) and a similar finding was found with the maternal subscales, that

70

is, maternal melings of support only correlated with the paternal promotion of
independence within the paternal attachment subscales (r = .23, p < .01).
Examining the variable maternal conflict with father, a significant negative
relationship was found with maternal feelings of support (r = -0.24, p < .05),
paternal feelings of affect (r = -0.29, p < .01), and paternal feelings of support (r =
-0.26, p < .05). All six attachment subscales were positively related to
adolescent well-being, four strong (r = 0.67, 0.51, 0.51, 0.61, p < .01) and two
moderately strong (r = 0.48 and 0.38, p < .01). Despite these correlations, no
relationship was found between maternal conflict with father and well-being and
between maternal feelings of support and paternal feelings of support.
Examining the correspondence between secure attachment to mother with
secure attachment to father within the Caucasian group, four of the nine
correlations among the subscales were significant. Paternal promotion of
independence positively correlated with maternal feelings of affect and maternal
promotion of independence (r = 0.32 and 0.32 respectively, p < .01). Paternal
feelings of affect correlated weakly with only maternal feelings of affect (r = 0.21,
p < .05). Paternal feelings of support correlated with only maternal promotion of
independence (r = .32, p < .01) (as was the case in the African American and
Hispanic group). Maternal feelings of affect did not correlate with any of the
paternal attachment subscales. Examining the variable maternal conflict with
father, a significant negative relationship was found with all three of the paternal

attachment subscales (r = -0.24, p < .05, r = -0.38 and —0.30, p< .01).

71

Table 8 Pearson Correlation Coefficients African American and Hispanic Group

Attachment Subscales

Well-bein , and Maternal conflict with Father

 

 

 

 

 

 

 

 

 

 

MIAFF MIIND MISUP PIAFF PIIND PISUP WB
M/AF F 1.00
MIIND .598“ 1.00
MISUP .843" .304“ 1 .00
P/AF F .250* .271 * .142 1.00
PIIND .425” .404" .231 * .542" 1.00
PISUP .108 .299“ .032 .878“ .381 ** 1 .00
WB .672” .506“ .476“ .508“ .612** .378“ 1 .00
MICON -.088 -.172 -.235* -.293** -.004 -.255* -.014

 

 

 

 

 

 

 

 

 

Maternal promotion of independence (MIIND), Maternal feelings of affect (M/AF F), Maternal

feelings of support (MSUP), Paternal feelings of affect (PIAF F), Paternal promotion of

independence (PIIND), Paternal feelings of support (PISUP), Well-being (WB), Ethnic Identity
(ETH), Social Desirability (SocD)
“ Correlation is significant at the 0.01 level (2-tailed)

' Correlation is significant at the 0.05 level (2 tailed)

Table 9 Pearson Correlation Coefficients Caucasian Group

Attachment Subscales, Well-being and Maternal conflict with Father

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

M/AFF MIIND MISUP PIAFF PIIND PISUP WB
M/AF F 1.00
MIIND .640" 1.00
MISUP .885" .522" 1 .00
P/AFF .211* .181 .107 1.00
P/IND .305" .317" .146 .751" 1.00
PISUP .180 .317" .146 .824“ .533" 1.00
WB .351“ .377" .377" .347" .276“ .333“ 1.00
MICON -.022 -.195 .069 -.238* -.377** -.295** -.377**

 

Maternal promotion of independence (MIIND), Maternal feelings of affect ( M/AF F), Maternal

feelings of support (MSUP), Paternal feelings of affect (PIAF F), Paternal promotion of

independence (PIIND), Paternal feelings of support (PISUP), Well-being (WB), Ethnic Identity
(ET H), Social Desirability (SocD)
** Correlation is significant at the 0.01 level (2-tailed)

” Correlation is significant at the 0.05 level (2 tailed)

72

 

Unlike the African American and Hispanic group, within the Caucasian
group, a moderately strong negative relationship was found between maternal
conflict with father and adolescent well-being (r = -0.38, p < .01). All six
attachment subscales had moderately strong positive relationships with
adolescent well-being (r = 0.35, 0.38, 0.38, 0.35, 0.28, 0.34, p < .01), somewhat
weaker relationships than found in the African American and Hispanic group. No
relationship was found between maternal conflict with father and the maternal
attachment subscales, thereby indicating that interparental conflict, even when
perceived to be instigated by the mother, does not appear to influence 1)
maternal attachment style nor 2) secure maternal attachment impact on
Caucasian adolescent well-being.

To summarize Tables 8 and 9, within the African American and Hispanic
group, there was a significant relationship between a) the degree of maternal
conflict with father and paternal feelings of affect and support, b) the subscales of
maternal attachment and adolescents well-being, and c) the subscales of
paternal attachment and adolescents well-being. Six of the nine correlations
between maternal attachment subscales and paternal attachment subscales
were significant. These findings support hypothesis one, that is, secure
attachment to mother and low levels of maternal conflict with father correspond
with secure attachment to father. Conversely, insecure attachment to mother
and high levels of maternal conflict with father correspond with insecure
attachment to father. These findings suggest more specificity is warranted in

defining the relationship between maternal attachment, paternal attachment, and

73

maternal conflict with father. The paternal attachment subscales appear to be
affected differently by the different maternal attachment subscales and by
maternal conflict with father.

These finding also support hypothesis two, namely, secure attachments to
parents are associated with higher levels of adolescent well-being and insecure
attachments to parents are associated with lower levels of adolescent well-being.
The findings suggest that the strength of some subscales of attachment is more
or less than that of other subscales in impacting adolescent well-being.

Within the Caucasian group, there is a significant relationship between a)
the degree of maternal conflict with father and paternal feelings of affect and
support, b) the subscales of maternal attachment and adolescent well-being, and
c) the subscales of paternal attachment and adolescent well-being. Four of the
nine correlations between maternal attachment subscales and paternal
attachment subscales are significant. These findings support hypothesis 1, but
suggest specific subscales of maternal attachment and maternal conflict affect
specific subscales of paternal attachment. In addition, the differences between
the two groups illustrated by these correlations suggest the hypothesis (3.a) that
Hispanic and African American adolescents’ parental attachment patterns are the
same as Caucasian adolescents’ parental attachment patterns may be
inaccurate. On the other hand, these differences do lend support to the
hypothesis (3.b) that Hispanic and African American adolescents possess more
adaptive coping strategies in relation to maternal conflict with father than do

Caucasian adolescents. Definitive conclusions require further analyses.

74

Path Analysis

Table 10 African American and Hispanic Group Pearson Correlation Coefficients
Endogenous and Exogenous Variables Used in Path 1

 

 

 

 

 

 

 

 

MIAtt M/Con P/Att WB ETH
M/Att 1 .00
MICON -.187 1.00
P/At’t .336* -.228* 1 .00
WB .644" -.014 .581“ 1.00
ETH .054 .147 .327" .488" 1 .00
$060 -.435** .243* -.182 -.543** -.053

 

 

 

 

 

 

 

Maternal Attachment (MlAtt), Maternal conflict with Father (MICON), Paternal
Attachment (PIAtt), Well-being (WB), Ethnic Identity (ETH), Social Desirability

(SocD).

** Correlation is significant at the 0.01 level (2-tailed)
* Correlation is significant at the 0.05 level (2 tailed)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3 Social
Desirability
-.435"
l Maternal
Attachment w
.336"
2 Maternal
conflict 4 Paternal
with Father @ Attachment
Figure 1

Analysis 1

75

 

 

 

 

F Well-
being

 

 

 

5 Ethnic
Identity

 

   

 

 

Input Path Diagram 1: African American] Hispanic Group Path

Path analysis was completed alter identifying the covariates and exploring
the correlates between the independent and dependent variables. Straight,
single-headed arrows, representing unidirectional paths, connect the boxes.
These straight arrows originate at the variable exerting the influence and point
towards the variable being affected. Curved, two headed arrows represent
correlations found between exogenous variables. The standardized path
coefficients (b) represent the effect of a given predictor variable on the
dependent variable after accounting for the remaining relationships in the model.
The variables and Pearson product-moment correlations for this analysis are
found in Table 10. The covariates of ethnic identity and social desirability were
included in the model.

Figure 1 illustrates the first path analysis for the African American and
Hispanic group. Path decomposition was used to examine goodness-of-fit. This
process involved the identification of all legitimate paths between the variables in
the model, resulting in a correlation coefficient equal to the product of all
coefficients in that path. Correlational decompositions were determined for all
possible bivariate correlations in the models, with the exception of those between
exogenous variables. The decompositions and calculations of reproduced
correlations for this path analysis can be found in Table 81 (p. 132). The
reproduced correlations are displayed adjacent to the observed correlations in
Table 11 (p. 78). Calculation of reproduced correlations through path
decompositions and subsequent comparison to the empirical correlations

indicated the model fits the empirical data.

76

For the African American and Hispanic group, a path analysis was
conducted to determine the causal effects among the variables maternal
attachment (MlAtt), maternal conflict with father (MICON), paternal attachment,
(PIAtt), and adolescent well-being (WB). Prior to the analysis, two covariates
were identified (Table 6, p. 67) and added to the analysis, specifically, ethnic
identity (ETH) and social desirability (SocD). Stepwise multiple regression was
used to establish the causal effect relationships (refer to Figure 81, p. 120). The
model was also tested without the inclusion of the ethnic identity and social
desirability and the results did not fit the model (e.g., MICON was positively
correlated with WB and MICON was not correlated with PlAtt). All path
coefficients were significant at or below the .05 level. Utilizing calculations from
Table B (p. 132), the direct, indirect and total causal effects of the model are
presented in Table 12, p. 78. R2 is noted for each endogenous variable. The
outcome of primary interest was adolescent well-being; the determinant with the
largest total causal effect was maternal attachment (.693). The remaining
determinants of adolescent well-being as indicated by the total causal effect were
paternal attachment (.523), social desirability (-.331), ethnic identity (.320), and
maternal conflict with father (-.096). This model explained approximately 78% of
variance in adolescent well-being. The primary determinant of paternal
attachment was maternal attachment (.336) with maternal conflict with father (-
.228) following. Approximately 26% of the variance in paternal attachment was
explained by the model. The primary determinant of ethnic identity was paternal

attachment (.327), followed by maternal attachment (.110) and maternal conflict

77

Table 11 African American and Hispanic Group Path 1 Observed and
Reproduced Correlations (Table 31)

 

 

 

 

 

 

M/A'I'l" MICON P/ATT WB ETH
Observed Correlations

MIATT 1.00

MICON -.187 1 .00

PIATI' .336 -.228 1 .00

WB .644 -.014 .581 1.00

ETH .054 .147 .327 .488 1 .00

SocD -.435 .243 -.182 -.543 -.053
Reproduced Correlations

MIATI' 1.00

MICON - 1.00

P/A'l‘l' .336 -.228 1 .00

WB .693 .096* .608 1 .00

ETH .109 -.075 .327 .424 1.00

SocD -.435 - -.146 -.393* -.048

 

*Difference >.07

Table 12 Summery for Causal Effects for African American and Hispanic Group

 

 

Path 1
flcome Determinant Direct Indirect TLal
PlAtt MIAtt 336* -- .336
(R2 = .257) MICON -.228* —— -.228
we M/Att .407" .286 .693
(R2 = .780) MICON --- -.096 -.096
PlAtt .318“ .105 .523
ETH .320" —— .320
$000 -.331** -- -.331
ETH PlAtt .327** —— .327
(R2 = .107) MIAtt -— .110 .110
MICON —- -.075 -.075
SocD MIAtt -.435** ~— -.435
(R2 = .189)

 

** Direct effect is significant at the .01 level.
* Direct effect is significant at the .05 level.

78

with father (-.O75). Approximately 11% of variance in ethnic identity was
explained by the model. The primary determinant of social desirability was
maternal attachment (—.435), which explained approximately 19% of the variance
in social desirability.

To summarize, among African American and Hispanic students, secure
maternal attachment and low levels of maternal conflict with father had a direct
causal influence on secure paternal attachment. This means that when the style
of maternal attachment was controlled, the degree of maternal conflict with father
uniquely predicted the style of paternal attachment. In addition, through its
impact on paternal attachment, maternal attachment had an indirect impact on
adolescent well-being, in addition to its direct influence. Maternal attachment did
not correlate with maternal conflict with father, which means the style of maternal
attachment did not moderate the relationship between the degree of maternal
conflict with father and the style of paternal attachment. However, both maternal
attachment style and maternal conflict with father uniquely predicted the
relationship between the style of paternal attachment and adolescent well-being
(i.e., in addition to its direct influence, paternal attachment had an indirect
influence on well-being stemming from the impact of maternal attachment and
maternal conflict with father). This path analysis also indicated that students with
secure maternal attachment tend to have more of an attitude of openness and
acceptance of shortcomings (i.e., maternal attachments’ impact on social

desirability) than students with insecure maternal attachment. Paternal

79

all;

he

81

attachment has a moderately strong influence on ethnic identity, which in turn,
has a moderately strong influence on adolescent well-being.

The first path analysis for the Caucasian group is presented in Figure 2, p.
81. Straight, single-headed arrows, representing unidirectional paths, connect
the variables. These straight arrows originate at the variable exerting the
influence and point towards the variable being affected. As illustrated, this path
diagram consists of only bivariate connections, that is, a variable is either
endogenous or exogenous, and so no path consists of more than two variables.
There are also no correlations found between the exogenous variables (i.e.,
maternal attachment, maternal conflict with father, and social desirability), thus
there are no curved, two headed arrows present. The standardized path
coefficients (b) represent the effect of a given predictor variable on the
dependent variable after accounting for the remaining relationships in the model.
The variables and Pearson product-moment correlations for this analysis are
found in Table 13, p. 81. The covariate social desirability was identified

previously and was included in the model.

80

Table 13 Caucasian Group Pearson Correlation Coefficients
Endogenous and Exogenous Variables Used in Path 1

 

 

 

 

 

 

 

MIAtt M/Con PlAtt WB
MIAtt 1 .00
MICON -.O40 1 .00
PlAtt 232* -.314** 1 .00
WB .399“ -.377** .350" 1 .00
8060 -.037 -.O74 -.304** -.074

 

 

 

 

 

 

Maternal Attachment (MIATT), Maternal conflict with Father (MICON), Paternal

Attachment (PlAtt), Well-being (WB), Ethnic Identity (ET H), Social Desirability (SocD).

** Correlation is significant at the 0.01 level (2-tailed)
‘ Correlation is significant at the 0.05 level (2 tailed)

 

 

5
Paternal

Attachment

 

 

0

 

l
Maternal
Attachment

 

 

 

 

2
Maternal
Conflict

 

 

’

Well-
being

 

 

 

3
Social
Desirability

 

 

 

Figure 2 Input Path Diagram 2: Caucasian Group Path Analysis 1

Table 14 Caucasian Group Path 1 Observed Correlations and Standardized

 

 

 

 

 

 

 

 

 

Path Coefficients

M/Att M/CON SocD
Observed r
PlAtt 0.232 -0.314 -0.304
WB 0.399 -0.377
Standardized b
PlAtt 0.231 -0.338 -0.329
WB 0.385 -0.362 0.000

 

 

 

 

 

Maternal Attachment (MlAtt), Paternal Attachment (PlAtt), Maternal conflict with father (MICON),
Social Desirability (SocD), Well-being (WB).
‘Indicates differences > .05.

81

 

Figure 2 illustrates the first path analysis for the Caucasian group. Path
decomposition is typically used to examine goodness—of-fit. This process
involves the identification of all legitimate paths between the variables in the
model, resulting in a correlation coefficient equal to the product of all coefficients
in the path. However, in this path analysis, all legitimate paths are only one-way,
bivariate paths. Each variable is either endogenous or exogenous and there are
no significant relationships between the exogenous variables. Therefore, the
standardized path coefficient (b) equates to the product of all coefficients in that
path. This means the standardized path coefficients were compared to the
observed correlations (r) and are displayed in Table 14. Comparison of the
coefficients in the table indicated no differences between observed correlation
and standardized path correlation exceeded the .05 level, except for paternal
attachment and adolescent well-being. The correlational analysis indicated a
moderately strong correlation (r = .35, 0< .01) between paternal attachment and
adolescent well-being. This relationship was fundamental to the current study
and was lost during the stepwise multiple regression (Figure C1, p. 134) once
maternal attachment and maternal conflict with father were controlled.
Therefore, this model did not fit the empirical data and further analysis was
warranted.

The correlational analysis of the attachment subscales (refer to Tables 8
and 9, p. 72) indicated that there were significant differences in how the
subscales related to each other and to the other variable of this study (e.g.,

maternal conflict with father, adolescent well-being, ethnic identity, and social

82

desirability). These findings prompted the path analysis to be revised so that the
attachment subscales were used in place of the full-scale measures. This
analysis was completed for both groups. The variables and Pearson product-
moment correlations for the Caucasian group analysis are found in Table 15, p.
84. The covariate of social desirability was included in the model, along with the
main exogenous and endogenous variables.

Figure 3 is the second path analysis for the Caucasian group. It illustrates
the superimposed path analyses of the three paternal subscales. The separate
paths for each of the paternal subscales are found in Figures D4, D5, and D6, p.
155. Path decomposition was used to examine goodness-of-fit. All legitimate
paths between the variables in the model were identified and resulted in a
correlation coefficient for each path, which equals the product of all coefficients in
the path. Correlational decompositions were determined for all possible bivariate
correlations in the models, with the exception of those between exogenous
variables. The decompositions and calculations of reproduced correlations for
this path analysis can be found in Table D, p. 153. The reproduced correlations
are displayed adjacent to the observed correlations in Table 16, p. 85.
Comparison of reproduced correlations and empirical correlations indicated the

model fit the empirical data.

83

Table 15 Pearson Correlation Coefficients Caucasian Group Endogenous and
Exogenous Variables Used in Path 2

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MIAFF MIIND MISUP P/AF F PIIND PISUP WB ETH SocD
M/AF F 1 .00
MIIND .640" 1 .00
MISUP .885“ .522“ 1 .00
PIAFF .211* .181 .107 1.00
PIIND .305” .317" .146 .751” 1.00
PISUP .180 .31 7** . 146 .824" .533" 1 .00
WB .351” .377“ .377“ .347” .276“ .333“ 1 .00
ETH .384“ .252" .51 2" .1 31 .053 .203 .372” 1 .00
SocD -.002 .054 .064 -.326** -.178 -.297** -.074 -.068 1.00
MICON -.022 -.195 .069 -.238* -.377** -.295** -.377** -.096 -.074

 

Maternal promotion of independence (MIIND), Maternal feelings of affect (MIAFF), Maternal
feelings of support (MSUP), Paternal feelings of affect (PIAFF), Paternal promotion of
independence (PIIND), Paternal feelings of support (PISUP), Well-being (WB), Ethnic Identity

(ET H), Social Desirability (8000), Maternal conflict with father (MICON)

** Correlation is significant at the 0.01 level (2-tailed)
* Conelation is significant at the 0.05 level (2 tailed)

 

 

 

 

MISUP

 

 

 

 

M/AF F

 

 

 

MICON

.685“

.205“

 

 

 

 

 

Soc/D

 

 

 

 

 

 

-.345"

-.437"

Q

.259“

Q

 

 

 

 

-.332*

 

 

 

 

-.265*

.283‘

 

 

 

 

PIIND
7
.377“
P/AFF
6 Well-
being
9
-.350"
PISUP
8
125 = .885
r23 = .640
r35 = .522

 

 

 

 

 

 

 

 

Figure 3 Input Path Diagram 3: Caucasian Group Path Analysis 2

84

 

Table 16 Caucasian Group Path 2 Observed and Reproduced Correlations

MIAFF M/er MISUP MICON

Observed Correlations

Pimp P/gspp PIAFF W48,

 

 

 

MIAFF 1.00

MIIND .640 1.00

MISUP .885 .522 1 .00

MICON -.022 -.195 .069 1.00

PIIND .305 .317 .146 -.377 1.00

PISUP .180 .317 .146 -.295 .533 1.00

PIAFF .211 .181 .107 -.238 .751 .824 1.00

WB .351 .377 .377 -.377 .276 .333 .347

SocD -.002 .054 .064 -.074 -.178 -.297 -.326 -.074
Reproduced Correlations

MIAFF 1.00

MIIND — 1.00

MISUP - - 1.00

MICON - - - 1.00

PIIND .298 .211* .169 -.332 1.00

PISUP .181 .283 .148 -.265 - 1.00

P/AF F .205 .131 .181 -.259 -- - 1.00

WB .380 .226* .378 -.408 .230 .280 .339

SocD - -— - - - -.33; -.345 -.077

 

*Difference >07

85

Table 17 Summem for Causal Effects for Caucasian Group Path 2 Outcome

 

 

 

 

D_et_ermina_nt [_)i_@ct Ingrid Tot_a_l
PIAFF MISUP .— .... ....+
(R2 = .217) MICON -.259* .... -259
MIAFF 205* _ .205
MIIND .... __ __ +
$000 -.345** -- -.345
PISUP MISUP .... .... .. +
(R2 + .262) MICON -.265* _.. -265
MIAFF __ __ __ +
MIIND 283* -- .283
SocD -.332* -- -.332
PIIND MISUP -.437* -- -.437 *
(r?2 = .270) MICON -.332* .... -332
MIAFF .685* __ __+
MIIND .... .... __ +
we MISUP .377** --- .377 *
(R2: .352) MICON -.350** -.058 -.408
MIAFF .— .046 .046 *
MIIND .... __ _ +
PIAFF .224 .... .224
SocD —- -.077 -.077

 

 

“Direct effect is significant at he .01 level.
*Direct effect is significant at the .05 level.

* Total effect my be incomplete due to unanalyzed components

For the Caucasian group, a path analysis was conducted to determine the
causal effects among the variables Maternal promotion of independence
(MIIND), Maternal feelings of affect (M/AF F), Maternal feelings of support
(MSUP), Maternal conflict with father (MICON), Paternal feelings of affect
(PIAF F), Paternal promotion of independence (PIIND), Paternal feelings of
support (PISUP), Well-being (WB), and Social Desirability (SocD). Stepwise
multiple regression was used to establish the cause-and-effect relationships

(refer to Figure D1, p. 142). All path coefficients were significant at or below the

86

.05 level. Recall that in the first path analysis, the relationship between paternal
attachment and adolescent well-being was no longer significant once maternal
attachment and maternal control toward father were controlled. However, in the
correlational analysis, PlAtt and WB positively correlated (r = .35, p< .01). In this
analysis, a direct relationship was found between paternal feelings of affect and
well-being (r = .224, p < .05), but not between paternal feelings of support or
paternal promotion of independence and adolescent well-being.

Utilizing calculations from Table D (p. 153), the direct, indirect and total
causal effects of the model are presented in Table 17. R2 is noted for each
endogenous variable. Once again, the outcome of primary interest was
adolescent well-being; the determinant with the largest total causal effect was
maternal conflict with father (-.408). The remaining determinants of adolescent
well-being as indicated by the total causal effect were maternal feelings of
support (.377), and paternal feelings of affect (.224), social desirability (-.077)
and maternal feelings of affect (.046). This model explained approximately 35%
of variance in adolescent well-being. The primary determinant of paternal
feelings of affect was social desirability (-.345) with maternal conflict with father (-
.259) and maternal feelings of affect (.205) following. Approximately 22% of the
variance in paternal feelings of affect was explained by the model. The primary
determinant of paternal feelings of support was social desirability (-.332),
followed by maternal promotion of independence (.283) and maternal conflict
with father (-.265). Approximately 26% of the variance in paternal feelings of

support was explained by the model. The primary determinant of paternal

87

promotion of independence was maternal feelings of support (-.437), followed by
maternal conflict with father (-.332). Note the negative correlation between
paternal promotion of independence and maternal feelings of support.
Approximately 27% of variance in paternal promotion of independence was
explained by the model.

The second Path analysis for the African American and Hispanic group
was completed, with the attachment subscales used in place of the full-scale
attachment measures. The variables and Pearson product-moment correlations
for the Caucasian group analysis are found in Table 18, p. 90. The covariates of
ethnic identity and social desirability were included in the model, along with the
exogenous and endogenous variables.

Figure 4, p. 91, is the second path analysis for the African American and
Hispanic group. It illustrates the superimposed path analyses of the three
paternal subscales. The separate paths for each of the paternal subscales are
found in Figures E4, E5, and E6, pp. 183-184. Path decomposition was used to
examine goodness-of-fit. All legitimate paths between the variables in the model
were identified and resulted in a correlation coefficient for each path. Correlation
decompositions were determined for all possible bivariate correlations in the
models, with the exception of those between exogenous variables. The
decompositions and reproduced correlations for can be found in Table E, p. 179.
The reproduced correlations are displayed adjacent to the observed correlations
in Table 19 (p. 90). The comparison of reproduced correlations and empirical

correlations indicated the model fit the empirical data.

88

For the African American and Hispanic group, a path analysis was
conducted to determine the causal effects among the variables Maternal
promotion of independence (MIIND), Maternal feelings of affect (MIAF F),
Maternal feelings of support (MSUP), Maternal conflict with father (MICON),
Paternal feelings of affect (PIAF F), Paternal promotion of independence (PIIND),
Paternal feelings of support (PISUP), Well-being (WB), Ethnic Identity (ETH) and
Social Desirability (SocD). Stepwise multiple regression was used to establish
the cause-and- effect relationships (refer to Figure E1, p. 157). All path
coefficients were significant at or below the .05 level. Utilizing calculations from
Table E (p. 179), the direct, indirect and total causal effects of the model are
presented in Table 20 (p. 93). R2 is noted for each endogenous variable. Once
again, the outcome of primary interest was adolescent well-being; the
determinant with the largest total causal effect was maternal feelings of affect
(.604). The remaining determinants of adolescent well-being as indicated by the
total causal effect were ethnic identity (.482), paternal feelings of support (.481 ),
social desirability (-.373), paternal promotion of independence (.293), maternal
promotion of independence (.164), paternal feelings of affect (-.146), and
maternal conflict with father (-.003). Note the negative impact of paternal

feelings of affect and social desirability.

89

Table 18 Pearson Correlation Coefficients African American and Hispanic Group
Endogenous and Exogenous Variables Used in Path 2

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MIAFF MIIND MISUP PIAFF PIIND PISUP WB ETH SocD
MIAFF 1 .00
MIIND .598" 1.00
MISUP .843“ .304” 1 .00
PIAFF 250* .271* .142 1.00
PIIND .425" .404” .231” .542” 1 .00
PISUP . 108 .299" .032 .878“ .381" 1 .00
WB .672“ .506“ .476“ .508" .61 2“ .378“ 1 .00
ETH .1 34 .023 -.057 .330" 256* .216* .488“ 1 .00
$000 -.494** -.223* -.394** -.233* -.243* .022 -. 543" -.053 1 .00
MICON -.088 -.172 -.235* -.293** -.004 -.255* -.014 .147 243*

 

Maternal promotion of independence (MIIND), Maternal feelings of affect (MIAFF), Maternal
feelings of support (MSUP), Paternal feelings of affect (P/AF F), Paternal promotion of
independence (PIIND), Paternal feelings of support (PISUP), Well-being (WB), Ethnic Identity
(ETH), Social Desirability (SocD)
** Correlation is significant at the 0.01 level (2-tailed)

* Conelation is significant at the 0.05 level (2 tailed)

Table 19 African American and Hispanic Group Path 2 Observed and

Reproduced Correlations

 

 

 

 

 

 

 

 

 

 

 

MAE F MIIND M/(QN PIIND P/S_UP P/A_FF W§ ETH
Observed Correlations

MIAFF 1.00

MIIND .598 1.00

MICON -.088 -.172 1.00

Pll ND .425 .404 —.004 1 .00

PISUP .108 .299 -.255 .381 1.00

PIAFF .250 .271 -.293 .542 .108 1.00

WB .672 .506 -.014 .612 .378 .508 1.00

ETH .134 .023 .147 .256 .216 .330 .488 1.00

SocD -.494 -.223 .243 -.243 .022 -.233 -.543 -.053
Reproduced Correlations

MIAFF 1.00

MIIND .598 1.00

MICON — — 1.00

PIIND .395 .417 .030 1.00

PISUP .150 .271 -.213 - 1.00

P/AF F .124* .209 -.311 - - 1.00

WB .683 .525 -.003 .539 .457 .117* 1.00

ETH .051 .085 .140 .217 .247 -.325 .374* 1.00

$99 -.476 -.2_§§i .201 -.187 439 -.12_2: -.6_04 1&4

 

 

 

*Difference >08

90

 

 

 

 

 

 

 

 

.598”

 

 

 

 

h0¥dqi

 

 

 

 

 

 

 

Figure 4 Input Path Diagram 4: African American/ Hispanic Group Path
Analysis 2

 

 

 

 

I

 

 

 

 

 

 

 

 

 

 

 

 

 

91

 

\Ndh
bemg

 

.481”

 

 

 

 

This model explained approximately 79% of variance in adolescent well-
being. The primary determinant of paternal feelings of affect was maternal
conflict with father (—.31 1) with maternal promotion of independence (.209)
following. Approximately 27% of the variance in paternal feelings of affect was
explained by the model. The primary determinant of paternal feelings of support
was maternal promotion of independence (.271) followed by ethnic identity (.247)
and maternal conflict with father (-.213). Approximately 19% of the variance in
paternal feelings of support was explained by the model. The primary
determinant of paternal promotion of independence was maternal feelings of
affect (.272), followed by maternal promotion of independence (.244), ethnic
identity (.217), and maternal conflict with father (.030). Approximately 26% of
variance in paternal promotion of independence was explained by the model.
The primary determinant of ethnic identity was paternal feelings of affect (.408),
followed by maternal conflict with father (.140) and maternal promotion of
independence (.085). Approximately 17% of the variance in ethnic identity was
explained by the model. The primary determinant of social desirability was
maternal feelings of affect (.476), followed by maternal conflict with father (.201).
Approximately 28% of the variance in social desirability was explained by the

model.

92

Table 20 Summary for Causal Effects for African American and Hispanic Group

 

 

 

Path 2
Outcome Determinant Direct Indirect Total
PIAFF MIIND 209* .— .209
(r?2 = .270) MICON -.311* —— -.311
MIAFF —— —- -- *
PISUP ETH 247* —— .247
(R2 = .191) MIIND 250* .021 .271+
MICON -.248* .035 -213
MIAFF —- —- —— *
PIIND ETH 217* -— .217
(R2 + .262) MIIND 244* -- .244 *
MICON -— .030 .030
MIAFF 253* .018 .272+
wa_ MIAFF .354** .250 .604 *
(R2 = .792) MIIND -.- .164 .164 *
MICON ——- -.003 -.003
PIIND .295** -- .295
PISUP .481** -—- .481
PIAFF -.369* .223 -.146
ETH .363** .119 .482
SocD -.373** -— -.373 *
SocD MIAFF -.476** -- -.476
(R2 = .282) MIIND —- .— —— *
MICON 201* -— .201
ETH MIAFF nu -— —— *
(R2 = .174) MIIND -— .085 .085
MICON 267* -.127 .140
PIAFF .408** —— .408

 

“Direct effect is significant at he .01 level.
*Direct effect is significant at the .05 level.
* Total effect my be incomplete due to unanalyzed components

93

Discussion

Parental Attachment, Interparental Conflict, and Adolescent Well-being

Four major findings emerged from this study. The first major finding
concerns the impact of maternal attachment style and overt maternal conflict with
fathers on paternal attachment style postdivorce. As predicted, maternal
attachment style and maternal conflict with father were found to influence the
style of paternal attachment within both groups. Contradictory to what was
anticipated, maternal attachment style and maternal conflict with father were
found to be unrelated. These findings are significant in that father-child
attachment style had been shown to be vulnerable to the relationship children
have with their mothers (Doyle, Markiewicz, Brendgen, Lieberman, & Voss, 2000;
Hojat, 1998) and with the cooperative parenting relationship (Hetherington, et al.,
1982; McKenry, et al., 1992; lhinger-Tallman, et al, 1993; Whiteside & Becker,
2000). Custodial mothers serve as primary caregivers (US. Census Bureau,
2001) and have more direct influence on children’s everyday activities and
psychosocial development. The suggestion that the mother-child relationship
may benefit from a lack of cooperative parenting and poor father-child
relationships (Whiteside & Becker, 2000) was not supported by this study. The
data gathered in this study indicated that the level of maternal conflict with father
uniquely predicted the style of paternal attachment. Nevertheless, the style of
maternal attachment did not impact the relationship between maternal conflict
with father and the style of paternal attachment. These findings indicate that it is

important to not just evaluate the relationship between maternal and paternal

94

attachment and adolescent well-being, but to do so within the context of family
process variables that may explain the parental divorce-child psychosocial
adjustment relationship. Interparental conflict has been identified consistently as
a major correlate of behavior problems in children across a wide array of family
structures and settings (for reviews Davies & Cummings, 1994; Erel & Burman,
1995). However, if the current model is accurate, it appears that within African
American, Hispanic, and Caucasian divorced families, interparental conflict
damages the father-child bond and thereby negatively affects adolescent well-
being. Among the Caucasian adolescents, perceived maternal conflict had a
direct negative affect on adolescent well-being, as well as on paternal
attachment. Despite racial/ethnic affiliation, the findings of this study suggest the
mother-child bond is resilient to and unaffected by perceived mother-initiated
interparental conflict. Secure maternal attachment promoted adolescent
psychosocial adjustment, independent of the level of maternal conflict with father.
The positive correlation between maternal attachment style and paternal
attachment suggests that the influence of family relations on the child cannot fully
be understood by studying one isolated relationship. When working with
adolescents from divorced families, a key component appears to be the maternal
style of attachment and how this style of attachment promotes or impedes the A
style of paternal attachment perceived by the adolescent (Davies 8. Cummings,
1994; Rothbaum & Weisz, 1994; Braver, et.al., 1991). Among African American
and Hispanic college students, the perception of mothers as fostering autonomy

corresponded with melings of availability, understanding, acceptance, care,

95

respect and facilitation of autonomy from fathers. In addition, the data revealed
an impact of the affective quality of the mother-adolescent relationship on the
perception of fathers as promoting independence. However, the data did not
reveal a relationship between the perception of mother as a source of support
and the style of paternal attachment. Nor did the data reveal a relationship
between the affective quality of the mother-adolescent relationship and perceived
paternal feelings of support. This suggests that adolescents respond differently
to mothers than they do to fathers, especially in divorced families. Among
African American and Hispanic college students, mothers treating their
adolescents as competent adults may open the door for adolescents and their
fathers to establish affective and emotionally supportive relationships. Likewise,
it appears that if mothers provide a secure foundation for the affective needs of
the adolescent, the adolescent is more receptive to father’s facilitation of
autonomy.

The second major finding of this study concerns the relative importance of
parental attachment and overt maternal conflict with fathers to adolescent well-
being postdivorce. Within the African American and Hispanic group, the
prediction that secure attachments to both mother and father would be
associated with higher levels of general, spiritual, and psychological well-being
was supported when the full-scale measures of attachment were analyzed (total
causal effects .693 and .523 respectively). The large effects of adolescent-
parent relationships on adolescent well-being were similar to other studies of

maternal and paternal attachment (Amato & Booth, 1996; Hojat, 1998; Whiteside

96

& Becker, 2000). Nevertheless, interesting results were produced when the
attachment subscales were analyzed instead of the full-scale measures. The
perception of mother as a source of support had no effect on well-being or on
any of the paternal attachment subscales, despite the moderately strong
relationship found between maternal feelings of support and well-being in the
correlational analysis (r = .377, p < .01). In addition, while maternal promotion of
independence appeared to influence paternal promotion of independence,
paternal affective quality of relationship, and fathers as source of support, all
three of which influence adolescent well-being, maternal promotion of
independence did not directly influence adolescent well-being. Correlational data
had suggested a moderately strong relationship between maternal promotion of
independence and adolescent well-being (r = .377, p < .01), but it was no longer
significant once paternal attachment subscales were entered into the analysis.
Conversely, maternal affective quality of relationship had a consistent influence
on well-being (b = .354, p < .01; r = .351, p < .01), yet it only impacted the
paternal attachment subscale of paternal promotion of independence (b = .244, p
< .05). The perception of father both as a source of support and as promoting
independence positively affected adolescent well-being (b = .481 and .295
respectively, p < .01). A finding of this study that contradicted previous research
findings drawn from primarily Caucasian samples (Hetherington, Cox, Cox, 1982;
Summers, Forehand, Arrnistead, & Tannenbaum, 1998) was that perceived
paternal feelings of affect had a negative direct impact on adolescent well-being

(b = -.369, p< .05).

97

These findings are indicative of how adolescents are influenced by their
perceptions of their relationships with their mothers and with their fathers
postdivorce. Noncustodial fathers have limited direct involvement in the daily
lives of their children, Gecas and Schwalbe (1986) suggested that when fathers
are involved, the importance of this involvement might be exaggerated in the
mind of adolescents and thus take on greater importance than the daily
involvement of the mother. Thus, African American and Hispanic college-
attending adolescents might perceive mother's facilitation of autonomy and
conveyance of support as less important than similar behaviors by fathers.
Paternal facilitation of autonomy and conveyance of support may be perceived
as indicative of a caring and secure relationship, thereby enhancing the father-
child relationship and adolescent well-being.

In regards to the affective quality of the relationship within the African
American and Hispanic group, the data indicated that perceptions of mother as
understanding, accepting, and considerate had a strong positive impact on
adolescent well-being, whereas similar perceptions of father had a moderately
strong negative impact on adolescent well-being. In secure relationships,
parental figures often serve as secure bases from which adolescents explore and
as sources of comfort in times of stress (Kems, Klepac, 8. Cole, 1996; Larson,
Richards, Moneta, Holmbeck, 8. Duckett, 1996). However, within divorced
families, where mothers have physical custody, the limited time with father may
consequently make the affective quality of the adolescent-father relationship

more labile and vulnerable, thus exerting a paradoxical impact on the

98

adolescent’s sense of well-being in conjunction with attending college. Several
studies indicate a precipitous decline in father contact after divorce, with 23% to
30% of sampled children and sampled fathers reporting no contact with each
other during the first year postdivorce (F urstenberg, Morgan, & Allison, 1987;
Mitchell, 1985; Seltzer, Schaeffer, & Chamg, 1989). Feelings or fear of
abandonment of the noncustodial fathers when the student transitions to college
may result in the secure father-adolescent affective relationship having a
negative impact on well-being (b = -.369). Conversely, in regard to the mother-
adolescent relationship that is based on daily contact, the secure mother-
adolescent affective relationship may facilitate the transition to college by
providing emotional assurance and stability (total effect = .604).

Among the Caucasian group, the absence of a relationship between
paternal attachment and adolescent well-being in the first path analysis indicates
the importance of including the different components of parental attachment in
any analysis of parental attachment style. Findings and conclusions based upon
the path analysis using the full-scale attachment measures would have been
much different than those that were drawn from the analysis using the
attachment subscales. These findings indicate that parent—child attachment
styles may be much more complicated than can be accurately represented with a
single measure. In addition, the correlations between maternal attachment style
and paternal attachment suggest that the influence of family relations on the

adolescent cannot fully be understood by studying one isolated relationship.

99

In regard to the first major finding of this study (i.e., the impact of maternal
attachment style and overt maternal conflict with fathers on paternal attachment
style) among Caucasian adolescents from divorced families, maternal promotion
of autonomy facilitated the father-adolescent relationship as a source of support.
A secure emotive relationship with mother facilitated a secure emotive
relationship with father, as well as greatly impacting the adolescent’s sense of
security in father’s promotion of autonomy. Maternal conflict with father uniquely
predicted the style of paternal attachment within the three attachment subscales.
Surprisingly, the perception of the mother-adolescent relationship as a source of
support significantly impeded a sense of security and comfort with the father's
promotion of autonomy. Thus, the hypothesis that low levels of maternal conflict
with father would correspond with secure attachment to father is supported.
Maternal attachment did not moderate the relationship between the degree of
maternal conflict toward father and the style of paternal attachment. Similar to
the African American and Hispanic group, maternal attachment and maternal
conflict toward father were not correlated. The hypothesis that secure
attachment to mother would correlate with secure attachment with father did not
hold true with the attachment subscales of mother as source of support and
paternal promotion of independence.

The interplay between the promotion of autonomy and source of support
subscales is interesting. Whereas secure maternal promotion of autonomy
positively correlated with secure paternal feelings of support, secure maternal

feelings of support negatively correlated with paternal promotion of autonomy.

100

Within this college-age population, these findings suggest that mothers’
responses to their child’s transitions from childhood to adulthood may
significantly impact not only their child, but also, their child’s relationship with
their fathers. When the mother-adolescent relationship is secure in the
promotion of autonomy, the adolescent experiences his/her father as more
emotionally supportive. However, the more adolescents perceive their mothers
as sources of emotionally support, the less secure they appear to be with their
fathers’ promotion of autonomy. Maternal affective quality had a stronger impact
on secure paternal promotion of autonomy than did maternal feelings of support,
however, neither paternal promotion of autonomy nor maternal affective quality
significantly influenced adolescent well-being. These findings provide some
preliminary evidence suggesting that there may be distinct differences between
how adolescents from divorced families are influenced by parental attachment
styles postdivorce and their subsequent well-being.

In regards to the second major finding regarding parental attachment and
adolescent well-being, within the Caucasian group, the prediction that secure
attachments to both mother and father would be associated with higher levels of
general, spiritual, and psychological well-being was not supported when the full-
scale measures of attachment were analyzed. Only secure maternal attachment
was significantly correlated with higher levels of general, spiritual, and
psychological well-being. This finding was not consistent with previous studies
that found large effects of adolescent-mother and adolescent-father relationships

on adolescent well-being (Amato 8: Booth, 1996; Hojat, 1998; Whiteside &

101

Becker, 2000). When the parental attachment subscales were used in the
analysis, adolescent well-being was enhanced by the perception of mothers as
sources of emotional support, positive affective qualities of paternal relationships,
and low levels of maternal conflict with father. These results support the
hypothesis regarding parental attachment and conflict, but provide greater
specificity than was theorized.

The third major finding of this study concerns the importance of ethnic
identity and a response style of social desirability to parental attachment and
adolescent well-being. The tendency to respond to self-report measures with a
style of social desirability is well documented in the research (Reynolds, 1982)
and was found to have a significant impact on the variables in this study. While
this response style is found and controlled in numerous research studies, few
researchers attempt to explain the relationship between a social desirability
response style and their main variables. In this study, a response style of social
desirability was found within both the African American and Hispanic group and
the Caucasian group and was entered into the path analysis first in order to
control for unwanted variance. The differences found between the groups in
relation to social desirability were interesting. Within the Caucasian group, social
desirability had a moderately strong impact on paternal feelings of affect and
paternal feelings of support, but did not correlate with any of the maternal
attachment subscales or impact adolescent well-being. Within the African
American and Hispanic group, social desirability was impacted by maternal

feelings of affect, maternal promotion of independence, and maternal conflict with

102

father. Social desirability also had a moderately strong negative impact on
adolescent well-being. These findings provide some preliminary evidence that
suggest the tendency to respond in a socially desirable manner is associated
with the parent-adolescent relationship, which may be indicative of adolescent
well-being, and may differ among racial/ethnic groups. The results of the current
study are consistent with Searle’s (1998) finding that social desirability had a
differential effect within attachment categories, but contradicted Seiffge’s (2003)
finding that securely attached persons reported less mental and physical
symptoms but higher social desirability. The relationship between parental-child
relationships and the willingness to present oneself in an unfavorable light
warrants clarification through further research.

Ethnic identity was also found to co-vary with paternal attachment styles
within the African American and Hispanic group and was thus included in the
path analysis. The data indicated that maternal conflict toward father and
perceived paternal feelings of affect positively influenced ethnic identity, while
ethnic identity in turn positively influenced paternal feelings of support, paternal
promotion of independence, and adolescent well-being. The moderately strong
effects between ethnic identity and paternal feelings of affect and adolescent
well-being suggest that ethnic identity might play a substantial role in mother-
child, father-child, and mother-father relationships, as well as significantly
impacting adolescent well-being. These findings suggest that ethnic identity
should be considered whenever studying family process variables within

racially/ethnically diverse samples.

103

The fourth major finding of this study concerns the differences in parental
attachment and adolescent well-being found between the African American and
Hispanic group and the Caucasian group. The findings of the current study did
not support the hypothesis that no significant differences would be found
differentiating Hispanic and African American adolescents’ from Caucasian
adolescents’ parental attachment patterns. Several differences in the data have
been illustrated thus far, but a few of the more significant include: 1) within the
African American and Hispanic group, MIAFF, MIIND, PIIND, PISUP, and P/AF F
causally effect adolescent well-being, whereas only MISUP and PIAF F causally
effect adolescent well-being within the Caucasian group; 2) maternal conflict with
father had a significant negative relation on adolescent well-being within the
Caucasian group, but not within the African American and Hispanic group; 3)
MIIND causally effect all three paternal attachment subscales within the African
American and Hispanic group, but only effects PISUP within the Caucasian
group; 4) MISUP had no relationship with PIAFF, PIIND, PISUP, or WB within the
African American and Hispanic group, yet MISUP had a moderately strong
negative impact on PIIND and a moderately strong positive impact on WB within
the Caucasian group; 5) P/AF F had a negative impact on WB within the African
American and Hispanic group, but a positive impact on WB within the Caucasian
group; 6) ethnic identity was a significant variable within the African American
and Hispanic group, but not within the Caucasian group; and 7) the model
accounted for 79% of the variance within the African American and Hispanic

group, but only 35% of the variance within the Caucasian group. These findings

104

suggest it would be ineffectual and unethical to attend to racially/ethnically
diverse family systems and adolescents in the same manner as majority group
family systems.

The findings from this study were consistent with prior research that found
father-adolescent and father-young adult relationships to be significant predictors
of young adult psychosocial well-being (Summers, Forehand, Arrnistead, &
Tannerbaum, 1998; Whiteside & Becker, 2000). As one of the first studies to
examine this relationship within the context of postdivorce parental attachment
among African American, Hispanic, and Caucasian college students, this study
extended the findings of other researchers who have examined the unique links
between 1) divorce and adolescent well-being, 2) parental attachment and
adolescent well-being, 3) parental conflict and adolescent well-being and 4)
maternal and paternal attachment styles. Clearly, the results of this study do
suggest that 1) mother-adolescent bonds and mother-father conflict have an
impact on the father-adolescent bonds and 2) mother-adolescent and father-
adolescent bonds play an important role in the development of adolescent self-
esteem. Clearly, differences exist in the postdivorce family process variables
and parental relationships between the racial/ethnic groups. Efforts should be
made to replicate this study to establish confidence in these findings and to
clarify the robustness of these initial interpretations

Limitations

There are several limitations inherent in this study. To start with,

combining the African American group and Hispanic group due to sample size

105

may limit the generalizability of the results. Future studies should include a more
diverse sample that includes a larger number of participants of different
races/ethnicities. Secondly, it is important to note that in the process of using
path analysis, causal inferences were drawn from correlational data. The degree
of confidence in the validity of causal inference from correlational data is typically
much weaker than inference drawn from data resulting from a well-designed
experimental study where the important concept of random assignment to
treatments has been incorporated (Tate, 1998). In path analysis with
correlational data there is no statistical test that will definitively indicate whether
the specifications, or misspecifications, represent true causal effects.

In addition, research such as this that relies entirely on self-report
measures of key constructs, and retrospective measure of variables, may be
vulnerable to memory and\or distortions. While attention to diversity is apparent
in the design, the study relied on a convenient sample from a large mid-westem
university identified largely as White. Therefore, appropriate caution should be
exercised in generalizing the findings. Efforts were made to control recognized
covariates, but there may be confounds such as personality characteristics or
other relationship-specific factors.

Lastly, the developmental stage of the subjects in this study may have
influenced the results. Psychoanalyst Erik Erikson has described the physical,
emotional and psychological stages of development and related specific issues,
or developmental work or tasks, to each stage. According to Erikson, the

subjects in this study were at the stage of development that he called Identity vs

106

 

 

 

Role
sat:
adc

inte

[I

Role Confusion. Individuals at this stage of life are learning how to answer
satisfactorily the question of "Who am I?” but even the most adjusted of
adolescents experience some role identity confusion. Attempts are made to
integrate many roles (e.g., child, sibling, student, athlete, worker) into a self-
image under role models and peer pressure. In later adolescence, the
adolescent seeks leadership, and gradually develops a set of ideals. Erikson
believes that, in our culture, adolescence do not yet have to "play for keeps,” but
can experiment, trying various roles, and thus hopefully find the one most
suitable for them. This means that individuals at this age are negotiating their
roles with their parents, their beliefs about life, and their ideas about who they
are. The impact of this process could have directly or indirectly influenced the
results in this study. According to Lewinsohn and Rosenbaum (1987), current
emotional states may affect perceptions of the parent-child relationship. In this
study, there was a statistically significant difference in the educational level of the
two groups (p < .01). Nearly 50% of the African American and Hispanic group
were freshman and 19% sophomores, whereas the Caucasian group was
dispersed fairly evenly across the educational levels, with 25% within each of the
four years. This study should be repeated with attention given to the educational
level/age of the participants.
Relevance to Practice

Findings from this study serve to highlight the changing and diverse nature
of parent-child relationships after divorce, as well as the complexity of family

dynamics among African American, Hispanic, and Caucasian families. Appraisal

107

of the mother-adolescent, father-adolescent, and mother-father relationships
serve as important predictors of multiple areas of adjustment for adolescents and
young adults. Adolescent attachment to parents seems to have important
implications for adolescent subjective well-being in a broad range of subscales.
A significant implication of this study resides in the difference found between the
African American and Hispanic group when compared to the Caucasian group.
These distinctions indicate that the same modality of treatment cannot be used
across racial and ethnic groups. We are once again reminded that, “one size
does not fit all.” Therapy needs to be sensitive to the physical and mental
features of the divorce process, including loyalty dilemmas, loss, renewal,
understanding of clients’ divorce-related experiences, and the reengagement of
adult children with their fathers within the cultural context of the individuals being
served. Racially/ethnically sensitive interventions need to be identified to help
parents handle conflict constructively, support and facilitate cooperative co-
parenting, and promote positive child attachment relationships with both parents.
Clinicians working with adolescents from divorced families with mother-custody
arrangements may want to select intervention strategies that help these
adolescents examine and understand their emotional relationships with their
fathers, perhaps aiding them in relinquishing a sense of responsibility for conflict
while helping them to focus on their ability to take responsibility for the positive
happenings in their lives. Findings of this study serve to emphasize specific

avenues for interventions to enhance the mother-adolescent relationship and the

108

father-adolescent relationship in distressed divorced families within different
racial/ethnic groups.

This study suggests that when working with Caucasian college students
from divorced families there may be great benefit in assessing paternal
relationships and parental conflict, past and present, and subsequent impact on
students’ well-being. If emotional support is missing from the student’s
relationship with his/her mother, the therapeutic relationship could serve as a
model. This relationship and the establishment of a support system will help the
student feel understood, build confidence and supply empowerment.
Compromised affective quality of the paternal relationship may manifest in a
need to control, fear of closeness, and a lack of reciprocity in interpersonal
relationships. The therapeutic challenge is to enhance the establishment of
secure attachment patterns of trust, affection, intimacy, communication, and
reciprocity. Students are assisted in learning to identify, manage, and express
emotions in a constructive manner.

Alternatively, this study suggests that when working with African American
or Hispanic students, compromised affective quality of maternal relationships
may result in feelings of isolation, alienation, and disconnect. Treatment may be
enhanced by focusing on establishing a sense of connection and belonging, as
well as the formulation of identity. Students need to be assisted in exploring the
environment with feelings of safety and security, which will then lead to healthy
cognitive and social development. A lack of paternal facilitation of autonomy or

affective support may challenge the basic qualities of trust, empathy,

109

cooperation, and reciprocity within relationships. The counselor can assist in the
modification of negative relationship dynamics, which will enhance stability and
support from inside and outside the family, and create a climate of hope,
positivity, and closeness. Core beliefs about the maternal relationship are
focused on in order to improve the paternal relationship and enable students to
create and maintain emotionally reciprocal relationships and cultivate a positive
and realistic sense of self and self-in-relation to the world.

Additionally, given 1) the increased concern for budgetary restrictions in
institutions across the country, 2) the percentage of students actually graduating
from college is declining (US. Department of Education, 1995), and 3) once at
college, students are reporting record high levels of emotional and psychological
stress (Sax, Astin, Korn, & Mahoney, 1999), the option of intervening with
students experiencing difficulties via psychoeducational groups focusing on
facilitating healthy attachment might be feasible and cost-efficient. Regardless of
the mode of intervention, it is important that clinicians working with college
students better understand the reasons why some students make it through
college successfully while other students become increasing emotionally
distressed, resulting in some students experiencing decreased school
performance, misconduct, depression, and difficulties with interpersonal
relationships. This research offers contributions in gaining understanding in this

area, especially in meeting the needs of African American and Hispanic students.

110

Future Research

Efforts should be made to replicate this study to establish confidence in
these findings and to clarify the robustness of these initial interpretations.
Postdivorce family process variables within different racial and ethnic groups
need further examination. Efforts should be made to use measures normed on
diverse samples or to norm the current measures on African American and
Hispanic samples. Future research may find it beneficial to explore the varying
levels of noncustodial father contact, the different possible dynamics of the in-
residence family composition on the out-of-residence relationship with father, and
whether similar findings exist between custodial fathers and noncustodial
mothers within and between different racial/ethnic groups. Future studies,
perhaps with longitudinal designs, could also yield support for the causal
inferences suggested in this study by further delineation of the unique effects of
maternal and paternal attachment styles on additional dimensions of adolescent
functioning postdivorce. Exploration of the role of ethnic identity in the parent-
child and mother-father relationships in the realm of attachment theory could

prove to be enlightening.

111

APPENDICES

112

APPENDIX A

Forms and Demographic lnforrnation

UCHRIS Approval
Recruitment Announcement
Consent Form

Table A1 Caucasian Group Descriptive Statistics for Demographic
Information

Table A2 Caucasian Group Descriptive Statistics for Subscales and
F ull-scale Parental Attachment Measures.

Table A3 African American and Hispanic Group Descriptive Statistics
for Demographic lnfonnation

Table A4 African American and Hispanic Group Descriptive Statistics
for Subscales and Full-scale Parental Attachment Measures.

Table A5 Cronbach’s Alpha

113

114
114
115

117

117

118

119

119

UCHRIS Approval
IRB #04-208 / APP # i019504
Approved 3I20l2004. Renewal approved 2l21l05.

Recruitment Script

Good day, my name is Janet Kinney and I am a doctoral candidate in
Counseling Psychology. I am here today to recruit volunteers to participate in my
dissertation research project. Your (instructor) has allowed me to speak with you
today about the study and offer an opportunity for those who are interested in
participating to complete a survey.

The purpose of this study is to learn more about the factors that affect
individuals from divorced families. Therefore, only undergraduates that
experienced the divorce of their biological parents in childhood are eligible to
participate. If you agree to participate in this study, you will be asked to complete
paper and pencil surveys that are designed to measure your perceptions of your
parents’ attitudes towards you, your relationships with your parents, and your
well being. Total time of participation is approximately 30-40 minutes.
Participants will have the opportunity to enter a drawing in which they will be
eligible to win one of three $100 awards. Those wishing to enter the drawing will
need to supply an email address on the consent form, along with their name.
Since this study is focusing on factors that affect students from divorced families,
there are some specific criteria that apply to participants. Potential subjects must
be: a) 18-23 years old, b) single with no children, c) parents divorced prior to age
16, d) mother had primary custody, e) postdivorce contact with both parents, f)
8”1 grade reading level.

Your responses to the survey questions will be kept confidential.
Furthermore, your name will not appear on any of the measures, and no one will
have any way of associating your name with any of the measures. If you
volunteer to participate, you are under no obligation to complete the
questionnaire packet. In a moment I will be passing out the informed consent
forms. If you would like to participate and you meet the study criteria, please
take one. If you do not qualify or would just rather not participate, please do not
take one. After we go over the informed consent form, I will answer questions
and then distribute the survey packet. I thank you for your time.

Are there any questions right now?

114

INFORMED CONSENT FORM

Postdivorce Parental Relationships and Well-being
Among African American, Hispanic, and Caucasian College Students

Procedures

If you agree to participate in this study, you will be asked to complete paper and
pencil instruments that are designed to measure your perceptions of your
parents’ attitudes towards you, your relationships with your parents, and your
well being. Total time of participation is approximately 30-40 minutes. All
participants will have the opportunity to enter a drawing in which they will be
eligible to win one of three $100 awards. Those wishing to enter the drawing
need to supply an email address on this consent form, along with their name.
Consent forms will be randomly selected to determine the winners. Once all data
is collected, winners will be asked via email and to supply an address where a
check can be mailed. Potential subjects must be: a) 18-23 years old, b) single
with no children, c) parents divorced prior to age 16, d mother had primary
custody, e) postdivorce contact with both parents, f) 8 grade reading level.

Risks/Discomforts
There are minimal risks involved in participating in the study. You might find that
publicly stating you are from a divorced family is uncomfortable or you may find
there are instrument questions that make you uncomfortable or upset. There are
several ways to deal with uncomfortable feelings that may arise:
a. Participation is strictly voluntary, so there is no pressure to participate.
b. You may consider speaking with a counselor about your reactions (e.g.,
you could contact the MSU Counseling Center at 355-8270).
c. You are free to decline to answer any questions you do not wish to answer
or to stop your participation at any time.

Privacy and Confidentiality

Your privacy will be protected to the maximum extent permissible under the law.
The following precautions will be taken to protect your confidentiality: a) no
individual names or other identifying information will be used in any reports or
publications that may result from this study, b) your name will not be on any of
the surveys, and 0) both the informed consent forms as well as the completed
instruments will be kept in a locked cabinet and only the primary researcher will
have access to these files.

Participation

Your participation in this study is voluntary; you may decline to participate without
penalty. If you decide to participate, you may withdraw from the study at anytime
without penalty and without loss of benefits to which you are otherwise entitled.

115

Contact

If you have questions at any time about the study or the procedures, you may
contact the principle investigator, Janet M. Kinney, MA, LLP, at

[mkimev3@aol. com or at 51 7-62 7-5490. If you have questions or concerns
regarding your rights as a study participant, or are dissatisfied at any time with
any aspect of this study, you may contact — anonymously, if you wish - Dr. Peter
Vasilenko, Chair of the University Committee on Research Involving Human
Subjects (UCRIHS) by phone: (51 7) 355—21 80, fax: (51 7) 432-4503, e-mail:
ucn'h§_@msu.edu, or regular mail: 202 Olds Hall, East Lansing, MI 48824. ”

Consent
I voluntarily agree to participate in this study. Please give the signed form to the
interviewer and keep the other copy of the consent form for your records.

Participant’s name (print)

 

Participant's signature

 

Email address
Date

 

116

Table A1 Caucasian Group Descriptive Statistics for Demographic Information

 

 

Std.

N Min Max Mean Deviation
Age 93 18.00 24.00 19.9677 1.30594
Education Level 93 1.00 5.00 2.4624 1.17547
Age at time of separation 68 1.00 16.00 8.1029 4.82264
Age at time of divorce 93 .00 16.00 7.9247 5.00920
Length of parents committed 93 2.00 26.00 11.6989 5.47878
relationship to each other in years
Number of siblings with same 93 .00 4.00 1.1183 .84506
parents
Frequency of contact with mother 93 1.00 4.00 1.5484 .71500
currently
Frequency of contact with mother 93 1.00 5.00 1.0860 .58339
prior to age 18
Frequency of contact with father 93 1.00 4.00 2.2903 .89176
currently
Frequency of contact with father 93 1.00 4.00 2.0430 .87121
prior to age 18
mother's educational level 93 1.00 5.00 2.9785 1.04235
father‘s educational I ev el 93 1.00 5.00 3.1828 1.12236

 

 

 

 

 

 

Table A2 Caucasian Group Descriptive Statistics for Subscales and Full-scale
Parental Attachment Measures.

 

 

 

 

I N Min Max Mean Std. Dev.

Maternal Attachmennota, 1 93 95.00 245.00 205.0753 33.38420
Maternal feelings of affect 93 37.00 119.00 99.4516 18.28489
Mt I ti 1
,n§,$:d§;22’° °" ° 93 37.00 80.00 65.5806 10.69978
Mother Wings of sumo“ 93 23.00 70.00 56.4516 11.67636
Paternal Attachment Total 93 90.00 246.00 184.3656 40.87435
Pat al ff ° 1' r
@3821?“ “any" 93 38.00 119.00 87.9677 23.65488
Pat I tio r
,,,;,,’2:J;fi’;2° " ° 93 36.00 81.00 64.5699 10.31820
F th r r rt

3 “teams" sum” 93 18.00 68.00 46.1398 14.80716

 

 

 

117

 

 

 

Table A3 African American and Hispanic Group Descriptive Statistics for

Demographic lnforrnation

 

 

 

 

 

 

 

 

N Min Max Mean Std. Dev.
Age 86 18.00 23.00 19.6512 1.31746
Education Level 86 1.00 5.00 1.9535 1.06171
Age at time of separation 54 1.00 16.00 6.2778 3.84814
Age at time of divorce 86 1.00 16.00 6.4651 3.75013
Length of parents committed 82 2.00 28.00 9.3415 5.03621
relationship to each other in years
Number of siblings with same 85 .00 4.00 1.2941 .99790
parents
Frequency of contact with mother 86 1.00 4.00 1.5581 .67918
currently
Frequency of contact with mother 86 1.00 3.00 1.1163 .41780
prior to age 18
Frequency of contact with father 86 1.00 5.00 2.7907 .99548
currently
Frequency of contact with father 1.00 4.00 2.6279 1.04089
prior to age 18
M oth er’s educational level 86 1.00 5.00 2.9419 .89908
86 1.00 5.00 2.7558 1.20719

Father’s educational level

118

 

Table A4 African American and Hispanic Group Descriptive Statistics for
Subscales and Full-scale Parental Attachment Measures.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

N Min Max Mean Std. Dev.
Maternal Attachment Tot 86 123.00 255.00 196.1744 31.21168
Maternal feelings of 86 55.00 118.00 95.2093 16.04194
affect
Maternal promotion of 86 40.00 87.00 63.3721 11.65287
independence
Mother feelings of 86 25.00 70.00 52.8605 12.01290
support
Paternal Attachment Tot 86 78.00 229.00 169.1395 39.04813
Paternal affective quality 86 40.00 1 15.00 80.2326 20.80423
of relationship
Paternal promotion of 86 22.00 78.00 61.0116 14.15003
independence
Father feelings of 86 18.00 66.00 41.1744 13.91334
support
Table A5 Cronbach’s Alpha
Cronbach’s Alpha N of Items
State Self-Esteem Scale .896 20
Life Attitude Profile .941 16
Marlowe-Crowne Social Desirability .844 1 3
Scale
Multigroup Ethnic Identity Measure .901 12
Perceived Wellness Scale .928 36
Schwarz Inter-Parental Conflict Scale .928 34
Maternal Attachment Scale .948 52
Paternal Attachment Scale .957 52

 

 

 

119

 

APPENDIX B

African American and Hispanic Group Path Analysis 1

Figure B1 African American and Hispanic Group Stepwise Multiple 121
Regression for Path Analysis 1

Figure B2 African American and Hispanic group Normal Probability Plot 131
of the Regression Standardized Residuals for Path Analysis 1

Figure B3 African American and Hispanic Residuals Scatter plot for Path 132
Analysis 1

Table B African American and Hispanic Group Path 1 Decompositions 133
and Calculations of Reproduced Correlations

120

Figure 81 African American and Hispanic Group Stepwise Multiple Regression
for Path Analysis 1

B.1.1 Step 1: Well-being regressed on ethnic identity and social desirability, then
maternal attachment, maternal conflict with father, and paternal attachment.

 

 

Model Summary

Adjusted Std. Error of
Model R R Square R Square the Estimate
1 .644a .415 .408 2.06080
2 .788b .621 .612 1.66894
3 835° .697 .686 1 .50009
4 .870d .756 .744 1.35385
5 883° .780 .767 1 .29391

 

 

 

 

 

 

 

a. Predictors: (Constant), Maternal Attachment Total

b. predictors: (Constant), Maternal Attachment Total, eth_
mean

0- Predictors: (Constant), Maternal Attachment Total, eth_
mean, Social Desirability Scale Mean
d. Predictors: (Constant), Maternal Attachment Total, eth_

mean, Social Desirability Scale Mean, Paternal
Attachment Total

9- Predictors: (Constant), Maternal Attachment Total, eth_
mean, Social Desirability Scale Mean, Paternal
Attachment Total, Schwarz inter-parental conflict scale
total

121

Allow!

 

 

 

 

 

 

 

 

 

 

 

 

Sum of
Model Squares df Mean Square F Sig.
_1 R' egression 252.979 1 252.979 59.568 .0003

Residual 356.740 84 4.247
Total 609.719 85

2 Regression 378.535 2 189.267 67.951 .000”
Residual 231.184 83 2.785
Total 609.719 85

3 Regression 425.196 3 141.732 62.984 000°
Residual 184.523 82 2.250
Total 609.719 85

4 Regression 461.252 4 115.313 62.912 .0006
Residual 148.467 81 1.833
Total 609.719 85

5 Regression 475.782 5 95.156 56.836 .0008
Residual 133.937 80 1.674
Total 609.719 85

 

a. Predictors: (Constant), Maternal Attachment Total

 

b- Predictors: (Constant), Matamal Attachment Total, eth_mean

C- Predictors: (Constant), Maternal Attachment Total, eth_mean, Social Desirability
Scale Mean

d- Predictors: (Constant), Maternal Attachment Total, eth_mean, Social Desirability
Scale Mean, Paternal Attachment Total

9- Predictors: (Constant), Maternal Attachment Total, eth_mean, Social Desirability
Scale Mean, Paternal Attachment Total, Schwarz inter-parental conflict scale total

f. Dependent Variable: sse_total

122

Coetflclenti

 

 

 

 

 

 

 

 

 

 

 

Unstandardized Standardized
Coefficients Coefficients

Model B Std. Error Beta t .
Moment) 2.247 1.422 1.580 .118
Maternal Attachment Totq .055 .007 .644 7.718 .000
2 (Constant) -5.696 1.651 3.449 .001
Maternal Attachment To .053 .006 .620 9.153 .000
eth_mean 2.552 .380 .454 6.714 .000
3 (Constant) J -1.000 1.807 -.553 .582
Maternal Attachment Tot .042 .006 .486 7.203 .000
eth_mean 2.501 .342 .445 7.316 .000
3223' Des'rab'w 3w" -3.406 .748 -.307 4554 .000
4 (Constant) -1.282 1.632 -.785 .435
Maternal Attachment Tot .035 .005 .403 6.315 .000
eth_mean 2.027 .327 .361 6.209 .000
3:? Des'rabm 8°“ -3.307 .675 -.298 4.895 .000
Paternal Attachment Tota .019 .004 .274 4.435 .000
5 (Constant) -1 .642 1 .565 -1 .049 .297
Maternal Attachment To .035 .005 .407 6.675 .000
eth_mean 1 .797 .322 .320 5.584 .000
3:2“ “gram" 3““ -3.672 .657 -.331 -5.585 .000
Paternal Attachment Tota .022 .004 .318 5.224 .000
imgwgrggfi‘a' .013 .004 .167 2.946 .004

 

 

 

 

a. Dependent Variable: sse_total

123

 

Excluded Variable!

 

 

 

 

 

 

 

 

 

 

 

 

Collinearity
Parfla' StatIStICS
Model 2 Beta In t Sig. Correlation Tolerance

1 Sihwalz inter-parental a
conflict scale total .110 1.296 .199 .141 .965
3”“ Des'rab'm 3"“ -.3248 -3.753 .000 -.381 .811

ean

eth_mean .4548 6.714 .000 .593 .997
Paternal Attachment Tota .4118' 5.347 .000 .506 .887

2 Schwarz inter-parental b
conflict scale total .036 .520 .605 .057 .940

. . .. b
3”” Des'rab'm’ sm'e -.307 .4554 .000 -.449 .810

can

Paternal Attachment Totq .284b 4.066 .000 .410 .791

3 Schwarz inter-parental ' c
conflict scale total .094 1.487 .141 .163 .906
Paternal Attachment To 274° 4.435 .000 .442 .790

4 Schwarz inter-parental d
conflict scale total .167 2.946 .004 .313 .851

 

 

3- Predictors in the Model: (Constant), Maternal Attachment Total
b- Predictors in the Model: (Constant), Maternal Attachment Total, eth_mean

0- Predictors in the Model: (Constant), Maternal Attachment Total, eth_mean, Social Desirability S
Mean

d- Predictors in the Model: (Constant), Maternal Attachment Total, eth_mean, Social Desirability S
Mean, Paternal Attachment Total

e. Dependent Variable: sse_total

3.12 Step 2: Paternal attachment regressed on ethnic identity and social
desirability, then maternal attachment and maternal conflict with father.

 

 

Model Summary
Adjusted Std. Error of
Model R R Square R Sgare the Estimate
1 .336a .113 .103 36.99038
2 .457b .209 .190 35.15191
3 .507° .257 .230 34.25979

 

 

 

 

 

 

 

a. Predictors: (Constant), Maternal Attachment Total

b. predictors: (Constant), Maternal Attachment Total, eth_
mean

6- Pmdictors: (Constant), Maternal Attachment Total, eth_
mean, Schwarz inter-parental conflict scale total

124

Allowr'

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sum of
Model Squares df Mean Square F Si .
1 egression 14668.133 1 14668.133 10.720 .0028
Residual 1 14936.2 84 1368.288
Total 1296043 85
2 Regression 27044.816 2 13522.408 10.943 000"
Residual 1025595 83 1235.657
Total 1296043 85
3 Regression 33358.196 3 1 1 1 19.399 9.474 .000°
Residual 96246.130 82 1 173.733
Total 1296043 85
a. Predictors: (Constant), Maternal Attachment Total
b- Predictors: (Constant), Maternal Attachment Total, eth_mean
c. Predictors: (Constant), Maternal Attachment Total, eth_mean, Schwarz
inter-parental conflict scale total
d- Dependent Variable: Paternal Attachment Total
Coefficient!
Unstandardized Standardized
Coefficients Coefficients
Model B Std. Error Beta t Si .
_1__—(Oonstant) 86.573 25.531 3.391 .001
Maternal Attachment Totq .421 .129 .336 3.274 .002
2 (Constant) 7.710 34.779 .222 .825
Maternal Attachment To .400 .122 .320 3.269 .002
eth_mean 25.340 8.007 .309 3.165 .002
3 (Constant) 24.368 34.649 .703 .484
Maternal Attachment To .344 .122 .275 2.832 .006
eth_mean 28.284 7.906 .345 3.577 .001
mmz';$fiem' -.252 .109 -.228 -2.319 .023

 

 

 

 

 

 

a. Dependent Variable: Patamal Attachment Total

125

 

Excluded Varlablal

 

 

 

 

 

 

 

 

 

 

 

Collinearity
Partial Statistics
”Odd Beta In t Sig. Correlation Tolerance
1 Schwarz inter-parental a
conflict scale total -.171 -1.654 .102 -.179 .965
Social Desirability a
Scale Mean -.044 *2384 .702 -,O42 .811
eth_mean .309a 3.165 .002 .328 .997
2 Schwarz inter-parental b
conflict scale total -.228 -2.319 .023 -.248 .940
Social Desirability 5
Scale Mean -.033 -.301 .764 -.033 .810
3 Social Desirability c
Scale Mean .014 .133 .895 .015 .780

 

a. Pmdictors in the Model: (Constant), Maternal Attachment Total

b- Predictors in the Model: (Constant), Maternal Attachment Total, eth_mean

C- Predictors in the Model: (Constant), Maternal Attachment Total, eth_mean, Schwarz inter-parer
conflict scale total

‘1 Dependent Variable: Paternal Attachment Total

B.1.3 Step 3: Paternal attachment regressed on ethnic identity, then maternal
attachment and maternal conflict with father.

 

 

 

 

 

 

Model Summary
Adjusted Std. Error of
Model R R Square R Square the Estimate
1 .336“ .113 .103 36.99038
2 .457b .209 .190 35.15191
3 .507° .257 .230 34.25979

 

 

 

3- Predictors: (Constant), Maternal Attachment Total
b. predictors: (Constant), Maternal Attachment Total, eth_

mean

c. predictors: (Constant), Matemal Attachment Total, eth_
mean, Schwarz inter-parental conflict scale total

126

 

ANOVA‘

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sum of
Model uares df Mean uare F SQ .
T—‘lfegression 148268.133 1 14628133 10.720 .002“ ‘
Residual 1 14936.2 84 1368.288
Total 1296043 85
2 Regression 27044.816 2 13522.408 10.943 .000b
Residual 1025595 83 1235.657
Total 1296043 85
3 Regression 33358. 196 3 1 1 1 19.399 9.474 .000°
Residual 96246130 82 1 173.733
Total 1296043 85
a. Predictors: (Constant), Maternal Attachment Total
b- Predictors: (Constant), Maternal Attachment Total, eth_mean
0- Predictors: (Constant), Maternal Attachment Total, eth_mean, Schwarz
inter-parental conflict scale total
d- Dependent Variable: Paternal Attachment Total
Coefficient!
Unstandardized Standardized
Coefficients Coefficients
Model B Std. Error Beta t Sig.
1 ( slant) 86.573 25.531 3.391 .001
Maternal Attachment Totq .421 .129 .336 3.274 .002
2 (Constant) 7.710 34.779 .222 .825
Maternal Attachment To .400 .122 .320 3.269 .002
eth_mean 25.340 8.007 .309 3.165 .002
3 (Constant) 24.368 34.649 .703 .484
Maternal Attachment T .344 .122 .275 2.832 .006
eth_mean 28.284 7.906 .345 3.577 .001
mwgggem' -252 .109 -.228 -2.319 .023

 

a. Dependent Variable: Paternal Attachment Total

127

 

Excluded Varlabla‘

 

 

 

 

 

 

 

 

 

 

Collinearity
Partial Statistics

Model Beta In t Sig. Correlation Tolerance
I gd'wa'z inter'pmma' 1718 1 654 102 179 965

conflict scale total '° ' ' ° " °

eth_mean .309a 3.165 .002 .328 .997
2 Schwarz inter-parental b

conflict scale total -.228 -2.319 .023 -.248 .940

 

a. Predictors in the Model: (Constant), Maternal Attachment Total
9 Predictors in the Model: (Constant), Maternal Attachment Total, eth_mean
C- Dependent Variable: Patemal Attachment Total

B.1.4 Step 4: Ethnic identity regressed on maternal attachment, maternal
conflict with father and paternal attachment.

Model Summary

 

R Square

Adjusted
R Square

Std. Error of l
the Estimate

 

[Model I R
L1 l

 

.327a

 

.107

 

.096

.45339 |

 

a. Predictors: (Constant), Paternal Attachment Total

ANOVA"

 

Model
1

 

Sum of
Squares

(It

Mean Square

Si

 

Regression
Residual
Total

2.064
17.267
19.331

 

 

1
34
85

 

2.064 10.041

.206

 

 

.002a

 

 

a. Predictors: (Constant), Paternal Attachment Total
P- Dependent Variable: eth_mean

Coefficient!

 

Model

 

Unstandardized
Coefficients

Standardized
Coefficients

 

Std. Error

Beta t

Sig.

 

1 (Constant)

Paternal Attachment Total

2.

 

.004

599

 

.219

.001 .327

 

 

11.893
3.169

.000
.002

 

 

a. Dependant Variable: eth_mean

128

 

 

Excluded Variable!

 

Model
1

 

 

 

Collinearity
Partial Statistics
Beta In t Sig. Correlation Tolerance
rz inter-parental a
conflict scale total .234 2.264 .026 .241 .948
Maternal Attachment Totq -.063° -.573 .568 -.063 .887

 

 

 

 

 

 

 

3- Predictors in the Model: (Constant), Paternal Attachment Total

b- Dependent Variable: eth_mean

B.1.5 Step 5: Social desirability regressed on maternal attachment, paternal
attachment, and maternal conflict with father.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Model Summary
[ I Adjusted Std. Error of 1
Model R R Square R Square the Estimate
l 1 | .4358 .189 .180 .21892 ]
a. Predictors: (Constant), Maternal Attachment Total
ANOVAb
Sum of
Model Squares df Mean Square F Sig. .
1 Regression .940 1 .940 1 9.624 .0003
Residual 4.026 84 .048
Total 4.966 85
a. Predictors: (Constant), Maternal Attachment Total
b- Dependent Variable: Social Desirability Scale Mean
Coefficients
Unstandardized Standardized
Coefficients Coefficients
Model B Std. Error Beta t Sig. ‘
1 (Constant) 1.332 .151 8.815 .000
Maternal Attachment Totq -.003 .001 -.435 -4.430 .000

 

 

 

 

 

 

 

 

 

 

a. Dependent Variable: Social Desirability Scale Mean

129

Excluded Variable!

 

 

 

 

 

 

 

 

Collinearity
Partial Statistics
Model Beta In t Sig. Correlation Tolerance
_"l51 atemal Attachment To -.040* -.384 .702 -.042 .887
Schwarz inter-parental a
conflict scale total .168 1.694 .094 .183 .965

 

 

a. Predictors in the Model: (Constant), Maternal Attachment Total
b- Dependent Variable: Social Desirability Scale Mean

B.1.6 Step 6: Maternal conflict with father regressed on ethnic identity and
social desirability first, then maternal attachment and paternal attachment.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Model Summary
I I Adjusted Std. Error of
Model R Square R Square the Estimate
| 1 I .2283 .052 .041 34.58256 |
a. Predictors: (Constant), Paternal Attachment Total
ANOVA"
Sum of
Model Squares df Mean Square F Sig. 1
1 Regression 551 1.622 1 551 1 .622 4.609 .035‘
Residual 1004601 84 1 195.954
Total 1059717 85
8- Predictors: (Constant), Paternal Attachment Total
b- Dependent Variable: Schwarz inter-parental conflict scale total
Coefficlents'
Unstandardized Standardized
Coefficients Coefficients
Model 8 Std. Errgr Beta t Sig.
onstant) 96.043 16.670 5.761 .000
Paternal Attachment Total -.206 .096 -.228 -2.147 .035

 

 

 

 

 

 

 

 

 

 

a. Dependent Variable: Schwarz inter-parental conflict scale total

130

Excluded Varlabled’

 

 

 

 

 

 

 

 

l Collinearity
Partial Statistics

Model Beta In t Sig. Correlation Tolerance
1 eth_mean .248.3 2.264 .026 .241 .893
Maternal Attachment Totall -.124‘I -1.100 .275 -.120 .887

 

 

8- Predictors in the Model: (Constant), Paternal Attachment Total
b- Dependent Variable: Schwarz inter—parental conflict scale total

Figure 82 African American and Hispanic group Normal Probability Plot of the
Regression Standardized Residuals for Path Analysis 1

Normal P-P Plot of Regression Standardized Residual

Dependent Variable: sse_total

 

    
 
 
 

 

 

 

1.0
0.8-
.o
2
o.
E 0.6—
:1
0
g. 0.4- '
0.2-
0.0—, I I l i
0.0 0.2 0.4 0.6 0.8 1.0
Observed Cum Prob

131

Figure 83 African American and Hispanic Residuals Scatter plot for Path

 

 

 

 

Analysis 1
Scatterplot
Dependent Variable: sse_total
34
Tu
3
E 2_ O O
1: 0 o
3
.5 _ o o O O
E 1 8 ° 0
.3 o O O O O
I: o o 0° 0
5 o— 0 O O
° 8 °
" O O 0
g -1"' O o O O m
3’ ° o
.2— O
V I I I I
-3 .2 -1 0 1

Regression Standardized Predicted Value

132

Table B African American and Hispanic Group Path 1 Decompositions and

Calculations of Reproduced Correlations

 

Reproduced r’s Path Decompositions r

r13

r14

r15

r16

 

931
(-.435)

(D)

9 41
(.336) = .336

(0)

941945
(.336)(.327)

(I)

961 "' 941964+ 941954965 "' 931963
(407) + (. 336)(. 318) + (. 336)(. 327)(. 320) + (-. 435)(-. 331)‘ - .693

(D) (I) (I) (I)

-.435

.109

 

 

 

 

r24 = p42
= (--228) = -.228
(D)
r25 = 942954
= (-.228) (.327) = --075
( | )
r 25 = 942964 + 942954965
= (-.228)(.318) + (-.228)(.327)(.320) = -.096
( l ) ( I )
r34 = 931941
= (-.435)(.336) = '-146
(S)
f35 = 931941945
= ( -.435)(.336)(.327) = ._Q43
(3)
’36 = 963 T 931941964 T 931941945965
= (- .331) + (- ..435)( 336)(. 318) + (- 435)(. 336)(. 327)(. 320) = -.393
(D) (S) (S)
r45 = 954
= (.327) = .327
(D)
’46 = 964 T 954965 "’ 941931963 "’ 941961
= (.318) + (.327)(.320) + (.336)(-.435)(-.331) + (.336)(.407) = .608
(D) ( l ) (S) (S)
['56 p65 + p54p64

(.320) + (.327) (.318) = .424
(D) (S)

 

133

APPENDIX C

Caucasian Group Path Analysis 1

Figure C1 Caucasian Group Stepwise Multiple Regression for Path
Analysis 1

Figure 02 Caucasian group Normal Probability Plot of the Regression
Standardized Residuals for Path Analysis 1

Figure C3 Caucasian Group Residuals Scatter plot for Path Analysis 1

134

135

140

141

Figure C1 Caucasian Group Stepwise Multiple Regression for Path Analysis 1

0.1.1 Step 1: Well-being regressed on social desirability, then maternal
attachment, maternal conflict toward father and paternal attachment.

 

 

 

Model Summary
R
group =
Caucasian Adjusted Std. Error of
Model (Selected) R Square R Square the Estimate
1 .3993 .160 .150 1.94619
2 .539” .290 .274 1.79837

 

 

 

 

 

 

 

8- Predictors: (Constant), Maternal Attachment Total

9- Predictors: (Constant), Maternal Attachment Total,
Schwarz inter-parental conflict scale total

 

 

 

Allowed
Sum of

Model Squares df Mean Square F Sig. ‘
1 egression 65.437 1 65.437 17.276 .0003

Residual 344.675 91 3.788

Total 410.1 12 92
2 Regression 119.040 2 59.520 18.404 , .000b

Residual 291.073 90 3.234

Total 410.1 12 92

 

 

 

 

 

 

 

 

a. Predictors: (Constant), Maternal Attachment Total

b- Predictors: (Constant), Maternal Attachment Total, Schwarz inter-parental conflict
scale total

C- Dependent Variable: sse_total
‘1 Selecting only cases for which group = Caucasian

135

Coelflclants‘r"

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Unstandardized Standardized
Coefficients Coefficients

Model 8 Std. Error Beta t Sig. ‘
1 (Constant) 8.195 1.263 6.490 .000

Maternal Attachment Tota‘ .025 .006 .399 4.156 .000
2 (Constant) 9.348 1 .201 7.786 .000

Maternal Attachment Tota .024 .006 .385 4.330 .000

Schwarz inter-parental

conflict scale total -.020 .005 -.362 -4.071 .000

3- Dependent Variable: sse_total
9- Selecting only cases for which group = Caucasian
Excluded Variables
Collinearity
Partial Statistics

Model Beta In t Sit Correlation Tolerance
1 Social sirability Scale 8

Mean -.089 -.922 .359 -.097 .999

Schwarz inter-parental a

conflict scale total -.362 4.071 .000 -.394 .998

Paternal Attachment Totq .2728 2.861 .005 .289 .946
2 Social Desirability Scale b

Mean -.116 -1.304 .195 -.137 .993

Paternal Attachment To .173b 1.820 .072 .189 .853

 

 

 

 

 

 

 

3- Predictors in the Model: (Constant), Maternal Attachment Total

b- Predictors in the Model: (Constant), Maternal Attachment Total, Schwarz inter-parental conflict :
total

9» Dependent Variable: sse__total

C.1.2 Step 2: Well-being regressed on maternal attachment and maternal

 

 

 

conflict.
Model Summary
R
group =

Caucasian Adjusted Std. Error of
Model (Selected) R Squ_are R Square the Estimate
1 .3998 .160 .150 1.94619
2 .539b .290 .274 1.79837

 

 

 

 

 

 

 

a. Predictors: (Constant), Maternal Attachment Total
b- Predictors: (Constant), Maternal Amachment Total,

Schwarz inter-parental conflict scale total

136

ANOVA“

 

 

 

Sum of
Model Squares df Mean Square F Sig.
1 Regression 65.437 1 65.437 17.276 .000“
Residual 344.675 91 3.788
Total 410.112 92
2 Regression 119.040 2 59.520 18.404 .000b
Residual 291 .073 90 3.234
Total 410.112 92

 

 

 

 

 

 

 

 

a. Predictors: (Constant), Maternal Attachment Total

b- Predictors: (Constant), Maternal Attachment Total, Schwarz inter-parental conflict
scale total

C- Dependent Variable: sse_total
d- Selecting only cases for which group = Caucasian

 

 

 

 

 

Coel'ficient9*b
Unstandardized Standardized
Coefficients Coefficients

Model B Std. Error Beta t Sig.
1 (Constant) 8.195 1 .263 6.490 .000

Maternal Attachment Totq .025 .006 .399 4.156 .000
2 (Constant) 9.348 1.201 7.786 .000

Maternal Attachment To .024 .006 .385 4.330 .000

Schwarz inter-parental '

conflict scale total -.020 .005 -.362 -4.071 .000

 

 

 

 

 

 

 

a. Dependent Variable: sse_total
9- Selecting only cases for which group = Caucasian

 

 

 

 

 

Excluded Variables
Collinearity
Partial Statistics
Model Beta In t Sig. Correlation Tolerance
1 chwarz inter-parental a
conflict scale total -.362 -4.071 .000 -.394 .998
Paternal Attachment T .272“ 2.861 .005 .289 .946
Ms°°'a' Des'rab'w scab 4089’ -.922 .359 -.097 .999
can
2 Paternal Attachment To .173” 1.820 .072 .189 .853
Social Desirability Scale b
Mean -.116 -1.304 .195 -.137 .993

 

 

 

 

 

 

 

a. Predictors in the Model: (Constant), Maternal Attachment Total

b- Predictors in the Model: (Constant), Maternal Attachment Total, Schwarz inter-parental conflict:
total

0- Dependent Variable: sse_total

137

0.13 Step 3: Paternal attachment regressed on social desirability, then

 

 

 

 

 

 

 

 

Model Summary
R
group =
Caucasian Adjusted Std. Error of
Model (Selected) R Squ_are R Square the Estimate
1 .314‘:1 .098 .088 39.02581
2 .454b .206 .189 36.81704
3 .510° .260 .235 35.75526

 

 

a. Predictors: (Constant), Schwarz inter—parental conflict
scale total

b- Predictors: (Constant), Schwarz inter-parental conflict
scale total, Social Desirability Scale Mean

0- Predictors: (Constant), Schwarz inter-pawntal conflict
scale total, Social Desirability Scale Mean, Maternal

maternal attachment and maternal conflict with father.

 

 

 

 

Attachment Total
ANOVK‘J
Sum of
‘ Model Squares df Mean Square F Sig.

1 Regression 15111.341 1 15111.341 9.922 .0023
Residual 1385942 91 1523.014
Total 1 53705.6 92

2 Regression 31711.042 2 15855.521 11.697 .000”
Residual 1219945 90 1355.495
Total 1 53705.6 92

3 Regression 39924.51 1 3 13308.170 10.410 .000°
Residual 1 1 3781 .1 89 1278.439
Total 1537056 92

 

 

 

 

 

 

 

 

a. Predictors: (Constant), Schwarz inter-parental conflict scale total

b- Predictors: (Constant), Schwarz inter-parental conflict scale total, Social Desirability
Scale Mean

0- Predictors: (Constant), Schwarz inter-parental conflict scale total, Social Desirability
Scale Mean, Maternal Attachment Total

d- Dependent Variable: Paternal Attachment Total
9- Selecting only cases for which group = Caucasian

138

W

 

 

 

 

 

 

Unstandardized Standardized
Coefficients Coefficients
Model B Std. ErroL Beta t Slg' .
1 (Constant) 200.535 6.537 30.679 .000
Schwarz inter-parental
conflict scale total -.333 .106 -.314 -3.150 .002
2 (Constant) 226.981 9.754 23.271 .000
Schwarz inter-parental
conflict scale total -.359 .100 -.338 -3.590 .001
18:.“ Des'rabmty scab -31.003 8.859 -.330 3499 .001
an
3 (Constant) 169.017 24.753 6.828 .000
Schwarz inter-parental
conflict scale total -.350 .097 -.329 -3.598 .001
“SAW” ms'mbmty scab 31.755 8.609 -.338 -3.689 .000
can
Maternal Attachment Total .283 .1 12 .231 2.535 .013

 

 

 

 

 

 

 

a. Dependent Variable: Paternal Attachment Total
b- Selecting only cases for which group = Caucasian

 

 

 

 

Excluded Variables
Collinearity
Parfia' StatIStICS
Model _ Beta In t Sig. Correlation Tolerance
1 Maternal Attachment To .220a 2.256 .026 .231 .998
SM°°'a' Des'rab'm sca'e -330" 3499 .001 -.346 .994
can
2 Maternal Attachment Total .231b 2.535 .013 .259 .997

 

 

 

 

 

 

 

a. Predictors in the Model: (Constant), Schwarz inter-parental conflict scale total

b- Predictors in the Model: (Constant), Schwarz inter-parental conflict scale total, Social Desirabilit
Scale Mean

0- Dependent Variable: Paternal Attachment Total

C.1.4 Step 4: Maternal conflict regressed on maternal conflict with father and
social desirability.

No relationship found.
0.15 Step 5: Social desirability regressed on maternal attachment and maternal
conflict with father.

No relationship found.

139

Figure C2 Caucasian group Normal Probability Plot of the Regression
Standardized

Residuals for Path Analysis 1

Normal P-P Plot of Regresslon Standardized Residual

Dependent Variable: sse_total

 

 
 
  
   
 
    

 

 

 

1.6
0.8-
.o
2
a ,,,,,,,
E 0.6-1
:1 .....
0
g 0.4-
02-
0-0 I l T I r
0.0 0.2 0.4 0.6 0.8 1.0
Observed Cum Prob

140

Figure C3 Caucasian Group Residuals Scatter plot for Path Analysis 1

Regression Standardized Resldual

Scatterplot

Dependent Variable: sse_total

 

 

 

o 0 o
o o
o
% 00 O O
o o 0
° 0
o 8 0,, o
O O 00 O
o o o oo o
o o oo
00 (g o
o o
o
I I l r
.2 -1 0 1

Regression Standardized Predicted Value

141

 

APPENDIX D

Caucasian Group Path Analysis 2

Figure D1 Caucasian Group Stepwise Multiple Regression for
Path Analysis

Figure D2 Caucasian Group Normal Probability Plot of the Regression
Standardized Residuals of Path Analysis 2

Figure D3 Caucasian Group Residuals Scatter plot for Path Analysis 2

Table D Caucasian Group Path 2 Decompositions and Calculations of
Reproduced Correlations

Figure D4 Caucasian Group Path 2 Paternal Promotion of Independence
Figure D5 Caucasian Group Path 2 Paternal Feelings of Affect

Figure D6 Caucasian Group Path 2 Paternal Feelings of Support

142

143

153

153

154

156
156

156

Figure Dl Caucasian Group Stepwise Multiple Regression for Path Analysis 2

0.1.1 Step 1: Well-being was regressed on social desirability, then on
maternal affective quality of relationship, maternal promotion of independence,
mother as source of support, maternal conflict with father, paternal feelings of
affect, father promotion of independence, and father feelings of support.

 

 

Modal Summary
Adjusted Std. Error of
Model R R Square R Square the Estimate
1 .3773 .142 .133 1.96596
2 .553b .306 .290 1 .77892
3 .593° .352 .330 1 .72800

 

 

 

 

 

 

 

a. Predictors: (Constant), Schwarz inter-parental conflict
scale total

9. Predictors: (Constant), Schwarz inter-parental conflict
scale total, mother as source of support

0- Predictors: (Constant). Schwarz inter-parental conflict
scale total, mother as source of support, paternal

affective quality of relationahip

 

 

 

 

 

 

 

 

 

Allowr'
Sum of
Model Squares df Mean Square F Sig.
1 Regression 58.398 1 58.398 1 5.1 10 .000‘I
Residual 351 .714 91 3.865
Total 410.1 12 92
2 Regression 125.302 2 62.651 19.798 .000”
Residual 284.810 90 3.165
Total 410.1 12 92
3 Regression 144.360 3 48.120 16.1 15 .000°
Residual 265.753 89 2.986
Total 410.112 92

 

 

 

a. Predictors: (Constant), Schwarz inter-parental conflict scale total

b- Predictors: (Constant), Schwarz inter-parental conflict scale total, mother as source
of support

C- Predictors: (Constant), Schwarz inter-parental conflict scale total, mother as source
of support, paternal affective quality of relationahip

d. Dependent Variable: sse_total

143

Coetflclantt

 

 

 

 

 

Unstandardized Standardized
Coefficients Coefficients
Model B Std. Error Beta t SE. ‘
I (Co—listant) 14.381 .329 43.674 .000
Schwarz inter-parental
conflict scale total -.021 .005 -.377 -3.887 .000
2 (Constant) 10.323 .932 1 1.082 .000
Schwarz inter-parental
conflict scale total -.022 .005 -.405 -4.603 .000
mother as source of
support .073 .016 .405 4.598 .000
3 (Constant) 8.702 1 .109 7.845 .000
Schwarz inter-parental
conflict scale total -.019 .005 -.350 . -3.968 .000
mother as source of
support .068 .016 .377 4.374 .000
paternal affective
quality of relationahip .020 .008 .224 2.526 .013

 

 

 

 

 

 

 

 

a. Dependent Variable: sse_total

144

Excluded Variables

 

 

 

 

 

 

 

 

 

 

 

Collinearity

Partial Statistics

Model Beta In t Sig Correlation Tolerance
1 2333:?“ -.102‘3 -1.052 .296 -.110 .994
:uaaiimImghip 343° 3.777 .000 .370 1.000
mdfpaezfifm' .315‘1 3.359 .001 .334 .962
gm” ”um °f 405° 4.598 .000 .436 .995
mfigaéfamgzamp 2738 2.835 .006 .286 .944
might” °f .1568 1.497 .138 .156 .858
323;? ”um °f .242la 2.449 .016 .250 .913
2 3333:?“ -.131b -1495 .138 -.157 .990
rfigymé'ngmp -.080b -.413 .681 -.044 .209
31:23:13?" .128b 1.199 .234 .126 .674
:jg’mfamfimp .224b 2.526 .013 .259 .928
Efilfigflmm °f .077b .798 .427 .084 .828
2:23;? ”um °f .174” 1.888 .062 .196 .335
3 33:33:23“ -.060° -.644 .521 -.068 .856
rufigmslraemxsmp -.191° -1.000 .320 -.106 .199
mfligirfiw 106° 1.015 .313 .108 .669
mgzzgimm °f -205° -1.507 .135 -.159 .386
21%;? ”um °f -.034c -.218 .828 -.023 .305

 

a. Predictors in the Model: (Constant), Schwarz inter-parental conflict scale total

 

b- Predictors in the Model: (Constant), Schwarz inter-parental conflict scale total, mother as souro
of support

9 Predictors in the Model: (Constant), Schwarz inter-parental conflict scale total, mother as sourc
of support, paternal affective quality of relationahip

d- Dependent Variable: sse_total

145

0.12 Step 2: Paternal feelings of affect was regressed on social desirability,
then on maternal feelings of affect, maternal promotion of independence,
mother feelings of support, and maternal conflict with father.

 

 

Model Summary
Adjusted Std. Error of
Model R R Square R Sware the Estimate
1 .3263 .106 .096 22.48567
2 .419b .175 .157 21.72001
3 466° .217 .191 21.28091

 

 

 

 

 

 

 

a. Predictors: (Constant), Social Desirability Scale Mean
b- Predictors: (Constant), Social Desirability Scale Mean,
Schwarz inter-parental conflict scale total

C- Predictors: (Constant), Social Desirability Scale Mean,
Schwarz inter-parental conflict scale total, Mattemal

 

 

 

 

affective quality of relationship
ANOVA"
Sum of
(m— Squares df Mean Square F Sig.__
1 Regression 5468.836 1 5468.836 10.816 .001a
Residual 46010068 91 505.605
Total 51478.903 92
2 Regression 9020.604 2 4510.302 9.561 .000”
Residual 42458.299 90 471 .759
Total 51478.903 92
3 Regression 11172.840 3 3724.280 8.224 000°
Residual 40306.064 89 452.877
Total 51478.903 92

 

 

 

 

 

 

 

 

 

a. Predictors: (Constant), Social Desirability Scale Mean

b- Predictors: (Constant), Social Desirability Scale Mean, Schwarz inter-parental
conflict scale total

c. Predictors: (Constant), Social Desirability Scale Mean, Schwarz inter-parental
conflict scale total, Mattemal affective quality of relationship

d- Dependent Variable: paternal affective quality of relationahip

146

 

Coefficientf

 

 

 

 

 

 

 

 

 

 

 

Unstandardized Standardized
Coefficients Coefficients
Model B Std. Error Beta t Slg' . 1
1 (find 102.382 4.964 20.624 .000
Social Desirability
Scale Mean 47.746 5.396 -.326 -3.289 .001
2 (Constant) 1 1 1.109 5.754 19.309 .000
Social Desirability
Scale Mean -18.813 5.226 -.346 -3.599 .001
Schwarz inter-parental
conflict scale total -.162 .059 -.263 -2.744 .007
3 (Constant) 84.629 13.392 6.320 .000
Social Desirability
Scale Mean -18.776 5.121 -.345 -3.667 .000
Schwarz inter-parental
conflict scale total -.159 .058 -.259 -2.751 .007
Mattemal affective
quality of relationship .265 .121 .205 2.180 .032

 

a. Dependent Variable: paternal affective quality of relationahip

147

 

 

Excluded Variables

 

 

 

 

 

 

 

 

 

 

 

Collinearity

Partial Statistics

Model Beta In t SIL Correlation Tolerance
1 Was ”me °f .128’ 1.295 .199 .135 .996
mflifigfm' -.263a 2744 .007 -273 .994
fifimflmfigm .2108 2.165 .033 .222 1.000
{figfl’gnzzfj'm' °f 200" 2.045 .044 .211 .997
2 310%” ”me °f .149b 1.556 .123 .163 .990
:figymgmxmp .205b 2.180 .032 .225 999
{figgnfifc‘j'ta‘m °f .155” 1.598 .114 .167 .960
3 $3338 ”we °f -1696 -.812 .419 -.086 .203
mgnzzfcimaw °f .033c .264 .792 .028 .556

 

a. Predictors in the Model: (Constant), Social Desirability Scale Mean

b- Predictors in the Model: (Constant), Social Desirability Scale Mean, Schwarz inter-parental
conflict scale total

C~ Predictors in the Model: (Constant), Social Desirability Scale Mean, Schwarz inter-parental
conflict scale total, Mattemal affective quality of relationship

d- Dependent Variable: paternal affective quality of relationahip

D.1.3 Step 3: Paternal promotion of independence was regressed on social
desirability, then on maternal feelings of affect, maternal promotion of
independence, mother feelings of support, and maternal conflict with father.

 

 

 

 

 

 

Modal Summary
Adjusted Std. Error of
Model R R Square R Square the Estimate
1 .377“ .142 .133 9.60837
2 .4800 .230 .213 9.15155
3 520° .270 .246 8.96181

 

 

 

a. Predictors: (Constant), Schwarz inter-parental conflict
scale total

b- Predictors: (Constant), Schwarz inter-parental conflict
scale total, Mattemal affective quality of relationship

C- Predictors: (Constant), Schwarz inter-parental conflict
scale total, Mattemal affective quality of relationship,

mother as source of support

148

 

Allowilt

 

 

 

 

 

Sum of

Model Squares df Mean Square F Sig' .

1 Regression 1393.598 1 1393.598 15.095 .000‘
Residual 8401.198 91 92.321
Total 9794.796 92

2 Regression 2257.215 2 1128.607 13.476 .000b
Residual 7537.581 90 83.751
Total 9794.796 92

3 Regression 2646.850 3 882.283 10.985 .000°
Residual 7147.946 89 80.314
Total 9794.796 92

 

 

 

 

 

 

 

a. Pmdictors: (Constant), Schwarz inter-parental conflict scale total

b- Predictors: (Constant), Schwarz inter-parental conflict scale total, Mattemal affective
quality of relationship

0- Predictors: (Constant), Schwarz inter-parental conflict scale total, Mattemal affective
quality of relationship, mother as source of support

(I. Dependent Variable: tamer as facilitator of independence

 

 

 

 

 

 

 

 

 

 

 

Coefficient:
Unstandardized Standardized
Coefficients Coefficients

Model B Std. Error Beta t Sig. ‘
1 (Eonstant) 69.480 1.609 43,173 .000

Schwarz inter-parental

conflict scale total -.101 .026 -.377 -3.885 .000
2 (Constant) 52.726 5.438 9,695 .000

Schwarz inter-parental

conflict scale total -.099 .025 -.371 -4.007 .000

Mattemal affective '

quality of relationship .168 .052 .297 3.211 .002
3 (Constant) 52.258 5.329 gene .000

Schwarz inter-parental

conflict scale total “039 025 --332 -3-596 .001

Mattemal affective

quality of relationship .386 .112 .685 3.459 .001

mother as source of

support -.386 .175 -.437 -2.203 .030

 

 

a. Dependent Variable: father as facilitator of independence

149

Excluded VariablesI

 

 

 

 

 

 

 

 

 

 

 

Collinearity

Partial Statistics

‘ Model Beta In t SIL Correlation Tolerance
1 5,; (i 33?th -.207° -2.174 .032 -223 994
was ”we °f .173“ 1.802 .075 .187 .995
rm'mgip .2978‘ 3.211 .002 .321 1.000
ngicffirm .2538‘ 2.644 .010 .268 .962
2 3'33?“ -.206'D -2.278 .025 -235 .994
mm“ ”um °f -.437b 2203 .030 -227 .208
Sfixéifiigm .099b .795 .429 .084 .557
3 $133:be -.179° -1.981 .051 -207 .969
mmpiizfifm .077c .629 .531 .067 .554

 

a. Predictors in the Model: (Constant), Schwarz inter-parental conflict scale total

 

b- Predictors in the Model: (Constant), Schwarz inter-parental conflict scale total, Mattemal affecti
quality of relationship

C- Predictors in the Model: (Constant), Schwarz inter-parental conflict scale total, Mattemal affecti
quality of relationship, mother as source of support

0. Dependent Variable: father as facilitator of independence

0.14 Step 4: Father feelings of support was regressed on social desirability,
then on maternal feelings of affect, maternal promotion of independence, mother
feelings of support, and maternal conflict with father.

 

 

Model Summary
Adjusted Std. Error of
Model R R Square R Square the Estimate
1 .2973 .088 .078 14.21765
2 .446” .199 .181 13.39791
3 .516° .266 .241 12.89598

 

 

 

 

 

 

 

a. Predictors: (Constant), Social Desirability Scale Mean
b- Predictors: (Constant), Social Desirability Scale Mean,

mother as facilitator of independence
0- Predictors: (Constant), Social Desirability Scale Mean.

mother as facilitator of independence, Schwarz
inter—parental conflict scale total

150

movnt

 

 

 

 

 

 

 

 

 

Sum of
Model Squares df Mean Square F Sig. .
Wression 1776.303 1 1776.303 8.787 004°

Residual 18394.879 91 202.142
Total 20171.183 92

2 Regression 4015.825 2 2007.912 1 1 .186 .000b
Residual 16155.358 90 179.504
Total 20171.183 92

3 Regression 5369.915 3 1789.972 10.763 000°
Residual 14801268 89 166.306
Total 20171.183 92

 

 

 

a. Predictors: (Constant), Social Desirability Scale Mean
b- Predictors: (Constant), Social Desirability Scale Mean, mother as facilitator of

independence

C- Predictors: (Constant), Social Desirability Scale Mean, mother as facilitator of
independence. Schwarz inter—parental conflict scale total

d- Dependent Variable: father as source of support

 

 

 

 

 

 

 

 

 

 

 

Coefficient!
Unstandardized Standardized
Coefficients Coefficients

Model B Std. Error Beta 1 Sig. .
1 (Constant) 54.354 3.139 17.316 .000

Social Desirability

Scale Mean -10.114 3.412 -.297 -2.964 .004
2 (Constant) 24.568 8.937 2.749 .007

Social Desirability

Scale Mean -10.727 3.220 -.315 -3.332 .001

mother as facilitator of

independence .462 .131 .334 3.532 .001
3 (Constant) 34.580 9.290 3.722 .000

Social Desirability

Scale Mean -11.305 3.106 -.332 -3.640 .000

mother as facilitator of

independence .392 .128 .283 3.055 .003

Schwarz inter-parental

conflict scale total -.102 .036 -.265 -2.853 .005

 

a. Dependent Variable: father as source of support

151

 

Excluded Variables

 

 

 

 

 

 

 

 

 

 

 

Collinearity

Partial Statistics

Model Beta In t Sig. Correlation Tolerance
1 $333338 some °f .166a 1.666 .099 .173 .996
imfiggfi';§’,;j°"°' -.319° -3352 .001 -333 .994
rfigmflraemxs‘fmp .180‘3 1.817 .073 .188 1.000
{fljfpjnfifc‘j'ta‘m °f .334a 3.532 .001 .349 .997
2 $323338 sour” °f -.011° -.102 .919 -011 126
fimflfitggfim -265b 2853 .005 -.290 .958
:‘uifigmg'gmgmp -.058b -.468 .641 -050 .589
3 2132333 source of .0540 .494 .622 .053 .695
Zglifymg'jfigfimp -.013° -.108 .914 -012 .578

 

3. Predictors in the Model: (Constant), Social Desirability Scale Mean

b- Predictors in the Model: (Constant), Social Desirability Scale Mean, mother as facilitator of
independence

0 Predictors in the Model: (Constant), Social Desirability Scale Mean, mother as facilitator of
independence, Schwarz inter-parental conflict scale total

d- Dependent Variable: father as source of support

015 Step 5: Maternal conflict with father was regressed on social desirability,

then the three maternal subscales.

No significant relations were found.

152

 

Figure 02 Caucasian Group Normal Probability Plot of the Regression
Standardized Residuals of Path Analysis 2

Normal P-P Plot of Regresslon Standardized Residual

Dependent Variable: m_tobel
1 .C

 

0.8-

0.6—

0.4 —

ExpectadCumProb

0.2-

 

 

 

°-° l l l T
0.0 0.2 0.4 0.6 0.8 1 .0

Observed Cum Prob

Figure D3 Caucasian Group Residuals Scatter plot for Path Analysis 2

Scatterplot

Dependent Variable: sse_total

 

 

 

 

2nd
3 o
g o o o o o 0 ° 0
14 o
t: o o 0 o
E d3 0 ° o
o
2 0 <5 0 0 °
2 °‘ °
5 o 0 9 c
at o o o o
r: o
o
s 9 ° 0
K
.2— O O O
I r I I I
'3 -2 -1 o 1

Regression Standardized Predicted Value

153

Table D Caucasian Group Path 2 Decompositions and Calculations of
Reproduced Correlations

Reproduced r’s Path Decompositions

r16

r17

r18

’19

981
(-.259)

(0)

991 + Papas
(-.350) + (-.259)(.224)

(D) (|)

page 1 of 2

r

= -.259

= -.332

= -.265

 

r 27

f 28

P62
(.205)

(0)

P72 "' r 25 P75
(.685) + (.885)(-.437)

( D) (U)

r23 Pas
(.640) (.283)

(U)

P62P96 + r25 P95
(.205)(.224) + (.885)(.377)

(I) (U)

.181

 

'37

T23P62
(.640)(.205)

(U)

rzapm r23P72
(.640)(.685) + (.522)(-.437)
(U) (U)

Pas
(.283)

(D)
raspes + T32P62P96

(.522)(.377)+ (.640)(.205)(.224)
( U ) (U)

154

= .131

= .211

 

 

 

 

 

Ta page 2 of 2
’46 = P64
= (-.345) = --345
(D)
r48 = PM
= (-.332) = -.332
(D)
r49
= P64P96
= (-.345)(.224) = --077
( l )
'56 = ''25 P62
= (.885)(.205) = .181
(U)
r 57 = P75 + r2:5P72
= (-.437) + (.885)(.685) = .169
(D) (U)
r58 : r53p83
= (.522)(.283) = .148
(U)
r59 = P95 + r52P62I396
= (.377) + (.885)(.205)+ (.224) = .378
(D) ( U)
’69 = 996 + 961p91+961951 995
= (.224) + (-.259) (-.350) + (-.259) (-.246)(.377) = .339
(D) (S) (S)
r 79 = 975995 + P71 P91 + P71P61996 + P72P62P96 + P75r25962996 + P72 r 25995
= (.-.437)(.377) + (-.332)(-.350) + (-.332)(-.259)(.224) *-
(.685)(.205)(.224)
(S) (S) (S) (S)
+ (-.437)(.885)(.205)(.224) + (.685)(.885)(.377) 3.230
(3) (S)
r 89 = P81P91+ PMPMPss + P81P61P96 + Paaraspes + P83r23P62P96

(-.265)(-.350) + (-.332)(-345)(224) + (-265)(-.259)(.224) +

(S) (S) (S

(283)(.522)(.377)+ (.283)(.640)(.205)(.224) -- .280
(3) (S)

155

Figure D4 Caucasian Group Path 2 Patemai Promotion of independence

 

 

 

M/SUP ..
@—fl Well-

5 being

 

 

 

9

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure D5 Caucasian Group Path 2 Patemai Feelings of Affect

 

M/CON

 

 

a Weil-

bemg

WAFF P/AFF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8WD @

 

 

 

Figure DB Caucasian Group Path 2 Paternal Feelings of Support

M/ ON

1 bemg

Soc/D P/ SUP
4 8

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

156

APPENDIX E

African American and Hispanic Group Path Analysis 2

Figure E1 African American and Hispanic Group Stepwise Multiple
Regression for Path Analysis 2

Figure E2 African American and Hispanic group Normal Probability Plot
of the Regression Standardized Residuals for Path Analysis 1

Figure E3 African American and Hispanic Residuals Scatter plot for
Path Analysis 1

Table E African American and Hispanic Group Path 2 Decompositions
and Calculations of Reproduced Correlations
Figure E4 African American and Hispanic Group Path 2 Patemai

Promotion of Independence

Figure E5 African American and Hispanic Group Path 2 Patemai
Feelings of Affect

Figure E6 African American and Hispanic Group Path 2 Patemai
Feelings of Support

157

158

178

179

180

184

184

185

Figure E1 African American and Hispanic Group Stepwise Multiple Regression
for Path Analysis 2

E11 Step 1: Well-being was regressed on ethnic identity and social
desirability, then on matemai feelings of affect, matemai promotion of
independence, mother feelings of support, matemai conflict with father, patemai
affective quality of relationship, paternal promotion of independence, and father
feelings of support. '

 

 

 

 

Model Summary
Change Statistics
Adjusted Std. Error of R Square
Model R R Sfluare R Sguare the Estimate Change—I F Change df1 df2 Sig. F Change
1 .543‘ .294 .286 2.26300 .294 35.059 1 84 .000
2 .711” .506 .494 1.90516 .211 35.518 1 83 .000
3 .821c .675 .663 1.55501 .169 42.587 1 82 .000
4 .864“ .746 .733 1 .38264 .071 22.720 1 81 .000
5 .880" .775 .761 1.31009 .029 10.220 1 80 .002
6 .690' .792 .777 1.26596 .018 6.675 1 79 .012

 

 

 

 

 

 

 

 

 

 

 

 

8- Predictors: (Constant). Social Desirability Scale Mean
b- Predictors: (Constant), Social Desirability Scale Mean, eth_mean
0- Predictors: (Constant), Social Desirability Scale Mean, eth_mean, Mattemal affective quality of relationship

d- Predictors: (Constant), Social Desirability Scale Mean, eth_mean, Mattemal affective quality of relationship, fathe
of independence

6- Predictors: (Constant), Social Desirability Scale Mean, eth_mean, Mattemal affective quality of reiationsh'p, fathe
of independence, father as source of support

f. Predictors: (Constant), Social Desirability Scale Mean, eth_mean, Mattemal affective quality of relationship, father
of independence, father as source of support, patemai affective quality of reiationah'o

158

ANOVA‘I

 

 

 

 

 

 

 

 

 

 

 

 

 

Sum of

Model Squares df Mean Square F Sig.

1 Regression 179.541 1 179.541 35.059 .000“
Residual 430.178 84 5.121
Total 609.719 85

2 Regression 308.459 2 154.229 42.492 .000b
Residual 301 .260 83 3.630
Total 609.719 85

3 Regression 411.438 3 137.146 56.717 .000c
Residual 198.281 82 2.418
Total 609.719 85

4 Regression 454.871 4 1 13.718 59.485 .000d
Residual 154.848 81 1.912
Total 609.719 85

5 Regression 472.412 5 94.482 55.049 000°
Residual 137.307 80 1.716
Total 609.719 85

6 Regression 483.109 6 80.518 50.240 .000f
Residual 126.610 79 1.603
Total 609.719 85

 

a. Predictors: (Constant), Social Desirability Scale Mean

b- Predictors: (Constant), Social Desirability Scale Mean, eth_mean

G. Predictors: (Constant), Social Desirability Scale Mean, eth_mean, Mattemal
affective quality of relationship

d- Predictors: (Constant), Social Desirability Scale Mean, eth_mean, Mattemal

affective quality of relationship, father as facilitator of independence

e. Predictors: (Constant), Social Desirability Scale Mean, eth_mean, Mattemal

affective quality of relationship, father as facilitator of independence, father as
source of support

f- Predictors: (Constant), Social Desirability Scale Mean, eth_mean, Mattemal affective

quality of relationship, father as facilitator of independence, father as source of
support, patemai affective quality of relationahip

9- Dependent Variable: sse_total

159

 

Coefficient!

 

 

 

 

 

 

 

 

 

 

Unstandardized Standardized
Coefficients Coefficients
Model 8 Std. Error Beta t 8L
1 (Constant) 17.124 .724 23.665 .000
gz'gzzamm’ -6.013 1.015 -.543 -5.921 .000
2 (Constant) 8.475 1 .574 5.385 .000
ng'fii‘mwi‘y -5.742 .656 -.518 -6.707 .000
eth_mean 2.586 .434 .460 5.960 .000
3 (Constant) .120 1.814 .066 .947
ggzlgzz'iwrability -3.166 .603 -.286 -3945 .000
eth_mean 2.298 .357 .409 6.438 .000
2'33: m3” .060 .012 .476 6.526 .000
4 (Constant) -.553 1.619 -.341 .734
ggz'figmww -3.005 .714 -.271 4.207 .000
eth_mean 1.952 .326 .348 5.996 .000
:31??er mghip .061 .012 .363 5.256 .000
gaggimm“ °f .057 .012 .303 4.767 .000
5 (Constant) -.659 1.534 -.430 .669
ggzlgzi‘rability -3.242 .661 -.293 -4.761 .000
eth_mean 1.819 .311 .324 5.846 .000
:fift’yms'mgip .061 .01 1 .366 5.569 .000
Eamflmm °f .044 .012 .231 3.604 .001
21%;? ”um °f .036 .011 .167 3.197 .002
6 (Constant) .155 1.516 .102 .919
38:33::be -4.138 .744 -.373 -5.563 .000
eth_mean 2.040 .313 .363 6.525 .000
23;": m§hip .059 .01 1 .354 5.566 .000
mmm" °f .056 .013 .295 4.418 .000
ms ”um °f .093 .024 .481 3.763 .000
fifigagfmgsamp -.046 .016 -.369 -2.584 .012

 

 

 

 

 

 

 

a. Dependent Variable: sse_total

160

Excluded Variables 9

 

 

 

 

 

conflict scale total

 

 

 

Model Beta In t Sig_.__
1 eth_mean .4603 5.960 .000
33:31:: Mimi, .534El 6.051 .000
mmtlm °f .405a 4.650 .000
m; as source of .3103 3.288 .001
fifigflfmfiahm .403‘3 4.813 .000
ngfltammf .5108 6.639 .000
:12; rats source of .3908 4.772 .000
ingfirxama' .1253 1.327 .166
2 rfgisym5'3aefxnp 476° 6.526 .000
{fififlnfiffi'm Of .399b 6.023 .000
31:63:; as source of 355D 4.732 .000
gfltaigaérmzfiamp -279b 3541 .001
$23,197,233?" of .417b 6.060 .000
2163;? source of .304b 4217 .000
afixfxzeml .0460 .566 .558
3 fifig‘pifnzsefi'fim' of 234° 3.099 .003
210$; as source of -.0560 -.453 .651
33273341317334". 226C 3-482 .001
finimanimmf 303° 4.767 .000
films source of .260c 4430 .000
Schwarz inter-parental .040c .604 .547

 

 

161

E.1.2 Step 2: Ran same stepwise multiple regression as step 1, excluding
mother as source of support, matemai promotion of independence, and matemai
conflict with father.

R2 and Beta’s of significant variables was unchanged.
E.1.3 Step 3: Father feelings of support was regressed on ethnic identity and

social desirability, then on matemai feelings of affect, matemai promotion of
independence, mother feelings of support, and matemai conflict with father.

 

 

Model Summary
Adjusted Std. Error of
Model R R Square R Square the Estimate
1 .2163 .047 .035 13.66470
2 .365b .133 .112 13.11028
3 .437° .191 .162 12.73888

 

 

 

 

 

 

 

a. Predictors: (Constant), eth_mean

b- Predictors: (Constant), eth_mean, mother as facilitator
of independence

0- Predictors: (Constant), eth_mean, mother as facilitator
of independence, Schwarz inter-parental conflict scale

 

 

 

 

total
ANOVA“
Sum of
Model Squares df Mean Square F Sig.
flagrasion 769.557 1 769.557 4.121 .0463
Residual 1 5684.827 84 186.724
Total 16454384 85
2 Regression 2188.385 2 1094.193 6.366 .003b
Residual 14265.999 83 1 71 .880
Total 16454.384 85
3 Regression 3147.506 3 1049.169 6.465 .001°
Residual 1 3306.877 82 162.279
Total 16454.384 85

 

 

 

 

 

 

 

 

8- Predictors: (Constant), eth_mean
b- Predictors: (Constant). eth_mean, mother as facilitator of independence

0- Predictors: (Constant), eth_mean, mother as facilitator of independence, Schwarz
inter-parental conflict scale total

d- DependentVariable: father as source of support

162

Coefficients"

 

 

 

 

 

 

 

 

 

 

 

Unstandardized Standardized
Coefficients Coefficients
Model 8 Std. Error Beta 1 Sig.
1 (Constant) 20.516 10.282 1.995 .049
eth_mean 6.309 3.108 .216 2.030 .046
2 (Constant) -1.052 12.396 -.085 .933
eth_mean 6.109 2.983 .209 2.048 .044
mother as facilitator of
independence .351 .122 .294 2.873 .005
3 (Constant) 4.627 12.270 .377 .707
eth_mean 7.205 2.933 .247 2.457 .016
mother as facilitator of
i I I noe .299 .121 .250 2.479 .015
Schwarz inter-parental
conflict scale total -.098 .040 -.248 -2.431 .017

 

 

a. Dependent Variable: father as source of support

163

Excluded Variabled‘

 

 

 

 

 

 

 

 

 

 

 

Collinearity

Partial Statistics

Model Beta In t Sig. Correlation Tolerance
1 332'3§§;“”"‘” .0343 .314 .754 .034 .997
2:333! gmgmp .060“ .745 .459 .031 .932
:mgmgmm °f .2943 2.673 .005 .301 .999
21,13” ”we °f .0453 .418 .677 .046 .997
30m“,;';‘,§';§;'f"‘a' 42938 -2.830 .006 -.297 .976
2 Efiz'figjflab‘m’ .104b .992 .324 .109 .943
glam: gfflsenrp -153" -1.188 .236 -.130 .626
2:11:23: as source of -.0506 -.462 .645 ‘051 .903
ingfir;$?m' -.248b -2.431 .017 -.259 .948
3 gfifi'figzgmbm‘y 166° 1.621 .109 .177 .901
21:1? 3: gimgmp -.154° -1 .234 .221 -.136 .623
$22; as source of 401:: -.948 .346 -.105 .872

 

8- Predictors in the Model: (Constant), eth_mean
b- Predictors in the Model: (Constant), eth_mean, mother as facilitator of independence

0- Predictors in the Model: (Constant), eth_mean, mother as facilitator of independence, Schwarz
inter-parental conflict scale total

d. DependentVariable: fatheras source ofsupport

164

 

E.1.4 Step 4: Patemai promotion of independence was regressed on ethnic
identity and social desirability, then on matemai feelings of affect, matemai
promotion of independence, mother feelings of support, and matemai conflict

 

 

with father.
Model Summary
Adjusted Std. Error of
Model R R Square R Square the Estimate
1 .2568 .066 .054 13.75966
2 .3448 .116 .097 13.44464
3 .4958 .245 .216 12.51636

 

 

 

 

 

 

 

3- Predictors: (Constant), eth_mean
b- Predictors: (Constant), eth_mean, Social Desirability
Scale Mean

C- Predictors: (Constant), eth_mean, Social Desirability
Scale Mean, mother as facilitator of independence

 

 

 

 

 

 

 

 

 

 

Allovrr‘
Sum of
Model Squares df Mean Square F Sig. ‘
1 Regression 1114.913 1 1114.913 5.889 .0178
Residual 15904.075 84 189.334
Total 17018.988 85
2 Regression 2016.044 2 1006.022 5.577 .0058
Residual 15002.944 83 180.758
Total 17018.988 85
3 Regression 4172.897 3 1390.966 8.879 .0000
Residual 12846.091 82 156.660
Total 17018.988 85

 

a. Predictors: (Constant). eth_mean

'1 Predictors: (Constant), eth_mean, Social Desirability Scale Mean
0 Predictors: (Constant), eth_mean, Social Desirability Sale Mean, mother as

facilitator of independence
4 Dependent Variable: father as facilitator of independence

165

 

Coefficienta‘

 

 

 

 

 

 

Unstandardized Standardized
Coefficients Coefficients
Model B Std. Error Beta t Sig.
1 (meant) 36.146 10.354 3.491 .001
eth_mean 7.594 3.130 .256 2.427 .017
2 (Constant) 46.382 11.107 4.176 .000
eth_mean 7.232 3.062 .244 2.362 .021
Social Desirability
Scale Mean -13.489 6.041 -.230 -2.233 .028
3 (Constant) 15.501 13.273 1.168 .246
eth_mean 7.106 2.851 .239 2.492 .015
Social Desirability
Scale Mean -8.735 5.768 -.149 -1.514 .134
mother as facilitator
of independence .443 .120 .365 3.710 .000

 

 

 

 

 

 

 

a. Dependent Variable: father as facilitator of independence

166

Excluded Variables"

 

 

 

 

 

 

 

 

 

 

 

Collinearity

Partial Statistics

1 Model Beta in t Sig._ Correlation Tolerance
1 sz'maxabiw -.230° -2233 .026 -.238 .997
2:66;? gmgmp 396° 4.066 .000 .406 932
mgrg'w“ °f .396. 4.119 .000 .412 .999
$923,” °°°r°° °f .247‘I 2.402 .019 .255 .997
migfimm' -.042° -.396 .693 -.043 .976
2 m“? :vsfixhip 374° 3.312 .001 .343 .745
:L‘geflzsefg'm' °f 365° 3.710 .000 .379 .950
33$,“ °°°°°° °° .164° 1.656 .102 .160 .839
3,122,333?“ 016° .164 .670 .018 .915
3 3:61:96": 323:” 203° 1.499 .136 .164 .495
2:12:12; as source of 095° .881 .381 .097 .789
iffiflfgfl“ 067° .656 .513 .073 .900

 

a. Predictors in the Model: (Constant), eth_mean

b- Predictors in the Model: (Constant), eth_mean, Social Desirability Scale Mean

0 Predictors in the Model: (Constant), eth_mean, Social Desirability Scale Mean, mother as

facilitator of independence

d- Dependent Variable: father as facilitator of independence

E.1.5 Step 5: Repeated step 4 excluding Social desirability.

 

 

 

Model Summary
Adjusted Std. Error of
Model R R Square R Square the Estimate
1 .2568 .066 .054 13.75988
2 .4738 .224 .205 12.61346
3 .511c .262 .235 12.38002

 

 

 

 

 

 

a. Predictors: (Constant), eth_mean
b- Predictors: (Constant), eth_mean, mother as facilitator

of independence

C- Predictors: (Constant), eth_mean, mother as facilitator
of independence, Mattemal affective quality of

relationship

167

 

ANOVA‘

 

 

 

 

 

 

 

 

 

 

Sum of
Model Squares df Mean Square F Sig. ‘
1 Regression 1114.913 1 1114.913 5.889 .0173
Residual 15904075 84 189.334
Total 17018.988 85
2 Regression 3813.692 2 1906.846 11.985 .000b
Residual 1 3205.297 83 1 59.1 00
Total 17018.988 85
3 Regression 4451.275 3 1483.758 9.681 .000c
Residual 12567.713 82 153.265
Total 17018.988 85

 

 

a. Predictors: (Constant), eth_mean

b. Predictors: (Constant), eth_mean, mother as facilitator of independence

0- Predictors: (Constant), eth_mean, mother as facilitator of independence, Mattemal

affective quality of relationship

d- Dependent Variable: tamer as facilitator of independence

 

 

 

 

 

 

 

 

 

 

Coefficient!I
Unstandardized Standardized
Coefficients Coefficients
Model B Std. Error Beta t Sig.
1 (Constant) 36.146 10.354 3.491 .001
eth_mean 7.594 3.130 .256 2.427 .017
2 (Constant) 6.400 1 1 .927 .537 .593
eth_mean 7.318 2.870 .247 2.550 .013
mother as facilitator
of independence .484 .117 .398 4.119 .000
3 (Constant) -.081 12.129 -.007 .995
eth_mean 6.450 2.848 .217 2.264 .026
mother as facilitator
of independence .307 .144 .253 2.130 .036
Mattemal affective
quality of relationship .215 .106 .244 2.040 .045

 

 

 

a. Dependent Variable: father as facilitator of independence

168

Excluded Variables

 

 

 

 

 

 

 

 

 

 

 

Collinearity

Partial Statistics

Model M Beta In a t Sig. Correlation Tolerance
1 ffi;’";';,afio;§mp .396 4.066 .000 .406 .982
mgg'mm °f 396° 4.119 .000 .412 .999
52:11:,” °°°°°° °f 247° 2.402 .019 .255 .997
imgiggfi'g‘m' -.042° -.396 .693 -.043 373
2 rufigmslglmfimp 244° 2.040 .045 .220 .628
Sum” °°°°°° °° .138b 1.360 .178 .146 .903
ifflfig'figfi'ggfma' 030° .300 .765 .033 .948
3 $33333 °°m°° °f -.203c -.938 .351 -.104 .192
ifi;‘§§2§§'gfi;§°"‘a' 031° .314 .755 .035 .946

 

a. Predictors in the Model: (Constant), eth_mean
b- Predictors in the Model: (Constant), eth_mean, mother as facilitator of independence

C- Predictors in the Model: (Constant), eth_mean, mother as facilitator of independence, Mattemal
affective quality of relationship

d- Dependent Variable: father as facilitator of independence

E.1.6 Step 6: Patemai affective quality of relationship was regressed on

ethnic identity and social desirability, then on matemai feelings of affect, matemai

promotion of independence, mother feelings of support, and maternal conflict

 

 

with father.
Model Summary
Adjusted Std. Error of
Model R R Square R Smiare the Estimate
1 .330‘3 .109 .098 19.75561
2 .3948 .156 .135 19.34641
3 .496c .246 .218 18.39827

 

 

 

 

 

 

 

a. Predictors: (Constant), eth_mean
b- Predictors: (Constant), eth_mean, Social Desirability
Scale Mean

0- Predictors: (Constant), eth_mean, Social Desirability
Scale Mean, Schwarz inter-parental conflict scale total

169

 

ANOVR’

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sum of
Model Squares df Mean Square F Sig.
1 egression 4005.472 1 4005.472 10.263 .0021I
Residual 32783.87? 84 390.284
Total 36789.349 85
2 Regression 5723.605 2 2661.903 7.646 .0018
Residual 31 065.544 83 374.284
Total 36789.349 85
3 Regression 9032.663 3 3010.888 8.895 .000c
Residual 27756.686 82 338.496
Total 36789.349 85
a. Predictors: (Constant), eth_mean
b- Predictors: (Constant), eth_mean, Social Desirability Scale Mean
0- Predictors: (Constant), eth_mean, Social Desirability Scale Mean, Schwarz
inter-parental conflict scale total
d- Dependent Variable: patemai affective quality of relationahip
Coefficient!
Unstandardized Standardized
Coefficients Coefficients
Model B Std. Error Beta 1 Sig._—
‘1'_—_(Constant) 33.102 14.665 2.227 .029
eth_mean 14.394 4.493 .330 3.204 .002
2 (Constant) 47.237 1 5.982 2.956 .004
eth_mean 13.894 4.406 .318 3.153 .002
ggz'gzjgab'my -18.627 6.694 -.216 -2.143 .035
3 (Constant) 46.785 15.200 3.078 .003
eth_mean 16.092 4.249 .369 3.787 .000
gm'gzzgabmw -11.843 6.547 -.138 -1.386 .170
imgigfirgggem' -.185 .059 -.313 -3.127 .002

 

 

a. Dependent Variable: patemai affective quality of relationahip

170

Excluded Variables

 

 

 

 

 

 

 

 

 

 

 

Collinearity

Partial Statistics

Model Beta In t Sig. Correlation Tolerance
1 $03,333:?“ -.216° -2.143 .035 -.229 .997
rfim'mEhip 209° 2.051 .043 .220 .982
fimgfijmm °° 263° 2.646 .010 .279 .999
m” °°°°°° °° .161° 1.579 .116 .171 .997
33:22:35?“ -.349° -3.576 .001 -655 .973
2 f;?§;"§';f;f§;§mp .135° 1.153 .252 .126 .745
{flgfp‘fnzifi'm' °° 226° 2.241 .026 .240 .950
31%;” °°°°°° °° 069° .611 .420 .069 .639
mmzrgfifm' -.313° -3.127 .002 -.326 .915
3 ruztliiymtfrlrflfaiichsenip “1° 1-271 207 .140 .744
$$§22£fl°°°°° °° .190° 1.949 .055 .212 .935
$3,” °°°°°° °° 043° .405 .666 .045 .621

 

a. Predictors in the Model: (Constant), eth_mean
b- Predictors in the Model: (Constant), eth_mean, Social Desirability Scale Mean

0- Predictors in the Model: (Constant), eth_mean, Social Desirability Scale Mean, Schwarz
inter-parental conflict scale total

‘1 Dependent Variable: paternal affective quality of relationahip

171

 

E.1.7 Step 7: Ran step 6 excluding social desirability.

 

 

Model Summary
Adjusted Std. Error of
Model R R Square R Sgare the Estimate
1 .330‘3 .109 .098 19.75561
2 .477b .228 .209 18.49994
3 520° .270 .243 18.09697

 

 

 

 

 

 

 

a. Predictors: (Constant), eth_mean

b- Predictors: (Constant), eth_mean, Schwarz
inter-parental conflict scale total

c. Predictors: (Constant), eth_mean, Schwarz
inter-parental conflict scale total, mother as facilitator

 

 

 

 

of independence
ANOVA“
Sum of
Model Squares df Mean Square F Sig. ‘
1 Regression 4005.472 1 4005.472 10.263 .002a
Residual 32783.877 84 390.284
Total 36789.349 85
2 Regression 8382.797 2 4191 .398 12.247 .000b
Residual 28406.552 83 342.248
Total 36789.349 85
3 Regression 9934.309 3 3311.436 10.111 000’
Residual 26855.039 82 327.500
Total 36789.349 85

 

 

 

 

 

 

 

 

a. Predictors: (Constant), eth_mean
b- Predictors: (Constant), eth_mean, Schwarz inter-parental conflict scale total

C- Predictors: (Constant), eth_mean, Schwarz inter-parental conflict scale total, mother
as facilitator of independence

(1 Dependent Variable: patemai affective quality of relationahip

172

Coefficbnta‘

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Unstandardized Standardized
Coefficients Coefficients
Model B Std. Error Beta t Si .
1 (Constant) 33.102 14.865 2.227 .029
eth_mean 14.394 4.493 .330 3.204 .002
2 (Constant) 38.327 13.997 2.738 .008
eth_mean 16.637 4.254 .381 3.911 .000
Schwarz inter-parental
flict scale total -.205 .057 -.349 -3.576 .001
3 (Constant) 14.849 17.430 .852 .397
eth_mean 16.184 4.167 .371 3.884 .000
Schwarz inter-parental
conflict scale total -.183 .057 -.311 -3.212 .002
mother as facilitator of
independence .373 .171 .209 2.177 .032
3- Dependent Variable: patemai affective quality of relationahip
Excluded Variabied‘
Collinearity
Model r Beta in t Sig. Correlation Tolerance
1 Mattemal affective a
quality of relationship .209 2.051 .043 .220 .982
mother as facilitator of a
independence .263 2.646 .010 .279 .999
mother as source of a
support .161 1.579 .118 .171 .997
Schwarz inter-parental a
conflict scale total -.349 -3.576 .001 -.365 .978
2 Mattemal affective b
quality of relation ship .173 1.792 .077 .194 .970
mother as facilitator of b
independence .209 2.177 .032 .234 .968
mother as source of b
support .087 .874 .385 .096 .944
3 Mattemal affective c
quality of relationship .076 .637 .526 .071 .628
mother as source of c
support .031 .301 .764 .033 .872

 

 

 

 

 

 

 

a. Predictors in the Model: (Constant), eth_mean
b- Predictors in the Model: (Constant), eth_mean, Schwarz inter-parental conflict sale total
C- Predictors in the Model: (Constant), eth_mean, Schwarz inter-parental conflict scale total, mother

as facilitator of independence

‘1 Dependent Variable: paternal affective quality of relationahip

173

 

 

E.1.8 Step 8: Ethnic identity was regressed on social desirability, then on
matemai affective quality of relationship, maternal promotion of independence,
mother as source of support, maternal conflict with father, patemai feelings of
affect, father promotion of independence, and father feelings of support.

 

 

 

 

 

 

Model Summary
Adjusted Std. Error of
Model R R Square R Square the Estimate
1 .3303 .109 .098 .45286
2 .417b .174 .154 .43663

 

 

 

a. Predictors: (Constant), patemai affective quality of
relationahip

b- Predictors: (Constant), patemai affective quality of
relationahip, Schwarz inter-parental conflict scale total

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ANOVR=
Sum of
Model Stuares df Mean Square F Si .
1 Regression 2.105 1 2.105 10.263 .002a
Residual 17.227 84 .205
Total 19.331 85
2 Regression 3.363 2 1.661 6.739 .0008
Residual 15.969 83 .192
Total 19.331 85
a. Predictors: (Constant), patemai affective quality of relationahip
b. Predictors: (Constant), patemai affective quality of relationahip, Schwarz
inter-parental conflict scale total
C- Dependent Variable: eth_mean
Coefficient!
Unstandardized Standardized
Coefficients Coefficients
Model B Std. Error Beta t Sig. ‘
1 (Constant) 2.667 .196 13.635 .000
patemai affective
quality of rel ati on ahip .008 .002 .330 3.204 .002
2 (Constant) 2.303 .237 9.721 .000
patemai affective
quality of relati on ahip .009 .002 .408 3.911 .000
Schwarz inter-parental
co nfli ct scale total .004 .001 .267 2.557 .012

 

 

 

 

 

 

 

 

3- Dependent Variable: eth_mean

174

Excluded Vel'leblei

 

 

 

 

 

 

 

 

 

 

Collinearity

pamaj Statistics

Beta In t Sig. Correlation Tolerance
Scaz'MeaL'a my 025° .236 .612 .026 .946
fifigms'rmxgp 055° .512 .610 .056 .936
fixefenfacim" °f -071° -.663 .509 -073 .927
3:32” ”me °° -.106° -1023 .309 -.112 .960
:megzcigfggfe'fma' 267° 2.557 .012 .270 .914
mic???“ °° .109° .690 .376 .097 .706
$2322” ”um °° -.319° -1.495 .139 -.162 .230
332:: 3221301013) -025° -.236 .814 -.026 .912
mfig'm3m 059° .569 .571 .063 .937
mnzg'w" °° -045° -.430 .668 -047 .917
mg” ”me °f -056° -.543 .588 -.060 .939
:g‘eifiginam °f 053° .433 .666 .046 .660
$23228 °°m°° °° -.321° -1557 .123 -.170 .230

 

 

a. Predictors in the Model: (Constant), patemai affective quality of relationahip

b. Predictors in the Model: (Constant), patemai affective quality of relationahip, Schwarz
inter-parental conflict scale total

0- Dependent Variable: eth_mean

E.1.9 Step 9: Social desirability was regressed on ethnic identity, then on
matemai affective quality of relationship, matemai promotion of independence,
mother as source of support, matemai conflict with father, patemai feelings of
affect, father promotion of independence, and father feelings of support.

175

Model Summary

 

 

 

 

 

 

Adjusted Std. Error of
Model R R Square R Square the Estimate
1 .494£‘ .244 .235 .21 143
2 .533” .284 .267 .20698

 

 

 

a. Predictors: (Constant), Mattemal affective quality of
relationship

*1 Predictors: (Constant), Mattemal affective quality of
relationship, Schwarz inter-parental conflict scale total

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ANOVAc
Sum of
Model Squares df Mean Square F Sig.
1 Regression 1.211 1 1.211 27.093 .0003
Residual 3.755 84 .045
Total 4.966 85
2 Regression 1.410 2 .705 16.460 .0008
Residual 3.556 83 .043
Total 4.966 85
a. Predictors: (Constant), Mattemal affective quality of relationship
b- Predictors: (Constant), Mattemal affective quality of relationship, Schwarz
inter-parental conflict scale total
0- Dependent Variable: Social Desirability Scale Mean
Coefficients|
Unstandardized Standardized
Coefficients Coefficients
Model B Std. Error Beta t S' .
1 (Constant) 1 .379 .1 38 9.995 .000
Mattemal affective
quality of relationship -.007 .001 -.494 -5.205 .000
2 (Constant) 1 .270 . 144 8.797 .000
Mattemal affective
quality of relationship -.007 .001 -.476 -5.107 .000
Schwarz inter-parental
conflict scale total .001 .001 .201 2.156 .034

 

 

 

 

 

 

 

a. Dependent Variable: Social Desirability Scale Mean

176

 

Excluded VariabieG

 

 

 

 

 

support

 

 

 

 

 

Collinearity
Partial Statistics

Model Beta In t Sig. Correlation Tolerance
1 eth_mean .0133 .137 .891 .015 .982

mother as facilitator of a

inl l .113 .953 .344 .104 .642

mother as source of a

support .077 .437 .663 .048 .290

Schwarz inter-parental a

conflict scale total .201 2.156 .034 .230 .992

patemai affective a

quality of rel ati on ahip -.117 -1.201 .233 -.131 .938

father as facilitator of a

independence -.041 -.390 .698 -.O43 .820

father as source of a

support .076 .797 .428 .087 .988
2 eth_mean -.020b -.208 .836 -.O23 .957

mother as facilitator of b

independence .154 1.316 .192 .144 .628

mother as source of

support .207 1.149 .254 .126 .264

patemai affective 9

quality of rel ati on ahip -.064 -.642 .523 -.O71 .864

father as facilitator of b

independence -.049 -.479 .634 -.053 .819

a
father as ”m” °f .134 1.401 .165 .153 .926

 

a. Predictors in the Model: (Constant), Mattemal affective quality of relationship

b- Predictors in the Model: (Constant), Mattemal affective quality of relationship, Schwarz
inter-parental conflict scale total

C- Dependent Variable: Social Desirability Scale Mean

177

 

Figure E2 African American and Hispanic group Normal Probability Plot of the
Regression Standardized Residuals for Path Analysis 1

Normal P-P Plot of Regression Standardized Residual

Dependent Variable: sse_total

 

Expected Cum Prob

 

 

 

 

0-07 i i l r
0.0 0.2 0.4 0.6 0.8 1.0

Observed Cum Prob

178

Figure E3 African American and Hispanic Residuals Scatter plot for Path

 

 

 

 

 

Analysis 1
Scatterplot
Dependent Variable: sse_total
3—
To
3 0
3 o
3 24 o o
l: o o
3
g 1" oo o o
a Q: ° 0
1a O
E O 0 ° 00 O
0- O O O
0:) o o o 8
.2 o
3 _1_ 0 ° o
g o o o o
O
n: o
2_ o
l I i i I
-3 -2 -1 0 1

Regression Standardized Predicted Value

179

Table E African American and Hispanic Group Path 2 Decompositions and

Reproduced r’s

r14

r15

r16

r17

r18

r19

Calculations of Reproduced Coglations page 1 of 4
Path Decompositions r

 

 

P41
(--476) = -.476

(D)

r12P62P65
(.598) (.29) (.408) = .051

(U)

P61 + r12P62
(.244) + (.598) (.253) = .395

(D) (U)

T12P12
(.598) (.250) = .150
(U)

T12P82
(.598) (.209) = .124

(U)

P91 "' P61P96 + P41P94 + r12P62P96 + r12P62P65P95 + r12P62P65P75P67
1’ T12P62P65P65P96 "’ r12P62P96 "' r12P72P97 683

(.354) + (.244)(.295 ) + (-.476)(-.373) + (.596)(.209)(-.369)
(D) ( l ) ( l ) (U)
+ (.598)(.209)(.408)(.363)+ (.598)(.209)(.408)(.247)(.481)
(U) (U)
+ (.598)(.209)(.408)(.217)(.295) + (.596)(.253)(.295)
U (U)

( )
+ (.598)(.250)(.480)
(U)

 

r 24

r 25

r 26

r27

r28

r12P42
(.598)(-.476) = -.285

(U)

P62P65
(.209)(.408) = .065

(|)

P62 + P62P65P65 + r12P62
(.253) + (.209)(.408)(.217) + (.598)(.244) = .417

(D) (l ) (U)

P72 ‘* P62P65P75
(.250) + (.209)(.408)(.247) = .271

(D) (|)
P62

(209) = .209
(D)

180

DQQEZOf4

 

 

r29 = P62P96+P62P65P95+P62P65P75P97+P62P65P65P96+P62P96+P72P97
+ r12P91 'l' r12P61P96 T r12P41P94
= (.209)(-.369) + (.209)(.408)(.363) + (.209)(.408)(.247)(.481)
( I ) ( l ) ( | )
+ (.209)(.408)(.217)(.295)+ (.253)(.295) + (.250)(.481)
( l ) ( l ) ( l )
+ (.598)(.354) + (.598)(.244)(.295) + (.598)(-.476)(-.373) = .525
(U) (U) U)
’34 = P43
= (.201) = .201
(D)
r35 = P53 "’ P63P65
= (.267) + (-.31)(.408) = .140
(D) ( l )
r35 = P53P65 + P63P65P65
= (.267)(.217) + (-.311)(.408)(.217) =.030
( l ) ( l )
f 37 = P73 + P53P75 4’ P63P65P75
= (.248) + (.267)(.247) + (-.311)(.408)(.247) = -.2‘l3
(D) ( l ) ( l )
r38 = P63
= (-.311) = -.311
(D)
r39 = P43P94 T P73Ps7 + P53P95 "' P53P75P97 "’ P35P56P96 T P63P96 "' P63P65P95
+ P63P65P75P97 + P63P65P65P96
= (.201)(-.373) + (-.248)(.481)(.267)(.363) + (.267)(.247)(.481)
( l ) ( l ) l
+ (.267)(.217)(.295) + (-.311)(-.369) + (-.311)(.408)(.363)
( | ) ( | ) ( l )
+ (-.311)(.408)(.247)(.481) + (-.311)(.408)(.217)(.295) = -.003
( r ) ( I )
'45 = P43P54 + P43P63P65 + P41 (12 P62P65
= (.201)(.267) + (.201)(-.311)(.408) + (-.476)(.598)(.209)(.408) = .004
(S) (S) (S)
’46 = P43P53P65 + P41P61 "’ P43P63P65P65 1’ P41r12P62 + P41T21P62P65P65
= (.201)(.267)(.217) + (-.476)(.244) + (.201)(-.311)(.408)(.217)
(S) (S) (S)
+ (-.476)(.598)(.253)+ (-.476)(.598)(.209) = -.187
(S) (S)
’47 P43P73 7’ P43P53P75 1' P43P63P65P75 + P41f12P72 + P41712P62P65P75

(.201)(—.248) + (.201)(.267)(.247) + (.201)(-.311)(.408)(.247)
s

(S) (S) ( )
+ (-.476)(.598)(.252) + (-.476)(.596)(.209)(.406)(.247) = -.120
(S) (S)

181

 

 

 

IL! E page 3 of 4
r43 = P43P63 T P41|'12P62
= (.201)(-.311) + (-.476)(.598)(.209) = -.122
(S) (S)
’49 = P94 T P43P73P97 T P43P53P95 T P43P63P96T P43P63P65P95 T P43P63P65P75P97
T P43P53P75P97 T P41P91 TP41P61P96 T P41712P72P97 TP41T12P62P96
T P41712P62P96 T P41r12P62P65P95 T P41712P62P65P57P97 604
= (-.373)+ (.201)(-.248)(.481) + (.201)(.267)(.363) + (.201)(-.311)(-.369)
(D) (S) (S) (S)
+ (.201)(-.311)(.408)(.363) + (.201)(-.311)(.408)(.247)(.481)
(S) (S)
+ (.201)(.267)(.247)(.481) + (-.476)(.354) + (-.476)(.244)(.295)
(S) (S) (S)
+ (-.476)(.598)(.250)(.481) + (-476)(.598)(.253)(.295)
(S) (S
+ (-.476)(.598)(.209)(-.369) + ( -.476)(.598)(.209)(.408)(.363)
(S) (S)
+(-.476)(.598)(.209)(.408)(247)(.481)+(-.476)(.598)(.209)(.408)(.217)(.295)
(S) (S)
’56 = P65
= (.217) = .217
(D)
r57 = P75
= (.247) = .247
r58 = P85 T P53P63
= (.408) + (.267)(-.311) = -.325
(D) (S)
r59 = P95 T P75P97 T 965me Despite T P53P43P94 T P53P63P96
= (.363) +(.247)(.481) + (.408)(-.369) +(.217)(.295) +(.267)(-.248)(.481)
(D) ( | ) (S) ( I ) (S)
+ (2.67)(.201)(-.373) + (.267)(-.311)(-.369) . .374
(S) (S)
’69 = P96 T P62P26P65P95 T P62P62P65P75P97 T P62P72P97 T P62P62P96

T P62r12P91 T P62712P41P97 T P61P91 T P61P41P94 T P61r12P72P97
T P61r12P62P96 T P61712P62P65P95 T P61r12P62P65P75P97 ' 539

(.295) + (.253)(.209)(.406)(.363) + (.253)(.209)(.408)(.247)(.481)
(D) (S) (S)
+ (.253)(.250)(.481) + (.253)(.209)(-.369) + (.253)(.596)(.354)

s s s

( ) ( ) ( )
+ (.253)(.598)(-.476)(-.373) + (.244)(.354) + (2.44)(-.476)(-.373)
(S) (S) (S)
+ (.244)(.598)(.250)(.481) + (.244)()(.596)(.209)(-.369)
s s

( ) ( )
+(.244)(.596)(.209)(.406)(.363)+ (.244)(.596)(.209)(.406)(.247)(.461)
(S) (S)

182

Table E page 4 of 4
’79 = P97 T P75P95 T P75P65P96T PnP53P95T P73P53P65P96T P73P63P65P95

T P73P63P96 T P72P62P96 TP73P63P65P65P96 T P72P62P65P95

T P72P62P65P65P95 T P72I12P61P96+ P72F12P91 T P72I12P41P94

= (.481)+(.247)(.363)+(247)(.217)(.295)+(-.248).267)(.363) +(-.248)(-.31 1)(-.369)
(D) (S) (S) (S) (S)
+ (-.248)(.267)(.217)(.295) + (-.248)(-.311)(.408)(.363) + (.250)(.209)(-.369)
S S (S)

 

( ) ( )
+ (-.248)(-.311)(.408)(.217)(.295)+ (.250)(.209)(.408)(.363)
(S) S
+ (.250)(.209)(.408)(.217)(.265) + (.250)(.598)(.244)(.295)
S (S)
+ (.250)(.596)(.354) + (.250)(.598)(-.476)(-.373) = .457
(S) (S)

 

P96 T P65P95 T P65P75P97 T P62P62P96 T P62P72P97 T P62I12P91 T P62T12P61P96
T P62r12P41P94 T P60P53P95 T P63P53P75P97 T P63P73P97 T P63P53P65P96

(-.369) + (.406)(.363) + (.408)(.249)(.481) + (.408)(.217)(.295)
(D) (I ) ( I ) (I )
+ (.209)(.253)(.295) +(.209)(.250)(.481) + (.209)(.596)(.354)
(S) (S) S)
+ (.209)(.598)(.244)(.295) + (.209)(.598)(-.476)(-373)
(S) (S)

r 69

+ (-.311)(.201)(-.373) + (-.311)(.267)(.363) + (-
.311)(.267)(.247)(.481)
(S) (S) (S)
+ (-3.11)(.267)(.217)(.295) + (-.311)(-.248)(.481) -- .117
(S) (S)

 

183

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

M/AFF I i
I -.476** SocD
4
.244‘ WELL-
BEING
M/IND 201 t l 9
' m8“) . 95"
2 .2538

 

 

 

 

M/c-QN ETH 1 +

 

 

 

 

 

 

Figure E4 African American and Hispanic Group Path 2 Patemai Promotion of

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

independence
.0 ,
l 4: all WELL-
-.476 BEING
M/iND
2
MICON

 

 

3

 

 

 

 

 

 

Figure E5 African American and Hispanic Group Path 2 Patemai Feelings of
Affect

184

 

 

 

 

 

 

Soc D a WELL-

BEING

: .48 I "
M/iND 25° P/SUP @

 

 

 

 

 

 

 

 

 

 

 

 

.0» 7

ETH 1
WC ON

> 5
t
3 @

 

 

 

 

 

 

 

 

 

Figure E6 African American and Hispanic Group Path 2 Father Feelings of
Suppon

185

References

Adams, T. A., Bezner, J. B., Drabbs, M. E. (2000). Conceptualization and
measurement of the spiritual and psychological dimensions of wellness in a
college population. Journal of American College Health. 4c8 (4). 165-173.

Adams, T.A., Bezner J. B., Steinhardt M. A. (1997). The conceptualization
and measurement of wellness: Integrating balance across and within dimensions.
American Joml of Health Promotion;11(3). 208-218.

Ahrons, C. (1994). The Good Divorce. New York, New York: Harper
Collins Publisher.

Ainsworth, M. D. S. (1989). Attachments beyond infancy. American
Psycholpgist, 44, 709-716.

Ainsworth, M. D. S. & Bowlby, J. (1991). An ethological approach to
personality development. American Psychologist. 46, 333-341.

Amato, P.R. & Gilbreth, JG. (1999). Nonresident fathers and Children's
well-being: a meta analysis. Journal of Marriage & FalImilv. 61, 557-573.

Amato PR, Booth A (1996), A prospective study of divorce and parent-
child relationships. Journal of Marriage and Famiu, 58, 356-365.

Amato, P. R. (1993). Children's adjustment to divorce: Theories,

hypotheses, and empirical support. Journal of Marriage and the Family 55, 23-
38.

Amato, P. R., & Keith, B. (1991). Parental Divorce and the Well-being of
Children: A Meta-analysis. Psychoipgical Bulletin, 110, 26-46.

Ardell D. (1986) High Level Wellness_. Ten Speed Press; Berkeley, CA.

Arditti, J. A. & Prouty, A. M. ( 1999). Change, disengagement, and
renewal: relationship dynamics between young adults and their fathers after

divorce. Journal of Marital and Family Therapy, 25, 61-81.

Arnstein, R. (1980). The student, the family, the university, and the
transition to adulthood. Adolescent Psychiatg, 8, 160-172.

Arnsden, G. C., & Greenberg, M. T. (1987). The inventory of parent and
peer attachment: Individual differences and their relationship to psychological
well-being in adolescence. Journal of Youth and Adolescence 16 427-454.

 

186

 

Asendorpf, J. B., 81 Wilpers, S. (2000). Attachment security and available
support: Closely linked relationship qualities. Jaamal of Social and Personal
Relationships. 17. 1 15-138.

Bakermans-Kranenburg, M. J., van lJendoom, M. H., & Kroonenberg, P.
M. (2004). Differences in attachment security between African American and
white children: Ethnicity or socio-economic status? Infant Behavior and
Development, 27, 41 7-433.

Baldwin, M. W., & Fehr, B. (1995). On the instability of attachment style
ratings. Personal Relationships, 2, 247-261.

Barnes, G. G., (1999). Divorce transitions: Identifying risk and promoting
resilience for children and their parental relationships. flrnal of Marital anq

Family Therapy, 25 (4), 425-441.

Bartholomew, K., & Horowitz, L. M. (1991). Attachment style among
young adults: A test of a four-category model. flimal of PersonalihLand Social
Psychology. 61. 226-244.

Belsky, J., Campbell, S. B., Cohn, J. F., & Moore, G. (1996). Instability of
infant-parent attachment security. Developmental Psvchoquy, 32, 921-924.

Billingham, R. E. & Abrahams, T. (1998). Parental divorce, body
dissatisfaction and physical attractiveness ratings of self and others among

college women. Collage Student Journal, 32, 148-152.

Billingsley, A. (1990). Understanding African American family diversity. In
J. Dewart (Ed.), The State of Black America 1990, New York: National Urban
League.

Brack, G, Gay, M. F ., Matheny, K. B. (1993). Relationships between
attachment and coping resources among late adolescents. Journal of Coll_e_ge
Stuciant development. May. 34 (3). 212-215.

Blain, M., Thompson, J., & Whiffen, V. (1993). Attachment and perceived
social support in late adolescence: The interaction between working models of
self and others. Jol_lrnal of flolescent Research. 8. 226-241.

Bowlby, J. (1969). Attachment and loss: Vol. 1. Attachment. London:
Hogarth.

Bradford, E., Lyddon, W. J. (1993). Current parental attachment: Its

relation to perceived psychological distress and relationship satisfaction in
college students. Journal of College Stadent Qavelopment. 34 (3). 256-260.

187

Braver, S., Wolchik, S., Sandler, l., Fogas, B., & Svetina, D. (1991).
Frequency of visitation by divorced fathers: Differences in reports by fathers and
mothers. American Journal of Qrthopgivchiatrvfit 448-454.

Carranza, L. V. & Kilmann, P. R. (2000). Links between perceived parent
characteristics and attachment variables for young women from intact families.
Aaolescence. 35 (138) 295-312.

Cavell, T., Jones, D., Runyan, R. D., Constantin-Page, L., 8. Velasquez, J.
(1993). Perceptions of attachment and adjustment of adolescents with alcoholic
fathers. Jaumal of Family Psvcholqgv. 7, 204-212.

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

Chapman, L. S. (1987). Developing a useful perspective on spiritual
health: Well-being, spiritual potential and the search for meaning. American
Journal of Health Pronflon, 1(3), 31-39.

Chickering, A. W. (1969). Education and identfiy‘ . San Francisco: Jossey-
Bass.

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

Cook, W. L. (1994). A structural equation model of dyadic relationships
with the family system. flirnal of Consglting and_Clinical Psychology. 6_2_, 500-
509.

Crawley, R. (1988). Black families in a neo-conservative era. Famiy
Relations. 37. 415-419.

Crowne, D. P., & Marlow, D. (1960). A new scale of social desirability
independent of psychopathology. Journal of Consult Psvcholpgy, 24, 349-335.

Cummings, E. M., & Watson-O’Reilly, A. (1997). Fathers in family context:
Effects of marital quality on child adjustment. In M. E. Lamb (Ed.), The role of
t_he fa_ther in child development (3rd ed). New York: Wiley.

D’Angelli, A. R. & Hershberger, S. L. (1993). African American
undergraduates on a predominantly White campus: Academic factors, social
networks, and campus climate. Journal of Negro Education, 62, 67-81.

Das Eiden, R., Teti, D. M., 8 Corns, K. M. (1995). Maternal working

models of attachment, marital adjustment, and the parent-child relationship.
Child Developmem, 66, 1504-1518.

188

Davies, P. T., & Cummings, E. M., (1994). Marital conflict and child
adjustment: An emotional security hypothesis. Psychological Bulletin, 116, 387-
41 1.

DeCindio, L. A., Floyd, H. H., Wilcox, J. A., McSevency, (1983). Race
effects in a model of parent-peer orientation. lI_tr_Ioles_c_ence, 18, 369-379.

Depken D. (1994) Wellness through the lens of gender: A paradigm shift.
Wellness Perspectives, 10, 54-69.

Dewitt-Parker, J. (2000). Predicting the successful adjustment of black
freshman to a predominantly white university: The contribution of parental
attachment and cultural traditionality. Dissertation-Abstracts International

Section A: Humanities and Social Sciences, 60 (8-A), 2812.

Donley, M. G. (1993). Attachment and the emotional unit. Family Process,
Q, 3-20.

Dombusch, S., Carlsmith, J., Bushwall, S., Ritter, P., Leiderman, H.,
Hastorf, A., & Gross, R. (1985). Single parents, extended households, and the
control of adolescents, Child Development, 56, 326-341.

Doyle, A. B., Markiewicz, M. B., Lieberman, M., & Voss, K. (2000). Child
attachment security and self-concept: Associations with mother and father

attachment style and marital quality. Merrill-Palmer Quartem, 46 (3), 514-539.

Dudley, J. R. (1991). The consequences of divorce proceedings for
divorced fathers. Journal of Divorce and Remarriage, 16, 171-193.

Duesterhaus, M. E. B. (2004). The predictive value of college students’
attachment styles on the development of intrinsically or extrinsically motivated life
goal. Dissertation Abstract International Section-A: Humanities and Social

Sciences, 65 (5-A), 1688.

Dunn, H.L. (1966) High-level wellness for man and society. In: FoItaJR,
Deck ES, eds. A sociolpgical framework for patient care. New York: John Wiley &
Sons, 213-219.

Egeland, B. & Farber, E. a. (1984). Infant-mother attachment: Factors
related to its development and changes over time. Child Development, 55, 753-
771.

Elicker, J., Englund, M., & Stroufe, L. A. (1992). Predicting peer
competence and peer relationships in childhood from early parent-child
relationships. In R. Parke & G. Ladd (Eds), Famiu-gr relations: Modes of
linkage (77-106). Hillsdale, NJ: Erlbaum.

189

Emery, R. E., & Forehand, R. (1994). Parental divorce and children’s
well-being: A focus on resilience. In R. J. Haggerty, L. R. Sherrod, N. Garmezy &
M. Rutter (Eds), imss. risk. and_rasilience in chflren and adolescents:
Processes. mechanisms and interventions (pp. 64-99). New York: Cambridge
University Press.

 

Erel, 0., 8. Burman, B. (1995). lnterrelatedness of marital relations and
parent-child relations: A meta-analytic review. Psycholpgical Bulletin, 118, 108-
1 32.

Famworth, M. (1984). Family structure, family attributes, and delinquency
in a sample 0 flow-income, minority males and females. Journal of Youth and
A_dolescence. 13, 349-364.

Feldman, S. S., & Eliott, G. R. (1990). At the threshold: The developipg
adolescent. Cambridge, MA. Harvard University Press.

Florian, V., Mikulincer, M., 8 Bucholtz, I. (1995). Effects of adult
attachment style on the perception and search for social support. Journal of
Psvchologv. 12a, 665-675.

Fox, N. A., Kimmerly, N. L., & Schafer, W. D. (1991). Attachment to
mother/ attachment to father: A meta-analysis. Child Development, 62, 210-225.

Furstenberg, F. F., Morgan, S. P., & Allison, P. D. (1987). Parental
participation and children's well-being after marital dissolution. American

Sociolpgical Review, 52, 695-701.
Gallagher, M. (1996). The Abolition of Marriage. Regnery Publisher.

Giordano, P. C., Cemkovich, S. A., & DeMaris, A. (1993). Family and

peer relations of Black adolescents. Journal of Marriage and the Famiy, 55,
277-287.

Goldberg, W. A. (1990). Marital quality, parental personality, and spousal
agreement about perceptions and expectations of children. Merrill-Palmer

Quarterly, 36, 531-556.

Gove, W. 8 Shin, H. (1989). The psychological well-being of divorced and
widowed men and women: An empirical analysis. Jofiumal of Family Issqes. 10.
122-144.

Grayson, P. (1998). A fast and bumpy ride: Students and therapy in an
ever changing environment. Journal of College Student anchotherapy, 13, 3-13.

190

Greenberg J. S. (1985) Health and wellness: A conceptual differentiation.
Journal of School Health 55 403-406.

 

Griffin, D., & Bartholomew, K. (1994). Models of the self and other:
Fundamental dimensions underlying measures of adult attachment. Journal of
Personality & Social Psvchologv. 67. 430-445.

Grossman, K. E., Grossman, K., Huber, F., & Warner, U. (1981) German
children’s behavior towards their mothers at 12 months and their fathers at 18
months in Ainsworth’s strange situation. International Jogrnal of Behavioral

Development, 4, 157-181 .

Grotevant, H. D. (1989). Child development within the family context. In
W. Damon (Ed), Child development today and tomorrow (34-51). San Francisco:
Jossey-Bass.

Grotevant, H. D., & Cooper, R. (1986). Individuation in family
relationships. Human Development, 29, 82-100.

Hannum, J. W. & Dvorak, D. M. (2004). Effects of family conflict, divorce,
and attachment patterns on the psychological distress and social adjustment of

college freshmen, Journal of Collage Student Development, 45(1), 27-42.

Hanson, T. L. ( 1999). Does parental conflict explain why divorce is
negatively associated with child welfare? Social Forces, 77 (4), 1283-316.

Harrison, A. 0., Wilson, M. N., Pine, C. J., Chan, S. Q., & Buriel, R.
(1990). Family ecologies of ethnic minority children. Child Development, 61, 347-
362.

Hayashi, G. M & Strickland, B. R. (1998). Long-terrn effects of parental
divorce on love relationships: Divorce as attachment disruption. Journal of Social

and Personal Relationships, 15 (1) 23-38.

Harris, I. & Salt, R.(1999) Patterns of paternity. Journal of Men’s Studies
7 (2), 245-264.

 

Hazan, C., 8. Shaver, P. R. (1994). Attachment as an organizational
framework for research on close relationships. Psycholagical Ingt_lirv. 5 (1). 1-22.

Hazan, C., & Shaver, P. R. (1987). Romantic love conceptualized as an
attachment process. Journal of Personalfly’ ad Social Psychology. 59, 270-280.

Heatherton, T.F. & Polivy, J. (1991). Development and validation of a
scale for measuring state self-esteem. Journal of Parsonalitv and Social

Pa ycholggy, 60, 895-910.

 

191

Hetherington, E. M. (1992). Coping with marital transitions: A family
systems perspective. In E. M. Hetherington & W. G. Clingempeel (Eds) Coping
with marital transitions (1-14).

 

Hetherington, E.M., Cox, M. & Cox, R. (1982) Effects of divorce on
parents and children. In: M. Lamb (Ed.), Nontracflonal families: Parenting ang
child development. Hilldale, NJ: Lawrence Erlbaum.

Hettler, W. (1984) Wellness: Encouraging a lifetime pursuit of excellence.

Health Values. 8 (4), 13-17.

Hinderlie, H. H., Kenny, M. (2002). Attachment, social support, and
college adjustment among black students at predominantly white universities.

Journal of Collage Student Development, 43 (3), 327-340.

Hines, A. (1997). Divorce-related transitions adolescent development, and
the role of the parent-child relationship: A review of the literature. Journal of
Marriage and the Family. 59, 375-388.

Hojat, M. (1998). Satisfaction with early relationships with parents and
psychosocial attributes in adulthood: Which parent contributes more?, Journal
(if—Genetic Psychology, 159 (2), 203-220.

Holmbreck, G. N., & Wandrei, M. (1993). Individual and relational
predictors of adjustment in first-year college students. Journal of @nseling
PsmhologyAO, 73—78.

Hughes, M., & Demo, D. H. (1989). Self-perceptions of Black Americans:
Self-esteem and personal efficacy. American Journal of Sociology, 95, 132-159.

Ihinger-Tallman, M., Pasley, K., & Buehler, C. (1993). Developing a
middle-range theory of father involvement postdivorce. Journal of Family Issues,
141, 550-571.

Jekielek, S. M. ( 1998). Parental conflict, marital disruption and children's
emotional well-being. SLocfil Force_§. 76. (3), 905-36.

Kelly, J. B. (2000). Children's adjustment in conflicted marriage and
divorce: A decade review of research. Journal of the American Academy of Child
and Adolescent Psvchiatm, 39 (8), p. 963-73.

Kelly, J. B. (1988). Longer-term adjustment in children of divorce:

Converging finding and implications for practice. Journal of Family Psychology,
_2_, 1 19-140.

192

Kenny, M. E. (1990). College seniors' perceptions of parental attachments:
The value and stability of family ties. Journal of Collage Studenthvelopment.
3_1, 3946.

Kenny, M. E. (1987). The extent and function of parental attachment
among first-year college students. Journal of Youth and Adolesgnce. 16. 17-29.

Kenny, M. E. (1987). The parental attachment questionnaire. (Available
from M. Kenny, Boston College.)

 

Kenny, M. E., & Donaldson, G. A. (1991). Contributions of parental
attachment and family structure to the social and psychological functioning of
first-year college students. Journal of Counseling Psycholpgy, 38, 479-486.

Kenny, M. E, Gallagher, L. A., Alvarez, S. R., & Silsby, J. (2002). Sources
of support and psychological distress among academically successful inner-city
youth. Adfolescence. 37 (145). 161-182.

Kenny, M. E., Lomax, R., Brabeck, M., & Fife, J. (1998). Longitudinal
pathways linking adolescent reports of maternal and paternal attachment to

psychological well-being. Journal of Early Adolescence, 18(3), 221-243.

Kenny, M. E. & Perez, V. (1996) Attachment and psychological well-being
among racially and ethnically diverse first-year college students. Journal of
CollegStggant Development. 37 (5), 527-535.

Kenny, M. E., 8 Rice, K. G. (1995). Attachment to parents and adjustment
in late adolescent college students: Current status, applications, and future
considerations. Counseling Psychologista 23, 433-456.

 

Kerns, K. A., Klepac, L., & Cole, A. K. (1996). Peer relationships and
preadolescents’ perceptions of security in mother-child relationship.
Developmental Psycholqu, 32, 457-466.

Klasner, L. & Pistole, M. C. (2003). College adjustment in a multiethnic
sample: attachment, separation-individuation, and ethnic identity. Journal of
Collage Stuident Develgpment. 44 (1). 92-109.

Kruk, E. (1993). Divorce and disengagement: Patterns of fatherhood
within and beyond marriage. Halifax, NS: Femwood.

Kurkowski, K., Gordon, D. A., & arbuthnot, J. (1993). Children caught in
the middle: A brief educational intervention for parents. Journal of Divorce and
Remarriage, 10, 139—151.

193

Lapsley, D. K., Rice, K. G., & Shadid, G. E. (1989). Psychological
separation and adjustment to college. Journal of Counseling Psychology, 36,
286-294.

Larose, S., & Boivin, M. (1998). Attachment to parents, social support
expectations, and socioemotional adjustment during the high school-college
transition. Journal of Research on Adolescence 8 1-27.

 

Larson,R. W., Richards, M. H., Moneta, G., Holmbeck, G., & Duckett, E.
(1996) Changes in adolescents’ daily interactions with their families from ages 10
to 18: Disengagement and transformation. QavelgqnentiPsvcholggy, 32, 744-
754.

Lawson, E. & Thompson, A. (1994). Historical and social conelates of
African American divorce: Review of the literature and implications for research.

The Western Journal of Black Studies, 18 (2), 91-103.

Lehr, R. 81 MacMillan, P. (2001). The psychological and emotional impact
of divorce: the noncustodial fathers' perspective. F_amilies in §ocietv. 82 (4), 373-
82.

Leonardi, A., & Kiosseoglou, G. (2000). The relationship of parental
attachment and psychological separation to the psychological functioning of

young adults. Journal of Social Psycholpgy, 140, 418-436.

Le Vrne, R. A. & Miller, P. M. (1990). Commentary. _l-leangevelopment,
1%) 73'80.

Lewis, M., Feiring, C., & Rosenthal, S. (2000). Attachment over time.
Child Development, 71 (3), 707-720.

Lindahl, K. M. & Malik, N. M. (1999). Marital conflict, family processes,
and boys’ externalizing behavior in Hispanic American and European American
families. Journal of (Enical Child Psvcholggy, 28, 12-24.

Little, M. A. (1992). The impact of the custody plan on the family: A five
year follow-up — executive summary. Family and Conciliation Courts Review, 30,
243-251.

Lopez, F. G. (1996). Attachment-related predictors of constructive
thinking among college students. Journal of Counseling & Development, 75, 58-
63.

Lopez, F. G., Melendez, M. C., & Rice, K. G. (2000). Parental divorce,
parent-child bonds, and adult attachment orientations among college students: A

194

comparison of three racial/ethnic groups. Journal of CounseLng Psvcholgv. 47
(2), 177-186.

Lytton, H. (1990). Child and parent effects in boys' conduct disorder: A
reinterpretation. Developmental Psycholagy, 26, 683-697.

Main, M., & Weston, D. R. (1981). The quality of the toddlers’ relationship
to mother and to father: Related to conflict behavior and the readiness to
establish new relationships. Child Develapment, 52, 932-940.

Mason, L. K. (2005). Attachment classification, parental bonding and
parental authority as they relate to anxiety, depression, and self-efficacy from a
multicultural perspective, Qisserta_ti<_>n Abstract International: Section B: The
Sciences and Engineering 65 (7Q), 3715.

Mathijssen, J. J., Koot, H. M., Verhulst, F. C., DeBruyn, E. J., & Oud, J. H.
(1998). The relationship between mutual family relations and child

psychopathology. Journal of Child Psycholggy and Psychiat_ry, 39 (4), 477-487.

Mattanah, J. F., Hancock, G. R., Brand, B. L. (2004). Parental
attachment, separation-individuation, and college student adjustment: A
structural equation analysis of mediational effects. Journal of Counseling
Psvchologv. 51 (2), 213-225.

McCombs, A., & Forehand, R. (1989). Adolescent school performance
following parental divorce: Are there family factors that can enhance success?
Adolescence, 24, 871-880.

 

McCormick, C. B., & Kennedy, J. H. (2000). Father-child separation,
retrospective and current views of attachment relationship with father and self-
esteem in late adolescence. Psycholggical Remus, 86, 827-834.

McCormick, C. B., & Kennedy, J. H. (1994). Parent-child attachment
working models and self-esteem in adolescence. Journal of Youth and
Adolescence 23, 1-18.

 

McCurdy, S. J. & Scherman, A. (1996). Effects of family structure on the
adolescent separation-individuation process. lilolescence. 31 (12_2), 307-319.

McKenry, R, Price, 8., Fine, M., & Servovich, J. (1992). Predictors of
single, noncustodial fathers' physical involvement with their children. The Journal
of Genetic Psychology. 1%, 305-319.

Menaghan, E. & Lieberman, M. (1986). Changes in depression following
divorce: A panel study. Journal of Marriage and the Family, 48, 319-328.

195

Meyer, S. A. (1998). Personality correlates of adult attachment. Journal
of Social Psycholagy, 138, 407-409.

Miller, J. B. & Stubblefield, A. (1993). Parental disclosure from the
perspective of late adolescents. Journal of Adolescence, 16 (4), 439-455.

Mitchell, A. (1985). Children in the middle. London: Tavistock.

Moos, R. H. (1974). Eamilygnvironment Scale. Palo Alto, CA: Consulting
Psychologists Press.

Myers, J. (1992) Wellness, prevention, development: The cornerstone of
the profession. Journal of Counseling and Development, 71, 136-139.

National Center for Health Statistics (2005). Monthly Vital Statistics
Report, 47 (1 ). US. Department of Health and Human Services: Hyattsville, MD
Ogbu, J. U. (1988). Cultural diversity and human development. New

Directions in Child Development, 42, 11-28.

Orbuch, T. L., Thornton, A., & Cancio, J. (2000). The impact of marital
quality, divorce, and remarriage on the relationships between parents and their
children. Marriage and Family Review. 29 (4). 221-246.

Papini, D. R., & Roggman, L. A. (1992). Adolescent perceived attachment
to parents in relation to competence, depression, and anxiety: A longitudinal

study. Journal of Higher Education, 57, 420-444.

Phinney, J. S. (1992). The multigroup ethnic identity measure. Journal of
Adolescent ResearChJ, 156-176.

Powers, S. l., Hauser, S. T., 8. Kilner, L. A. (1989). Adolescent mental
health. American Psychologist. 44 (a), 200-208.

Raymond, J. (2001). The ex-files, American Demggraphics, 23, (2), 60—64.

Reker, G. T. (1992). Mangal of the Life Attituge Profile-Revised.
Peterborough, ON: Student Psychologists Press.

 

Reker, G.T. & Peacock, E. J. (1981). The Life Attitude Profile-R (LAP-R):
A multidimensional instrument for assessing attitudes toward life. Canadian
Journal of Behavioral Scienceg13. 264-273.

Reynolds, W. M. (1982). Development of reliable and valid short forms of

the Marlowe-Crowne social desirability scale. Journal of Clinical Psycholggy, 38,
1 19-125.

196

Rice, K. G., & Cummins, P. N. (1996). Late adolescent and parent
perceptions of attachment: An exploratory study of personal and social well-

being. Journal of Counseling and Development, 75, 50—57.

Rice, K. G., Cunningham, T. J., Young, M. B. (1997). Attachment to
parents, social competence, and emotional well-being: A comparison of black
and white late adolescents. Journal of Counseling Psvchologva44, 89-101.

Rice, K. G., Fitzgerald, D. P., Whaley, T. J., & Gibbs, C. L. (1995). Cross-
sectional and longitudinal examination of attachment, separation-individuation,
and college student adjustment. Journal of Counseling & Development, 73, 463-
474.

Rothbaum, F. & Weisz, J. R. (1994). Parental caregiving and child
externalizing behavior in nonclinical samples: A meta-analysis. Psycholpgical
Bulletin 116, 55-74.

 

Rothbaum, F., Weisz, J., Pott, M., Mijake, K., & Morelli, G. (2000).
Attachment and culture: Security in the United States and Japan. American
Psychologist, 55, 1093-1 104.

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

Child Development, 60, 340-356.

Sax, L., Astin, A., Korn, W., & Mahoney, K. (1999). The American
freshman: National norms for 1999. Los Angeles: Higher Education Research
Institute, University of California.

Schlack, H. M. (2003). Attachment style and health behaviors.
Dissertation Abstract lntemational: Section E he Sciences and Engineering, 63

(8-B), 3971.

Schultheiss, D., & Blustein, D. (1994). Role of adolescent-parent
relationships in college student development and adjustment. Journal of
Counseling Psvchologv. 41, 248-255.

Schwarz, J. C. (1999). Schwarz Inter-Parental Conflict Scale (IPC).
(abstract) in J. Touliatos, B. P. & Straus, M. (Eds), Handbook of family

measurement techniques, Thousand Oaks, CA: Sage Publications, Inc. 485-
486.

Scott, D. J. (1999). The effect of parental divorce on the career
development process of young adults: A test of separation/attachment theory.
Qi_sserta1ign Abstracts International Section B: The Sciences and Engineering. 60

(4-B), 1989.

197

Seaward, B. L. (1995) Reflections on human spirituality for the worksite,
American Journal of Health Promotion 9, 165-168.

 

Selzer, J. (1998). Father by law: effects of joint legal custody on
nonresident fathers' involvement with children. Demggraphy, 35, 135-146.

Seltzer, J. (1994). Consequences of Marital Dissolution for Children.
Annual Review of Sociolggy 20, 235-66.

Seltzer, J. A., Schaeffer, N. C., & Chamg, H. (1989). Family ties after
divorce: The relationship between visiting and paying child support. Journal of

Marriage and Family, 51, 1013-1032.

Shaw, D. S., Winslow, E. B., & Flanagan, C. (1999). A prospective study
of the effects of marital status and family relations on young children’s
adjustment among African American and European American families. Child

Development, 70(3), 742-755.

Stevenson, M. R., & Black, K. N. (1995). How (fivorce affects offspring: A
research approach. Madison, WI: Brown and Benchmark.

Stone, G., & McKenry, P. C. ( 1998). Nonresidential father involvement: a
test of a mid-range theory. Journal of Genetic Psychology, 159 (3), 313-36.

Summers, P., Forehand, R., Annistead, L., & Tannenbaum, L. (1998).
Parental divorce during early adolescence in Caucasian families: The role of
family process variables in predicting the long-term consequences for early adult
psychosocial adjustment. Journal of COtglseling and Clinical Psycholggy, 66 (2),
327-336.

 

Tate, R. L. (1998). An introduction to modeling outcomes in the behavioral
and social sciences (2nd ed.) Edina, MN: Burgess Publishing.

Taub, D. J. (1995). Relationship of selected factors to traditional-age
undergraduate women’s development of autonomy. Journal of Collage Student
Developmellt, 36(2), 141-151.

Thompson, R.A, Lamb, M., & Estes, D. (1982). Stability of infant-mother
attachment and its relationship to changing life circumstances in an unselected
middle-class sample. Child Development, 53, 144-148.

Travis J. W., Callander M. G. (1990). Wellness for Helping Professionals.
Mill Valley, CA: Wellness Associates.

198

US. Census Bureau. (2005). US. report on census population, 2004.
Washington, DC: Government Printing Office.

US. Census Bureau. (2003). US. remrt on census population, 2002.
Washington, DC: Government Printing Office.

US. Census Bureau. (2001). US. remit on census population, 1996.
Washington, DC: Government Printing Office.

Van Ijzendoom, M. H., & DeWolff, M. S. (1997). In search of the absent
father — meta-analyses of infant-father attachment: A rejoinder to our
discussants. Child Development. 68 (4). 604-609.

Van IJzendoom, M. H., & Kroonenberg, P. M. (1988). Cross-cultural
patterns of attachment: A meta-analysis of the strange situation. Child

Development, 59, 147-156.

Van IJzendoom, M. H., & Sagi, A. (1999). Cross-cultural patterns of
attachment. In J. Cassidy & P. R. Shaver (Eds.), Handbook of Attachment (713-
734). New York: Guilford.

Vaughn, 3., Egeland, B., Sroufe, L. A., & Waters, E. (1997). Individual
differences in infant-mother attachment at twelve and eighteen months: Stability
and change in families under stress. Child Development, 50, 971-975.

Veneziano, R. A. & Rohner, R. P. (1998). Perceived patemai acceptance,
patemai involvement, and youths” psychological adjustment in a rural, biracial

southern community. Journal of Marriage and the Famiy, 60, 335-343.

Vivona, J. (2000). Parental attachment styles of late adolescents: Qualities
of attachment relationships and consequences for adjustment. Journal of
Qaunseling Psvcholm 47. 316-329.

Wallerstein, J.S.. & Kelly, J.B. (1980). Surviving the break-up: How
children and parents cog with t_livorca New York: Basic Books.

Weiss, R. (1975). Marital Separation. New York: Basic Books.
Whiteside, M. F. & Becker, B. J. (2000). Parental factors and the young

child’s postdivorce adjustment: A meta-analysis with implications for parenting
arrangements. Journal of Family Psvcholggy, 14, 5—26.

 

Whiteside, M. F. (1998). The parental alliance following divorce: Literature
review and policy implications. Journal of Marital and Famiu Therapy, 24, 3-24.

199

Whitmer, J. M. & Sweeney, T. J. (1992). A holistic model for wellness
prevention over the life span. Journal of Counseling and Development, 71, 140-
148.

Woodward, L., Pergusson, D. M., & Belsky, J. (2000). Timing of parental
separation and attachment to parents in adolescence: Results of a prospective
study from birth to age 16. Journal of Marriage and the Family, 62, 162-174.

Zayas, L. H. 81 Solari, F. (1994). Early childhood socialization in Hispanic
families: Context, culture, and practice implications. Professional Psychology;

Research and Practice, 25(3), 200-206.

Zayas, L. H., & Palleja, J. (1988). Puerto Rican familism: Implications for
family therapy. Family Relations, 37, 260-264.

Zimmerman, M., Salem, 0., & Maton, K. (1995) Family structure and

psychosocial correlates among urban African-American adolescent males. Child
Develmment, 66, 1 598-161 .

200

I);illjiirjjjijjgii