.UBRARY f
Michigan State
University

 

 

 

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

 

DATE DUE DATE DUE DATE DUE

 

JAN 0 9 2511

 

 

M

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/DateDuopssopJS

 

 

ETHNICITY, ATTACHMENT, AND GENDER ROLE SOCIALIZATION AS
PREDICTORS OF LEVEL OF VOICE IN ADOLESCENT GIRLS:
THE CUMULATIVE IMPACT ON RELATIONSHIPS
AND PSYCHOLOGICAL WELL-BEING
By

Sally A. Theran

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Psychology

2003

ABSTRACT
PREDICTORS OF LEVEL OF VOICE IN ADOLESCENT GIRLS:

THE CUMULATIVE IMPACT ON RELATIONSHIPS
AND PSYCHOLOGICAL WELL-BEIN G

BY

Sally A. Theran

Gilligan and colleagues (1990) argue that girls experience “loss of voice” in adolescence,
which can be understood as girls’ inability to voice their own opinions. They propose
that this is the result of societal pressure, and is manifested by inauthentic relationships.
The purpose of this study was to empirically examine this theory, and test a theoretical
model for loss of voice in an ethnically diverse and lower SES sample. Thus, this study
examined both predictors (parental attachment and gender role socialization) and
outcomes (quality of friendship, romantic relationship, dating conflict, and psychological
well-being) of level of voice. One-hundred and eight 14-year-old girls were administered
measures of parental attachment, gender role socialization, level of voice, quality of
friendship, quality of romantic relationship, dating conflict, depression, self-esteem, and
social support. Structural equation modeling results indicated, overall, that the data was a
good fit for the proposed models, suggesting support for a theoretical model of level of
voice. Specifically, gender role socialization predicted level of voice, and parental
attachment only significantly predicted level of voice with authority figures, but not level
of voice with peers. The results supported the androgyny model Of gender role
socialization for level of voice with authority while the results supported the masculinity

model for level of voice with peers. Level of voice predicted the quality of friendships

and psychological well-being, and approval and emotional social support with authority
figures moderated the relationship between level of voice with authority figures and
depressive symptomatology. Level of voice with authority figures was a significant
predictor of quality of romantic relationships, while it did not predict dating conflict.
Finally, girls who attended a school with a higher percentage of ethnic minority students
had higher levels Of generalized peer voice than girls who attended a school with fewer
ethnic minority students, while across schools, girls of color had higher levels of voice
than Caucasian girls with classmates and with teachers. The results from this study help
contribute to empirical understanding of level of voice, and elucidate the mechanisms that

contribute to loss of voice in the context of intimacy and mutuality.

ACKNOWLEDGEMENTS

I am so grateful to so many people for their assistance and support during this
long process. First, my advisor, Anne Bogat, was supportive and encouraging of my
independent research from the beginning. I very much appreciated her wise suggestions
and feedback during this process, and I am incredibly grateful that she mentored me
during the past six years. Second, I could not have asked for a better committee. Alytia
Levendosky has been an invaluable support throughout graduate school, and her
insightful comments into the importance of mutuality in relationships made this a better
product. I also owe Anne and Alytia a great debt for their mentoring of me during the
past four years on the Mother-Infant Study; they have both taught me how to do
important, responsible, and enjoyable research. Fred Oswald was always willing to help
me with my statistics, and gave generously of his time. Kelly Klump’s enthusiasm for my
dissertation was very encouraging, and I greatly benefited from her expertise on puberty
and adolescence.

I am deeply appreciative for the support of the Lansing School District, and the
personnel of the Pattengill, Gardner, and Dwight Rich Middle Schools. All of the
principals and teachers I met and interacted with were so kind to help me. I am especially
grateful to Ms. Fletcher and Ms. Washburn who worked tirelessly to help me recruit my
participants.

My dissertation would not have been possible without the financial support of the
Blue Cross/Blue Shield Foundation, The Woodrow Wilson Dissertation Grants in

Women’s Studies, The Graduate School at MSU, and the MSU Department of

iv

Psychology. I thank all of these organizations for their tremendous and generous
financial support.

Many of my colleagues and friends provided so much practical and emotional
support. My husband, John Bergeron, listened to my initial ideas years ago, and has been
unwavering with his support ever since, in so many ways. I have loved being colleagues
together! Andrea Kleiver assisted me with data collection and data entry, and I have been
proud to mentor her in the process. The members of my dissertation study group, Anat
Barlev, Lisa Blaskey, and Cheryl-Lynn Podolski were so important to me over the last
several years. I learned much from the three of them, and their consistent enthusiasm was
wonderful. I absolutely loved working with my fiiends and colleagues from the Mother-
Infant Study - Alissa Huth-Bocks, Jenny Trotter, Robin Weather-ill, Erika DeJonghe,
Kerry Leahy, and Shallimar Jones. I cannot imagine a lab with better cooks than ours,
and their ability to laugh and be silly was crucial. I want to thank my former yoga
buddies who are now scattered all over the country — Jessica Goodkind, Anne Mauricio,
and Elaine Shpungin. Our yoga classes and subsequent coffee talks at Cappuccino Café
are some of my fondest and most dear memories of graduate school, and I miss you all.
My family and friends were consistently supportive, and have waited a long time to be
able to call me Dr. Sally!

Finally, I am most grateful to the girls who participated in this study, and gave so
generously of their time and selves. I loved hearing about their stories and lives, and

thank them so much.

TABLE OF CONTENTS

LIST OF FIGURES ............................................................................................................ ix
LIST OF TABLES .............................................................................................................. x
KEY FOR THE MEASUREMENT MODELS ................................................................. xi
INTRODUCTION ................................................................................................................ 1
Level of Voice ................................................................................................................ 5
Identity Theory ......................................................................................................... 5
Feminist Critiques of Identity Theory ...................................................................... 9
Relational Theory ................................................................................................... 12
Theory and Research on Level of Voice ................................................................ 14
Predictors of Level of Voice ........................................................................................ 24
Attachment ............................................................................................................. 24
Research on Attachment .................................................................................. 25
Connection between Loss of Voice and Attachment ....................................... 3O

Child’s Gender Role Socialization ......................................................................... 36
Outcomes of Level of Voice ........................................................................................ 39
Quality of Friendships ............................................................................................ 39
Romantic Relationships ......................................................................................... 45
Quality of romantic relationships ..................................................................... 45

Violence in Romantic Relationships ................................................................ 47
Psychological Well-Being ...................................................................................... 50
Depression ........................................................................................................ 50
Self-Esteem ...................................................................................................... 54

Social Support ........................................................................................................ 56
Rationale ...................................................................................................................... 59
Hypotheses ................................................................................................................... 64
Predictors of Voice Model ..................................................................................... 64
Friendship and Well-being Outcomes of Voice Model ......................................... 64
Romantic Relationship Outcomes of Voice Model ............................................... 65
Additional Hypotheses ........................................................................................... 65
METHODS ........................................................................................................................ 70
Participants ................................................................................................................... 70
Procedure ..................................................................................................................... 73
Measures ...................................................................................................................... 74
Demographics ........................................................................................................ 74
Demographics Measure .................................................................................... 74

Puberty Screening .................................................................................................. 74

Pubertal Development Scale -— Revised ........................................................... 74

Attachment ............................................................................................................. 75
Behavioral Systems Questionnaire .................................................................. 75
Gender Socialization .............................................................................................. 77
Children’s Sex Role Inventory ......................................................................... 77
Voice ...................................................................................................................... 78
Inauthentic Self in Relationships Scale ............................................................ 78
Teenage Voice (Harter, 1995) .......................................................................... 79
Friendship .............................................................................................................. 80
Friendship Qualities Scale ............................................................................... 80
Romantic Relationships ......................................................................................... 81
Conflict in Adolescent Dating Relationships ................................................... 81
Romantic Relationship Satisfaction ................................................................. 82
Psychological Outcomes ........................................................................................ 83
Children’s Depression Inventory — Short Form ............................................... 83
Rosenberg Self-Esteem Questionnaire ............................................................. 83
Social Support ........................................................................................................ 84
The Self-Perception of Social Support — Approval subscale — Revised ......... 84
The Perceived Social Support Scale ................................................................ 84
RESULTS .......................................................................................................................... 86
Results of the Hypotheses ............................................................................................ 92
Predictor Model for Authority Figures .................................................................. 93
Hypothesis 1 ..................................................................................................... 93
Hypothesis 2 ..................................................................................................... 94
Predictor Model for Peers ...................................................................................... 97
Hypothesis 1 ..................................................................................................... 97
Hypothesis 2 ..................................................................................................... 97
Friendship and Well-Being Outcome Model for Authority Figures .................... 100
Hypothesis 3 ................................................................................................... 100
Hypothesis 4 ................................................................................................... 100
Friendship and Well-Being Outcome Model for Peers ........................................ 103
Hypothesis 3 ................................................................................................... 103
Hypothesis 4 ................................................................................................... 103
Romantic Relationship Outcome Model for Authority Figures ........................... 106
Hypothesis 5 ................................................................................................... 107
Hypothesis 6 ................................................................................................... 107
Romantic Relationship Outcome Model for Peers .............................................. 1 ll
Hypothesis 5 ................................................................................................... 111
Hypothesis 6 ................................................................................................... 111
Results of the Additional Hypotheses ........................................................................ 114
Hypothesis 7 ......................................................................................................... 114
Hypothesis 8 ......................................................................................................... 117
Hypothesis 9 ......................................................................................................... l 19
Hypothesis 10 ....................................................................................................... 120

DISCUSSION .................................................................................................................. 123

Predictors of Level of Voice ...................................................................................... 124
Attachment ........................................................................................................... 124
Gender Role Socialization ................................................................................... 128

Outcomes of Level of Voice ...................................................................................... 131

Ethnic Differences in Level of Voice ......................................................................... 134

Summary .................................................................................................................... 138

Strengths and Limitations .......................................................................................... 140

Directions for Future Research .................................................................................. 142

REFERENCES ................................................................................................................ 147
APPENDICES ................................................................................................................. 165

Appendix A: Parent Consent F orm ............................................................................ 165

Appendix B: Child Consent Form ............................................................................. 167

Appendix C: Demographic Form ............................................................................... 169

Appendix D: Pubertal Development Scale ................................................................ 172

Appendix E: Behavioral Systems Questionnaire ....................................................... 174

Appendix F: Children’s Sex Role Inventory .............................................................. 178

Appendix G: Inauthentic Self in Relationships Scale ................................................ 180

Appendix H: Teenage Voice ...................................................................................... 182

Appendix I: Friendship Qualities Questionnaire ....................................................... 186

Appendix J: Conflict in Adolescent Dating Relationships ........................................ 189

Appendix K: Romantic Relationship Satisfaction ..................................................... 194

Appendix L: Children’s Depression Inventory — Short Form .................................... 198

Appendix M: Rosenberg Self-Esteem Questionnaire ................................................ 200

Appendix N: The Self-Perception of Social Support — Approval subscale ............... 202

Appendix 0: The Perceived Social Support Scale .................................................... 206

viii

LIST OF FIGURES

Figure 1: Conceptual Model ................................................................................................ 4
Figure 2: Structural model of predictors of voice .............................................................. 67
Figure 3: Structural model of fiiendship and well-being outcomes of voice ..................... 68
Figure 4: Structural model of romantic relationship outcomes of voice ............................ 69
Figure 5: Measurement model of predictors of voice with authority figures ..................... 95
Figure 6: Structural model of predictors of voice with authority figures ........................... 96
Figure 7: Measurement model of predictors of voice with peers ....................................... 98
Figure 8: Structural model of predictors of voice with peers ............................................. 99
Figure 9: Measurement model of outcomes of voice with authority figures ................... 101
Figure 10: Structural model of outcomes of voice with authority figures ....................... 102
Figure 11: Measurement model of outcomes of voice with peers ................................... 104
Figure 12: Structural model of outcomes of voice with peers ......................................... 105

Figure 13: Measurement model of romantic outcomes of voice with authority figures ..109

Figure 14: Structural model of romantic outcomes of voice with authority figures ........ 110

Figure 15: Measurement model of romantic outcomes of voice with peers .................... 112
Figure 16: Structural model of romantic outcomes of voice with peers .......................... 113
Figure 17: Level of voice and social support interaction ................................................. 116

ix

Table 1.

Table 2.

Table 3.

Table 4.

Table 5.

Table 6.

Table 7.

Table 8.

Table 9.

Table 10.

Table l 1.

Table 12.

Table 13.

Table 14.

Table 15.

LIST OF TABLES

Demographic characteristics of participants ....................................................... 72
Skew and kurtosis values for transformed variables ........................................... 87
Psychometric properties of measures .................................................................. 88
Correlation table of all variables ......................................................................... 89
Factor loadings for level of voice variables ........................................................ 91
Standardized regression weights for attachment measurement model ............... 92
Correlations among the eight positive subscales of the RS ........................... 107
Factor loadings for social support variables ..................................................... 114
Regression analysis for social support with authority figures as a moderator of
level of voice with authority figures and depression ............................... 116
Distribution of gender role socialization categories ....................................... 117
Means and standard deviations for level of voice across the gender role
socialization categories ............................................................................ 118
Correlations of voice with masculinity and femininity ................................... 118
Means and standard deviations for relationship variables with high/low levels
of voice ..................................................................................................... 120
Frequencies of different ethnic groups across middle schools ....................... 121

Means and standard deviations for level of voice for Caucasian girls versus
girls of color ............................................................................................. 122

KEY FOR THE MEASUREMENT MODELS

Predictors of voice

Attachment

BSQ

Security — attachment system secatt
Security — caregiving system seccare
Security - affiliation system secafl
Dismissing -— attachment system disatt
Dismissing — caregiving system discare
Dismissing — affiliation system disafir
Gender Role Socialization
CSRI
Masculinity masc
Femininity fem
Voice
Voice
ISR Self in Relationships inauthself
TV
Mother mvoice
Father fioice
Teacher (voice
Best friend bfloice
Classmates clvoice
Outcomes of Voice
Friendship
FQS
Conflict conflict
Security sec
Closeness close
Psychological Well-being
CDI Depression depress
RSEQ Self-esteem selfest

xi

Outcomes of Voice cont.

Romantic Relationships

CADRI

Social Support

SPSS

PSS

Threatening behavior
Relational Abuse
Physical Abuse

Verbal emotional abuse

Satisfaction w/relationship
Togethemess
Possessiveness

Grth

Communication
Toleration

Appreciation

Emotional support
Painfulness

Specialness

Mother

Father

Best friends
Classmates
Teachers
Romantic Interest

Family
Friends

threat
rel
phys
verbal

relsat
together
possess
grow
comm
toler
appre
emosupt
pain
spec

moapprov

faapprov
bfapprov
clapprov
teapprov
roapprov

faemot
fiemot

INTRODUCTION

The current study examined predictors and outcomes of loss of voice. Loss of
voice can be understood as girls’ difficulty in discussing their own opinions, and
negotiating conflicts with others. Research has generally neglected to examine why some
girls and not others might experience loss of voice, and what might be the possible effects
of loss of voice. Therefore, this study examined two predictors of loss of voice:
attachment to parents and gender role socialization. In addition, the relationship between
level of voice and several relational outcomes (i.e., quality of fiiendships, romantic
relationships, and dating conflict) and an individual-level outcome (i.e., psychological
well-being) was examined.

AS girls enter adolescence, they experience greater depression and lower self-
esteem than do boys during this time (Hill & Lynch, 1983; Steinberg & Morris, 2001).
While much has been written lately in the popular literature about the emotional risks that
pervade early adolescence for girls (Pipher, 1994; Spangler, 1999), to date, little empirical
research has examined these risks and the context of the risks. Theorists suggest that
early adolescence can be a time of uncertainty for girls; as their bodies change, so do the
societal rules for all kinds of relationships, including familial, platonic, and romantic
(L.M. Brown, 1998; L.M. Brown & Gilligan, 1992). In general, girls value relationships,
and may define themselves by their relationships (Miller, 1991).

As girls search for their own identities, they may find that what they knew to be
true as young girls is no longer true as adolescents. Girls may discover that societal
norms fi'own upon puffing their own needs before others, and they may receive implicit

messages from those around them to subvert their own interests to preserve their

relationships. This “silencing of the self,” (Jack, 1991) “loss of voice,” (Gilligan et al.,
1990), or “false self” (Harter, 1997), can lead to negative consequences for girls, such as
dissociation fi'om their own feelings, which may lead to depression and lowered self-
esteem (L.M. Brown, 1998). Thus, some girls may sacrifice their own sense of self, and
be nice and compliant with others, in order to preserve their relationships.

However, not every girl loses her sense of self during early adolescence, or even
later. In addition, research has demonstrated that girls who do self-silence do not present
with the same self-silencing behavior across different contexts (Harter, Waters, Whitesell,
& Kastelic, 1998b). Thus, girls may display different levels of self-silencing with
parents, teachers, friends, or classmates.

Little research has examined what might differentiate girls who experience a
silencing of the self, or a loss of voice, fiom those who do not. Attachment between
parents and daughter would be likely to affect how much an adolescent silences herself.
For example, individuals who are ambivalently attached are thought to be preoccupied
with relationships (Salzrnan, 1997). Therefore, a girl who is ambivalently attached might
be more likely to self-silence, in order to preserve relationships that are important to her.
In contrast, a girl who is securely attached may be less likely to self-silence, because she
has a secure base from which to relate to others, and thus may be more likely to test
societal norms for women (see Pathway A in Figure 1). In addition, it was hypothesized
that girls with a feminine orientation would be more likely to self-silence (see Pathway B
in Figure 1), as they also would be more likely to accept cultural norms of women as

caretakers (Jack, 1991).

While some researchers have started to investigate loss of voice in adolescent girls
(Harter, Waters, & Whitesell, 1997c; Harter et al., 1998b), to date, this research has only
examined the impact of loss of voice on psychological well-being, but no one has
demonstrated the immediate repercussions for loss of voice upon the quality Of
relationships in girls’ lives. Given Gilligan and colleagues’ assertions that girls with
lower levels of voice tend to have inauthentic relationships (Gilligan et al., 1990), it is
possible that girls who self-silence will have less intimate fiiendships (see Pathway C in
Figure 1) and lower quality of romantic relatiOnships (see Pathway E in Figure 1). In
addition, it is likely that girls who self-silence would be more likely to experience dating
conflict (see Pathway D in Figure 1). Finally, it is likely that girls pay a price for self-
silencing, and this suppression of voice may be associated with depressive
syrnptomatology and low self-esteem (see Pathway F in Figure 1). However, the
relationship between level of voice and psychological well-being may be moderated by
emotional support and approval support from significant persons in their lives (see
Pathway G in Figure 1).

In order to understand the construct of level of voice, which involves one’s
feelings of identity and the construction of the self, it is important to review traditional
definitions of the self. The following sections will review identity theory, feminist
critiques of identity theory, and will explore relational theorists’ definitions of the female

self.

Gender

Role
Socialization

 

Level
of
Voice

 
     
   
 

Attachment
with
Parents

Figure I : Conceptual Model

  
   

 
 

Quality of
Friendships

 
   

 

   
   
 

C Dating
Conflict
D
Quality of
E Romantic
Relationships
Social
Support
F
G

 

 
     
 

Psychological
Well-Being

Level of Voice
Identity Theory

Historically, both developmental psychologists and psychoanalytic theorists (e.g.,
Erikson, Freud, and Blos) have viewed the establishment of individuation and identity as
essential elements in the construction of the self. Jack (1991) discussed Freud’s theory Of
the self as autonomous and his idea that relationships are not valuable in and of
themselves but only in that they serve basic drives and needs.

Erikson argued that once identity has been established, intimacy can develop in
later adulthood (Erikson, 1963, 1968). Erikson’s popular theory of the self involves eight
developmental stages across the life span in which the initiation of the next stage is
dependent upon successful completion of the developmental “crisis” of the current stage.
After the first stage, trust versus mistrust, the goal of the next four stages, infancy through
late adolescence, is the individuation fiom the family of origin, and emotional autonomy.

Traditionally, adolescence was viewed as a time of separation from the family of
origin; Blos Viewed detachment from parents as necessary to overcome oedipal struggles
(Hill & Holmbeck, 1986). In addition, evolutionary views of intimacy emphasize the
importance of separation fiom family, and, specifically for males, the formation of
autonomy is paramount (Steinberg, 1988, as cited in W.A. Collins & Repinski, 1994).

Researchers have critiqued traditional views of identity that mandate detachment
in adolescence as a necessary prerequisite for identity formation (Frank, Avery, & Larnan,
1988; Hill & Hohnbeck, 1986; Ryan & Lynch, 1989). Hill and Holmbeck argue that
underlying the concept of autonomy are assumptions of detachment from parents and

freedom from parental influence, and note that “autonomy, as a label for freedom from

parental attachments and influence is misleading. Close relationships with parents not
only continue to exist during the second decade of life but parents continue to influence
their offspring” (p. 181).

More current conceptualizations of emotional autonomy in adolescence
acknowledge the importance of the family structure. While some researchers (Adams &
Marshall, 1996; Grotevant & Cooper, 1985) support Erikson’s traditional views that
identity formation involves the formation of a distinct self, and note that the individual
needs to feel significant and separate from the family of origin as part of the socialization
process, they also note that the self is constructed within a relational context. Indeed,
findings fi'om their own research support the importance and significance of both
individuality and connectedness within the family of origin (Grotevant & Cooper, 1985).
As a result, Grotevant and Cooper argue that emotional autonomy involves the
transformation of the adolescent’s relationship with parents, rather than the rejection of
the relationship.

The concept of emotional autonomy remains controversial and can be positively
associated with parental rejection, and negatively associated with attachment (W .A.
Collins & Repinski, 1994). Some researchers have argued that emotional autonomy, or
separation-individuation, in later adolescence is in fact two different factors: positive
separation feelings and independence from parents (Rice, Cole, & Lapsley, 1990).

The definition of emotional autonomy was reconceptualized by Steinberg and
Silverberg (1986) as different from detachment from parents; they argued that it can be
construed as a process of individuation, in which the adolescent “relinquishes childish

dependencies on, and conceptualizations of them [the parents]” (p. 843). Ryan and Lynch

(1989) argued that this definition of emotional autonomy was very negative, and was
analogous to detachment. Their research indicated that emotional autonomy was
associated with avoidant attachment to parents (Ryan & Lynch, 1989). Research
addressing Ryan and Lynch’s (1989) concerns that the context of the family was not
addressed in conceptualizations of emotional autonomy (e. g., Steinberg & Silverberg,
1986) demonstrated that adolescents who reported feeling supported by their parents were
better adjusted than those who did not feel such support (Lambom & Steinberg, 1993).
Therefore, despite theoretical assertions that increased emotional autonomy is related to
greater psychological well-being and decreased behavior problems (Steinberg &
Silverberg, 1986), research findings are more complicated, and suggest that for healthy
adolescent adjustment, a balance between individuation and support is optimal (Lambom
& Steinberg, 1993).

Hill and Holmbeck (1986) posited that traditional psychoanalytic theory, which
emphasizes detachment from the family of origin, is based primarily on males’
socialization experiences; separation from families may be more advantageous to boys
than girls. There are conflicting results about gender differences in emotional autonomy;
some researchers found that girls were more autonomous than boys (Larnbom &
Steinberg, 1993; Steinberg & Silverberg, 1986), while other researchers found that boys
were more emotionally autonomous than girls (Frank et al., 1988; Frank, Pirsch, &
Wright, 1990; Ryan & Lynch, 1989). One study found that although time spent with
family decreases over time during high school, time spent talking with family, especially
mothers, did not decrease. Indeed, for girls, the time spent discussing interpersonal issues

with their mothers increased (Larson, Richards, Moneta, Holmbeck, & Duckett, 1996).

This suggests that girls have different socialization experiences in terms of identity
development than do boys, and that separation from the family is not a prerequisite for
development.

Hill and Lynch (1983) proposed a “gender-intensification hypothesis” (p. 201).
They defined this as an acceleration of gender-differential socialization during early
adolescence, specifically around the time of puberty. According to this hypothesis, the
adoption of gender-specific beliefs and behaviors becomes increasingly apparent during
adolescence (Jones & Costin, 1995). The gender intensification hypothesis suggests that
“. . .behavioral, attitudinal, and psychological differences between adolescent boys and
girls increase with age and are the result of increased socialization pressures to conform
to traditional masculine and feminine sex roles” (Galambos, Almeida, & Petersen, 1990,
p. 1905). Hill and Lynch (1983) argued that this occurs especially for girls, and that girls
may feel increased pressure to conform to societal norms in areas such as achievement,
aggression, and emotional well-being.

Research generally supports this idea that in early adolescence, girls increasingly
conform to societal norms for women. Boldizar (1991) found that femininity declines in
the second through the sixth grade, and then increases in seventh grade, a time of
transition for many adolescents. Another study found that older adolescents, as compared
to younger adolescents, manifested greater differences between boys’ and girls’
endorsement of their relational characteristics, such as feeling responsible for the welfare
of others, and experiencing difficulty in being assertive in relationships (Aube, Fichman,
Saltaris, & Koestner, 2000). This supports Hill and Lynch’s (1983) premise that girls

adopt gender-specific beliefs in adolescence. Other studies have found no gender

differences in depression or self-esteem in childhood, however, in early adolescence, girls
experience an increase in depressive symptoms and a decrease in self-esteem (Simmons
& Rosenberg, 1975; Wichstrom, 1999). In addition, while one study found that the
relationship between self-image and achievement decreased from sixth to seventh grade,
supporting the gender intensification hypothesis, the relationship between self-image and
achievement increased fi'om seventh to eighth grade, which does not support the gender
intensification hypothesis (L.R. Roberts, Sarigiani, Petersen, & Newman, 1990). The
authors suggest that the gender intensification hypothesis may be temporary, and may be
especially salient in a time of transition.

As indicated earlier, identity in early adolescence does not necessarily involve
separation from the family of origin. In addition, the gender-intensification hypothesis
suggests that girls and boys have different socialization experiences in early adolescence
based upon society’s expectations for them. The next section will discuss feminist
critiques of identity theory and review relational theory, an alternative way of Viewing
females’ ways of relating to others. In addition, theory and research related to level of
voice will be discussed.

Feminist Critiques of Identity Theory

Some theorists have argued that theories of identity and emotional autonomy are
based on men’s experiences only (Gilligan, 1982; Jack, 1991; Jordan, 1997; Jordan,
Kaplan, Miller, Stiver, & Surrey, 1991). They argue that males and females have
different ways of understanding their identity, based upon both unconscious identification
with the mother as well as societal expectations. As a result of both of these processes,

girls and boys are socialized differently, and thus have different relational experiences.

Feminist theorists and researchers have also argued that Erikson does not fully
account for the development of intimacy in his hierarchical eight-stage lifespan theory of
development (Franz & White, 1985; Jordan, 1997). They argue that while Erikson’s
theory of development may neglect a female perspective, the most problematic weakness
is its neglect of interpersonal attachment as a crucial element of each stage of
development, and that a comprehensive theory of personality and identity development
must account for both individuation and interpersonal connectedness.

In addition, with the exception of the first stage, trust versus mistrust, the goal of
each subsequent stage until young adulthood is increased separation or individual self-
development (Miller, 1991). Intimacy is predicated upon the establishment of identity
(Jordan et al., 1991). Thus, when the individual reaches the sixth stage (intimacy vs.
isolation) in early adulthood, s/he is supposed to be intimate with another person, and
enter into a mutual relationship, while having spent the earlier stages of development
striving for individuation (Miller, 1991).

Franz and White (1985) argued that, with the exception of the trust and intimacy
stages, when Erikson refers to the “connectedness” of the individual, he interprets these
connections as institutional connections (e. g., family, school, society, etc.) rather than as
individual interpersonal attachments. Therefore, because Erikson neglects these
interpersonal attachments throughout the stages of his developmental theory, he fails to
adequately explain how an individual could become truly intimate and generative as an
adult (Franz & White, 1985).

Chodorow (1987) argued that as a result of the mother as the primary

identification object, boys and girls have different experiences in developing gender

10

identity. While boys must develop their gender identity as the “not-mother,” and thus
differences become salient to their identities, girls do not have to do this, and have more
of a sense of continuity of self.

Therefore, as a byproduct of their unbroken identification with their primary
caregiver, girls feel less separate than do boys, which results in girls “experiencing
another’s needs or feelings as one’s own” (Chodorow, 1987, p. 167). Furthermore,
because girls do not need to separate from their mothers to form a gender identity, they
retain their earliest identification with their mothers, “[they] not only reproduce
mothering but also are prepared to repeat the specific behaviors, attitudes, and values of
their mother’s nurturing role and to take on the cultural devaluation Of her female gender”
(Jack, 1991, p. 12).

Chodorow contended that autonomy does not have to be based on a self-other
distinction. It “can be based on the fundamental interconnectedness, not synonymous
with merger, that grows out of our earliest unconscious developmental experience and
that enables a nonreactive separateness” (Chodorow, 1987, p. 62). Jack (1991) argued
that individuals do not need to separate from others in order to form a sense of self. In
contrast, intimacy is necessary across the lifespan; as opposed to the traditional
psychoanalytic idea of separation, closeness with others enhances the development of the
self. Individuals are differentiated by the assertion of one’s agency as defined by needs,
feelings, and capacities within relationships (Jack, 1991).

One study tested Chodorow’s theory that development for males is equivalent to
separation fi'om parents, while development for women is akin to differentiated levels Of

interpersonal connectedness (Frank et al., 1988). The researchers found support for

11

Chodorow’s theory; overall, results indicated that young women were more connected to
their parents than were men, and the most difference in connectedness was between
daughters and sons in relation to their mothers. Daughters experienced significantly less
emotional autonomy in relation to their mothers than did sons. Women were more likely
to describe competent/connected relationships, while men described individuated
relationships more ofien.

Relational Theory

Psychologists who study women from a developmental perspective, a clinical
orientation, or a psychoanalytic point of View all agree that women’s orientation to
relationships is the central component of female identity and emotional activity (Jack,
1991). According to the relational view, the self is embedded in relationships.

These researchers propose that for women, the primary experience of the self is
relational. Women define themselves within the context of relationships, rather than
within the context of external accomplishments. The self-in-relation model suggests that
aspects of the self (e. g., creativity, autonomy) develop from within the context of
relationships, and that separation is not necessary to enhance and create the self (Jordan et
al., 1991).

Thus, for women, the ultimate ideal of being in relationships is the full realization
of relationships, involving intimacy, trust, and authenticity. Researchers and clinicians at
the Stone Center describe the importance of mutuality in relationships, and assert that to
have meaningful, authentic relationships, individuals must be able to represent their own
inner experiences in relationships, and, in turn, encourage authenticity in the other person

(Miller, Jordan, Kaplan, Stiver, & Surrey, 1997). This authentic mutual relationship

12

involves a balance of accommodation (self-sacrificing) and self-involvement (other-
sacrificing). They note that “mutuality involves commitment to engage in the
development and support of both people; it involves respectfully building a relationship
together that both sustains and transcends the individuals engaged in it” (Miller et al.,
1997, pp. 31-32).

Research indicates that individuals who identified themselves as having a mutual
style in their intimate relationship and identified their partners as also having a mutual
style had the highest levels of authenticity in their relationships (Harter et al., 1997b).
Individuals who identified themselves as having a style in which they were “other”
focused (i.e., their partner’s needs had a higher priority than their own), valuing intimacy
above all other aspects, and identifying their partner as having a “self” orientation in
which the partner was dominant in the relationship, reported the lowest levels of
authenticity within the relationship (Harter et al., 1997b).

While relational theory elegantly elucidates the importance of relationships for
females, relational theorists (e. g., Jordan, 1997; Miller etal., 1997), never explain the
process by which a mutual relationship occurs. While they make an important distinction
between being authentic in relationships, where mutuality is essential for a healthy
relationship, and being inauthentic in relationships, where mutuality breaks down and the
self is lost, they fail to detail the process by which one obtains, and hence maintains,
authentic and mutual relationships. Therefore, their theory does not address possible
factors that may help women maintain mutual relationships. As mentioned earlier, it is
possible that secure attachment and non-stereotypical gender role socialization could

result in girls being less likely to lose their voices. The following review on theory and

13

research related to level of voice, a manifestation of inauthenticity in relationships, will
discuss this construct, and then the following sections will discuss predictors and
outcomes of level of voice.

Theon and Research on Level of Voice

Relational theory explains the importance of relationships for females, yet girls
may struggle with maintaining them during adolescence (Miller, 1991 ). Miller argues
that given this conflict, girls will begin to lose “more of the sense that they can bring their
agency and sexuality, as they experience it, into the relationship” (Miller, 1991, p. 21).

Thus, girls may arrive at puberty with a different interpersonal orientation, from
identification with their mother (Chodorow, 1978), and a different range of social
experiences than boys, as a result of societal expectations, both of which may contribute
to the loss of voice in early adolescence. There are various motives (Harter et al., 1998b)
for losing one’s voice. F erninists assert that one motive is that our society values a “good
woman” who is unassertive, polite, pleasant and quiet (Gilligan, 1990). This creates
conflict for adolescent girls wherein they must resist this archetype in order to be faithful
to themselves, but must accept it to remain in their relational role, because the “good
woman” put others’ needs and desires ahead of her own.

Theorists elaborate on the connection between loss of voice and the “good woman
stereotype” and argue that a girl may be “caught between speaking what she knows from
experience about relationships and increased pressure to negate this knowledge for an
idealized and fraudulent view of herself and her relationships — the View rendered
credible by the possibility of being a perfect gir ” (L.M. Brown & Gilligan, 1992, p. 61).

Brown and Gilligan (1998; 1992) suggest that the powerful appeal of this “good woman”

14

stereotype is clear in that girls may feel that it promises them love and attention, and they
may bask in the recognition of being good girls. However, the cost of adopting this
stereotype is high; girls must subvert their sense of self, and may feel trapped in this role.
One girl noted tha “. .. at the end of the day I go home and I usually just yell because I
can’t — we’re always so quiet during the day” (L.M. Brown, 1998, p. 95).

Gilligan and others argue that girls hit a relational impasse at adolescence (L.M.
Brown & Gilligan, 1993; Taylor, Gilligan, & Sullivan, 1995). Girls have been socialized
all their lives to emphasize the importance of intimacy and relationships. When they
reach adolescence, it is assumed that they will continue to nurture and value relationships;
however, society does not value this, and does not reward them for their relational
approach. Instead, society values individualism, assertiveness, and independence. Thus,
girls are taught to both devalue relationships and achieve independence and autonomy
(Steiner-Adair, 1990). An example from one of Brown’s qualitative studies (1998)
elucidates the tension between relationships and achievement when two best friends take
the same exam, and one girl gets a better grade than the other. The girl who received the
lower grade says that she feels jealous of her fiiend, and wishes that her fi'iend had also
gotten a lower grade. Therefore, authentic relationships may suffer at the cost of
achievement, and Brown (1998) notes tha “. .. the girls move out Of psychological
relationship with one another in their search for individual merit and the highest
honors...” (p. 127).

Thus, according to level of voice theory, girls lose either way, in that they either

lose their relationship with their inner self to preserve relationships, or they give up

15

relationships with others to become self-sufficient (Taylor et al., 1995). The fiiction and
conflict between the two results in girls’ ambivalence and subsequent loss of voice.

Therefore, Gilligan and her colleagues argue that during this relational impasse,
girls have to decide to be selfish (i.e., become independent and self-sufficient) or selfless
(if they remain responsive to others), and girls often choose to be selfless (L.M. Brown &
Gilligan, 1992; Gilligan, 1990). According to theory, to remain in relationships, the
“good woman” puts others’ needs ahead of her own, and the “good woman” ends up
dissociating from her voice to live in society, and gives up herself to maintain these
relationships (L.M. Brown, 1998). Adolescent girls may recognize that this is
paradoxical, and struggle with how to stay in touch with themselves and others.

Thus, this theory suggests that girls, as the result of both a relational impasse in
early adolescence and identification with their mothers, may compromise their true selves
in relationships for the sake of preserving relationships, and inauthentic relationships
result (Gilligan et al., 1990). Girls may recognize that to speak their minds might
threaten relationships which might lead to abandonment (Harter et al., 1998b). An
example of this is a girl in one of Brown and Gilligan’s (1992) studies who at age 12 was
not afiaid to voice her opinion, and was savvy about her relationships with others, but at
age 13, when faced with a conflict with a peer, did not know how to act because she “had
to be nice,” since the peer was fiiends with her family. Unlike during childhood, when
she spoke her opinions and could distinguish between what authorities said and what she
knew to be true, in adolescence, she did not know how to express her true feelings and
how to balance her inner conflicts with the priorities of relationships. Ultimately, she

removed herself from the relational conflict to preserve relationships, and stifled her

16

feelings of dislike. By removing herself from the conflict, she dissociated from her true
self, and inauthentic relationships resulted (L.M. Brown & Gilligan, 1993).

The theory grounding the loss of voice in adolescence is both dynamic in nature,
in terms of its origins in the identification with the early maternal figure (Chodorow,
1987), and systemic in nature, in terms Of girls’ relational impasse (L.M. Brown &
Gilligan, 1992; Steiner-Adair, 1990; Taylor et al., 1995). Gilligan’s work (L.M. Brown &
Gilligan, 1992; Gilligan et al., 1990) involves girls discussing how the loss of voice
affects their involvement in relationships, specifically their literal voice. However,
Gilligan’s work is controversial, in that her data is not accessible to others (Sommers,
2000), she does not give any empirical evidence of her findings, and her research has
primarily involved middle-class to upper-class Caucasian girls. Thus, it is difficult to
generalize her theories to the population of adolescent girls at large.

Harter and colleagues are some of the first researchers to examine the idea of loss
of voice both empirically and quantitatively, which has improved our understanding of
level Of voice theory, given the above critiques of Gilligan’s work. Harter and her
colleagues View the loss of voice, or false-self behavior, as manifested by the presence or
absence of specific verbal behavior, such as not saying what one thinks, not expressing
opinions, or saying what one thinks others want to hear (Harter et al., 1998b).

During mid-adolescence, Harter et a1. (Harter, Bresnick, Bouchey, & Whitesell,
1997a) found that both boys and girls have more of the cognitive skills to list possibly
contradictory attributes about the self, but still do not have the maturity to integrate these

into a unified self. During late adolescence, individuals had the capacity to integrate

17

possible contradictions within the self, and perhaps even found it desirable to act
differently with different people.

In addition, it was found that the most commonly reported conflictual attributes
across roles were with mothers versus fathers (Harter et al., 1997a). Harter and her
colleagues also found that there was conflict between attributes related to self with
parents versus self with peers. Interestingly, some gender differences were found in terms
of emotions related to conflictual self-descriptions; girls were more upset than boys,
across age groups, over conflicting attributes. Harter et al. (1997a) reported that
adolescent girls, particularly those with a more feminine orientation, may struggle over
whom is the “real self.” While those in early adolescence (i.e., sixth grade) reported that
they were always their “true selves,” seventh graders can understand the salience of the
true self versus the false self. Adolescents’ descriptions of their true selves included “‘the
real me inside,’ ‘my true feelings,’ ‘what I really think and feel,’ ‘behaving the way I
want to behave and not how someone else wants me to be.’ False selves were described
as ‘being phony,’ ‘putting on an act,’ ‘expressing things you don’t really believe or feel,’
‘changing yourself to be something that someone else wants you to be.”’ (Harter, Marold,
Whitesell, & Cobbs, 1996a, p. 360).

Rather than construing voice as a universal construct, Harter’s research indicates
that voice is context dependent (Harter et al., 1997c; Harter et al., 1998b). Harter and
colleagues (Harter et al., 1997c) found that adolescents reported the highest levels of self-
reported false self behavior with their fathers (30-40% Of attributes), lower levels of false
self behavior (20-25%) with classmates, teachers, and their mothers, and the least false

self behavior with close friends (10-1 5%). Another study by Harter and colleagues

18

(1998b) found that girls, compared to boys, had higher levels of voice with female
classmates and close fiiends; for boys and girls, level of voice with close friends was
higher than other relationships. In addition, results fiom a qualitative study indicated that
girls had higher levels of voice in private rather than public settings. One girl noted that
she learned from her mother tha “. . .when I’m at home I can be myself but when I’m at
school I’m supposed to behave and stuff” (L.M. Brown, 1998, p. 151).

When asked why they engaged in false self behavior, adolescents reported three
main reasons. First, they noted that they wanted to please others, impress others, and/or
gain acceptance from others. Second, they described an alienation process from their true
self, due to lack of validation from others. Finally, they reported that they were
experimenting with different versions of themselves, akin to trying on different styles of
clothing, and were trying to figure out which self was the best fit (Harter et al., 1996a).
When asked to choose one of the above reasons as the most important explanation for
false self behavior, approximately 60% of early adolescents chose the first reason, 30%
chose the second, and 10% chose the third. Interestingly, results were similar across
gender, and were similar for false self behavior with parents and peers. In addition,
adolescents who reported false self behavior because they or others devalued their true
selves reported more false self behavior than did adolescents who endorsed the other two
explanations for false self behavior.

Harter and colleagues (1998b) disagree with Gilligan that this phenomena occurs
with all girls, and argue that there is more variation in voice within gender than between
genders. Harter et al.’s (1998b) study found that femininity and sex-role socialization,

rather than gender, were more salient aspects of voice. Fenrinine girls reported lower

19

levels of voice in public versus private settings, as opposed to androgynous girls, whose
level of voice did not differ within different contexts. Thus, gender itself does not seem
to predispose one to risk loss of voice; rather, gender orientation, specifically femininity,
may be the more salient risk factor.

Finally, several qualitative studies have indicated that loss of voice may be less
likely to occur with girls from ethnic minority groups (Clark, 1999; Way, 1995) and girls
from lower socioeconomic backgrounds (L.M. Brown, 1998). Way (1995) found that by
junior or senior year in high school, ten of the 12 adolescent girls interviewed, all of
whom were ethnic minorities and working class, felt that they were open and honest in
many of their relationships. Way notes that “these outspoken voices were focused not
only on expressing anger and disagreement in relationships, but also on voicing care and
connection” (p. 107). Way suggests that these girls may have less difficulty speaking
their minds because they were not raised to be as passive as White girls. Way also notes
that ethnic minority parents may socialize their daughters differently, and raise them to be
strong and outspoken, because they recognize that “if their daughters are passive and
quiet, they may simply disappear in a society in which they and their daughters are
already pushed to the margins” (p. 124).

Brown’s (1998) study indicates that girls from lower socioeconomic backgrounds
were less likely to lose their voices, and were more likely to express their anger in school
settings. However, similar to the relational theorists’ inability to demarcate the ways for
women to have mutual relationships, Brown does not analyze why girls from lower

socioeconomic backgrounds may be less likely to lose their voices, which is a significant

20

weakness of her study. As indicated earlier, research in general has failed to examine
individual and societal factors that might precipitate loss of voice.

Thus, research indicates that Gilligan’s idea of the “good woman” seems to appear
only in certain contexts, andwith feminine girls. In addition, it may be less likely to
occur in girls of different ethnic and socioeconomic backgrounds. However, Harter’s
research has been focused on middle-class white girls, and the studies by Brown (1998),
Clark (1999), and Way (Way, 1995) had very small samples. In addition, in Way’s work,
the majority of the girls had gained their voices by late adolescence, and she did not talk
to them when they were in early adolescence, at a time of transition.

While all of the preceding theory and research on level of voice involves
adolescent girls, Jack (1991) argues that the relational conflicts that girls experience in
adolescence continue to affect women through adulthood. Jack’s (1991) silencing the self
theory suggests that women who have inauthentic relationships as a result Of loss of voice
may be more vulnerable to depression in adulthood. She suggests that their inner conflict
between their actual and ideal self contributes to depression. From her qualitative
interviews with women, Jack developed a scale, the Silencing the Self Scale (STSS) to
quantitatively examine self-silencing (Jack & Dill, 1992).

There have been inconsistent gender differences for adults when the STSS was
utilized. Several studies (Duarte & Thompson, 1999; Gratch, Bassett, & Attra, 1995;
Koutrelakos, Baranchik, & Damato, 1999) found that men displayed higher levels of self-
silencing than did women, depending upon the subscale, while other research indicates
that men and women do not differ (Hart & Thompson, 1996; Spratt, Sherman, & Gilroy,

1998). However, Duarte and Thompson (1999) hypothesized that men and women self-

21

Silenced for different reasons. They suggested that boys are socialized to suppress
emotion and intimacy, and that this suppression has less negative consequences for men
than for women. Duarte and Thompson (1999) found that for women, the subscales Care
as Self-sacrifice and the Divided Self were significantly correlated, providing preliminary
support that when women sacrifice their own needs for the sake of others’ needs, they
may experience a loss of self. However, the relationship between these two subscales
was not significant for men, suggesting that the consequences for men’s self-sacrificing
behavior are less negative than for women.

Ethnic differences are also evident from the STSS. One study found that when
comparing African-Americans, Asians, Caucasians, and Hispanics, Asians were far more
likely to indicate self-silencing than were any other ethnic group (Gratch et al., 1995).
Yet another study found that while scores on the STSS for Caucasian and Afiican-
American women did not differ significantly from each other, STSS scores were not
related to depression for African-American women, but were related significantly to
depression for Caucasian women (Carr, Gilroy, & Sherman, 1996). This finding suggests
that the process of losing one’s self is manifested differently in different cultural contexts;
given that the loss of self may be a cultural and societal phenomenon, women in more
collectivist societies may experience loss of voice as a normative process. Carr et a1.
(1996) suggest that the negative repercussions from the loss of self may, in part, be
specific to Western patriarchal cultures, with the emphasis on independence and
separation. Therefore, it is clear that to date, research studies have lacked a thorough
understanding of the antecedents of loss of self, and how this is differentiated in different

ethnic populations. It is apparent that the concept of loss of self is heterogeneous, and

22

cannot be generalized to every ethnic group and population. In addition, as stated earlier,
all research utilizing the STSS has been conducted with adults.

In summary, research indicates that not all girls and women are vulnerable to loss
of voice and self-silencing. However, the majority of both qualitative and quantitative
research has involved upper-middle-class and middle-class white girls (L.M. Brown &
Gilligan, 1992; Gilligan et al., 1990; Harter et al., 1998b), with a few exceptions in the
qualitative literature (L.M. Brown, 1998; Taylor et al., 1995; Way, 1995). The study of
white middle-class girls makes it difficult to generalize findings to the general population.
Therefore, the current study contributes to the field by using a population of girls in
which ethnic minorities are represented, and all the participants are from lower
socioeconomic backgrounds.

In addition, those who displayed the highest levels of false self behavior reported
lower self-esteem and higher depressive symptomatology (Harter et al., 1996a).
Therefore, those who experience high levels of false self behavior are more at risk for
negative mental health outcomes, thus supporting Gilligan and colleagues’ (1992; 1990)
argument that loss of voice has negative repercussions. The relationship between loss of
voice and psychological well-being will be discussed later.

Finally, research on loss of voice has been promising, however, little work has
been done to examine predictors and outcomes Of voice; this lack of research results in a
vague understanding of what might cause loss of voice, and how loss of voice might
affect girls’ lives. There is little understanding of why some girls and not others are
vulnerable to loss of voice, which is especially important given that qualitative and

quantitative research both indicate that not all girls lose their voice in early adolescence

23

(Harter et al., 1998b; Way, 1995). The next two sections will discuss attachment and
gender role socialization as predictors of level of voice.
Predictors of Level of Voice
Attachment

Attachment theory suggests that infants are motivated to seek nurturance from the
mother, and that the success of the reciprocal relationship is thought to deternrine later
responses to security threats (Bowlby, 1969/ 1982). Thus, when a caregiver, primarily the
mother, is available and sensitive to an infant’s needs, the infant develops a sense of the
mother as a secure base from which to explore the outside world, and the infant is
comforted by the mother when upset. As adolescents and adults, those with secure
attachment will feel comforted by those around them, value caregiving, and enjoy and
value reciprocal and mutual relationships (F urman, Simon, Shaffer, & Bouchey, 2002).

However, infants without the foundation of a healthy relationship with their
mother develop insecure attachment, which includes both insecure-avoidant and insecure-
ambivalent attachment. Those in the third insecure subtype, disorganized attachment,
lack a coherent behavioral strategy as infants; this classification will not be discussed in
this review as it is a small minority of those assessed [as little as 4% of an adolescent
sample (Furman et al., 2002)] and most of the adolescent and adult attachment research
has been conducted using the three group model (F eeney, 1999).

Children in the insecure-avoidant subtype lack confidence in their caregivers, and
have expectations of rejection; in adolescence and adulthood, this is termed “dismissing”
attachment, and an adolescent may have little interest in seeking help from parents, and

not perceive fiiends and family members as reassuring in stressful times (Furman, 1999).

24

Children in the insecure-ambivalent subtype typically have caregivers who were
inconsistently and unpredictably responsive and available to their child’s needs, and
consequently may distort their attachment needs to receive a response fi'om a caregiver
(Ainswortlr, Blehar, Waters, & Wall, 1978). In adolescence and adulthood, this is termed
‘preoccupied” attachment, and someone with this attachment may find it difficult to feel
comforted when upset, may be too concerned about others’ problems, and “may
overinvest in relationships in a self-sacrificing manner” (Furman, 1999, p. 142). The
preoccupied group is so emotionally hungry for nurturance that they will adopt a strategy
of compliance to feel accepted (Feeney, 1999).
Research on Attachment

Bowlby (1969/1982) argued that interpersonal connectedness plays an important
role throughout development, not just in infancy. Research does indicate that parent-
child attachment is important throughout the lifespan (Greenberg, Siegel, & Leitch,
1983), and the quality of parent-child attachment at infancy has been shown to predict
later social functioning in childhood and adolescence (Elicker, Englund, & Sroufe, 1992;
Shulman, Elicker, & Sroufe, 1994; Sroufe, Carlson, & Shulman, 1993; Weinfield,
Ogawa, & Sroufe, 1997).

Sroufe and his colleagues have conducted a longitudinal study since the early
1970’s that has examined the effects of infant attachment on aspects of functioning with
teachers, peers, and romantic partners. They argue that personality is developed from,
and organized by, attachment relationships (Sroufe & Fleeson, 1986). In addition, they
suggest that the consistency and continuity of attachment over time is due to the

transactional nature of attachment, and stated “. . .a child who has positive expectations

25

concerning others, feelings of worth and confidence, and a conception of relationships as
responsive and mutually enhancing engages and responds to peers in positive ways,
expects and elicits positive and age-appropriate support from adults, and stretches his or
her abilities in setting goals and meeting challenges” (Sroufe et al., 1993, p. 334).
Therefore, insecurely attached girls may be more likely to have low levels of voice, in
that they do not have a secure base of relationships which would allow them to be honest
and open about threats to relationship.

Secure attachments in infancy lead to intimacy in later relationships for three main
reasons. First, a relationship in which the primary caregiver is responsive and empathic
toward the child would lead the child to expect others to react similarly (W .A. Collins,
Hennighausen, Schnrit, & Sroufe, 1997). Second, such caregivers are role models for
how to be empathic and have mutually reciprocal interactions (Sroufe & Fleeson, 1986).
Third, responsive care and autonomy support strengthens a sense Of self-efficacy, which
would make an individual attractive to others (Elicker et al., 1992). Thus, adolescents
with secure attachment may be less likely to lose their voice. Given their experiencse
with responsiveness, empathy, and mutually reciprocal relationships, these adolescents
may not feel threatened by challenging assertions of gender stereotyping, and may have
safe mutual relationships that allow them to be honest in relationships.

Research with children supports Bowlby’s (1969/1982) theory that secure
attachment predicts later self-reliance, and that insecurely attached children are more
dependent (Sroufe et al., 1993; Turner, 1991). In addition, during preschool, the
friendships of avoidant children were less likely to be characterized by mutuality,

responsiveness, and affective involvement. In later childhood, avoidant and resistant

26

children were rated as more dependent than secure children in behavioral observations
with peers. Research indicates that attachment security at infancy was positively
correlated with social skills in middle childhood (W einfield et al., 1997). Therefore, it
seems that since insecure children lack mutuality and responsiveness in their peer
relationships, they may have lower levels of voice given that these two elements are
essential for mutual relationships (Miller et al., 1997).

In middle-childhood, children who were securely attached were more socially
competent, and were more likely to form fiiendships (Elicker et al., 1992; Shulman et al.,
1994). In descriptions Of dyads of children with the same attachment history, there were
remarkable differences in children’s conflict management and peer interaction styles
(Shulman et al., 1994; Sroufe et al., 1993). Secure children were found to have a “semi-
perrrreable membrane” in that they played near others, invited others to join their groups,
and were not threatened by a friend engaging in activities with others (Sroufe et al.,
1993). With a secure-secure pair of children, when one of the children felt abandoned by
the other child, she talked to this friend and asked her if they were still fiiends (Shulman
et al., 1994). Shulrnan and colleagues (1994) described the interaction, and noted that
“they argued, they exchanged points of View, each directly expressing emotion” (p. 350).
In this way, the girls they describe are demonstrating high levels of voice, and are
indicating their ability to discuss relational issues. Irnportantly, the conflict ended with
the second child reassuring the first, and both girls appeared to be happy, and to have
attained a new level of closeness.

In contrast, avoidant children were found to have a “non-permeable membrane” in

that they only played together, were jealous of others, and when one was absent, the other

27

seemed lost (Sroufe et al., 1993). A pair of children with insecure-avoidant attachment
histories appeared to be close, but there was little affect between them, and no conflict
arose (Shulman et al., 1994). Shulman and colleagues (1994) describe this pair as having
“a ‘raw’ less sophisticated need for intimacy” (p. 356) that attracts them to each other.
However, there was little exploration of the relationship, and the avoidant pair was
described as having a “protected closeness” (p. 356) that obfuscated the emergence, let
alone the resolution, of conflicts.

Resistant children were found to have a “totally permeable membrane” in that
they had problems sustaining relationships in groups; one might become absorbed by the
group and leave the other behind, or one might become overwhelmed by the difficulty of
coordinating multiple relationships and become separated from the group (Sroufe et al.,
1993). The lack of conflict in the resistant pair is similar to the avoidant pair, but in
contrast, the resistant pair is described as having a sporadic friendship, in which genuine
closeness is apparent, but is not maintained (Shulman et al., 1994). Because of the
sporadic nature of their relationship, the children never fought, and thus experienced
neither conflict resolution, nor the closeness post-conflict that the secure pair
experienced. This supports the concept that adolescents with insecure attachment may
have lower levels of voice, in part due to their inability to discuss conflict. In addition,
this inability to discuss conflict may hamper their intimacy with others.

These studies (Shulman et al., 1994; Sroufe et al., 1993) support assertions
(Sroufe & F leeson, 1986) that patterns of relationships with early caregivers continue into

development. Children with insecure attachment did not develop long-term meaningful

28

relationships, and displayed unrelatedness, while the avoidant pair described above had
little personal agency, and demonstrated a clear inability to explore their relationship.
There is little research on attachment in adolescents; this is a significant deficit
given that adolescence is a time of transition and change (L.S. Brown & Wright, 2001).
Research that has been conducted has found that secure adolescents had a greater capacity
for emotional vulnerability and sustained dating relationships (W .A. Collins & Sroufe,
1999). Secure attachment from infancy was significantly related to intimacy in peer
relationships in adolescence (W .A. Collins & Sroufe, 1999). In mid-adolescence, at age
15, secure adolescents had the capacity to be vulnerable with others; that is, they were
able to be open to feelings, and were open to a range Of experiences. Adolescents’
capacity for vulnerability distinguished insecure and secure groups (Sroufe et al., 1993).
This capacity for vulnerability seems to be an essential element of being able to express
one’s voice; one must be able to trust in another, and feel accepted. Therefore,
adolescents who do not have this may be more vulnerable to loss of voice.
Cross-sectional research indicates that attachment to caregivers is salient in terms
of psychosocial adjustment in adolescence. Attachment to parents is positively correlated
with general self-concept and self-esteem (Armsden & Greenberg, 1987; Kems, Klepac,
& Cole, 1996; Salzman, 1996), especially in a time of transition (Papini & Roggrnan,
1992). Attachment to mothers and fathers is negatively related to depressive symptoms
(Armsden, McCauley, Greenberg, Burke, & Mitchell, 1990; Cole-Detke & Kobak, 1996;
Kenny, Lomax, Brabeck, & Fife, 1998; Kobak, Sudler, & Gamble, 1991; Salzman, 1996),
anxiety (Armsden et al., 1990; Kenny et al., 1998), and eating disorders (Cole-Detke &

Kobak, 1996).

29

Attachment is also salient in respect to adolescents’ relationships with parents,
peers, and romantic partners. Attachment to parents is positively correlated with parent
support (Cotterell, 1992). Security is positively correlated with social acceptance, as
reported by peers (Allen, Moore, & Kuperminc, 1997), and, in addition, securely attached
adolescents show greater social competence with peers, and are more likely to make
fiiends, have securely attached friends, and be more popular (Doyle & Markiewicz,
1996). Attachment security is related to satisfaction with best fiiend (Levendosky, Huth-
Bocks, & Semel, 2002), and adolescents who perceive their mothers as available and
have a sense of autonomy have closer, supportive, and more secure friendships
(Lieberman, Doyle, & Markiewicz, 1999). In addition, attachment to caregivers is
somewhat predictive of later attachment to romantic partners (Blain, Thompson, &
Whiffen, 1993; N.L. Collins & Read, 1990; Owens et al., 1996). Therefore, this evidence
supports assertions that attachment to caregivers continues to be salient throughout
development (Sroufe et al., 1993; Sroufe & F leeson, 1986).

Attachment may be especially salient during adolescence; Armstrong and Roth
(1989) argued that a sense of security is made up of layers of interactions, and, thus, a
secure attachment is beneficial during the separation anxiety period during adolescence
when struggles with identity formation and intimacy issues are salient.

Connection between Loss of Voice and Attachment

The mechanisms for transmission of voice may be societal (e.g., Gilligan et al.,
1990), in terms of girls’ experiences of a relational impasse, but loss of voice may also be
related to insecure attachment. As Bowlby (1969/1982) suggests, the initial mother-

infant relationship creates an Internal Working Model (IWM), which is the unconscious

30

model of the self and other that represents the early attachment relationship between the
infant and the caregiver. IWMs affect other relationships, and help determine the ability
to be intimate. Thus, the mother-daughter relationship represents the initial model of a
relationship in a child’s developmental trajectory, and affects all subsequent relationships
in the child’s life (Surrey, 1991).

Harter and colleagues (Harter et al., 1997a) suggest that the origins of early false
self behavior may be found in early attachment relationships. They note that parents help
children construct a narrative of the child’s inner experience, and create an
autobiography. Harter and her colleagues speculate that if parents convey to their
children that certain experiences should be forgotten or erased from their memories, such
a distortion may contribute to a child’s false self behavior.

No empirical research has been conducted to examine the relationship between
attachment and loss of voice in adolescence, although two studies that examined this
relationship in adults found that preoccupied attachment was associated with silencing of
the self, using Jack’s STSS scale (Remen, Chambless, & Rodebaugh, 2002; J .M.
Thompson & Hart, 1996). Thus, at least with adults, this is preliminary evidence of a link
between insecure attachment and loss of voice.

In one of Carol Gilligan’s studies (Gilligan et al., 1990), a researcher conducted
qualitative interviews with nine girls with insecure attachment. Girls who were in a
subgroup analogous to preoccupied attachment, as assessed by the author through a
qualitative interview, were harsh judges of themselves and had low self-confidence
(Salzrnan, 1990). These girls struggled to admit to feelings of anger, and were reluctant

to admit that their needs to be taken care of remained unfulfilled. Preoccupied girls often

31

had to sacrifice their own needs to take care of others, especially their mothers, to the
extent that they were no longer aware of their own burgeoning needs. Another qualitative
study (Salzman, 1997) found that girls who were preoccupied reported that the most
undermining quality of their relationship with their mothers was the strength and
unpredictability of their mothers’ attacks on their self-esteem. In a sense, they lost their
own voices, and according to Salzrnan, in adolescence, “listen(ed) to my mother’s voice,”
(p. 254) which was extremely negative and denigrating.

Salzrnan (Salzman, 1990) wrote that girls who were in the dismissing group
learned “to fear revealing themselves to people whom they care about, people with the
power to hurt them. As a result, they have cultivated the habit of censoring certain truths
about themselves within relationships, of holding a lot inside, of hiding their inner
vulnerability” (p. 139). Thus, they were less likely than the preoccupied group to have
the ability to have truly intimate relationships, and place value on relationship. One girl
noted “I am not honest even to those closest to me, because I don’t want to get that
involved in a relationship. I don’t want to trust that much in a relationship, so I don’t
really allow them to see the real me, and I am constantly lying about the person that I am
and how I feel” (Salzman, 1990, pp. 139-140).

Despite the paucity of research linking attachment and level of voice, there are
several empirical findings fi'om studies examining attachment in adolescents that support
the connection between attachment and level of voice. These studies indicate that
insecure female adolescents are less assertive, and have more gender-stereotyped
behavior, and that insecure adolescents overall have more rejecting and critical mothers,

and have poorer communication and conflict management skills than secure adolescents.

32

Research indicates that insecure girls are less assertive than secure girls (Turner,
1991). Specifically, one study found that college-age students who were classified as
preoccupied were less assertive than dismissing or secure adults (Kobak & Sceery, 1988).
In addition, secure adolescents also had balanced aggressiveness with their mothers in
problem-solving discussions (Kobak, Cole, Ferenz-Gillies, Fleming, & Gamble, 1993),
and were not dominated by their mothers as were the insecure adolescents.

Traditional gender role socialization for girls has been associated with lower
levels Of voice (Harter et al., 1998b). Interestingly, one study found that pairs of fiiends,
where one was secure and one was insecure, demonstrated more gender-stereotyped
patterns (i.e., they complained more about their mothers) of self-disclosure than did pairs
of fiiends who were both secure (Kerns et al., 1996). Another study found that insecure
girls also had higher levels of positive expressive behavior than did secure girls (T urner,
1991). Turner interpreted this expressive behavior as a way of avoiding conflict, and
suggested that insecure girls may do this to maintain peer groups. This provides more
evidence for the relationship between attachment and voice, as girls with insecure
attachment display similar communication patterns (i.e., more traditional gender-
stereotyping) as do girls with low levels of voice.

Gilligan notes that some girls may not resist the relational crisis in adolescence
because they do not have a significant adult figure with whom to connect, thus resulting
in lower levels of voice (Gilligan, 1990). One research study found that insecure
adolescents who experience their attachment figure as rejecting deactivate from
attachment, thereby minimizing their distress (Cole-Detke & Kobak, 1996). As a result

of deactivating from attachment, girls may also lower their voices to maintain

33

relationships. Another research study found that insecure attachment mediated the
relationship between maternal coldness and self-criticism (R. Thompson & Zuroff, 1999).
Therefore, insecure attachment is related to girls’ being more self-critical. This self-
critical behavior may be related to level of voice, in that girls who are more self-critical
would be more susceptible to low level of voice

Research indicates that secure adolescents are better at communicating than are
insecure adolescents, and that security was significantly related to reports of
companionship and validation in their best fiiendships (Kerns, 1996). Secure adolescents
had constructive problem-solving strategies that were less avoidant and were
characterized by less dysfunctional anger (Kobak et al., 1993). In addition, pairs of
friends where both were secure were observed to be more responsive and less critical in
behavioral observations than were pairs of fiiends where one was insecure and one was
secure (Kerns et al., 1996). In addition, insecure attachment was related to poorer conflict
management skills (Schneider & Younger, 1996); this indicates that insecurely attached
girls would be likely to have lower level of voice, as they would not be able to discuss
problems and negotiate conflicts.

Kerns and her colleagues noted that secure attachment involves responsive care
from an attachment figure, and Open and relaxed communication between attachment
figure and child, involving both negative and positive affect (Kerns, Tomich, Aspelmeier,
& Contreras, 2000). Therefore, since these children are able to integrate both positive
and negative affect, they will be more likely to speak their mind about relationships, and

be less likely to lose their voices. Also, securely attached children reported feeling less

34

lonely (Kerns, 1996; Kerns et al., 1996); this is another connection between attachment
and voice in that if girls have low levels of voice, they will likely be more lonely.

Girls who are insecurely attached in Gilligan’s previously mentioned qualitative
studies (Gilligan et al., 1990) Viewed the expression of anger as either dangerous or futile,
with the consequence of being unable to resolve conflict in relational situations (Salzman,
1990). These girls deal with anger by either denying the existence of conflict or
removing themselves from the conflictual situation. Given their problematic attachment
histories, they are unable to use family attachment as a context for resolving conflict.

In summary, it is clear that attachment is important throughout the lifespan
(Greenberg et al., 1983; Sroufe & Fleeson, 1986), and that quality of attachment is related
to intimacy, vulnerability, and mutuality in peer and romantic relationships in childhood
and adolescence (W.A. Collins & Sroufe, 1999; Greenberg et al., 1983; Shulman et al.,
1994; Sroufe & Fleeson, 1986). In addition, studies regarding attachment in adolescence
have found that insecure female adolescents were less assertive and more likely to have
gender-stereotyped behavior (Kerns, 1996; Turner, 1991), both of these characteristics are
similar to manifestation of low levels of voice in girls (L.M. Brown & Gilligan, 1993;
Harter et al., 1998b). Finally, research indicates that insecure adolescents have poorer
communication styles and poorer conflict management skills than do secure adolescents
(Kerns, 1996; Kerns et al., 1996; Kerns et al., 2000; Kobak et al., 1993; Schneider &
Younger, 1996). Therefore, while no studies to date have examined the relationship
between attachment and loss of voice in adolescents, there is preliminary evidence that

the two may be related. The current study contributes to the field by being the first to

35

examine the relationship between attachment and loss of voice in adolescence (see
Pathway A in Figure 1).

As discussed earlier, loss of voice may be related to insecure attachment. In
addition, there may be a societal influence of voice upon girls, as a result of the relational
impasse discussed by Gilligan and her colleagues (L.M. Brown & Gilligan, 1993;
Gilligan et al., 1990). The next section will discuss the relationship between gender role
socialization and level of voice (see Pathway B in Figure 1).

Child 's Gender Role Socialization

Gender role socialization can be defined as the way in which we are socialized to
internalize different personality attributes that have been culturally defined to represent
either masculinity or femininity (Boldizar, 1991). In addition, there are individual
differences in the extent to which individuals internalize attributes stereotypically related
to their own gender or the other gender. Research does support the existence of
masculinity and femininity as two separate components (Spence, Helrnreich, & Stapp,
1975), although they are not necessarily dichotomous (Unger & Crawford, 1992).
Androgyny is attained when someone intemalizes attributes stereotypically related to both
genders (Unger & Crawford, 1992).

Researchers have examined three different models of gender role socialization as
they relate to psychological well-being in adolescents (Cate & Sugawara, 1986; Larnke,
1982; Whitley, 1983). The congruence model suggests that the best outcome is when
one’s gender role is congruent with one’s gender. The androgyny model suggests that the

best outcome is when one is high on both masculinity and femininity. Finally, the

36

masculinity model suggests that the best outcome is when one has a masculine sex role
orientation (Whitley, 1983).

Research has generally supported the masculinity and androgyny models, both of
which involve the creation of categories Of gender role socialization, based upon level of
masculinity and femininity scores (Markstrom-Adams, 1989; Whitley, 1983). However,
in general, when the androgyny model is examined with continuous scores, it is found
that the contribution of masculinity is the most salient predictor of psychological well-
being, rather than the dual contribution of both masculinity and femininity (Boldizar,
1991; Cate & Sugawara, 1986; Lamke, 1982; Wells, 1980). Thus, research indicates that
it is best to utilize both categorical and continuous methods of classifying gender role
socialization (Cate & Sugawara, 1986; Lamke, 1982; Wells, 1980).

One interesting study found that masculinity is more salient for females than for
males (Cate & Sugawara, 1986). For males, being high or low on masculinity did not
affect their self-esteem, but for females, masculinity was correlated with greater self-
worth. The authors suggested that for females, higher masculinity makes them feel
valued by society, but for males, simply the status of being male in society is sufficient
enough to impact their self-esteem positively.

As discussed earlier, Gilligan argues that female adolescents experience a crisis of
identity when they are pressured by society to accept the “good woman” stereotype,
which emphasizes caring and relatedness even in the face of self-sacrificing. However, as
discussed earlier in relation to the literature on gender role socialization, adolescent girls
vary in terms of their gender role socialization, and it is likely that not all girls accept the

“good woman” stereotype. Indeed, one study found that only 38% of adolescent girls

37

could be classified as feminine (Rose & Montemayor, 1994). Therefore, it is possible,
given the societal pressure that underlies loss of voice, that high femininity, rather than
gender per se, might predict level of voice and that girls would differ in their level of
voice as a function of their gender role socialization.

Harter et al. (1998b) found this to be the case. In addition, the context of voice
mattered; in private contexts with close friends and family, feminine girls and
androgynous girls did not differ in their levels of voice, while in public settings, such as
in school, feminine girls had significantly lower levels of voice than androgynous girls
(Harter et al., 1998b). However, given the literature that suggests it is best to examine
gender role socialization as two separate, continuous variables (masculinity and
femininity) and that information and statistical variance are lost by categorizing gender
role socialization (Boldizar, 1991; Hoffman & Borders, 2001; Spence et al., 1975; Wells,
1980), it seems surprising that Harter and her colleagues did not examine the independent
contributions of masculinity and femininity.

In one of the few studies examining the connection between level of voice and
gender role socialization, Tolrnan and Porche (2000) found that, for eighth grade girls,
there was a connection between the femininity scale of the Bern Sex Role Inventory
(BSRI) and their measure of voice, although it was only a trend. Another study found
that both the femininity scale and the masculinity scale from the BSRI were correlated
with silencing of self in adolescents (Hart & Thompson, 1996). However, it is likely the
BSRI is not the best means of measuring gender role socialization in adolescents for two
reasons. First, the BSRI was normed on adults, and second, it was created in the 1970’s;

society’s conceptions of gender roles have changed significantly since this time period

38

(Hoffinan & Borders, 2001 ). The use of the BSRI in these studies may have affected
their results (Hart & Thompson, 1996; Tolman & Porche, 2000).

Therefore, there is some preliminary evidence that gender role socialization may
impact level of voice. That is, if girls have internalized the societal ideal of the “good
woman,” they may be more likely to lose their voice during adolescence, and feel that
they cannot have both close relationships and honest relationships. However, it is clear
that more research needs to be done to examine this relationship, given the
methodological problems of the extant research. The current study improved upon
previous research in that a more current measure of gender role socialization was used
(Boldizar, 1991), the independent contributions of masculinity and femininity were
examined (see Pathway B in Figure 1), and the differential contributions of masculinity,
femininity, and androgyny were examined.

While the previous two sections have focused on possible predictors of voice,
both intraindividual and societal, it is also important to examine how level of voice might
influence different areas of girls’ lives, such as their relationships with peers, romantic
partners, and their psychological well-being. The next section will examine these
possible outcomes of level of voice.

Outcomes of Level of Voice
Quality of Friendships

As discussed earlier, low levels of voice in adolescence may result in inauthentic
relationships; thus, it is likely that level of voice would affect the quality of friendships.
Peer relationships are certainly important during adolescence; friendships can buffer

against stress, and are related to positive psychological adjustment and developmental

39

outcomes (Bukowski, Newcomb, & Hartup, 1996; Wenz-Gross, Siperstein, Untch, &
Widaman, 1997).

Traditionally, research has neglected the study of adolescent’s friendships, and
research that was conducted lacked empirical data (Inderbitzen-Pisaruk & Foster, 1990).
Theory about adolescent fiiendship has been neglected as well. One of the few theories
about fiiendship was conceptualized by Harry Stack Sullivan; recent research on
friendship and intimacy has been based upon Sullivan’s theories (Furman, 1993).
Sullivan (1953) noted that true, reciprocal love first appears in children’s relationships
with a same-sex fiiend, and that preadolescence is characterized by the need for personal
intimacy. Sullivan argued that intimacy is the most salient aspect of adolescent
fiiendship, and that intimacy can be defined as closeness. Sullivan suggested that
intimate fiiendships provide the essential experience of being understood and validated.
He stated,

“intimacy is the type of situation involving two people which permits validation

of all components of personal worth. Validation of personal worth requires a type

of relationship which I call collaboration, by which I mean clearly formulated
adjustments of one’s behavior to the expressed needs of the other person in the
pursuit of increasingly identical — that is, more and more nearly mutual -
satisfactions, and in the maintenance of increasingly similar security operations”

(Sullivan, 1953, p. 246).

Therefore, according to Sullivan, friendships in adolescence are the first relationships to
involve mutual collaboration, and the establishment of intimacy in fiiendships provides

the basis for meaningful romantic intimate relationships later in life. Given Sullivan’s

40

theory, it seems essential that these beginning intimate relationships are authentic; if false
relationships develop in early adolescence, then these false relationships will be the basis
of all subsequent relationships, and will affect the quality and intimacy of firture intimate
relationships. Therefore, the possible loss of voice in early adolescence has important
implications for the intimacy of fiiendships, which may then impact future relationships,
and possibly render them more inauthentic.

Shulman and Knafo (1997) proposed a developmental model of adolescent
friendship. In their model, adolescent fiiendships begin at the stage of “closeness
orientation” in which adolescents have chosen their fiiends in terms of attraction to each
other, and they understand commitment and obligations to the friendship. The second
stage, “negotiating differences,” involves an awareness of each other’s needs, and fiiends
become skilled at negotiating conflicts that arise. The final, most integrated stage is
“mutuality-intimacy.” Shulman and Knafo also argue that attachment history is
especially relevant to the negotiation of each of these stages; they stipulate that
adolescents with secure attachment are more able than those with insecure attachment to
process conflicts in a mutually satisfactory manner. In contrast, adolescents with an
insecure attachment are more likely to refiain from conflict.

Thus, Shulman and Knafo’s (1997) model of fiiendship as moderated by
attachment is analogous to how level of voice would affect the quality of fiiendship; it is
likely that adolescents with a low level of voice would act similarly to adolescents with
an insecure attachment, and avoid conflict with close friends. Therefore, it seems

possible that girls with lower levels of voice will be arrested at the second stage,

41

“negotiating differences,” which will be represented by lower levels of intimacy in
fiiendships.

Kerns (1996) offers several reasons why a secure attachment would facilitate peer
relationships. First, a secure attachment may promote increased exploration, including
peer fiiendships. Second, caretakers of securely attached children are responsive to their
needs, which may promote the formation of a responsive interaction style that carries over
to other relationships. Third, the development of a positive internal working model
allows the child to form positive views of others. Kerns suggests that the internal
working model also allows the child to see both sides of the attachment relationship, and
she argues that, for example, children with a sensitive mother are likely to be sensitive
towards others, and treat others sensitively. Therefore, children with secure internal
working models are likely to try to establish close responsive peer relationships similar to
those developed with the primary attachment figure.

Some research has indicated that the quality of the fiiendship is a salient predictor
of psychological adjustment (Bukowski, Hoza, & Boivin, 1993). Therefore, the quality
of the fiiendship, both positive and negative aspects (Bemdt, 1996), must be assessed for
a full understanding of peer relationships (Aboud & Mendelson, 1996). In addition,
children’s perceptions of their fiiendships and their understanding of their peer
fiiendships are fundamental determinants of the quality and features of the relationship
(Bukowski, Hoza, & Boivin, 1994; Furmarr, 1996).

Parker and Asher (1993) proposed that quality of friendship, which includes
elements of companionship, supportiveness, intimate exchange, and level of conflict, was

an important construct of friendship adjustment. Other researchers (e.g., Mendelson &

42

Aboud, 1999) stipulate as well that intimacy is an essential element of adolescent
friendship. Intimacy in fiiendship involves sensitivity and openness to honest
expressions of thought, feelings, and personal information (Mendelson & Aboud, 1999).

Intimacy is salient in adolescence as adolescents experience developmental
changes. As children progress fi'om preadolescence to adolescence, their fiiendships are
characterized by an increase in intimacy (Kerns, 1996) and an increase in complexity
(Grazyk & Henry, 2001). In addition, adolescents report more support fiom their fiiends
than do children (F urman & Buhrmester, 1992). Intimate fiiendships have positive
effects upon psychological adjustment; for example, one study found that supportive best
fiiendships contributed significantly to students’ involvement in school (Bemdt & Keefe,
1995).

It is important that within the context of the relationship, the adolescent feels the
opportunity to feel valued and appreciated by the fiiend (Bukowski et al., 1994).
Buhrmester (1990) suggested that there are several theoretical reasons why intimacy is
important in adolescence. First, adolescence can be a stressful time, and intimacy may
help buffer the stress, and second, close friendships provide a sense of validation, which
is helpful.

There are gender differences in fiiendship experiences. Girls report more
intimacy in their fiiendships than do boys (Buhrmester, 1990); they also report higher
levels of the quality of the fiiendship (Jones & Costin, 1995). Evidence also suggests that
girls are more attuned to subtle differences in features of friendships than are boys; for

girls, fiiendship may be a more multi-dimensional relationship (Grazyk & Henry, 2001).

43

In addition, the quality of best friendships affected the adjustment of girls more than boys
(Berndt & Keefe, 1995).

In summary, Sullivan’s (1953) theory of intimacy of fiiendships in adolescence
involves mutuality and increasingly secure operations, and research indicates that
intimacy and quality of fiiendships become increasingly salient in adolescence
(Buhrmester, 1990; Inderbitzen-Pisaruk & Foster, 1990; Kerns, 1996). In addition, some
research suggests the maturity and intimacy of friendship is related to attachment
(Shulman & Knafo, 1997). Given this, it seems possible that level of voice would also
affect intimacy of fiiendship. According to level of voice theory, low levels of voice
inherently involve inauthentic relationships, which might be related to the lack of
intimacy of fiiendship. In addition, research indicates that mutuality of relationships is
important; this is another important aspect of level of voice. Finally, it is likely that girls
might be especially susceptible to level Of voice affecting their intimate relationships,
given evidence that they are more attuned than boys to differences in fiiendship (Grazyk
& Henry, 2001) and that the quality of their fiiendships influences their adjustment
(Berndt & Keefe, 1995).

Therefore, the current study hypothesized that level of voice would influence the
intimacy of friendships (see Pathway C, Figure 1). For example, it is likely that
adolescents with a low level of voice would perceive themselves as having close
friendships, in that they have compromised their own selves for relationships in their
lives. Thus, while these adolescents would appear to rate high on elements of fiiendship
such as loyalty and availability, their friendships would be lower on dimensions of

intimacy, which involve honest expression.

The next section will review research on romantic relationships in adolescence,
and discuss how level of voice may be related to both the quality of romantic
relationships and abuse during romantic relationships.

Romantic Relationships

Quality of romantic relationships

Adolescents become increasingly intimate with both fiiends and romantic partners
during adolescence (Furman & Buhrmester, 1992), and adolescent romantic relationships
are an important mechanism through which adolescents sort through issues of identity
(B.B. Brown, Feiring, & Furman, 1999). Romantic relationships have recently begun to
be examined by researchers, although most researchers have examined the differences in
adolescents who are dating and not dating (Connolly & Johnson, 1996), and have
neglected the qualitative aspects of adolescent romantic relationships.

Wyndol Furman and his colleagues (Furman & Simon, 1999; Furman & Wehner,
1994, 1997) have proposed a behavioral systems theory of adolescent relationships akin
to the concept of the internal working model proposed by attachment theorists, but more
encompassing of additional aspects of relationships. Furman suggests that adolescents
have views of relationships. Views are defined as “representations Of a particular type of
relationship, the self in that type of relationship, and the partner in that type of
relationship” (Furman et al., 2002). Views of relationships are similar to internal working
models in that they are thought to guide behavior, but differ in that views are thought to
incorporate all of the behavioral systems that are active at that time.

Furman and his colleagues propose that adolescents’ Views of others encompass

four main behavioral systems: attachment, caretaking, affiliation, and sexuality (Furman

45

& Simon, 1999; Furman & Wehner, 1994). Furman proposes that in the parent-child
relationship, attachment is the primary behavioral system, whereas affiliation is the
primary behavioral system in peer relationships. Therefore, an individual’s attachment to
the mother may influence later Views of affiliation with peers; Furman suggests that this
is borne out by evidence that a child’s parental attachment predicts later peer competence,
a non-attachment process (W einfield et al., 1997).

Furman also suggests that fiiendships are related to romantic relationships in that
the affiliative process with a peer may shape romantic relationships, which encompass
views of all four behavioral systems (Furman et al., 2002). Furrnan’s theoretical link
between friendships and romantic relationships is supported by evidence that early
adolescents perceived romantic relationships more in terms of friendship and '
companionship than did older adolescents (Feiring, 1996; Shulman & Scharf, 2000).

Therefore, there is evidence of a relafionéhip between adolescents’ peer
relationships and romantic experiences, and it is possible that level of voice would affect
the quality of both types of affiliative relationships. While no one has examined the
relationship between level of voice and the quality of romantic relationships in
adolescence, Furman and Shaffer (in press) have alluded to the connection between the
two constructs. They suggested that adolescents are likely to act in ways that they feel
will attract romantic partners; for example, heterosexual girls, may act in stereotypically
feminine ways to attract males (F urman & Shaffer, in press).

Thus, the current study examined the connection between loss Of voice and quality
of romantic relationships. It may be that some girls feel pressure to attract males by

acting stereotypically feminine, and that this behavior would result in loss of voice, since

46

they might not be able to voice their own opinions. Given that level of voice is a
relational construct, it is possible this loss of voice would impact relationships with
romantic partners. The current study examined this relationship (see Pathway E, Figure
1). If girls do subvert their voice in romantic relationships, this may put them more at
risk for abuse in relationships, specifically psychological abuse, which is relational in
nature (see Pathway D, Figure 1). The next section will review research on violence in
adolescent romantic relationships.

Violence in Romantic Relationships

Attitudes toward dating violence and experience of psychological, physical, or
sexual violence in a relationship may also affect the quality of romantic relationships.
Violence in adolescent romantic relationships, other than college-age samples, has been
rarely researched (Feiring & F urrnan, 2000). Research that has been conducted has found
that adolescent girls were as likely as boys to be the instigators of abusive behavior in
intimate relationships; however, the actions of boys were more severe than those of girls
(Molidor & Tolman, 1998). Indeed, girls’ use of violence in a relationship is more
accepted and perceived as more justified than boys’ use of violence (Price & Byers, 1999;
Smith Slep, Cascardi, Avery-Leaf, & O'Leary, 2001).

In addition, Molidor and Tolman (1998) found that boys were as likely as girls to
experience some physical violence in a dating relationship. This indicates that while the
level of prevalence may not differ, the level of severity does differ, which may impact
boys and girls in different ways. Girls were more likely to report negative physical and
emotional effects of the violence than were boys, and girls were more likely to physically

fight back when they experienced physical violence from their partner, which may

47

account for some of the physical abuse reported by boys. However, another study
(Foshee, 1996) found that even after controlling for self-defense, girls reported
perpetrating more physical violence than boys toward their partners.

As understood by the research on adult domestic violence, psychological abuse
may be experienced more fi'equently than physical abuse, and may have more detrimental
and long-lasting effects (Campbell & Lewandowski, 1997; Ellington & Marshall, 1997).
Psychological abuse can include isolation, threatening behavior, relational abuse (e.g.,
spreading rumors about one’s partner), and verbal and emotional abuse (Molidor, 1995;
Wolfe et al., 2001). Indeed, one study found that while 59% of high school students had
at least one experience with physical violence in a relationship, 96% had experienced
psychological abuse in a relationship (J ezl, Molidor, & Wright, 1996).

One barrier for investigating Violence in adolescent romantic relationships has
been the lack of appropriate measures (Smith Slep et al., 2001; Wolfe et al., 2001).
Research on adolescents has traditionally used versions of the Conflict Tactics Scale, a
widely used measure with adult populations (Straus, 1979). Not only has the CTS been
normed on college-age and adult populations, but some of the items are clearly
inappropriate for an adolescent sample, as they relate to economic abuse, or use of
children as coercion (Wolfe et al., 2001).

Attachment problems may pose a risk factor for dating Violence (W ekerle &
Wolfe, 1998; Wolfe & Feiring, 2000). Some research (Flanagan & Furman, 2000) has
attempted to link relational Views to sexual victimization in adolescence and early
adulthood. The authors found that college-age women who were victims of sexual

assault were more likely to be preoccupied than women who were not victimized;

48

however, the same pattern was not true for high school women (Flanagan & Furman,
2000).

Another study that examined the roles of child maltreatment and attachment in
predicting dating violence found that while maltreatment was the most salient predictor of
male abusive behaviors, the interaction of attachment history and maltreatment was
significant for both avoidant and ambivalent males, suggesting that attachment is a
moderator of the relationship between child maltreatment and Violent behavior towards a
partner in adolescence (W ekerle & Wolfe, 1998). For females, while maltreatment
history was the most salient predictor of being a victim of abuse, being avoidantly
attached was also a significant predictor (Wekerle & Wolfe, 1998).

Given the paucity of research on level of voice, it is not surprising that only one
study has examined the relationship between voice and abuse. A qualitative study by
Salzrnan (1990) found that girls who had both low levels of voice and insecure
attachment were more likely to be in harmful romantic relationships than girls without
both of these risk factors. However, as previously mentioned, this study only involved
nine girls, and was not empirical, both of which make it difficult to generalize the
findings.

There is a dearth of research on the relationship between level of voice and
violence in dating relationships; if some girls do present with a loss of voice in early
adolescence, and, as a result, feel uncomfortable speaking their mind, and feel that
relationships are to be valued above their sense of self, they may very well be at risk for
staying in a relationship that is no longer healthy, due to physical or emotional abuse.

Therefore, the current study examined this relationship (see Pathway D, Figure 1).

49

Psychological Well-Being
Depression

Jack’s (1991; Jack & Dill, 1992) theory suggests that women use self-silencing to
cultivate and maintain relationships, and, as a result, may experience depression as they
inhibit their own true feelings about relationships. While Bowlby (1980) proposed that
depression is connected to feelings of past and present loss of an Object, Jack (1991)
suggests that for women, rather than the loss of an object, depression is connected to a
loss of self in relation to others.

Other theorists also maintain that the experience of depression may be connected
to experiences in relationships (Blatt & Maroudas, 1992). Blatt’s work suggests that
diverse psychological theories, such as interpersonal, attachment, psychoanalytic ego
psychology, and cognitive behavioral, all propose two types of depression: first, those
preoccupied with, and dependent upon, interpersonal relationships, and second, those
preoccupied with achievement and self-worth (Blatt & Maroudas, 1992).

It is well-established that while there are no gender differences in depression in
childhood, after early adolescence and into adulthood, females have higher rates of
depression than do males (Hart & Thompson, 1996; Kenny, Moilanen, Lomax, &
Brabeck, 1993; Petersen, Sarigiani, & Kennedy, 1991; Renouf & Harter, 1990;
Wichstrom, 1999). Nolen-Hoeksema (1994) proposed that adolescent girls may have
higher rates of depression than do adolescent boys because of three factors. First,
biological changes occur in early adolescence that cause girls to have increased body fat,
and boys to have increased muscle; these distinctly different changes cause girls to be

more self-conscious, whereas boys tend to be proud of their body changes. Second, girls

50

are at increased risk for sexual victimization (e. g., sexual exploitation) in adolescence
(Sroufe & Rutter, 1985). Third, in early adolescence, gender intensification (Hill &
Lynch, 1983) may occur, in which girls are pressured to adopt stereotypically feminine
gender roles, and may feel frustrated at having their roles restricted by society.

Nolen-Hoeksema and Girgus (1994) proposed a model of gender differences in
adolescent depression in which girls have more risk factors, such as a relational style,
than boys do before adolescence, and that those risk factors lead to depression when girls
are faced with challenges that increase in early adolescence, as discussed above.
Therefore, they suggest that preexisting gender differences in early adolescence, such as
gender differences in relational styles, interact with challenges in adolescence to engender
gender differences in depression. This is consistent with research that indicates that the
challenges that girls faced in adolescence were predictive of gender differences in
depression in later adolescence (Petersen et al., 1991; Wichstrom, 1999).

Several studies have examined the effects of aspects of a relational orientation
upon depression. One study found that several relational characteristics, such as feeling
too responsible for the welfare of others and having difficulty being assertive were
predictors Of depressive symptoms in girls, rather than gender or femininity per se (Aube
et al., 2000). A similar study found that other negative aspects of an extreme relational
orientation, such as ruminating, and silencing the self, accounted for 59% of the variance
in depressive symptoms (Hart & Thompson, 1996). Finally, one study found that after
controlling for gender, adolescents’ abilities to communicate assertively were negatively

related to depressive symptoms (Kobak & F erenz-Gillies, 1995). It is interesting to note

51

that gender itself did not predict depressive symptoms in any of the studies cited above
(Aube et al., 2000; Hart & Thompson, 1996); (Kobak & Ferenz-Gillies, 1995).

Thus, it may be that feminine traits themselves do not put adolescents at risk for
depressive symptomatology, but, rather, the negative aspects of a relational orientation
(e.g., when one sacrifices the self for the sake of relationships), are risk factors for
depression. Therefore, if the negative aspects of relational characteristics are predictive
of depressive symptoms, it is likely that loss of voice, when girls lose their sense of
selves, will predict depressive symptomatology.

Research has also demonstrated that during adolescence, boys and girls
experience depression in different ways, and with different symptoms. One study that
performed a factor analysis on the Children’s Depression Inventory (Kovacs, 1992) found
gender differences on each of the six factors; the most notable difference was that
adolescent girls experienced more symptoms of depression related to low self-concept
than did adolescent boys (Donnelly & Wilson, 1994). In addition, when adolescents were
asked if they experienced anger when depressed, and, if so, who was the target of their
anger, girls were more likely than boys to report more anger toward themselves than
towards others, indicating that girls are more likely than boys to internalize their anger
(Renouf & Harter, 1990).

While only three studies have examined the relationship between level of voice
and depression in adolescence, these studies have all found that depressive symptoms
were negatively correlated with level of voice; that is, as level of voice declined,
depressive symptoms increased (Hart & Thompson, 1996; Harter et al., 1996a; Tolman &

Porche, 2000).

52

In summary, there is abundant evidence that girls experience greater depressive
symptomatology in adolescence than do boys (Hart & Thompson, 1996; Petersen et al.,
1991; Renouf & Harter, 1990; Wichstrom, 1999); some theory suggests that for women,
depression may result from the loss of connection with others, and dependence upon
personal relationships (Blatt & Maroudas, 1992; Jack, 1991). Studies have found that
aspects of a lack of mutuality in relationships (e.g., ruminating and silencing of the self)
were predictors of depressive symptomatology (Aube et al., 2000; Hart & Thompson,
1996), as was the lack of ability to communicate assertively (Kobak & F erenz-Gillies,
1995). In addition, preliminary results from the three studies examining the relationship
between loss of voice and depressive symptomatology in adolescent girls found that the
two constructs were significantly related to each other (Hart & Thompson, 1996; Harter et
al., 1996a; Tohnan & Porche, 2000).

Therefore, given evidence Of a relationship between level of voice and depressive
symptomatology discussed above, it seems possible that level of voice may be related to
depressive symptomatology. The current study examined this relationship (see Pathway
F, Figure 1). While there is preliminary evidence, as discussed above, of a relationship
between level of voice and depression, the current study contributed by examining
different settings and contexts of voice. Another measure of psychological well-being,
self-esteem, may also be affected by level of voice, given preliminary evidence from
current research on gender differences in self-esteem (Harter & Marold, 1994; Rose &

Montemayor, 1 994).

53

Self-Esteem

Self-esteem is highly relevant to the construct of level of voice. Some theory
suggests that when girls enter adolescence, and become aware of the conflict between
being “good women” and the pressure to achieve, both academically and emotionally,
they subsequently become less able to voice their anger, especially in public settings
(L.M. Brown, 1998; L.M. Brown & Gilligan, 1992; Gilligan et al., 1990). Girls learn that
their anger is not socially acceptable, and, furthermore, their anger could threaten their
relationships (L.M. Brown, 1998). Therefore, it is possible that if girls have low levels of
voice, then they will internalize their feelings of being silenced, and have lower self-
esteem.

Brown (1998) notes that this anger is often justified, and thus when girls are
unable to voice their anger, they internalize their feelings, and start to doubt themselves.
In her book, Brown gives an example of this mechanism. Middle-school boys receive
preferential treatment in the classroom over girls. Although the girls in her study, within
their all-female small group, all discussed how they do not feel listened to by their
teachers, when asked if they felt that the teachers discriminated in the classroom based on
gender, the girls were unable to acknowledge the gender discrimination, and instead, told
Brown that it must be their fault, and that they should speak louder in the classroom. By
doing this, these girls then created a double-bind for themselves: it would be difficult to
speak up in the classroom if they did not feel as though they would be heard, and when
they did not speak up, they internalized the blame. Brown (1998) argued that this

internalization leads to dissociation from valid feelings, and it is also plausible that this

54

leads to lower self-esteem as girls may continue to doubt their own abilities to be self-
advocates, and to be efficacious and create systems change.

Following Brown’s (1998) theories, girls should have lower self-esteem than
boys, given that she suggests that boys do not silence their own anger to the extent that
girls do. Research lends supports to this theory; in early adolescence, girls do have lower
self-esteem than boys (Lintunen, Leskinen, Oinonen, Salinto, & Rahkila, 1995; Simmons
& Rosenberg, 1975). When different arenas of self-esteem are measured, however,
findings differ. One study found that adolescent boys had greater physical competence
than adolescent girls, while girls had greater social competence than boys (Cate &
Sugawara, 1986). However, the study used a measure of self-esteem that was normed for
children, which may have affected the results (Cate & Sugawara, 1986). Overall,
research indicates that adolescent boys have higher self-esteem than adolescent girls
regarding their social behavior and appearance, and girls have higher self-esteem than
boys regarding their higher perceived romantic self-worth (Harter & Marold, 1994; Rose
& Montemayor, 1994). This research is consistent with theory regarding level of voice; if
girls have low levels of voice, they will be likely to have low self-esteem regarding their
social behavior, as described by Brown’s (1998) example.

In summary, theory suggests that as result of the relational impasse experienced in
early adolescence, girls may stifle their feelings, and especially anger (L.M. Brown, 1998;
L.M. Brown & Gilligan, 1992; Gilligan et al., 1990). Brown (1998) suggests that girls
then become dissociated from themselves, and consequently may experience lower self-
esteem. Empirical research supports this theory; girls do have lower self-esteem than

boys, specifically in respect to their social behavior (Harter & Marold, 1994; Rose &

55

Montemayor, 1994). Therefore, given theory that suggests a relationship between level of
voice and self-esteem, it seems possible that level of voice would predict self-esteem.
The current study examined this relationship (see Pathway F, Figure 1).

Social Support

While research indicates intimacy of relationships is salient in adolescence, social
support is another important aspect of peer relationships. The next section will review
different types of social support that are important in adolescents’ lives, and will discuss
the relationship between social support and level of voice.

Social support has been associated with better physical and psychological health
(Cohen & Wills, 1985; Flannery & Wiemarr, 1989; Kessler & McLeod, 1985). Most of
the research on social support has been conducted with adults, and only recently has
research begun to examine social support in the lives of adolescents (Cauce, Mason,
Gonzales, & Hiraga, 1996). Social support is certainly relevant in the lives of
adolescents; research has found that social support is predictive of psychological
adjustment (Moran & Eckenrode, 1991). Studies have found that emotional support is
related to school competence, peer competence, and perceived self-competence (Cauce,
1986; Canoe et al., 1996; Robinson, 1995). In addition, social support can provide
emotional support around developmental issues salient in adolescence (e.g., puberty,
changes in familial relationships) (Cauce et al., 1996).

One type Of support, approval from family and peers, was more salient to self-
worth than either emotional or instrumental support (Robinson, 1995). In general,
research has indicated that approval social support from parents, classmates, and close

friends predicts depression (Harter, Marold, & Whitesell, 1992), and is a strong predictor

56

of self-esteem (Harter et al., 1992; Harter, Stocker, & Robinson, 1996b). Approval from
others is important in that “any form of support that clearly communicates that others
approve of the self may be internalized as positive feelings toward the self’ (Robinson,
1995, p. 255). Therefore, self-worth (or self-approval) is derived from internalization of
approval fiom others (Harter et al., 1992).

Approval support, or feeling validated, is related to true self behavior; one study
found that the ability to be authentic in relationships was significantly correlated to
feeling validated within the relationship (Harter et al., 1997b). In addition, as discussed
earlier, lack of validation from others was a reason that adolescents cited for false-self
behavior (Harter et al., 1996a). Research also indicates that individuals who identified
themselves as having a mutual style in their intimate relationships and identified their
partners as also having a mutual style had the highest levels of validation within their
relationships (Harter et al., 1997b).

In adolescence, it is important to distinguish between different sources Of social
support, such as friends and family, given the evidence that friends become increasingly
more important in adolescence (Furman & Buhrmester, 1992; Procidano & Heller, 1983).
Studies have found differential effects for different sources of social support. One study
found that family social support was related to parental attachment (Cauce et al., 1996).
Harter and her colleagues examined validation support from four different sources of
support. They found that validation support was predictive Of self-worth, and that this
was significantly stronger when sources, or contexts, were the same, indicating context-

specific social support (Harter, Waters, & Whitesell, 1998a).

57

Research has indicated that social support is particularly salient in terms of loss of
voice; Harter and her colleagues found a relationship between hope about future support
and true/false self behavior (Harter et al., 1996a). Their research indicated that hope
about future support mediated the relationship between quality of support and true/false
self behavior in that hopelessness about future support was predictive of false self
behavior (Hatter et al., 1996a). In addition, the majority of adolescents cited “approval of
others” as the most important reason for their false self behavior, and the second most
common reason cited was an alienation process from their true self, due to lack of
validation from others. These two reasons accounted for 90% of false self behavior
(Harter et al., 1996a), indicating the relevance of social support as a protective factor
against loss of voice.

In summary, social support is associated with better physical and psychological
health, and is salient during adolescence (Cauce, 1986; Canoe et al., 1996; Cohen &
Wills, 1985; Flarmery & Wieman, 1989; Kessler & McLeod, 1985; Robinson, 1995).
Therefore, although Harter and colleagues (1996a; 1998a; 1997b) did not test for
moderating effects of social support on an outcome such as psychological well-being,
approval support may be especially salient to psychological well-being (Robinson, 1995),
and research indicates that “approval of others” was the most important reason that
adolescents displayed false self behavior (Harter et al., 1996a). This suggests that social
support may have a role as a moderator between level of voice and psychological well-
being. Thus, it was hypothesized that if social support was a moderator of this
relationship, it would be particularly helpful for those with low levels of voice. The

current study will assess this relationship (see Pathway G, Figure 1).

58

Rationale

Adolescent girls may be more at risk than are adolescent boys for negative
outcomes in adolescence (Petersen, 1988; Steinberg & Monis, 2001). Relational theory
proposes that for women, the primary experience of the self is relational, and that females
place a high value on relationships. However, relational theory also suggests that it is
important that relationships are mutual and reciprocal (i.e., a balance of accommodation
and self-involvement) (Jordan et al., 1991; Miller et al., 1997). Gilligan, Brown, and
their colleagues propose that girls lose their voice in early adolescence (L.M. Brown,
1998; L.M. Brown & Gilligan, 1992, 1993; Gilligan et al., 1990). They suggest that loss
of voice is manifested by girls’ difficulty in expressing their own opinions, and is
indicated by the preservation of relationships that may be inauthentic. This theory of loss
of voice in adolescent girls has been scarcely researched.

The current study contributes to the current research in a number of ways. First,
most of the research has involved qualitative studies, and there is a need for quantitative
research to corroborate the qualitative findings (Harter et al., 1998b).

Second, most of the work on loss of voice theory has involved studies of white,
middle-class girls (Taylor et al., 1995); this makes it difficult to generalize the theory and
its findings to adolescent girls at large. This is salient given the evidence from several
qualitative studies that girls from diverse ethnic backgrounds and lower SES may be less
vulnerable to loss of voice (L.M. Brown, 1998; Clark, 1999; Way, 1995). Therefore,
there is a need for quantitative research that involves girls from lower SES and who are

ethnically diverse. The current study direcfly addressed this need in that it involved a

59

group of girls who are ethnically diverse and fi'om low socioeconomic backgrounds; this
is the first quantitative study on level of voice to involve such a group of girls.

Third, while researchers acknowledge that loss of voice does not occur in all girls
(L.M. Brown, 1998; Taylor et al., 1995), there is little understanding of why some girls
and not others may be vulnerable to loss of voice in adolescence. Given this deficiency in
our understanding of loss of voice, there is a need for research that addresses predictors of
voice. The current study directly addressed this need by examining two possible
predictors of voice: attachment to parents and gender role socialization.

Research on attachment theory indicates that attachment continues to be salient
throughout the lifespan, and that the quality Of parent-child attachment is related to
competency and intimacy in later relationships in childhood and adolescence (Elicker et
al., 1992; Shulman et al., 1994; Sroufe et al., 1993; Weinfield et al., 1997). Although no
one has quantitatively examined the relationship between attachment and loss of voice,
results fi'om a small qualitative study suggest that quality of attachment would impact loss
of voice, and that insecure attachment would be a risk factor for loss of voice (Salzman,
1990). In addition, insecure adolescents, and girls in particular, display a constellation of
behaviors that are analogous to loss of voice. Irrsecurely attached female adolescents,
compared to those are who securely attached, are less assertive, exhibit more gender-
stereotyped behavior than secure female adolescents (Kerns, 1996; Turner, 1991), and
have poorer conflict management skills (Kerns, 1996; Kerns et al., 1996; Kobak et al.,
1993; Schneider & Younger, 1996). These are all qualities that might impact level of

voice.

60

Therefore, given the connection between attachment and intimacy in adolescence,
and the outcomes of insecure attachment in adolescence, it is important to examine the
relationship between attachment and loss of voice. As previously explained, while there
is some preliminary evidence for a connection between the two constructs (Salzrnan,
1990), no studies to date have empirically examined this relationship. Understanding the
connection between attachment and level of voice is an important step in beginning to
comprehend why some girls and not others may be vulnerable to loss of voice. Therefore,
the current study contributes to the extant literature on voice by being the first to
empirically examine the relationship between attachment and loss of voice (see Pathway
A in Figure 1)

Another possible predictor of voice is gender role socialization; Gilligan’s
(Gilligan et al., 1990) discussion of the “good woman” archetype suggests that girls who
are more feminine would be more likely to lose their voice. In addition, the gender
intensification hypothesis suggests that girls become more feminine in early adolescence
(Hill & Lynch, 1983); some research indicates girls start to feel pressure to be “perfect”
(e.g., soft-spoken, intelligent, etc.) in early adolescence (L.M. Brown, 1998), and that this
is related to lower levels of voice.

There is preliminary quantitative evidence that gender role socialization may
impact level of voice (Harter et al., 1998b). However, studies that have examined the
relationship between gender role socialization and loss of voice had methodological
problems in that they either used an outdated measure of gender role socialization (Hart &
Thompson, 1996; Tolman & Porche, 2000) or did not examine the independent

contributions of masculinity and femininity (Harter et al., 1998b). The current study

61

improved upon previous research in that it used a more current measure of gender role
socialization (Boldizar, 1991) and examined the independent contributions of masculinity
and femininity to level of voice (see Pathway B in Figure 1).

In addition, it is important to determine the consequences of loss of voice, if any.
The current study examined the impact of level of voice upon three different aspects of
adolescent girls’ lives: quality of friendships, dating conflict, quality of romantic
relationships, and psychological well-being (see Pathways C, D, E, and F in Figure 1).

While research has examined the relationship between voice and psychological
well-being, no studies to date have examined the effects of loss of voice upon
relationships. Given that level of voice is an inherently relational construct, loss of voice
may affect the quality of relationships in adolescents’ lives. As discussed earlier, girls in
a qualitative study who experienced loss of voice felt that their fiiendships were less
intimate as a result of loss of voice (L.M. Brown, 1998). This study contributes to the
current body of research on voice in that it is the first study to empirically examine the
relationship between loss of voice and fiiendship (see Pathway C in Figure 1).

In addition, just as girls’ low levels of voice may compromise the intimacy of
their platonic fiiendships (L.M. Brown, 1998), their low levels of voice may also
compromise the intimacy of their romantic relationships (Jack, 1991) (see Pathway E in
Figure 1). It may also be that level of voice will impact abusive relationships in
adolescence; there is some preliminary evidence that loss of voice in women is related to
unhealthy romantic relationships (Jack, 1991), and it is possible that girls who are at risk
for loss of voice would be more wary of leaving an unhealthy relationship. Therefore, the

current study was the first to empirically examine the relationship between loss of voice

62

and dating conflict (see Pathway D in Figure 1) and loss of voice and the quality of
romantic relationships (see Pathway E in Figure 1).

There is preliminary evidence to suggest that loss of voice is related to depression
and self-esteem (Hart & Thompson, 1996; Harter et al., 1996a; Tolman & Porche, 2000),
although more research is needed. The current study examined the relationship between
loss of voice and these two aspects of psychological well-being (see Pathway F in Figure
1).

Research on approval support and emotional support suggests that these
constructs are highly relevant to loss of voice (Harter et al., 1996a; Harter et al., 1997b;
Robinson, 1995). The ability to be authentic in relationships was related to feeling
validated in the relationship (Harter et al., 1997b), and lack Of validation was cited as a
reason for false-self behavior (Harter et al., 1996a). It may be that this tends to be more
true for girls who have lower levels of voice than girls with higher levels of voice,
suggesting that social support may moderate the relationship between level of voice and
psychological well-being. The current study was the first to examine the role of social
support as a moderator between level of voice and psychological well-being (see Pathway
G in Figure 1).

The current study also has possible practical implications for adolescent
development. Given that previous research indicates that loss of voice is more prevalent
in public settings, such as the school, than private settings (L.M. Brown, 1998; Harter et
al., 1998b), if girls are at risk for loss of voice, schools could possibly intervene, and

create preventative mechanisms to buffer girls against loss of voice.

63

Hypotheses
Predictors of Voice Model

Given that voice is assessed in five different contexts, with mother, father,

teacher, classmates, and best fiiends, and given differential findings of voice based upon

the private context of voice versus the public context of voice (Harter et al., 1997c; Harter

et al., 1998b), each model was run iwice, for the different settings. The structure of these

different settings was determined by an exploratory factor analysis of the different

contexts of voice. Please see Figure 2 for the structural model of predictors of voice.

1.

Children’s attachment to their parents, based upon their self-report, would both be
significantly and positively related to their level of voice within any given context of
voice. Specifically, secure attachment would be positively related to higher levels of
voice.
Children’s gender role socialization, based upon their self-report, would be
significantly related to their level of voice. Given the structure of the latent construct
of gender role socialization, it was impossible to know the specific relationships of
masculinity and femininity to voice through the structural modeling, and so additional
multivariate analyses were performed to examine the relationship between gender role
socialization and level of voice (see Hypothesis 8).
Friendship and Well-being Outcomes of Voice Model

Please see Figure 3 for this structural model.
Level of voice would be significantly and positively related to quality of fiiendship.
Level of voice would be significantly and positively related to psychological well-

being.

64

Romantic Relationship Outcomes of Voice Model
Given that not all participants will have had romantic relationships, this outcome
model will be run separately. Please see Figure 4 for this structural model.

5. Level of voice would be significantly and positively related to quality of romantic
relationships.

6. Level of voice would be significantly and negatively related to dating violence.

Additional Hypotheses
Given the paucity of research on level of voice, several additional hypotheses
regarding level of voice were considered. These explored the relationship between voice
and social support, gender role socialization, dating violence, romantic relationships, and
ethnicity. These hypotheses were as follows:

7. Multiple regressions tested the hypothesis that social support would moderate the
relationship between voice and depression and self-esteem, corroborating the
buffering model of social support.

8. The median-split method of using the CSRI was used, and using the categories of
androgynous, undifferentiated, masculine, and feminine, it was anticipated that
feminine girls would have lower levels of voice than would masculine or
androgynous girls.

9. Level of voice was dichotomized based upon the median-split method into high voice
and low voice, and it was anticipated that girls in the low voice category would have
more emotional abuse in their relationships and have lower satisfaction in their

relationships than girls in the high voice category.

65

10. Caucasian girls would tend to have lower levels of voice than girls of color, and this
would not be a function of SES. In addition, given the different ethnic/racial make-up
of the three middle schools (see Table 13), the effect of school was covaried, by

controlling for the different schools.

66

     
       
 

Gender Role
Socialization

Attachment
with
Parents

Figure 2: Structural model of attachment and gender role socialization as predictors of

67

Quality of

 

Friendships
+
Social
Level Support
of Voice
+ +

Psychological

Well-Being

 

Figure 3: Structural model of fiiendship and well-being as outcomes of voice

68

Dating
Conflict

Level
of Voice

 
    
   
 

  

Quality of
Romantic
Relationship

Figure 4: Structural model of romantic relationship outcomes of voice

69

METHODS
Participants

One-hundred and eight girls in the 8th grade were recruited to participate in the
current study (see Table 1 for demographic information). Girls were recruited through
the Lansing School District (LSD) in three middle schools.1 The overall response rate
was 57%. The inclusion criteria were that participants had to be at least in early puberty,
to ensure that girls were in approximately the same stage of development, because
research indicates that with the onset of puberty, girls may experience lower self-esteem,
increased depressive symptomatology, changes in familial and peer relationships,
decreased body satisfaction, and increase in cognitive abilities (Dom, Crockett, &
Petersen, 1988; Hill & Lynch, 1983; Petersen, Leffert, & Graham, 1995). Thus, to ensure
that the findings were related to individual differences, and not pubertal differences, all
participants had to be at least in early puberty. All participants met inclusion criteria as
determined by the Pubertal Development Scale (PDS-R, Petersen, Crockett, Richards, &
Boxer, 1988). The average age was 14.16 years (SD = .58). Thirty-seven percent of
participants were Caucasian, 31% were African-American, 14% were Biracial, 11% were
Latina, 4% were Asian-American, and 3% were Native-American. Three percent Of
participants’ mothers had no formal schooling, 15% had less than a high school
education, 31% had graduated high school, 23% had some college, 8% had an associates
degree, 8% had a college degree, 3% had some graduate school, and 9% had a graduate
degree. One percent of participants’ fathers had no formal schooling, 17% had less than a

high school education, 31% had graduated high school, 29% had some college, 4% had

70

an associates degree, 8% had a college degree, 1% had some graduate school, and 10%
had a graduate degree. Eighty-two percent of mothers worked outside the home, and 83%

of fathers worked outside the home.

 

' There are four middle schools in the LSD; one school did not respond to recruitment phone calls within
the circumscribed period of time, and therefore was not part of the current study.

71

Table 1.

Demographic characteristics of participants

Ethnic Background
Caucasian 37%
African-American 31%
Biracial 14%
Latina 11%
Asian-American 4%
Native-American 3%
School
Pattengill Middle School 53%
Gardner Middle School 12%
Dwight Rich Middle School 35%
Mother ’s Educational Level (11 = 100)
No formal schooling 3%
Less than a high school education 15%
Graduated high school 31%
Some college 23%
Associates degree 8%
College degree 8%
Some graduate school 3%
Graduate degree 9%
Father ’s Educational Level (II = 84)
No formal schooling 1%
Less than a high school education 17%
Graduated high school, 31%
Some college 29%
Associates degree 4%
College degree 8%
Some graduate school 1%
Graduate degree 10%
Mother works outside home (n = 106) 82%
Father works outside home (n = 98) 83%
Age (mean) 14.16

72

Procedure

Teachers of 8th grade classrooms in three different middle schools in the Lansing
School District sent consent forms to their students’ homes for a parent/ guardian (see
Appendix A) to give permission for his/her child to fill out questionnaires assessing
relationships and self-worth during a 45- to 50-minute class period. The principal at each
school determined the class in which data collection occurred, and this was dependent
upon each teacher’s willingness to participate. At Pattengill Middle School, data
collection occurred during either English or Social Studies classes, and the two teachers
involved reminded their students daily to return their permission slips. The response rate
for this data collection site was 50%. At Gardner Middle School, data collection occurred
during a Life Skills Class, and the teacher reminded the students several times over two
weeks to return their permission slips. The response rate for this data collection site was
47%. At Dwight Rich Middle School, it was initially planned that the entire 8th grade
would be involved in the data collection, but when it became clear that this was too large
for the teachers to coordinate, it was determined that two different girls’ gym classes,
taught by one teacher, would be eligible to participate. The response rate at this data
collection site was 75%.

Two weeks after permission slips were distributed, self-report instruments were
group administered in a classroom setting, with the current researcher supervising the
completion of measures with an undergraduate assistant. All students completed a
consent form (see Appendix B). No identifying information about the students was
collected. Each participant received a $10 gift certificate to McDonald’s for her

participation, and each school received $100 for school supplies for their cooperation.

73

Measures
Demographics
Demographics Measure

Information on participants’ grade, birthdate, ethnic background, religion, gender,
parents’ occupation, parents’ educational level, and grade point average, both overall, and
in specific classes, was collected (see Appendix C for this measure).

Puberty Screening
Pubertal Development Scale - Revised (PDS-R, Petersen et al., 1988)

This 7-item scale was used to screen participants for onset of puberty. This scale
differs for males and females; both versions inquire about weight, height, growth spurt in
height, growth of body hair, and skin changes (i.e., onset of pimples). However, the
female version asks about breast growth and onset of menstruation, while the male
version asks about deepening of voice and hair growth on the face. Participants were
asked to choose between four stages of development for each marker of puberty; for
example, for skin changes, they were asked to choose between not yet started showing
changes, have barely started showing changes, sla'n changes are definitely underway, and
skin changes seem completed (see Appendix D). The original scale was modified to
delete the following question: “do you think that your development is any earlier or later
than most girls your age?” In samples of 6th to 8th graders, Petersen and her colleagues
found coefficient alphas ranging from .68 to .78 for the PDS for boys and .76 to .83 for

the PDS for girls. The coefficient alpha for the current study was .60.

74

Petersen and her colleagues created a five-category classification of pubertal
stage: pre-pubertal, early pubertal, mid-pubertal, late pubertal, and post-pubertal
(Crockett, 1988). These five stages are based on the level of development reported on the
three indices of pubertal change thought to be most salient for girls and boys. For girls,
these were pubic hair growth, breast development, and menarche. For the current study,
girls were considered pre-pubertal if they reported no development on any of the three
indices, early pubertal if they reported no menarche along with some development of
either breasts or pubic hair but not both (i.e., a combined score of 3 for the last two
indicators), mid-pubertal if they reported no menarche along with some development of
either breasts or pubic hair or more development on at least one of these two
characteristics (i.e., a combined score of 4 or more for breasts and body hair), late
pubertal if they reported menarche with less than complete breast and/or pubic hair
growth (i.e., a combined score of 7 or less for breasts and body hair), and post-pubertal if
they reported menarche along with completed development Of both pubic hair and breasts
(i.e., a combined score of 8 for breasts and body hair).

Attachment
Behavioral Systems Questionnaire (BSQ, Wehner & F urman, 2000)

This 45-item scale was used to assess participants’ behavioral systems with their
parents. These systems are attachment, caregiving, and affiliation. Participants were
asked to rate on a 5-point Likert type scale how much they agree or disagree with each
question about their relationship with their parents, where 1 = strongly disagree and 5 =
strongly agree. Sample items include, “my parents act as if I count on them too much,”

“I find it easy to be understanding of my parents and their needs,” and “truthfully, my

75

relationships with my parents are just not that important to me” (see Appendix E).
Within each behavioral system, there were five items that assessed secure attachment,
dismissing attachment, preoccupied attachment, respectively. Nine subscales were
created by averaging the five items for each combination of the different attachment
subtypes and behavioral system (e. g., security-attachment system, security-caregiving
system, security-affiliation system, dismissing-attachment system, etc.).

Previous research utilizing the BSQ has found good internal consistency with
coefficient alphas greater than .85 for each of the scales. The BSQ is related to measures
of adult attachment, such as the Adult Attachment Scales (N .L. Collins & Read, 1990),
and Hazan and Shaver’s (1987) Attachment Style Prototypes measure (Furman, 1996).
To assess the validity of the BSQ, it was administered to a sample of adolescents along
with the Adult Attachment Interview (AAI) (George, Kaplan, & Main, 1985), a semi-
structured interview designed to assess adults’ states of mind in relation to their working
models (F urman et al., 2002). The authors found very similar patterns of results for the
AA] and the BSQ; therefore, while the BSQ is a self-report measure, and may not tap into
internal working models, research indicates that it may be an adequate substitute for
assessing attachment, given that the AAI is a time-consuming measure that is laborious to
administer and code. In the present study, coefficient alphas for security in each Of the
behavioral systems were .77, .73, and .87, respectively, alphas for dismissing were .79,
.69, and .87 for each system respectively, and alphas for preoccupied were .67, .68, and

.77 for each system respectively.

76

Gender Socialization
Children '5 Sex Role Inventory (CSRI, Boldizar, 1991)

The CSRI is based on the Bern Sex Role Inventory, a widely used measure of
gender-role socialization. This 30-item scale was used to measure participants’ gender-
role socialization. The scale is comprised of 10 masculine items, 10 feminine items, and
10 neutral items that are used as fillers. An example of a masculine item is “I can control
a lot of the kids in my class;” an example of a feminine item is “it makes me feel bad
when someone else is feeling bad;” and an example of a neutral item is “I’m always
losing things” (see Appendix F). Each item is worded as a statement about the self, and
participants were asked to rate each item according to how true it was for them on a 4-
point scale (1 = not true at all of me, 2 = a little true of me, 3 = mostly true of me, and 4 =
very true of me). Separate masculinity and femininity scores were computed by averaging
the responses to the scale items; higher scores indicated higher masculinity or femininity.
Boldizar (1991) reported that the coefficient alpha for the masculine scale was .75, and
the coefficient alpha for the feminine scale was.84, indicating adequate reliability. In
addition, the masculinity and femininity scales were not significantly correlated with each
other, indicating independence of the two scales. In Boldizar’s study, for girls whose
mean ages ranged from 9-13 years of age, the mean score for the masculinity subscale
was 2.69 and for the femininity subscale was 3.10.

In the current study, the CSRI was examined in two different ways: first, the
averages of the masculinity subscale and the femininity subscale were used in the
structural equation modeling. Because findings based on classifications may mask the

individual contributions of masculinity and femininity, it is important to examine the

77

individual continuous scores, as these may be more meaningful (Hoffman & Borders,
2001). Indeed, in her initial validation study, Boldizar (1991) first examined masculinity
and femininity on a continuum, and then examined the CSRI using the median-split
methods; while she found significant results that indicated that androgynous children
were better adjusted, she suggested that their high scores in masculinity were the
foundation of the significant results.

Second, the CSRI was scored using the median-split method first proposed by
Bern and used frequently (e.g., Spence et al., 197 5), in order to parse out the individual
contributions of masculinity and femininity, and to compare the results of the current
study with previous research (e. g., Harter et al., 1997b). Those that scored above the
median for both the masculine and feminine scales were classified as Androgynous, those
below the median for both the masculine and feminine scales were classified as
Undifferentiated, those above the median for the feminine scale and below the median for
the masculine scale were classified as Feminine, and those above the median for the
masculine scale and below the median for the feminine scale were classified as
Masculine. In the current scale, the coefficient alpha was .78 for the masculinity subscale
was .78 and .79 for the femininity subscale.

Voice

Inauthentic Self in Relationships Scale (ISR, T olman & Porche, 2000)

This ten-item measure is a subscale from the Adolescent F emininity Ideology
Scale (AF IS), which is composed of the ISR and another subscale measuring the extent of
the objectification of one’s own body. The subscales are completely independent;

therefore, the utilization of the ISR by itself does not compromise the integrity of the

78

AF IS (Tolman & Porche, 2000). The ISR measured the extent to which participants’
have internalized inauthentic relationships with others, a significant part of adolescent
female development (Tolman & Porche, 2000). Sample items include, “I express my
opinions only if I can think of a nice way of doing it,” “often I look happy on the outside
in order to please others, even if I don’t feel happy on the inside,” and “I wish I could say
what I feel more often than I do” (see Appendix G). Participants were asked to respond
on a 6-point Likert scale ranging from 1 = strongly disagree to 6 = strongly agree.
Scores were obtained by summing the responses for each item, and dividing by the total
number of items. The items were reverse-scored prior to summing so that lower scores
indicate higher levels of internalization of the inauthentic self. The ISR was negatively
related to a scale of mutuality in relationships, thus supporting construct validity (Tohnan
& Porche, 2000). Coefficient alphas for the subscale ranged fi'om .67 to .81, depending
upon the age of the respondents (T olman & Porche, 2000). In the current study, the
coefficient alpha was .62.
Teenage Voice (Harter, 1995)

This five-item questionnaire was used to measure participants’ levels of voice.
The five items measured the extent to which respondents were able to “express their
opinions,” “share what they are really thinking,” “let others know what is important to
them,” “say what is on their mind,” and “express their points of View.” It was
administered five times, to measure level of voice in five different contexts: with mother,
father, teacher, classmates, and best friends (see Appendix H). The item format for this
measure was first developed by Harter (1985a) to reduce socially desirable responses. An

example item is: “some teenagers share what they are really thinking with their mothers

79

BUT other teenagers find it hard to share what they are thinking with their mothers.” For
each item, participants were asked to read both statements and decide which statement
best described them, the one on the left, or the one on the right. After choosing the
statement that best described them, they were asked to mark whether the chosen
statement was really true or sort of true for them. The responses were then scored on a
four-point scale ranging fiom highest level of voice (4) to lowest level Of voice (1),
summed, and averaged across the number of items. Thus, lower scores indicated lower
levels of voice.

Harter reported that the internal consistency reliabilities across the relational
contexts that she measured ranged fiom .86 to .89 (Harter et al., 1998b), and she
supported the use of additional relational contexts (Harter, 1995). Tests of construct
validity indicated that low levels of voice represented false self behavior, rather than
shyness or reluctance to voice private opinions. She reported that the majority of
adolescents (69% - 78%) perceived lack of voice as the result of false self behavior
(Harter et al., 1998b). In the current study, the coefficient alpha for level of voice was .84
with mother, .90 with father, .88 with teacher, .82 with classmate, and .81 with best
fiiend.

Friendship
Friendship Qualities Scale (F QS, Bukowski et al., 1994)

This 22-item scale measured participants’ perceptions of the qualities of their
closest fiiendship. Sample items included “my fiiend and I spend a lot of our fi'ee time
together,” “even if my fiiend and I have an argument we would still be able to be fiiends

with each other,” and “when I have not been with my fiiend for a while I really miss

80

being with him/her” (see Appendix I). The current study used three subscales of
fiiendship quality: conflict, security (including aspects of continuity of intimacy despite
problems in the relationship and trust of friendship), and closeness (including aspects of
affective bond within the fiiendship, and reflected appraisal, or how much the adolescent
feels important to his/her fiiend). Each participant wrote the initials of her closest fiiend,
and then rated how true an item was of this friend, using a 5-point Likert scale where 1 =
not true and 5 = really true. Scores were obtained by summing the responses for each
item within each subscale, and dividing by the total number of items in the subscale.
Higher scores indicate higher levels of conflict, security, and closeness. In the current A
study, the coefficient alphas for the subscales were .66, .86, and .89, respectively.
Romantic Relationships
Conflict in Adolescent Dating Relationships (CADRI, Wolfe et al., 2001)

This 62-item scale measured participants’ conflict with a romantic partner. There
are four scales of abuse: threatening behavior (four items), relational abuse (three items),
physical abuse (four items), and verbal emotional abuse (ten items) (see Appendix J).
There is an additional scale of positive conflict resolution behavior; these items are
typically endorsed at a high rate and are unrelated to the abuse questions, yet the authors
noted that focus groups felt that it was important to include positive items in the measure
(Wolfe et al., 2001). Respondents were asked to rate how often each of the items may
have occurred to them with a boy/girlfriend during an argument, and rated the items on a
4-point Likert scale ranging from 1 = never to 4 = ofien. An example of a threatening
behavior question is “I threatened to hurt him/her,” an example of a relational abuse item

is “I tried to turn his/her fiiends against him/her,” and example of physical item is “I

81

pushed, shoved, or shook her,” and an example of a verbal emotional abuse item is “I did
something to make him/her feel angry.” The coefficient alphas for the scales in the
current study are .66, .52, .83, and 82, respectively. Scale scores were computed for each
scale separately by dichotomously coding each item if it had occurred or not, and then
summing across all of the items in a given subscale.

Romantic Relationship Satisfaction (Levesque, I 993)

This 58-item measure assessed participants’ satisfaction with their romantic
relationships (if they had a current or former romantic relationship) and measured their
relationship experiences on a 6-point Likert scale ranging from 1 = Strongly Disagree to
6 = Strongly Agree. A sample item from the satisfaction scale is “our relationship has
met my best expectations”; the coefficient alpha reported for this scale was .88
(Levesque, 1993), and was .92 in the current study. The scale score was computed by
averaging the item responses. After the 5-item satisfaction scale, twelve different
dimensions Of relationships are measured: togetherness, possessiveness, growth,
communication, exhilaration, toleration, passion, appreciation, emotional support,
painfulness, commitment, and Specialness (see Appendix K). The questions for each
dimension assess what the individual gives and gets from the relationship. A sample item
fi'om the Specialness subscale is “I feel that s/he was meant for me” whereas a sample
item from the growth scale is “I am pleased when s/he pursues her/his own interests.”
Separate scores for each subscale were computed by averaging the item responses. All
items include both gender pronouns for inclusion of heterosexual and homosexual

relationships. A factor analysis supported the scale structure (Levesque, 1993).

82

Coefficient alphas for the twelve subscales ranged from .61 to .92, indicating adequate
internal reliability.

Psychological Outcomes
Children ’3 Depression Inventory — Short Form (CDI-S, Kovacs, 1992)

This 10-item self-report scale measures a variety of depressive symptoms and is a
shortened version of the original CDI. For each item, participants were asked to pick one
of three written statements that best describes their recent feelings in the past two weeks,
(e.g., “I am sad once in a while,” “I am sad many times,” or “I am sad all the time”). See
Appendix L. A sum score was obtained and averaged across the number of items, with
higher scores indicating more severe depressive symptomatology. For the structural
equation modeling, the scale was reverse-scored so that higher scores indicated less
depressive symptomatology, in order to be consistent self-esteem, and create a construct
of psychological well-being. The coefficient alpha for the current study was .82; this
scale has been widely used and has demonstrated good reliability and validity.
Rosenberg Self-Esteem Questionnaire (RSEQ, Rosenberg, I 965)

This measure consists of 10 items measuring self-worth and self-acceptance.
Examples of items include “I feel that l have a number of good qualities,” “1 take a
positive attitude toward myself,” and “I certainly feel useless at times” (see Appendix M).
Respondents were asked to rate their agreement on a 4-point Likert scale ranging from 1
= Strongly Agree to 4 = Strongly Disagree. A sum score was obtained and averaged
across the number of items, with higher scores indicating higher self—esteem. This scale
is widely used and has demonstrated good reliability and validity; the coefficient alpha in

the current study was .83.

83

Social Support
The Self-Perception of Social Support -— Approval subscale — Revised (SPSS-A, Harter &
Robinson, 1988).

This subscale is part of the revised version of the Social Support Scale for
Children (Harter, 1985b). It assessed participants’ perceptions of how much people in
their life approve of them and display positive regard toward them. Participants rated
their perception of approval support from six sources: mother, father, best fiiends,
classmates, teachers, and a person with whom they are romantically interested (see
Appendix N). A sample item is: “some teenagers have teachers who are really proud of
their students BUT other teenagers have teachers who are n_ot very proud of their
students.” For each item, participants read both statements and decide which statement
best describes them, the one on the left, or the one on the right. After choosing the
statement that best describes them, they mark whether the chosen statement was really
true or sort of true for them. This response format was structured to reduce socially
desirable responses. The responses were then scored on a four-point scale ranging from
least supportive (l) to most supportive (4), and averaged across items for a separate score
for each source. Higher scores indicate higher levels of perceived support The coefficient
alphas for this scale in the current study were .72 for classmate approval, .82 for best
friend approval, .83 for romantic interest approval, .84 for both mother and teacher, and
.90 for father approval.

The Perceived Social Support Scale (PSS-FR, PSS-FA, Procidano & Heller, 1983)

This 40—item scale was used to assess participants’ perceived emotional social

support from fiiends and family. Sample items include “my fiiends give me the moral

84

support I need” and “my family enjoys hearing about what I think” (see Appendix 0).
The scale was altered from the original response format of “yes,” “no,” and “don’t know,”
to a 4-point Likert scale ranging from 1 = strongly disagree to 4 = strongly agree in order
to allow for greater variance within the scale and to remove the neutral point. Scores
were summed separately for friends and for family, and averaged across the number of
items; higher scores indicate higher levels of perceived support. The researchers reported
that validity was established by indicating significant relationships between the scales and

measures of social competence and psychopathology. In the current study, the coefficient

alpha was .88 for fiiend support and .93 for family support.

85

RESULTS

All of the participants in the current study were determined to be at least in the
early puberty stage, the criteria for inclusion in the study. One percent of participants
were in early puberty, 7% were in mid-puberty, 72% were in late puberty, and 18% of
participants were post pubertal. Two participants did not fully complete the puberty
questionnaire, but one of those was determined to be at least in early puberty from the
items to which she responded, and the other participant was observed to be physically
developed, and consultation with the participant’s teacher confirmed that she was at least
in mid-puberty.

Data were initially collected on 125 participants; however, during data collection,
15 girls reported that they did not have a significant father figure for whom to report level
of voice, and two girls reported that they did not have a significant mother figure for
whom to report level of voice. Given that these relationships did not exist for these
participants, it was determined that the data could not be estimated appropriately.
Further, comparisons of the correlations between the remaining subscales for level of
voice (i.e., with classmates, teachers, and best fiiends, and the existing parent) were
examined, and were found to be markedly different for girls who did not report on one of
their parents versus the remaining 108 girls. Therefore, it was decided to exclude the data
for the 17 girls, as they reflected a nonrandom sample, and did not affect the focus of the
current study.

All data were screened for missing values, and significant kurtosis and skew.
PRELIS was used to screen for missing data. The EM Algorithm was used to estimate

missing data (1.74%). Several variables had significant negative kurtosis and skew; that

86

is, the ratio of kurtosis and skew to their respective standard errors was much greater than
21:2 (level of voice with best fiiend, mother approval support, father approval support, best
friend approval support, and fiiend support). Accordingly, these variables were log-
transformed (base 10). Subsequent analysis indicated that the log transformation greatly
reduced the kurtosis and skew (see Table 2). See Table 3 for the psychometric properties
of all of the variables and Table 4 for the correlation table of all of the variables.

Table 2.

Skew and kurtosis values for transformed variables

 

 

 

 

 

 

 

 

 

 

Before transformation After transformation
Skew (SE) Kurtosis (SE) Skew (SE) Kurtosis (SE)
Best fiiend voice -2.00 (.23) 5.66 (.46) -l.12 Q3) 0.42 (.46)
Mother approval -1.38 (.23) 1.78 (.46) -0.64 (.23) -0.59 (.46)
Father approval -1.4fl.23) 1.53 (.46) -0.80 (.23) -0.44 (.46)
Best friend approval -1.59 (.23) 2.85 (.46) -0.80 L23) -0.23 (.46)
Friend support -.62 (.23) 1.13 (.46) 0.20 (.23) 0.08 (.46)

 

 

 

 

87

Table 3.

Psychometric properties of measures: Means, standard deviations, and range

 

 

Measure Mean SD Range of Possible
Scores Range
BSQ
Security — attachment system 3.17 .89 1-5 1—5
Security — caregiving system 3.28 .72 1-5 1-5
Security — affiliation system 3.40 .92 1-5 1-5
Dismissing — attachment system 3.03 .93 1-5 1-5
Dismissing —— caregiving system 2.85 .69 1.4-5 1-5
Dismissing — affiliation system 2.23 .90 1-5 1-5
Preoccupied — attachment system 2.25 .72 1-4.4 1-5
Preoccupied — caregiving system 2.52 .72 1-4 1-5
Preoccupied — affiliation system 2.54 .83 1-5 1-5
ISR 3.96 .79 2.1-5.6 1-6
Teenage Voice
Mother voice 2.83 .89 1-4 1-4
Classmate voice 2.91 .79 1-4 1-4
Teacher voice 2.63 .89 1-4 1-4
Father voice 2.56 1.01 1-4 1-4
Best friend voice 3.65 .52 1-4 1-4
CSRI
Masculine subscale 2.69 .54 1.4-3.8 14
Feminine subscale 2.99 .54 1.7-4 1-4
Friendship Scale
Security 4.15 .77 2-5 1-5
Conflict 2.34 .86 l -5 1 -5
Closeness 3.92 .88 1 .63-5 1-5
CDI 1.64 .36 0-1.5 0-2
RSEQ 3.16 .55 1.8-4 1-4
CADRI
Physical abuse .69 1.01 0-4 0-4
Relational abuse .52 .81 0-3 0-3
Emotional abuse 4.07 2.81 0-10 0-10
Threat .76 1.16 0-4 0-4
Romantic Relationship Satisfaction
Relationship satisfaction subscale 4.20 1.40 1-6 1—6
Growth subscale 4.60 1.25 1-6 1-6
Specialness subscale 4.32 1.46 1-6 1-6
Composite Social Support with Peers .00 .75 -2.11-1.27 n/a
(z-score)
Composite Social Support with Authority .00 .83 -2.23-1.29 n/a
Figures Escore)

 

88

m0. V a «

38mm 3:053 I 98.33 380m .mN 83a I «8&8 33m . 5N 333:» 3533mm .ON 3333. 5380 .N 3333- 83838 OEEOOa—ox .VN

ugh MN 38% 08288m .NN 38% Queue—ox ._N 33a 3335 .ON Ommm .O_ 50 .3 8083.20 .5_ 86800 .2 553m .n _ 0888a 08883 .3 £333
2:39.32 .2 38> 035 “mom .2 38> Bauh— .: 38> 333.5 .0_ 38> 8.8820 .O 38> 3502 .w «E .5 8893 nova—Ea I magma .O 829$
3.5883 I wagfla .n 88m? 808.833 I 8388me .v 889?. comma-ma I bra—com .m 839$ 8883 I bra—com .N 883m 808883 I .0853 .—

«5mO «NNO «O NO «3.0 «NmO O_ .0 «3.0 _ _ .0 «OvO- «ch. «NvO- « _0.0 «NmO «va wN
«OvO «mVO «me n — .0 «OmO «ON-.0 «ONO w _ .0 3.0- « _ NO. «NO. «ONO MOO 5_ .0 5N
«OmO «OmO NOO O0.0 «VNO 3.0 O_ .0 OOO N_ .0. _0.0. 00.0. «m NO OOO O_ .0 ON
«3.0 «va No.0 3.0 «ONO 3.0 :O O—O :O. 3.0. 3.0. «3.0 3.0 OOO ON
«OmO «mm .0 3.0 m _ .0 «OmO _ _.0 O— .0 3 .0 OOO. NOO. mOO «VNO O_ .0 wOO VN
OOO. _ _.0. O— .0. 3.0 3.0. OOO OOO. m0.0. OOO- 3.0 3.0 3.0. 50.0 50.0 MN
3 .0. O0.0 00.0 00.0 00.0 E .0 3.0. ~ _.0 g _.0. m _ .0. NOO N_ .0 no.0 NOO NN
_ _.O. O_ .0- 3.0- NOO 50.0- MOO «NNO- wOO- OOO N H .0 m ~ .0 OOO- NOO. 3.0. N
MOO- wOO. 2.0. 3.0 N_ .0. OOO- 3 .0. OOO. OOO. 50.0. MOO 3.0. 3.0 NOO- 0N
«0m.0 «NvO «ONO «ONO «NmO «nvO «NmO «OVO 5~ .O- «ONO. m 0 .O- «NmO O~ .0 «VNO O—
«wNO- «ON-.0. w _ .O. «OmO- «MNO- «NmO- «NO. «NO. «VNO «NNO 5 ~ .0 « _ mO- m _ .0. «ONO. w—
«mwO «N10 «NmO m g .0 « g mO O_ .0 NO .0 OOO m0.0. 3.0- 3.0. NO 3.0 3.0 5_
3.0. 50.0 n _ .0 OOO WOO- 50.0- wOO NOO- mOO OOO 0_ .0 OOO- 3.0. 3.0- O—
« 3x0 « 3.0 «3.0 w—O «ONO «ONO 3.0 O0.0 3.0- 2.0. 3.0- 5.0 N_.0 m0.0 m _
00._ «NvO N~O «NNO «NmO 0~ .0 «OmO NOO «MNO. «ONO- « _ NO- «OVO «ONO «OmO 3
00._ «5N.O «vNO «OvO « _ n .0 « _ NO «Om-.0 E .O- O _ .0. m _ .0- NNO 3.0 O— .0 Q
00._ m _ .0 «ONO «3.0 «mmO «5N.O O~ .0. «NNO. 5 _ .0. O_ .0 O0.0 50.0 N—
00. _ «0m .0 «an «NvO O_ .0 «OmO. « _ NO. «ONO- « _ mO «5mO «NO I
004 «5vO «me 5~ .0 « _ mO. «me. «NNO. «3.0 «5N.O «NmO O—
00; «ONO « O OO N_ .0. « _ NO. m0O- m _ .0 3.0 3.0. O
00. _ « O NO «NvO. «OmO. «mmO. «5v.0 «ONO « 5.0 w
00._ O ~O. v.— .0- OOO OOO NOO- O0.0- 5
00; « ~ nO «OmO « _ 5.0- «mmO. «O10. O
00; «OmO «WOO. «MOO- «OeO. m
00; « :VO- «NNO. « OOO- v
00._ «5O .0 «On .0 m
004 «NmO N
00.— _

3 2 N_ ~ ~ 0~ O m 5 O O v m N —

sauce, 8 .8 use 8888 4. 22¢

89

004

ON

«OvO
004

5N

04.0
«VOO
00.4

ON

04.0
«5m.0
«55.0

00.4

ON

ONO
«me
«w5.0
«m5.0

00.4

VN

«NO.
«NO.

3.0
O 4.0-
O4 .0.
00.4

ON

OOO.
m4 .0.
50.0
OOO.
w 4 .0-
LOO
004

NN

«OOO-
04.0-
50.0.
O4 .0.

«5N.0-

«5OO

«wOO
004

4N

O 4 .0.
O4.0-
OOO.
O 4 .0-
m4 .0-
«OOO
«OOO
«OOO
004

ON

«510
«3&0
« 4 mO
«4N.0
«mmO
OOO
OOO
4 4.0-
3.0
004

O4

«OOO-
«VOO-
«NNO-
«ONO-
«OOO-
OOO.
wOO.
OOO
N4 .0.
«OOO.
00.4

04

04.0
«MOO
O4.0
«5mO
«MOO
00.0.
N0.0
50.0
NOO
«4NO
4V4 .0.
00.4

54

04 .0-
OOO-
OOO-
«ONO-
04.0.
3.0
O 4 .0
NOO-
OOO
40.0-
OOO
«NNO
004

O4

3. v a .
mega—.4 584853 I toga 483m .wN 83 I .8843 483m . 5N 8338 3888qu .ON 04333 5380 .ON 04333 338828 8448848434 .3

838.4. N 3.5a 488488m .NN 3245 48848434 .4N 38% 4393.4 .ON Ommm .O4 490 .3 $33340 .54 84980 .2 383m .O 4 04333 8484804 .3 04335
084382 .2 38> 4805 83m .N4 38> 4843-4 .44 38> 3.403.4- .04 38> 88340 .O 38> 344842 .0 ~54 .5 889$ 34844448 I wfimmEmE .O 889$
3883 I 938899 .O 889$ 868888 I 8484889 .4. 889$ caveman I 544.88% .m 889$ 88803 I bracem .N 889$ 86884083 I 383m .4

« 4 NO
«0O .0
«VNO
«5nd
«OvO
N4 .0-
OOO-
OOO-
40.0
«5N.0
O4 .0.
«O OO
«NNO
004

O4

wN
5N
ON
ON
ON
ON
NN
4N
0N
O4
O4
54
O4
O4

OF‘NMV
w-I—tu—nv—rv—I

“NMVWOFOOO‘

.38 883 8 .8 use 8388 .4 2.3

90

Structural equation modeling (SEM) was used to test the hypothesized models
shown in Figures 3, 4, and 5. The following fit indices were examined to assess model
fit: Chi-Square test, the Goodness-of-fit Index (GFI), and the Root-mean-square error of
approximation (RMSEA). The guidelines for interpreting the GF I are: 1.00 = ideal fit,
.90-.99 = excellent fit, .85-.89 = average fit, < .85 = poor fit. The guidelines for
interpreting the RMSEA are: .00-.05 = excellent fit, 05-08 = moderate fit, .08-.10 =
acceptable fit, > .10 = poor fit (Bollen, 1989). The SEM analyses informed modifying the
model, as needed, to improve overall model fit, guided by the constraints of theoretical
consistency.

As discussed previously, level of voice functions differently depending upon the
setting (Harter et al., 1998b). To determine if this was the case for the current study, and
if so, how setting categories should be constructed, an exploratory principal axis factor
analysis, varimax rotation (SPSS), was conducted on the six level of voice indicators (see
Table 5 for factor loadings). The factor analysis supported a two-factor solution with
mother voice, father voice, and teacher voice loading on Factor 1 (Authority Voice), and
the ISR scale, classmate voice, and best fiiend voice loading on Factor 2 (Peer Voice).
Therefore, each of the structural models was run twice, for authority figures and for peers.
Table 5

F actor loadings for level of voice variables

 

 

Factor 1 Factor 2
Mother voice .51 .26
Father voice .75 .1 1
Teacher voice .63 .28
ISR .01 .57
Classmate voice .33 .79
Best fiiend voice .20 .49

 

91

Results of the Hypotheses

All of the structural equation models were tested using maximum likelihood (ML)
estimation. As is customary, the measurement models were tested prior to testing the
structural models. In the interests of parsimony and of obtaining appropriate parameter
estimates, most of the measurement models were tested by estimating the fit of the entire
model, with the latent constructs covarying with each other, and the direct paths
eliminated (Schurnacker & Lomax, 1992). There was one exception to this. When the
overall measurement model was tested for predictors Of voice, and nine attachment
indicators comprised the averages of subtype of attachment for each of the three
behavioral systems, it was found that none of the indicators for the preoccupied subtype
significantly loaded on the latent construct of attachment (see Table 6 for factor loadings).
Table 6.

Standardized regression weights for attachment measurement model

 

 

 

 

 

 

 

 

 

 

 

 

 

Estimate
Security — attachment system 0.65"
Security — caregiving system 0.64"
Security — affiliation system 0.84“
Dismissing — attachment system -0.58**
Dismissing — caregiving system -0.71**
Dismissing —- affiliation system -0.78**
Preoccupied - attachment system 0.10
Preoccupied — caregiving system -0.05
Preoccupied — affiliation system -0.06

 

** p < .001

In addition, in another attempt to test a three-indicator measurement model of
attachment by collapsing the subtype of attachment scores across the systems,
preoccupied attachment still did not load significantly on the latent construct of

attachment, and a negative error variance was present, possibly indicating other problems

92

with the model. As a result, the decision was made to only use the secure and dismissing
subscale scores for each of the three behavioral systems. The initial measurement model
indicated poor fit for the data [x2 (9, N= 108) = 52.72, p < .01, GFI = .87, RMSEA =
.21], and the modification indices suggested correlating several of the error terms. This
was theoretically sound given that the same subtype of attachment was measured in
different behavioral systems, and that each behavioral system was represented twice, one
per each attachment subtype. After correlating three error terms, the model provided an
excellent fit for the data [x2 (6, N = 108) = 6.17, p = .41, GFI = .98, RMSEA = .02]. This
was the final measurement model used for attachment in the predictor models.
Predictor Model for Authority Figures

The measurement model for attachment, gender role socialization, and level of
voice with authority figures indicated a moderately good fit for the data [x2 (39, N = 108)
= 56.77, p = .03, GFI = .92, RMSEA = .07]. Please see Figure 5 for the measurement
model with the standardized estimates. All of the indicators loaded significantly on their
respective latent constructs.

The structural model was also a good fit for the data [x2 (39, N = 108) = 56.77, p =
.03, GFI = .92, RMSEA = .07]. Please see Figure 6 for the structural model with the
standardized pathways.
Hypothesis 1

As predicted, a direct, positive relationship was demonstrated between attachment
and level of voice with authority figures (Path A, Figure 6). The standardized regression
weight was .55 (z = 3.80, p < .001, one-tailed test). Participants with higher security were

more likely to have higher levels of voice.

93

Hypothesis 2

As predicted, a direct, positive relationship was demonstrated between gender role
socialization and level of voice with authority figures (Path B, Figure 6). The
standardized regression weight was .43 (z = 2.58, p < .01, one-tailed test). Participants

with higher gender role socialization were more likely to have higher levels of voice.

94

 

 

 

 

 

 

1 1
O O O
0.64“ 0.66“ 1 1 1

 

  
 
     
   
 
 
    
   
  

 

 

 

 

Invoice fvoice tvoice

 

 

 

Gender Role

063* 0 .67“
Socialization

0.71“ 068“

  

Level
of Voice -
Authority

0.50"

  
   
 

Attachment

0.77“
with Parents

0.62" ,,
.62 o.92** -0.48 *-o.s "-0.77"

 

 

 

$60811 seccar e

secaff disatt discare disaff

1 1

 

 

 

 

 

 

 

 

 

11 11
999999
A]

'0.41** 4125* 0.41 **

*p<.05, "p<.001

Figure 5. Measurement model for attachment, gender role socialization, and level of
voice with authority figures.

95

 

masc fem

 

 

 

 

 

0.64" 0.66"

 

mvoice fvoice tvoice

   
     
   
 
 
    
   
  

 

 

 

 

 

 

 

0.63* " .67"
0.68"

Gender Role
Socialization

  

Level
of Voice -
Authority

 
    
 

Attachment
with Parents

0.62“ ,, ...
'52 .92“ -0.48 ..o. -o.77**

 

secatt seccare secaff disatt discare disaff

 

 

 

 

 

 

 

 

 

 

 

*p<.05, **p<.001

Figure 6. Structural model of attachment and gender role socialization as predictors of
level of voice with authority figures.

96

Predictor Model for Peers

The measurement model for attachment, gender role socialization, and level of
voice with peers indicated a moderately good fit for the data [x2 (39, N = 108) = 68.52, p
< .01, GFI = .90, RMSEA = .08]. Please see Figure 7 for the measurement model with
the standardized estimates. All of the indicators loaded significantly on their respective
constructs.

The structural model was also a moderately good fit for the data [3809, N = 108)
= 68.52, p < .01, GFI = .90, RMSEA = .08]. Please see Figure 8 for the structural model
with the standardized pathways.

Hypothesis 1

There was not a significant relationship between attachment and level of voice
with peers (Path A, Figure 8). The standardized regression weight was -.11 (z = -.82, p =
.21, one-tailed test).

Hypothesis 2

As predicted, a direct, positive relationship was demonstrated between gender role
socialization and level of voice with peers (Path B, Figure 8). The standardized
regression weight was .55 (z = 3.30, p < .001, one-tailed test). Participants with higher

gender role socialization were more likely to have higher levels of voice.

97

 

 

 

 

 

 

 

1 1
masc fem 9 Q e1 1
0.68“ 0.63“ 1 1 1
bfvoice clvoice isr

  
 
     
   
 
 
   
   
 
 

 

 

 

 

 

 

 

Gender Role

0.45* 0.99“ 0,51M
Socialization

0.51“

  

Level
of Voice -

0.47 Peers

 
    
 

Attachment
with Parents

0.61“ “.3” .91” -0.4 *-0.6** -0.78**

 

 

 

 

 

 

 

 

 

 

secatt [Seccare secaff disatt discare disaff

 

 

-0.41** -0,25* 0.41"

*p< .05, **p< .001

Figure 7. Measurement model for attachment, gender role socialization, and level of
voice with peers.

98

0.47“

 
 
   
 

 

masc

 

 

fem

 

 

 

0.68” 0.63"

   
       
 

0.61“ “.3" .91“

Attachment
with Parents

Gender Role
Socialization

-0.4

* -0.6 ‘ ** -0.78**

 

secatt seccare

 

 

 

secaff

 

 

disatt

 

 

discare

 

B: .56“

disaff

 

 

bfvoiee clvoice isr

 

 

 

 

 

 

 

 

*p< .05, **p< .001

0.45* 0.99" 051"

Figure 8. Structural model of attachment and gender role socialization as predictors of
level of voice with peers.

99

Friendship and Well-Being Outcome Model for Authority Figures

The measurement model for level of voice with authority figures, fiiendship, and
psychological well-being indicated an excellent fit for the data [12 (17, N = 108) = 14.15,
p = .66, GFI = .97, RMSEA = .00]. Please see Figure 9 for the measurement model with
the standardized estimates. All of the indicators loaded significantly on their respective
constructs.

The structural model was also an excellent fit for the data [x2 (l 8, N = 108) =
15.44, p = .63, GF I = .97, RMSEA = .00]. Please see Figure 10 for the structural model
with the standardized pathways.
Hypothesis 3

As predicted, a direct, positive relationship was demonstrated between level of
voice with authority figures and quality of fiiendship (Path C, Figure 10). The
standardized regression weight was .35 (z = 2.72, p < .01, one-tailed test). Participants
with higher levels of voice reported higher quality of friendship.
Hypothesis 4

As predicted, a direct, positive relationship was demonstrated between level of
voice with authority figures and psychological well-being (Path F, Figure 10). The
standardized regression weight was .52 (z = 3.13, p < .001, one-tailed test). Participants
with higher levels of voice reported less depressive symptomatology and higher self-

esteem.

100

 

1’?

e6
1
conflict close

 

 

 

 

 

 

 

0.95” 023* o_90**

 

   
   
   

 

 

 

 

 

 

 

Friendship
1 0.32*
IVOiCG 0.70.".
1 , o.71** Level
fVOIce of Voice - 0.28”
0.58” Authority
e1 1 mvoice
0.50*

   
   

Psychological
Well-being

0.74" 0.92“

 

 

depress selfest

1 1

 

 

 

 

 

 

*p<.05,**p<.001

Figure 9. Measurement model for level of voice with authority figures, fiiendship, and
psychological well-being

101

 

security conflict close

 

 

 

 

 

 

 

0.91 * 023* (193Mr

 

 

   
   
   

 

 

 

Friendship
C: .35*
tVOiCB 0.70**
. 0,71" Level
fVOIce of Voice -
o_57** Authority
mvoice F: .52"

 

 

 

 
     

Psychological
Well-being

0.76" 0.91"

 

 

 

 

 

 

depress selfest

 

 

*p< .05, **p< .001

Figure 10. Structural model of fiiendship and psychological well-being as outcomes of
level of voice with authority figures.

102

Friendship and Well-Being Outcome Model for Peers

The measurement model for level of voice with peers, friendship, and
psychological well-being indicated a very good fit for the data [x2 (17, N = 108) = 24.73,
p = .10, GFI = .95, RMSEA = .07]. Please see Figure 11 for the measurement model with
the standardized estimates. All of the indicators loaded significantly on their respective
constructs.

The structural model was also an excellent fit for the data [x2 (18, N = 108) =
25.74, p = .11, GFI = .95, RMSEA = .06]. Please see Figure 12 for the structural model
with the standardized pathways.
Hypothesis 3

As predicted, a direct, positive relationship was demonstrated between level of
voice with peers and quality of friendship (Path C, Figure 12). The standardized
regression weight was .35 (z = 3.03, p < .001, one-tailed test). Participants with higher
levels of voice reported higher quality of fiiendship.
Hypothesis 4

As predicted, a direct, positive relationship was demonstrated between level of
voice with peers and psychological well-being (Path F, Figure 12). The standardized
regression weight was .60 (z = 3.61 , p < .001, one-tailed test). Participants with higher

levels of voice reported less depressive symptomatology and higher self-esteem.

103

 

63 bfvoiee

 

 

clvoice

 

 

 

 

isr

 

f?

 

 

*p< .05, **p< .001

 

security

 

 

 

conflict

 

close

 

 

 

0.98 022" 0.86“

Friendship

0.65“
0.81 **
0.61 **

0.70“

 

 

 

depress

 

Psychological
Well-being

     

0.98”

 

 

selfest

 

 

Figure I 1. Measurement model for level of voice with peers, friendship, and

psychological well-being

104

 

security conflict close

 

 

 

 

 

 

 

0-97 '-23* 0.87“

 

 

 

 

 

Friendship
C: .35*
bfvoiee 0 559*
, 0.78“
clvorce
0.61 **
F: .60**
isr

 

 

 

 

 
     

Psychological
Well-being

0.70“ 0.98"

 

depr ; : ; selfest

 

 

 

 

*p< .05, **p< .001

Figure 12. Structural model of fiiendship and psychological well-being as outcomes of
level of voice with peers.

105

Romantic Relationship Outcome Model for Authority Figures

As anticipated, not all of the participants had been involved in a romantic
relationship at the time of data collection. Seventy-five participants indicated that they
had at least started dating, and therefore they completed the CADRI and RRS measures.

There were originally ten subscales for the RS, but this would have made 17
indicators for the entire measurement model, Violating the rule of approximately ten
subjects per indicator (Schurnacker & Lomax, 1992). In order to reduce the number of
indicators for quality of romantic relationships, the two negative scales, pain and
possessiveness, were dropped because they had the least conceptual meaning, based on
development of romantic relationships, for 14-year-olds. In addition, these subscales had
the lowest coefficient alphas (.74 and .67, respectively). Correlations among the
remaining eight subscales were high, indicating possible redundancy among the subscales
(see Table 7). In addition, the measurement model for the confirmatory factor analysis
indicated that the model was not a good fit for the data [x2 (20, N = 75) = 57.94, p < .001,
GFI = .83, RMSEA = .16]. Before assessing alternative measurement models for the
RS, with fewer indicators in the interest of parsimony, it was decided to retain the
satisfaction with relationship scale, given that it was the only scale that specifically
measured relationship satisfaction. Alternative models were tested, and the best fit for
the data was the model that retained the growth and Specialness scales. This model was
an adequate fit for the data [x2 (1, N= 75) = 2.56,p = .11, GFI = .98, RMSEA = .15], and
the standardized factor loadings for the three indicators were .80, .83, and .90, all

significant at p< .001.

106

Table 7. Correlations among the eight positive subscales of the RRS

l

l. Satisfaction w/rel. 1.00

2. Togethemess

3. Growth

4. Communication
5. Toleration

6. Appreciation

7. Emot. support
8. Specialness

** p < .001

0.85"
0.74"
0.77**
0.69"
0.76"
0.81"
0.78”

2

1.00

0.74'Ml
0.72"
0.69"
0.71 *‘1'
0.78"
0.81"

3

1.00

0.82“
0.77“
0.84"
0.78"
0.77"

4 5 6 7 8
1.00
0.82" 1.00
0.83“ 0.86" 1.00
0.82" 0.84" 0.85" 1.00

0.80" 0.74" 0.76" 0.78" 1.00

The measurement model for level of voice with authority figures, dating conflict,

and quality of romantic relationship indicated a good fit for the data [x2 (32, N = 75) =

40.86, p = .14, GFI = .91, RMSEA = .06]. Please see Figure 13 for the measurement

model with the standardized estimates. All of the indicators loaded sigrrificanfly on their

respective constructs.

The structural model was also a good fit for the data [x2 (33, N= 75) = 41.91,p =

.14, GFI = .91, RMSEA = .06]. Please see Figure 14 for the structural model with the

standardized pathways.

Hypothesis 5

There was not a significant relationship between level of voice with authority

figures and dating conflict (Path D, Figure 14). The standardized regression weight was

-.08 (z = -.57, p = .57, one-tailed test).

Hypothesis 6

As predicted, a direct, positive relationship was demonstrated between level of

voice with authority figures and quality of romantic relationship (Path E, Figure 14). The

standardized regression weight was .31 (z = 2.11, p < .05, one-tailed test). Participants

107

with higher levels of voice with authority figures reported higher quality of romantic

relationships.

108

 

threat rel phys verbal

 

 

 

 

 

 

 

 

 

0.82” 0.80 0.77

Dating
Conflict

 

1"°‘°° 0.76“

 

   
     
     
   
 

 

 

 

, 0.49" Level
‘ ”0'09 of Voice -
0,74" Authority
mvoice

if?

 

 

 

Quality of
Romantic
Relationship

o.9o** 086* 0.87”

 

 

spec grow relsat
1 l 1 1
0 0

Figure 13. Measurement model for level of voice with authority figures, dating Violence,
and quality of romantic relationship.

 

 

 

 

 

 

 

 

*p< .05, **p< .001

109

 

tvoice

 

 

fvoice

 

 

Invoice

 

 

 

 

 

 

   
   
   

 
     
   

 

‘ threatJ I rel l ‘ phys [verbal
0.82“ 0.80* ‘ 0.77*
0.69“
Dating
Conflict
D: -.08
0.76“
o_49** Level
of Voice -
0.73“ Authority
E: .31*
Quality of
Romantic
Relationship
0.90“ 0.87”
0.85”
spec grow relsat

 

 

 

 

 

 

 

*p< .05, **p< .001

Figure 14. Structural model of dating violence and quality of romantic relationship as
outcomes of level of voice with authority figures.

110

Romantic Relationship Outcome Model for Peers

The measurement model for level of voice with peers, dating conflict, and quality
of romantic relationship indicated a good fit for the data [x2 (32, N = 7 5) = 40.86, p = .14,
GFI = .91, RMSEA = .06]. Please see Figure 15 for the measurement model with the
standardized estimates. All of the indicators loaded significantly on their respective
constructs.

The structural model was also a good fit for the data [x2 (33, N = 75) = 38.93, p =
.22, GFI = .92, RMSEA = .05]. Please see Figure 16 for the structural model with the
standardized pathways.
Hypothesis 5

There was not a sigrificant relationship between level of voice with peers and
dating conflict (Path D, Figure 16). The standardized regression weight was .04 (z = .34,
p = .73, one-tailed test).
Hypothesis 6

There was not a significant relationship between level of voice with peers and
quality of romantic relationships (Path E, Figure 16). The standardized regression weight

was .11 (z = .87, p = .38, one-tailed test).

111

1111

threat phys verbal

 

 

 

 

 

 

 

 

 

 

0.82” 0'80 0-77 0.70"

Dating
Conflict

 

. 0.05
DIVOIOe 042”.

 

 

. .97" Level
clvouce Of Voice - .015
0.47" Peers

 

 

111

 

isr

 

 

 

 
    

Quality of
Romantic
Relationship

 
   
 

0.90" 0.86“ 037"

 

 

Figure 15. Measurement model for level of voice with peers, dating violence, and quality
of romantic relationship.

 

 

 

 

 

 

 

 

*p< .05, **p< .001

112

 

threat rel phys verbal

 

 

 

 

 

 

 

 

 

 

 

   
     

 

 

 

0.82” 080* 0.77* 070»
Dating
Conflict
_ D: .04
bfvonce 0421M
,g7** Level
clvoice of Voice -
0.47“ Peers
E: .11
isr

 

 

 

 
 
 

Quality of
Romantic
Relationship

 
     

0.91“ 0.85 0-86"

 

 

 

 

spec grow relsat

 

 

 

 

 

 

 

 

*p<.os,**p<.001

Figure 16. Structural model of dating violence and quality of romantic relationship as
outcomes of level of voice with peers.

113

Results of the Additional Hypotheses
Hypothesis 7
As previously described, several of the social support variables were log
transformed to reduce kurtosis and skew. Following this, all of the social support
variables were transformed to z-scores to be on the same metric. An principal axis
exploratory factor analysis using varimax rotation was performed in order to determine if
the social support variables loaded on similar constructs, with the goal to reduce the
number of social support variables that would subsequently be entered into the regression
analyses. Please see Table 8 for the factor loadings of the social support variables.
Table 8

F actor loadings for social support variables

Factor 1 Factor 2
Mother approval support .69 .37
Father approval support .77 .04
Family emotional support .64 .34
Teacher approval support .50 .53
Best friend approval support .18 .64
Classmates approval support .24 .64
Romantic Interest approval support .09 .67
Friends’ emotional support .24 .48

The results of the factor analysis supported the use of two factors of social
support: authority figure support (mother approval support, father approval support, and
family emotional support) and peer support (best friend approval support, classmates’
approval support, romantic interest approval support, and friends’ emotional support).
Teacher approval support loaded on both factors, and thus was dropped from further

analyses. Composite scores of social support were created by averaging the z-scores for

114

the relevant indicators. Please see Table 3 for the psychometric properties of the
composite social support scales.

In addition, given the factor analysis that supported two different factors of level
of voice (see Table 5 for a review), composite variables of authority level of voice and
peer level of voice were created by converting the relevant indicators to z-scores,
summing the relevant indicators for each scale, and then averaging across the number of
indicators.

Determination of social support as a moderator between level of voice and
depression and self-esteem was tested using the method suggested by Baron and Kenny
(1986). Average level of voice was entered in the first step, average social support was
entered in the second step, and the interaction term was entered in the third step. In order
to minimize the number of regressions performed, it was decided to only examine the
relationship between level of voice and social support in the same relationship (i.e., only
regressions with interactions between voice with authority figures and social support with
authority figures and interactions between voice with peers and social support with peers
were performed). Each combination of voice and social support was performed twice,
once predicting depression, and once predicting self-esteem, for a total of four different
regression analyses.

Only the interaction between level of voice with authority and social support with
authority was significant, predicting depression (see Table 9). As level of voice
decreased, those with higher scores on approval and emotional support from authority
figures had lower scores on depression (t = 2.10, p < .05), while those with lower scores

on support had higher scores on depression. At higher levels of voice, there was no

115

difference in participants’ scores on depression (see Figure 17). Thus, there was partial
support for social support as a moderator between level of voice and depression.

Table 9

Regression Analysis for Social Support with Authority Figures as a Moderator of Level of
Voice with Authority Figures and Depression

 

 

Variable Adjusted R_2 A 33 E for B: A 13
Step 1: Average level of voice .11 .11 12.73" -.33**
Step 2: Approval and emotional .17 .07 8.54”“ -.30*

social support composite

Step 3: Interaction term .21 .03 4.40”“ .19“

 

*p< .05. **p< .001

0.6 T—_F+' —————- - ’_ . ”w- ____.- .—

Mina,

 

 

 

 

0.4 — -—~ ~——»— — ———
.2 i
m 'C‘ k_r,,<,__.‘
€03i~——— — 'HOHhiss
8' ‘ |+lossl
G 02 L-sk — A“ _,_ -—

 

 

.0
A
. +___.___3t.— -.

 

low high
Level of Voice with Authority Figures

Figure I 7. Level of voice and social support interaction

116

Hypothesis 8

In order to parse out the relationship of gender role socialization of level of voice,
the CSRI was divided into categories using the median-split method. The median of the
masculinity scale was 3.00, and the median of the femininity scale was 2.70. Participants
above the median of both scales were classified as androgynous, below the median of
both scales were classified as undifferentiated, above the median of the masculinity scale
but below the median of the femininity scale were classified as masculine, and below the
median of the masculinity scale but above the median of the femininity scale were
classified as feminine. Please see Table 10 for the distribution of the gender role
socialization categories for the current sample.
Table 10

Distribution of gender role socialization categories

 

 

% N
Androgynous 36 39
Undifferentiated 22 24
Feminine 23 20
Masculine 19 25

 

Following the categorization of participants, a multivariate analysis of variance
(MANOVA) was performed with gender role socialization category as the independent
variable and the average levels of voice in the peer voice composite variable and the
authority voice composite as the dependent variables (see Table 11 for mean levels of
voice across the gender role socialization categories). Planned contrasts were conducted

comparing the androgynous and feminine participants, and masculine and feminine

117

participants. When Levene’s test for homogeneity of variance was significant, equal
variances across groups were not assumed, and appropriate statistics were used.

Table 11

Means and standard deviations for level of voice across the gender role socialization
categories

 

 

Androgynous Masculine Feminine
Mean of authority voice .36” (.76) -.07 (.70) -.03"“ (.64)
Mean of peer voice .33“ (.55) .328” 1.58) -.57"" (.73)

 

Cells with different superscripted letters are significantly different from each other
*p < .05, **p < .001

Table 12

Correlations of voice with masculinity and femininity

 

 

Masculinity Femininity
Mean of authority voice .39** .38**
Mean of peer voice .50'MI .10

 

** p < .001

Results indicated that androgynous girls had significantly higher level of voice
with authority figures than did feminine girls (t = 1.96, p = .05), but there was no
difference between feminine and masculine girls on level of voice with authority figures.
In addition, results indicated that both androgynous and masculine girls had significantly
higher level of voice with peers than did feminine girls (t = 4.87, p < .001, and t = 4.39, p
< .001, respectively). Finally, mean level of voice with authority figures was positively
correlated with both masculinity and femininity while mean level of voice with peers was
positively correlated with masculinity (see Table 12). Thus, Hypothesis 8, that feminine
girls would have lower levels of voice than either androgynous or masculine girls was
supported for level of voice with peers, and partially supported for level of voice for

authority figures.

118

Hypothesis 9

In order to parse out the relationship of level of voice and dating conflict and
relationship satisfaction, the two composite level of voice variables were divided into
categories using the median-split method. The median of level of voice with authority
figures was .01, and the median of level of voice with peers was .23. Participants above
the median of the scale were classified as having high level of voice, and below the
median as having low level of voice. There were 54 participants in each category.

Following the categorization of participants, a multivariate analysis of variance
(MANOVA) was performed with high/low level of voice as the independent variable and
relationship satisfaction, physical abuse, relational abuse, emotional abuse, and threats as
the dependent variables (see Table 13 for the means and standard deviations for the
dependent variables across the high/low level of voice categories). Results indicated that
no difference in dating conflict across the level of voice categories, and participants with
high level of voice with authority figures did not tend to have higher relationship
satisfaction than participants with low levels of voice with authority figures. There was a
trend for participants in the high level of voice with peers category to have higher
relationship satisfaction than participants in the low level of voice with peers category
[F(l,73) = 3.52, p = .07]. Thus, Hypothesis 9 was not supported. There was no
difference in dating conflict for participants in the low voice category versus the high
voice category, although there was partial support for participants in the low voice with
peers category versus the high voice category with peers having lower relationship

satisfaction.

119

Table 13

Means and standard deviations for relationship variables with high/low levels of voice

 

 

 

 

Relationship Physical Relational Emotional Threats
Satisfaction Abuse Abuse Abuse
Voice with authority figures
High 4.32 (1.49) .58 (.79) .39 (.39) 3.92 (2.77) .63 (1.10)
Low 4.08 Q32) .81 (1.20) .65 (.65) 4.22 (2.88) .89 (1.22)
Voice with peers
High 4.49 (1.27) .66 (.85) .53 (.76) 4.45 (2.78) .68 (.93)
Low 3.90 (1.48) .73 (1.17) .51 (.87) 3.68 (2.81) .84 (1.36)
Hypothesis 10

In order to determine if there were ethnic differences in level of voice, a
multivariate analysis of variance (MANOVA) was performed with ethnicity as the
independent variable and level of voice with peers (ISR), level of voice with mother,
level of voice with classmates, level of voice with teachers, level of voice with father, and
level of voice with best friend as the dependent variables. To compare level of voice in
Caucasian versus girls of color, ethnicity was collapsed across the non-white ethnic
groups. In addition, because the ethnic composition of the middle schools are different,
school was used as a covariate (see Table 14 for fiequencies of different ethnic groups
across the three schools). Finally, given that there were only 13 participants from
Gardner Middle School, these participants were excluded from the analyses, for a total of

95 participants.

120

Table 14

Frequencies of diflerent ethnic groups across middle schools

 

 

 

 

School
Pattengill Gardner Dwight Rich
Caucasian 38% 43% 26%
Afiican-Arnerican 32% 40% 59%
Latina 22% 12% 9%
Asian-American 6% 4% 4%
Native-American 2% 1% 2%

 

 

 

Results indicated a significant effect for ethnicity with level of voice with
classmates [F (l ,93) = 8.20, p < .01], with girls of color having higher voice with
classmates than Caucasian girls. In addition, there was a trend for girls of color to have
higher voice with teachers than Caucasian girls [F (1,93) = 3.13, p = .08] (see Table 15 for
mean levels of voice for Caucasian girls versus girls of color).

Voice with peers (ISR) was the only variable that had a significant effect for
school [F (1 ,93) = 6.61, p < .05]. An additional analysis of variance with school as the
independent variable and ISR average as the dependent variable confirmed that
participants at Dwight Rich Middle School had higher levels of voice than girls at

Pattengill Middle School (M = 4.23 versus M = 3.73).

121

Table 15

Means and standard deviations for level of voice for Caucasian girls versus girls of color

 

 

 

 

 

Caucasian Girls @= 34) Girls of Color (N = 61)

Mean SD Mean SD
ISR 3.72 .80 4.05 .75
Mother level of voice 2.81 .88 2.92 .88
Classmate level of voice" 2.59 .93 3.11 .61
Teacher level of voice‘l' 2.40 .88 2.81 .85
Father level of voice 2.58 1.00 2.60 1.02
Best fiiend level of voice 3.65 .49 3.70 .45

 

* Mean difference p < .01
1' Mean difference p = .08

Finally, in order to confirm that the ethnic differences were not a result of
ethnicity as a proxy for socioeconomic status, a chi-square analysis comparing mother’s
level of education and ethnicity was performed. Results were not significant [x2 (8, N =
87) = 5.13, p = .74], indicating that mother’s education did not differ across ethnicity.
Therefore, there was partial support for Hypothesis 10; girls of color had higher level of
voice with classmates and teachers than Caucasian girls, consistent with the hypothesis.
In addition, girls who were in a school with a greater concentration of ethnic minorities
tended to have higher levels of voice with the ISR scale than did girls at a school with

fewer minority students.

122

DISCUSSION

The purpose of the current study was to empirically examine Gilligan’s (L.M.
Brown & Gilligan, 1992; Gilligan et al., 1990) theory that girls experience loss of voice
in early adolescence, and to test the proposed theoretical model for predictors and
outcomes of level of voice. Specifically, it was hypothesized that attachment with parents
and gender role socialization would predict level of voice, and that level of voice would
predict quality of fiiendships, psychological well-being, quality of romantic relationships,
and dating conflict. Finally, it was hypothesized that approval and emotional social
support would moderate the relationship between level of voice and psychological well-
being, feminine girls would have lower levels of voice than would masculine or
androgynous girls, girls with lower levels of voice would have more emotional abuse in
their relationships and have lower satisfaction in their relationships than girls with higher
levels of voice, and that Caucasian girls would have lower levels of voice than girls of
color.

Overall, the results supported the hypotheses, and the data was a good fit for the
proposed models, suggesting support for a theoretical model of level of voice. In
addition, the results fiom the structural equation models suggest that each of the
constructs in each model fit well in relation to each other; that is, parental attachment and
gender role socialization both fit well as predictors of voice, and fiiendship quality and
psychological well-being and dating conflict and quality of romantic relationship fit
together as outcomes of voice, respectively. The results will be discussed in the
following sections, and then the strengths and limitations of the current study and

suggestions for future research will follow.

123

Predictors of Level of Voice

The current study examined level of voice in two different arenas: with authority
figures and with peers. Research supports the concept of differences in level of voice
across different contexts (e.g., private and public) (Harter et al., 1998b). However, while
3 Harter and colleagues (1998b) constructed their peer and private contexts of voice based
on theory and mean differences, the current study based the analyses of voice in authority
and peer contexts on both theory and the results of an exploratory factor analysis, thus
providing additional empirical support for the existence of different levels of voice in
different contexts.

Attachment

In the current study, attachment with parents was a significant predictor of level of
voice with authority figures. The finding that attachment is predictive of level of voice is
consistent with the theoretical literature. Salzrnan (1990) found, in a small qualitative
study, that girls who had dismissing attachment were less likely to have truly intimate
relationships, suggesting that perhaps they might have lower levels of voice. In addition,
this finding supports the suggestion that parents help construct a child’s autobiography,
and may foster lower levels of voice by helping to create a construction of a false
narrative (Harter et al., 1997a). Finally, this is consistent with research with adults that
found a relationship between aspects of attachment and self-silencing (Rernen et al.,
2002; J.M. Thompson & Hart, 1996).

As previously discussed, researchers have found that insecurely attached female
adolescents were less assertive, had poorer communication styles and poorer conflict

management skills than securely attached adolescents (Kerns, 1996; Kerns et al., 1996;

124

Kerns et al., 2000; Kobak et al., 1993; Schneider & Younger, 1996); these characteristics
are similar to ways in which low levels of voice would be manifested in relationships.

This is the first study to empirically demonstrate a connection between parental
attachment and level of voice with authority figures. This suggests that the adolescent’s
primary attachment relationship with her parents serves to bolster level of voice; secure
parental attachment was positively related to level of voice with authority figures. Thus,
this suggests that the internalization of a secure working model, and the perception of
caregivers as responsive and caring, may help girls resist the relational crisis that they
may experience in early adolescence. The process of an authentic and intimate
relationship with a caregiver may act as a template for intimate and mutual relationships
with other authority figures.

Conversely, research indicates that insecurely attached adolescents, who
experience their attachment figure as rejecting, deactivate from attachment, thereby
minimizing their distress (Cole-Detke & Kobak, 1996). The current study found that
dismissive parental attachment was negatively related to level of voice with authority
figures, suggesting that as a result of deactivating from attachment, adolescents may
lower their voices and thus preserve inauthentic relationships.

However, attachment to parents was not a significant predictor of level of voice
with peers, contrary to the original hypothesis. There are several possible reasons for this
finding. First, as previously discussed, during early adolescence teenagers begin to spend
more time with peers, and spend less time with parents (Buhrmester, 1990).
Developmentally, the attachment relationship with the parent may be less salient at this

point in terms of influencing relationships with peers. In addition, as part of this

125

 

developmental growth, there is a press for autonomy in early adolescence, which also
may help explain the lack of relationship between attachment with parents and level of
voice with peers. Steinberg (2001) noted “that early adolescence is an important period
for the negotiation of autonomy-related changes in the parent-child relationship” (p. 7).
Thus, as adolescents press for autonomy, they may be less influenced by their caregivers
and more influenced by peers. Allen and colleagues (Allen, Porter, Tencer, & Williams,
2003) found that even by age 13, the quality of relationship with a best fiiend may have
striking similarities to the attachment relationship with the parent, and they suggested that
by early adolescence, relationships with peers may be influencing adolescent security
even more than relationships with parents. In addition to the normative developmental
process of spending more time with peers, the context of the actual family may affect the
extent of the influence of family versus peers; research indicates that adolescents in less
cohesive and adaptive families are more influenced by peers than parents (Gauze,
Bukowski, Aquan-Assee, & Sippola, 1996).

The second reason that parental attachment may not have predicted level of voice
with peers is that the impact of the attachment relationship with parents may be specific
to level of voice with similar figures. As previously discussed, there is evidence that
adolescents see each part of their self differently with peers versus parents and teachers
(Harter et al., 19983). Thus, it may be that the attachment relationship is specific to that
system, i.e., attachment with peers might impact level of voice with peers. There is
preliminary support for this in the friendship literature; Way and Chen (2000) found that
girls reported supportive fiiendships even when they did not receive similar support at

home. In addition, authors of another study that categorized types of adolescents’

126

fiiendships based upon different qualities of the friendship (e.g., closeness, intimacy,
conflict resolution, etc.) noted that the resulting typologies were strikingly similar to the
different attachment categories; that is, in one category, adolescents described feeling
close to fiiends and able to resolve conflict, while in another, they were preoccupied with
the friendship (Way, Cowal, Gingold, Pahl, & Bissessar, 2001). Therefore, adolescents’
friendships may have similar aspects of the attachment relationship, and, as previously
discussed with the results of Allen et al.’s (2003) study, may be more salient in respect to
level of voice with peers.

The third reason that parental attachment may not have predicted level of voice
with peers is related to measurement issues. It may be that the BSQ, which was used to
measure current parental attachment in the current study, is measuring the quality of the
current relationship to caregivers, which is not as salient a predictor of level of voice with
peers as it is with parents. Currently, attachment is measured behaviorally until middle
childhood, and then is measured at the level of representation in late adolescence in
adulthood. Cassidy (2003) noted that it is difficult to assess behaviorally current parental
attachment in adolescence, and that aside from self-report measures, there are no current
ways to measure attachment during this time period. An alternative explanation is that
parental attachment that is assessed earlier in childhood would be more predictive of level
of voice with peers than current attachment. This would be consistent with longitudinal
research that indicates that the quality of maternal attachment at infancy is a strong
predictor of social competence in later childhood and early adolescence (Elicker et al.,
1992; Shulman et al., 1994; Weinfield et al., 1997), and consistent with research,

previously discussed, that indicates that in adolescence, attachment to peers may be more

127

salient than attachment to parents. Thus, one explanation is that early attachment to
parents impacts attachment to peers, but that by adolescence, peer attachment is more
salient than parental attachment in terms of influencing level of voice with peers.
Gender Role Socialization

Gender role socialization predicted level of voice with both authority figures and
peers. With authority figures, androgynous girls had higher levels of voice than did
feminine girls, although there was no difference between feminine girls and masculine
girls on level of voice. In addition, both masculinity and femininity were positively
correlated with authority levels of voice. This suggests that with authority figures, the
dual contribution of higher levels of femininity and masculinity contribute to higher
levels of voice, supporting the androgyny model of gender role socialization. In contrast,
with peers, both androgynous and masculine girls had higher levels of voice than did
feminine girls, and only masculinity, but not femininity, was correlated with level of
voice with peers. Thus, for peers, this suggests that higher levels of masculinity drive
higher levels of voice, supporting the masculinity model of gender role socialization.

The findings fi'om the current study substantially contribute to our understanding
of how gender role socialization impacts level of voice. First, the findings support
current research that indicates that gender itself does not impact level of voice, but rather
the extent of gender role socialization impacts level of voice (Harter et al., 1998b).
Second, while other research on gender role socialization has used either continuous or
categorical methods of analyzing the relationship between gender role socialization and
level of voice (e.g., Harter et al., 1998b; Tolman & Porche, 2000), the current study

utilized both methods; this is advantageous in terms of parsing out the independent

128

contributions of masculinity and femininity in that using just the median-split method of
examining gender role socialization may mask the independent overriding contributions
of each (Cate & Sugawara, 1986).

Third, while previous researchers have argued that higher levels of femininity
contribute to lower levels of voice (Harter et al., 1998b; Smolak & Munstertieger, 2002),
the current study found that for peers, higher levels of masculinity contributed to higher
levels of voice. The findings of the current study add to the literature by redefining and
clarifying masculinity as the salient agent that predicts voice with peers. In addition,
these findings depathologize girls’ internalization of femininity. This finding is similar to
research that indicates that the contribution of masculinity is the most salient predictor of
psychological well-being (Boldizar, 1991; Cate & Sugawara, 1986; Lamke, 1982; Wells,
1980).

In addition, with authority figures, it was the contribution of bgth masculinity and
femininity that led to higher levels of voice. This supports the androgyny model of
gender identity (Whitley, 1983). The androgyny model “defines individuals as possessing
both positive masculine traits such as assertiveness and autonomy and positive feminine
traits such as nurturance and communication skills” (Rose & Montemayor, 1994, p. 580).

Therefore, while context for voice does matter in terms of peer versus authority
voice, and it is clear that gender role socialization is salient in terms of level of voice, the
current study did not support Harter ct al.’s (1998b) previous findings that in private
contexts with close fiiends and family, feminine girls and androgynous girls did not differ
in their levels of voice, while in public settings, such as in school, feminine girls had

significantly lower levels of voice than androgynous girls.

129

By examining just the categorical results from the current study, it would appear
that feminine girls have lower levels of voice with peers than androgynous girls, which
would partially confirm Harter’s (1998b) findings. However, given that masculine and
androgynous girls do not differ on level of voice with peers, and that femininity is not
correlated with level of voice with peers, this suggests that the lack of masculinity
contributes to low levels of voice. Interestingly, in her study, Harter (1998b) only
examined level of voice with the feminine and androgynous categories. The masculinity
classification in girls has not always been examined because of the small size of the
groups (e. g., Hatter et al., 1998b; Rose & Montemayor, 1994), and thus the current study
contributed by using all three salient categories of gender role socialization.

Thus, this study does not support Gilligan and colleagues’ theory that if girls have
internalized the societal ideal of the “good woman,” they may be more likely to lose their
voice during adolescence, and feel that they cannot have both close relationships and
honest relationships (L.M. Brown, 1998; L.M. Brown & Gilligan, 1992; Gilligan et al.,
1990). This is also contrary to research with a college population which did find that
level of voice was negatively correlated with femininity (Smolak & Munstertieger, 2002).
It is notable that the participants in the current study are more ethnically diverse than
those in either Harter’s (1998b) or Smolak and Munstertieger’s (2002) studies; only 15%
and 11%, respectively, of their participants were ethnic minorities, while 67% of
participants in the current study were from ethnic minority groups. In addition, their
participants were at least middle-class, which is different from the current study. It may
be that girls from lower socioeconomic backgrounds are less likely to internalize the ideal

of what Gilligan (1990) argues society views as “good women” and are instead socialized

130

to not be passive, and to adopt more masculine traits. This may be because, for girls fiom
low socioeconomic backgrounds, their families are aware that these girls are more at risk,
and less likely to be recognized and acknowledged by society because they are already
disadvantaged by their social class, and thus are socialized to be more outspoken, and to
internalize masculine traits with peers (Way, 1995).

Finally, in the current study, 25% of girls were in the “rmdifferentiated” gender
role socialization group. However, because previous researchers who utilized the CSRI
with adolescents either did not utilize the groups used in the current study (Aube et al.,
2000; Obeidallah, McHale, & Silbereisen, 1996) or did not report their group sizes (e.g.,
Boldizar, 1991), it is difficult to compare the results of the current study to previous
research with the CSRI. However, the means from the current study on the femininity
and masculinity subscales were similar to those from Boldizar’s initial study.

Outcomes of Level of Voice

Level of voice with authority figures and level of voice with peers were each
significant predictors of fiiendship quality. This finding emphasizes the dilemma that
Brown and Gilligan (1992) discuss; while girls may strive to maintain false relationships
for the purpose of maintaining relationships, the quality of these relationships are
compromised by their inauthentic qualities. Thus, higher levels of voice predicted higher
quality of friendship.

More importantly, the results of this study are among the first to empirically
demonstrate that the more that girls can be authentic and resolve conflict, the more
intimate their relationships will be as a result. This supports both theory and research that

indicates that intimacy in friendship is increasingly salient as children enter adolescence

131

(e.g., Buhrmester, 1990; Kerns, 1996), and that the intimacy of friendship involves
mutuality in relationships (Sullivan, 1953).

Level of voice with authority figures and level of voice with peers were also
significant predictors of psychological well-being, comprised of self-esteem and low
depressive symptomatology. This is consistent with previous research that found a
negative relationship between level of voice and depressive symptomatology (Hart &
Thompson, 1996; Harter et al., 1996a; Tolman & Porche, 2000) and level of voice and
self-esteem (Tolman & Porche, 2000). In addition, the finding that level of voice is
predictive of self-esteem supports the theory that mutuality in relationships would lead to
feeling more positive about one’s self (Jack, 1991).

In addition, approval and emotional social support fiom authority figures
moderated the relationship between level of voice with authority figures and depression.
At low levels of voice, girls with high levels of social support had significantly less
depressive symptomatology than did girls with lower levels of support, but at higher
levels of voice, there was no difference in depressive symptomatology for high versus
low approval and emotional support.

As discussed earlier, adolescents cited “approval of others” and “alienation
process from their true self, due to lack of validation from others” as the two most
important reasons for their false self behavior (Harter et al., 1996a); these two reasons
accounted for 90% of false self behavior. Thus, for participants who did not have high
levels of voice, the validation that they received from others in the form of approval and
emotional social support buffered, or protected, them from the depressive

symptomatology experienced by those with lower levels of support. However, at high

132

levels of voice, the approval support did not protect against depressive symptomatology
because the approval support was internalized, in the form of high level of voice.

These findings help clarify the relationship between social support and level of
voice. While one study found that social support and level of voice were correlated
(Harter et al., 1998b), this finding does not add to our understanding of how social
support may influence level of voice. In contrast, the findings from the current study
indicate that approval from others is salient for those with low levels of voice; this
external support helps protect against depressive symptomatology. However, for those
with high levels of voice, or internalized approval, social support does not influence
depressive symptomatology.

Level of voice with authority figures, but not level of voice with peers, was
predictive of quality of romantic relationships, and level of voice was not predictive of
dating conflict. Level of voice with authority figures was a significant predictor of
quality of romantic relationships. It is possible at this early developmental phase in
romantic relationships that authority figures are more influential than peers, especially
with the modeling of intimate relationships. Thus, it may be that the ability to express
one’s self with authority figures, and the mutuality that accompanies higher levels of
voice, influences quality of romantic relationships.

While level of voice with peers was a predictor of quality of fiiendship, the
finding that it was not a predictor of quality of romantic relationships suggests that
perhaps during early adolescence, romantic relationships are not as fully developed and
complex as are fiiendships, and thus the lack of mutuality in relationships with peers

would not impact the quality of romantic relationships, because the same intimacy does

133

not exist with romantic partners that exists with peers. In addition, there is still little
research on adolescent romantic relationships (Steinberg & Morris, 2001), so it is difficult
to parse out what factors influence romantic relationships in early adolescence.

While level of voice with peers did not predict romantic relationships, there was a
trend for girls with high levels of voice with peers to have more romantic relationship
satisfaction than girls with low levels of voice. This suggests that although level of voice
with peers may not predict intimacy and growth in romantic relationships, there is
preliminary support that adolescents with higher levels of voice would report more
satisfaction in their intimate relationships, potentially because their feelings of honesty
and mutuality in their relationship may lead to their conscious satisfaction with
relationships.

There could be several reasons for the lack of relationship between level of voice
and dating conflict. First, the sample size was small for these analyses, as only 69% of
participants had started dating, and second, there was not much variance on several of the
dating conflict subscales, such as relational abuse and threats. That is, there was a
restriction of range because so many participants indicated that they had not experienced
abuse. This suggests that perhaps l4-year-olds are not experiencing types of dating
conflict that older adolescents might experience.

Ethnic Differences in Level of Voice

The current study indicated some ethnic differences in terms of level of voice.
Girls who were Caucasian had lower levels of voice with classmates than did girls of
color, and there was a trend for Caucasian girls to have lower levels of voice with

teachers than girls of color, irrespective of the ethnic composition of the school. This is

134

preliminary evidence that in the school setting, girls of color are less at risk for loss of
voice than Caucasian girls.

One explanation for the results is that all of the participants attended schools at
which students of color were the majority, and that perhaps being a member of the
majority within the school setting was influential on girls’ abilities to speak their mind in
this context. This is consistent with research that shows that ethnic minority students
benefit, in terms of well-being, from being in schools in which they are the majority (e.g.,
Lysne & Levy, 1997). Thus, it would be helpful to compare ethnic minority and
Caucasian adolescent students in schools at which they are the majority and the minority
to see how level of voice may differ in different contexts across school composition.

The results suggest that Caucasian girls have lower levels of voice than girls of
color in the school setting. As previously mentioned, it would be helpful to assess level
of voice for Caucasian girls and girls of color in schools in which they are the majority
and minority, and with different ratios of white to non-white students, in order to
determine if there is a “tipping point” for when the actual level of being a minority or
majority becomes beneficial. It may be that for Caucasian girls, being a minority
negatively impacted their level of voice with teachers and classmates.

The current study also found that with the more generalized scale of voice, the
ISR, girls who were in a school that had a higher percentage of non-white students had
higher levels of voice than girls in a school that had a lower percentage of non-white
students. This is partial support for the hypothesis that the school context itself may be a

protective factor against loss of voice.

135

Although there has been little research on the influence of the ethnic and racial
composition of schools on adolescent identity, Lysne and Levy (1997) built upon Phinney
and Rosenthal’s (1992) suggestion that ethnic identity might vary with the cohesiveness
of the community, and were the first to examine whether adolescents’ ethnic identity
might vary With ethnic composition of the school as a result of identification with the
same ethnic group. They found that Native-American adolescents who were in a school
that was predominantly Native-American had stronger ethnic identity than Native-
Americans in a predominantly Caucasian school regarding their exploration and
commitment of ethnic identity (Lysne & Levy, 1997). Lysne and Levy (1997) suggested
that perhaps the school context could parallel or substitute for an ethnic neighborhood.
Thus, in the current study, it is possible that girls of color who were in the school with a
higher composition of non-white students had higher levels of general peer voice perhaps
as a result of enhanced feelings of group inclusion. In addition, it could be that girls in a
school with a similar racial make-up are more protected against negative external events
such as racism (Lysne & Levy, 1997)

However, there were no ethnic differences with level of voice with parents or best
fiiends. As previously discussed, the girls in the current study are fi'om similar low
socioeconomic backgrounds. While previous qualitative research found that girls from
ethnic minority backgrounds were less at risk for loss of voice (Clark, 1999; Way, 1995),
it may be that this is also the case for girls of all ethnicities from low socioeconomic
backgrounds. For example, Brown (1998) discussed that girls fiom lower socioeconomic
backgrounds may be less at risk for loss of voice than girls from middle-class or upper-

middle class backgrounds, even though she did not discuss why this might occur. In

136

addition, in arguing that developmental models of identity that are appropriate for middle-
class and upper-middle class white girls may not generalize to other girls, Erkut and
colleagues noted that “for girls of color, culturally and linguistically different girls,
working-class girls, and girls living in poverty, gender is not the only site for struggle and
negotiation, nor is it necessarily the most salient site” (Erkut, Fields, Sing, & Marx, 1996,
pp. 56-57).

The comparisons of the mean scores of level of voice with parents in Harter’s
(1998b) study versus the current study indicate that there may be ethnic differences and/or
socioeconomic differences in level of voice in schools and public settings, but not with
parents. The results from the current study with best fiiend voice and parent voice can be
compared to previous studies that also used the Teenage Voice scale (Harter, 1995) that
were conducted with primarily white, middle-class populations, and the ISR scale, that
was conducted with a diverse urban sample (Tolman & Porche, 2000). The means for
best fiiend voice from Harter’s study (Harter et al., l998b) was 3.73, and it was 3.65 in
the current study. Although these are similar, there is also likely a ceiling effect, given
that the highest possible value is 4.00. The mean for parent voice was 2.77 from Harter’s
study (Harter et al., 1998b) and 3.17 from Smolak and Munstertieger’s (2002) study (note
that neither study differentiated mother voice fi'om father voice). In comparison, in the
current study, the mean for mother voice was 2.83 and for father voice was 2.56. The
mean parent voice fiom Smolak and Munstertieger’s (2002) study is slightly larger than
the means from the current study, which could be attributed to the age of their college-age
sample. It is likely that level of voice would increase in college, as girls move out of

relational crises.

137

The mean for the ISR scale was 3.96 in the current study, while it was 4.81 in
Tolman and Porche’s (2000) study. For the current study, when school was a factor, girls
of color at the school that had a larger percentage of ethnic minority students had
significantly higher mean ISR scale score of 4.23 versus 3.73 for the girls of color at the
school with a smaller percentage of ethnic minority students. The pattern of these
findings suggests that, with generalized level of voice with peers, perhaps the setting
itself can serve as a protective factor, as previously discussed. That is, the girls in
Tolman and Porche’s (2000) study, who are in a more urban setting than the girls from
the current study, had higher scores on the ISR scale, and in the current study, when girls
of color attended a school with a higher ethnic minority population, their ISR scores were
significantly higher than girls of color from a school with a lower proportion of ethnic
minority students.

Summary

The results of the current study found that gender role socialization predicted level
of voice, and that parental attachment only significantly predicted level of voice with
authority figures, but not level of voice with peers. Level of voice predicted the quality of
fiiendships and psychological well-being, and approval and emotional social support with
authority figures moderated the relationship between level of voice with authority figures
and depressive symptomatology. Level of voice with authority figures was a significant
predictor of quality of romantic relationships, while it did not predict dating conflict.

The results supported the androgyny model of gender role socialization for level
of voice with authority figures; that is, both masculinity and femininity were salient for

higher levels of voice. In contrast, the results supported the masculinity model for level

138

of voice with peers; higher scores on masculinity contributed to higher levels of voice
with peers, rather than lower scores on femininity contributing to higher levels of voice
with peers. Finally, there were some ethnic differences in the current study. Girls of
color who attended a school with a higher percentage of ethnic minority students had
higher levels of generalized peer voice (ISR), while across schools, girls of color had
higher levels of voice than Caucasian girls with classmates and with teachers. There were
no ethnic differences in level of voice with best fiiends, mothers, or fathers.

In their qualitative work, Brown and Gilligan (1992; 1993) suggested that girls
have two choices: to preserve relationships and be “selfless” and ultimately lose their
voices or to resist the relational impasse that they experience, and be selfish, which they
argue will ultimately create conflict in relationships. This dichotomy, or the choice to be
selfless or selfish, was not corroborated by the current study. Rather, girls with higher
levels of voice had mg; intimate fiiendships. Thus, the results from the current study
suggest that it is possible to both retain a sense of self and have healthy, honest, and
intimate relationships. The dichotomy proposed by Brown and Gilligan has unfortunate
implications for girls and women, and was not supported by the current study. In
contrast, the importance of mutuality as emphasized by Stone Center theorists was
supported (Jordan et al., 1991).

The current study contributes significantly to understanding of level of voice
theory. Gilligan and colleagues (1992; 1990) have argued that girls experience loss of
voice in a certain context, i.e., our society that promotes a “good woman” stereotype.
However, the results of the current study do not support her theories. Rather, the

empirical results from the current study suggest that for level of voice with peers, the dual

139

contribution of masculinity and femininity contribute to fights; levels of voice. This is
consistent with Chodorow’s (1978; 1987) theories about girls’ maintaining their identities
by identification with their mothers, and is also consistent with research that indicates that
masculinity contributes to better mental health. In addition, with level of voice with
peers, 9M masculinity was correlated with level of voice, suggesting that the 112m of
masculinity contributes to loss of voice with peers. Therefore, while the current study
supports Gilligan’s concept of loss of voice as context-specific, the results indicate that
the contexts are different than she proposed, and that the contributions of masculinity and
femininity are more distinct than previously supposed.

Overall, the current study supports a theoretical model for understanding level of
voice that involves gender role socialization, parental attachment, fiiendship, and
psychological well-being. The current study elucidates the mechanisms that contribute to
loss of voice in the context of intimacy and mutuality. There was limited support for
previous theoretical, qualitative, and quantitative work on level of voice as related to
gender role socialization (Gilligan et al., 1990; Harter et al., 1998b; Smolak &
Munstertieger, 2002). The results of the study also indicate the importance of mutual and
intimate relationships, and the resulting impact on fiiendships and psychological well-
being. Finally, the results of the current study suggest that for ethnic minority girls, being
the etlmic majority in school positively influences level of voice.

Strengths and Limitations

The current study has a number of strengths. First, the participants are ethnically

diverse, and from primarily low-socioeconomic backgrounds, and participants of these

ethnic and socioeconomic backgrounds are rarely studied (Way & Pahl, 2001). In

140

addition, the participants in the current study differ from participants in previous research
studies on level of voice which have utilized older, more homogenous samples and, as a
result, make it difficult to generalize from the results (e.g., Harter et al., l998b; Smolak &
Munstertieger, 2002).

Second, the current study utilized methodologically rigorous methods of testing
the model for predictors and outcomes of level of voice. The structural equation
modeling combined with the additional analyses provide strong methodological support
for the results. For example, the use of both methodologies helped elucidate the
individual and dual contributions of masculinity and femininity to level of voice.

Without the additional multivariate analyses, it would not have been clear exactly how
gender role socialization influenced level of voice.

Third, the current study is the first to propose a theoretical model that can be
quantitatively tested to understand predictors and outcomes of level of voice, whereas
previous studies have neglected to thoroughly examine what factors might influence level
of voice (e.g., L.M. Brown, 1998; Gilligan et al., 1990; Harter et al., l998b).

However, as in any study, there are some limitations. First, the BSQ, the
questionnaire designed to assess parental attachment, had some measurement
inconsistencies. The subscale designed to measure preoccupied attachment did not load
with secure and dismissive attachment, and was not a good fit for the data. This
methodological issue suggests that the attachment measure should possibly be interpreted
with caution.

Second, the sample of participants involved in a romantic relationship (n=75) was

small, and limited the power of the structural equation modeling, as well as the variance

141

of the dating conflict scale. Overall, the study would have benefit from a larger sample
size in order to have enough power to test the overall goodness of fit of both the
predictors and outcomes of level of voice in the same structural model. In addition, the
cross-sectional design of the current study makes it difficult to infer causation from the
predictors of level of voice. For example, while in the current study it was predicted and
confirmed that level of voice positively predicted intimacy of friendship, it could also be
that more intimate fiiendships lead to higher levels of voice.

Third, the current study only utilized self-report measures. As previously
discussed, this may have been a factor with assessing attachment to parents. In addition,
girls may have underreported dating conflict, and may have had a positive bias in terms of
reporting on fiiendship quality. Therefore, it would have been helpful to have
observational measures of fiiendship, and alternative measures of assessing attachment.

Directions for Future Research

There are several different ways that future research could expand upon current
knowledge of level of voice. First, it would be ideal to conduct longitudinal studies so as
to examine level of voice over time. If these studies began in early or middle childhood,
they could assess parental attachment with both behavioral (e.g., Strange Situation) and
representational measures of attachment (e. g., doll play). In addition, it would be helpful
to examine level of voice over time with both girls and boys, in order to parse apart
gender differences in level of voice. It would also be helpful to utilize not only self-
report measures of fiiendship quality, but to use observational measures of fiiendship that

could assess resolution of conflict (e.g., Kerns et al., 1996).

142

Second, in the current study, level of voice did not predict dating conflict, and
only level of voice with authority figures predicted quality of romantic relationships. As
previously discussed, there may be some methodological and theoretical issues with
assessing both dating conflict and quality of romantic relationships in 14-year-old girls.
An alternative would be to examine the relationship between level of voice and sexual
desire. Tolman (1991; 1994) has argued that our society tells girls that it is taboo to feel
sexual, or have sexual desires. Tolman notes that girls’ sexuality is only spoken about in
schools in ways that discourage sexual exploration. Sexuality is often discussed in the
negative contexts of victimization, disease, pregnancy, and morality. In addition,
developmental theories about girls have muted girls’ sexual desire as a dynamic in
adolescent development. Even some feminist authors such as Pipher (1994) ignore the
idea of sexuality as a driving force when describing girls’ experiences with sexuality. As
previously discussed, level of voice may have an important relationship to romantic
relationships, but one that may be difficult to assess in this age group. Therefore, future
research on the relationship between level of voice and sexual desire would contribute to
the understanding of how loss of voice may impact romantic relationships.

Third, it would be interesting to examine the relationship between personality and
level of voice. Recently, researchers have begun investigating the F ive-Factor Model
(FFM) of personality in adolescence (e.g., Branje, van Aken, van Lieshout, & Mathijssen,
2003). In the FFM, also knows as the “Big Five,” there are five distinct personality
dimensions: extraversion, agreeableness, conscientiousness, neuroticism, and openness to
experience (Steinberg, 2002). Research suggests that, overall, these personality

characteristics are stable over time (Asendorpf & Van Aken, 2003; Lamb, Chuang,

143

 

Wessels, Broberg, & Hwang, 2002), and, of the Big Five, agreeableness, or how kind and
sympathetic one is, may be the most related to peer outcomes (J ensen-Campbell et al.,
2002; Jensen-Campbell & Graziano, 2001).

Specifically, research indicates that agreeableness was negatively related to
fiiends’ rating of aggressiveness (Asendorpf & Van Aken, 2003), positively associated
with peer acceptance and reciprocity of fiiendship (J ensen-Campbell et al., 2002),
increases over time (Lamb et al., 2002), and of the Big Five, was most closely associated
with processes and outcomes of peer conflict (J ensen-Campbell & Graziano, 2001). In
addition, for girls, agreeableness was associated with better outcomes of conflicts,
including compromise, and teachers rated girls as more agreeable than boys (J ensen-
Campbell & Graziano, 2001). Interestingly, J ensen-Campbell and Graziano (2001)
suggested that the gender differences in their study might be attributed to girls’ different
definitions of conflict, as compared to boys’ definitions, and that agreeableness may be
more normative for girls, in terms of their gender role socialization. Thus, from the
research discussed above, it seems that for girls, the manifestation of agreeableness,
which is associated with motives to maintain positive relationships with others (J ensen-
Campbell et al., 2002), and its connection with compromise in interpersonal conflicts,
may be similar to ways in which low levels of voice would be manifested in relationships.
Thus, future research could examine the FFM of personality in relation to level of voice.

Fourth, when future research examines ethnic differences, it would be beneficial
to go beyond skin color, and examine the intersection between level of voice and ethnic
identity. Phinney (1990) argued that although an ethnic label, or how one identifies one’s

ethnic background, may appear to be simple, it is actually very complex. In addition,

144

Phinney (1996) noted that ethnic labels and categories are problematic in general. She
wrote that “labels themselves are not consistent indicators of group membership; rather,
they vary across time and situations, carry different connotations among individuals and
groups, and gloss over within-group variation. To get beyond simplistic ethnic
categories, we need to examine the meanings associated with ethnicity, specifically
culture, identity, and minority status” (Phinney, 1996, p. 920)

Thus, it would be helpful to examine the relationship between level of voice and
ethnic identity. Ethnic identity, which is multi-dimensional, can be defined as “the
psychological relationship of ethnic and racial minority group members with‘their own
group” (Phinney, 1990, p. 499). Ethnic identity is crucial to the functioning of group
members, and is related to self-esteem and self-concept (Phinney, 1990), and thus a more
developed ethnic identity can be construed as a protective factor. Research indicates that
ethnic identity is positively related to psychological well-being and self-esteem for
adolescents in different ethnic groups (Phinney, Cantu, & Kurtz, 1997; RE. Roberts et
al., 1999). Thus, researching the intersection between level of voice and ethnic identity
over time would allow for further elucidation of the contributing factors to loss of voice,
and how different girls experience level of voice.

The current study assessed attachment to parents and gender role socialization as
predictors of loss of voice, and quality of fiiendship, dating conflict, quality of romantic
relationships, and psychological well-being as outcomes of loss of voice. In addition,
social support was examined as a moderator of voice and psychological well-being. This
study has several strengths, including the theoretical model, rigorous methodology, and

under-studied sample, which, along with the independent and combined contributions of

145

the predictor and outcome constructs, help contribute to our empirical understanding of

level of voice.

146

REFERENCES

Aboud, F.E., & Mendelson, M.M. (1996). Determinants of fiiendship selection and
quality: Developmental perspectives. In W.M. Bukowski, A.F. Newcomb &
W.W. Hartup (Eds.), The company they keep: Friendship in childhood and
adolescence (pp. 87-114). New York: Cambridge University Press.

Adams, G.R., & Marshall, SK. (1996). A developmental social psychology of identity:
Understanding the person in context. Journal of Adolescence, 19, 429-442.

Ainsworth, M.D.S., Blehar, M.C., Waters, E., & Wall, S. (1978). Patterns of attachment.
Hillsdale, NJ: Erlbaum.

Allen, J.P., Moore, C.M., & Kuperminc, GP. (1997). Developmental approaches to
understanding adolescent deviance. In 8.8. Luthar & J.A. Burack (Eds.),
Developmental psychopathology: Perspectives on adjustment, risk, and disorder.
New York: Cambridge University Press.

Allen, J .P., Porter, M., Tencer, H., & Williams, F. (2003, April). Attachment as a
relational construct beyond childhood: Parent and peer relationship qualities
linked to attachment security. Paper presented at the Biennial Meeting of the
Society for Research in Child Development, Tampa, FL.

Armsden, G.C., & Greenberg, MT (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.

Armsden, G.C., McCauley, E., Greenberg, M.T., Burke, P.M., & Mitchell, JR. (1990).
Parent and peer attachment in early adolescent depression. Journal of A bnormal
Child Psychology, 18, 683-697.

Armstrong, J.G., & Roth, D.M. (1989). Attachment and separation difficulties in eating
disorders: A preliminary investigation. International Journal of Eating Disorders,
8, 141-155.

Asendorpf, J.B., & Van Aken, M.A.G. (2003). Validity of big five personality
judgements in childhood: A 9 year longitudinal study. European Journal of
Personality, I 7, 1-17.

Aube, J ., Fichman, L., Saltaris, C., & Koestner, R. (2000). Gender differences in
adolescent depressive symptomatology: Towards an integrated social-
developmental model. Journal of Social and Clinical Psychology, 19, 297-313.

Baron, R.M., & Kenny, DA. (1986). The moderator-mediator variable distinction in
social psychological research: Conceptual, strategic, and statistical considerations.
Journal of Personality and Social Psychology, 51 , 1 173-1182.

147

Bemdt, T.J. (1996). Exploring the effects of friendship quality on social development. In
W.M. Bukowski, A.F. Newcomb & W.W. Hartup (Eds.), The company they keep:
Friendship in childhood and adolescence (pp. 346-365). New York: Cambridge
University Press.

Bemdt, T.J., & Keefe, K. (1995). Friends' influence on adolescents' adjustment to school.
Child Development, 66, 13 12-1329.

Blain, M.D., Thompson, J.M., & Whiffen, V.E. (1993). Attachment and perceived social
support in late adolescence: The interaction between working models of self and
others. Journal of Adolescent Research, 8, 226-241.

Blatt, S.J., & Maroudas, C. (1992). Convergences among psychoanalytic and cognitive-
behavioral theories of depression. Psychoanalytic Psychology, 9, 157-190.

Boldizar, JP. (1991). Assessing sex typing and androgyny in children: The Children's Sex
Role Inventory. Developmental Psychology, 27, 505-515.

Bollen, K. (1989). Structural equation modeling with latent constructs. New York:
Wiley.

Bowlby, J. (1969/1982). Attachment and loss, Vol. I : Attachment (2nd ed.). New York:
Basic Books.

Bowlby, J. (1980). Attachment and loss, Vol. 3: Loss, separation and depression. New
York: Basic Books.

Branje, S.J.T., van Aken, M.A.G., van Lieshout, C.F.M., & Mathijssen, J .J .J .P. (2003).
Personality judgments in adolescents' families: The perceiver, the target, their
relationship, and the family. Journal of Personality, 71, 49-81.

Brown, B.B., F eiring, C., & Furman, W. (1999). Missing the love boat: Why researchers
have shied away from adolescent romance. In W. Furman, B.B. Brown & C.
F eiring (Eds.), The development of romantic relationships in adolescence (pp. 1-
16). Cambridge, England: Cambridge University Press.

Brown, L.M. (1998). Raising their voices: The politics of girls' anger. Cambridge, MA:
Harvard University Press.

Brown, L.M., & Gilligan, C. (1992). Meeting at the crossroads: Women's psychology and
girls' development. Cambridge, MA: Harvard University Press.

Brown, L.M., & Gilligan, C. (1993). Meeting at the crossroads: Women's psychology and
girls' development. Feminism & Psychology, 3, 11-35.

148

Brown, L.S., & Wright, J. (2001). Attachment theory in adolescence and its relevance to
developmental psychopathology. Clinical Psychology and Psychotherapy, 8, 15-
32.

Buhrrnester, D. (1990). Intimacy of fiiendship, interpersonal competence, and adjustment
during preadolescence and adolescence. Child Development, 61, 1101-1 1 l l.

Bukowski, W.M., Hoza, B., & Boivin, M. (1993). Popularity, fiiendship, and emotional
adjustment during early adolescence. In B. Laursen (Ed.), Close friendships in
adolescence (V 01. 60, pp. 23-39). San Francisco: Jossey-Bass Publishers.

Bukowski, W.M., Hoza, B., & Boivin, M. (1994). Measuring friendship quality during
pre- and early adolescence: The development and psychometric properties of the
Friendship Qualities Scale. Journal of Social and Personal Relationships, 11,
485-488.

Bukowski, W.M., Newcomb, A.F., & Hartup, W.W. (1996). Friendship and its
significance in childhood and adolescence: Introduction and comment. In W.M.
Bukowski, A.F. Newcomb & W.W. Hartup (Eds.), The company they keep:
Friendship in childhood and adolescence (pp. 1-15). New York: Cambridge
University Press.

Campbell, J.C., & Lewandowski, LA. (1997). Mental and physical health effects of
intimate partner violence on women and children. Psychiatric Clinics of North
America, 20, 353-374.

Carr, J .G., Gilroy, F.D., & Sherman, M.F. (1996). Silencing the self and depression
among women. Psychology of Women Quarterly, 20, 375-392.

Cassidy, J ., Ziv, Y., Rodenberg, M., & Woodhouse, S. (2003, April). Adolescent
perceptions of parents: Associations with adolescent (AAI) attachment and
interaction with parents. Paper presented at the Biennial Meeting of the Society
for Research in Child Development, Tampa, FL.

Cate, R., & Sugawara, A.l. (1986). Sex role orientation and dimensions of self-esteem
among middle adolescents. Sex Roles, 15, 145-158.

Cauce, A.M. (1986). Social networks and social competence: Exploring the effects of
early adolescent fiiendship. American Journal of Community Psychology, 14,
607-628.

Cauce, A.M., Mason, C., Gonzales, N., & Hiraga, Y. (1996). Social support during
adolescence: Methodological and theoretical considerations. In K. Hurrelman &
S.F. Hamilton (Eds.), Social problems and social contexts in adolescence:
Perspectives across boundaries (pp. 131-151). New York: Aldine de Gruytcr.

149

Chodorow, N. (1978). The reproduction of mothering: Psychoanalysis and the study of
gender. Berkeley, CA: University of California Press.

Chodorow, N. (1987). Feminism and difference: Gender, relation, and difference in
psychoanalytic perspective. In M.R. Walsh (Ed.), The psychology of women:
Ongoing debates (pp. 246-264). New Haven, CT: Yale University Press.

Clark, M. (1999). Adding the voices of Jewish girls to developmental theory. Women &
Therapy, 22, 55-68.

Cohen, S., & Wills, T.A. (1985). Stress, social support, and the buffering hypothesis.
Psychological Bulletin, 98, 310-357.

Cole-Detke, H., & Kobak, R. (1996). Attachment processes in eating disorder and
depression. Journal of Consulting and Clinical Psychology, 64, 282-290.

Collins, N.L., & Read, 8.]. (1990). Adult attachment, working models, and relationship
quality in dating couples. Journal of Personality and Social Psychology, 58, 644-
663.

Collins, W.A., Hennighausen, K.C., Schrnit, D.T., & Sroufe, L.A. (1997). Developmental
precursors of romantic relationships: A longitudinal analysis. In S. Shuhnan &
W.A. Collins (Eds.), Romantic relationships in adolescence: Developmental
perspectives (Vol. 78, pp. 69-84). New York: J ossey-Bass Publishers.

Collins, W.A., & Repinski, DJ. (1994). Relationships during adolescence: Continuity
and change in interpersonal perspective. In R. Montemayor, G.R. Adams & T.P.
Gullotta (Eds.), Personal relationships during adolescence (pp. 7-36). Thousand
Oaks, CA: Sage Publications.

Collins, W.A., & Sroufe, L.A. (1999). Capacity for intimate relationships: A
developmental construction. In W. Furman, B.B. Brown & C. Feiring (Eds.), The
development of romantic relationships in adolescence (pp. 125-147). Cambridge,
England: Cambridge University Press.

Connolly, J .A., & Johnson, A.M. (1996). Adolescents' romantic relationships and the
structure and quality of their close interpersonal ties. Personal Relationships, 3,
1 85-1 95.

Cotterell, J .L. (1992). The relation of attachment and supports to adolescent well-being
and school adjustment. Journal of Adolescent Research, 7, 28-42.

Crockett, L.J. (1988). Pubertal Development Scale: Pubertal Categories. Unpublished
manuscript.

Donnelly, M., & Wilson, R. (1994). The dimensions of depression in early adolescence.
Personality and Individual Diflkrences, I 7, 425-430.

150

Dom, L.D., Crockett, L.J., & Petersen, AC. (1988). The relations of pubertal status to
intrapersonal changes in young adolescents. Journal of Early Adolescence, 8, 405-
419.

Doyle, A.B., & Markiewicz, D. (1996). Parents' interpersonal relationships and children's
fiiendships. In W.M. Bukowski, A.F. Newcomb & W.W. Hartup (Eds.), The
company they keep: Friendship in childhood and adolescence (pp. 115-136). New
York: Cambridge University Press.

Duarte, L.M., & Thompson, J .M. (1999). Sex differences in self-silencing. Psychological
Reports, 85, 145-161.

Elicker, J ., Englund, M., & Sroufe, L.A. (1992). Predicting peer competence and peer
relationships in childhood from early parent-child relationships. In R.D. Parke &
G.W. Ladd (Eds.), Family-peer relationships: Modes of linkage (pp. 77-106).
Hillsdale, N.J.: Erlbaum.

Ellington, J.B., & Marshall, LL (1997). Gender role perceptions of women in abusive
relationships. Sex Roles, 36, 349-369.

Erikson, EH. (1963). Childhood and Society. New York: W.W. Norton.
Erikson, EH. (1968). Identity: Youth and Crisis. New York: W.W. Norton.

Erkut, 8., Fields, J .P., Sing, R., & Marx, F. (1996). Diversity in girls' experiences: Feeling
good about who you are. In B.J.R. Leadbeater & N. Way (Eds.), Urban girls:
Resisting stereotypes, creating identities (pp. 53-64). New York: New York
University Press.

F eeney, J .A. (1999). Adult romantic attachment and couple relationships. In J. Cassidy &
P.R. Shaver (Eds.), Handbook of attachment: Theory, research, and clinical
applications (pp. 355-377). New York: The Guilford Press.

F eiring, C. (1996). Concept of romance in lS-year-old adolescents. Journal of Research
on Adolescence, 6, 181-200.

F eiring, C., & Furman, W.C. (2000). When love is just a four-letter word: Victimization
and romantic relationships in adolescence. Child Maltreatment, 5, 293-298.

Flanagan, A.S., & Furman, W.C. (2000). Sexual victimization and perceptions of close
relationships in adolescence. Child Maltreatment, 5, 350-359.

Flannery, R.B., & Wieman, D. (1989). Social support, life stress, and psychological
distress: An empirical assessment. Journal of Clinical Psychology, 45, 867-872.

Foshee, V.A. (1996). Gender differences in adolescent dating abuse prevalence, types and
injuries. Health Education Research, 11, 275-296.

151

Frank, S.J., Avery, C.B., & Larnan, MS. (1988). Young adults' perceptions of their
relationships with their parents: Individual differences in connectedness,
competence, and emotional autonomy. Developmental Psychology, 24, 729-737.

Frank, S.J., Pirsch, L.A., & Wright, V.C. (1990). Late adolescents' perceptions of their
relationships with their parents: Relationships among deidealization, autonomy,
relatedness, and insecurity and implications for adolescent adjustment and ego
identity status. Journal of Youth and Adolescence, 19, 571-588.

Franz, C.E., & White, KM. (1985). Individuation and attachment in personality
development: Extending Erikson's theory. Journal of Personality, 53, 224-256.

Furman, W. (1993). Theory is not a four-letter word: Needed directions in the study of
adolescent fiiendship. In B. Laursen (Ed.), Close friendships in adolescence (V 01.
60, pp. 89-103). San Francisco: Jossey-Bass Publishers.

F urman, W. (1996). The measurement of friendship perceptions: Conceptual and
methodological issues. In W.M. Bukowski, A.F. Newcomb & W.W. Hartup
(Eds.), The company they keep: Friendship in childhood and adolescence (pp. 41-
62). New York: Cambridge University Press.

Furman, W. (1999). Friends and lovers: The role of peer relationships in adolescent
romantic relationships. In W.A. Collins & B. Laursen (Eds.), Relationships as
developmental contexts: The 30th Minnesota Symposium on Child Development
(pp. 133-154). Hillsdale, NJ: Lawrence Erlbaum Associates.

F urman, W., & Buhrmester, D. (1992). Age and sex differences in perceptions of
networks of personal relationships. Child Development, 63, 103-115.

Furman, W., & Shaffer, L. (in press). The role of romantic relationships in adolescent
development. In P. Florsheim (Ed.), Adolescent romantic relations and sexual
behavior: Theory, research, and practical implications. Mahwah, NJ: Lawrence
Erlbaum Associations.

F urrnan, W., & Simon, V.A. (1999). Cognitive representations of adolescent romantic
relationships. In W. Furman, B.B. Brown & C. Feiring (Eds.), The development of
romantic relationships in adolescence (pp. 75-98). Cambridge, England:
Cambridge University Press.

Furman, W., Simon, V.A., Shaffer, L., & Bouchey, H.A. (2002). Adolescents' working
models and styles for relationships with parents, fiiends, and romantic partners.
Child Development, 73, 241-255.

Furman, W., & Wehner, EA. (1994). Romantic views: Toward a theory of adolescent
relationships. In R. Montemayor, G.R. Adams & T.P. Gullotta (Eds.), Personal
relationships during adolescence (pp. 168-195). Thousand Oaks, CA: Sage
Publications.

152

Furman, W., & Wehner, EA. (1997). Adolescent romantic relationships: A
developmental perspective. In S. Shulman & W.A. Collins (Eds.), Romantic
relationships in adolescence: Developmental perspectives (V 01. 78, pp. 21-36).
New York: J ossey-Bass Publishers.

Galarnbos, N.L., Almeida, D.M., & Petersen, AC. (1990). Masculinity, femininity, and
sex role attitudes in early adolescence: Exploring gender intensification. Child
Development, 61 , 1905-1914.

Gauze, C., Bukowski, W.M., Aquan-Assee, J., & Sippola, L.K. (1996). Interactions
between family environment and fiiendship and associations with self-perceived
well-being during adolescence. Child Development, 67, 2201-2216.

George, C., Kaplan, N., & Main, M. (1985). An adult attachment interview. Unpublished
manuscript, University of California at Berkeley.

Gilligan, C. (1982). In a difierent voice: Psychological theory and women's development.
Cambridge, MA: Harvard University Press.

Gilligan, C. (1990). Teaching Shakespeare's sister: Notes fiom the underground of female
adolescence. In N.P. Lyons & T.J. Hammer (Eds.), Making connections: The
relational worlds of adolescent girls at Emma Willard School (pp. 6-29).
Cambridge, MA: Harvard University Press.

Gilligan, C., Lyons, N.P., & Hammer, T.J. (1990). Making connections: The relational
worlds of adolescent girls at Emma Willard School. Cambridge, MA: Harvard
University Press.

Gratch, L.V., Bassett, M.E., & Attra, S.L. (1995). The relationship of gender and ethnicity
to self-silencing and depression among college students. Psychology of Women
Quarterly, 19, 509-515.

Grazyk, P.A., & Henry, DB. (2001). A developmental perspective on the qualitative
aspects of adolescent best fiiendships, biennial meeting of the Society for
Research in Child Development. Minneapolis, MN.

Greenberg, M.T., Siege], J .M., & Leitch, C.J. (1983). The nature and importance of
attachment relationships to parents and peers during adolescence. Journal of
Youth and Adolescence, 12, 373-386.

Grotevant, H.D., & Cooper, CR. (1985). Patterns of interaction in family relationships
and the development of identity exploration in adolescence. Child Development,
56, 415-428.

Hart, B.I., & Thompson, J .M. (1996). Gender role characteristics and depressive
symptomatology among adolescents. Journal of Early Adolescence, I 6, 407-426.

153

Harter, S. (1985a). The Self-Perception Profile for Children: Revision of the Perceived
Competence Scale for Children, Manual. Denver, CO: University of Denver.

Harter, S. (1985b). The Social Support Scale for Children, Manual. Denver, CO:
University of Denver.

Harter, S. (1995). Teenage voice. Unpublished manuscript, University of Denver.

Harter, S. (1997). The personal self in social context: Barriers to authenticity. In R.D.
Ashmore & L.J. J ussirn (Eds.), Self and identity: Fundamental issues (V 01. 1).
New York: Oxford University Press.

Harter, S., Bresnick, S., Bouchey, H.A., & Whitesell, N.R. (1997a). The development of
multiple role-related selves during adolescence. Development and
Psychopathology, 9, 835-853.

Harter, S., & Marold, DB. (1994). The directionality of the link between self-esteem and
affect: Beyond causal modeling. In D. Cicchetti & S.L. Toth (Eds.), Disorders and
Dysfunctions of the Self (V 01. 5, pp. 335-3 69). Rochester, NY: University of
Rochester Press.

Harter, S., Marold, D.B., & Whitesell, NR. (1992). Model of psychosocial risk factors
leading to suicidal ideation in young adolescents. Development and
Psychopathology, 4, 167-188.

Harter, S., Marold, D.B., Whitesell, N.R., & Cobbs, G. (19963). A model of the effects of
perceived parent and peer support on adolescent false self behavior. Child
Development, 6 7, 360-374.

Harter, S., & Robinson, N. (1988). The social support scale for older children and
adolescents. Unpublished manuscript, University of Denver.

Harter, S., Stocker, C., & Robinson, N.S. (1996b). The perceived directionality of the link
between approval and self-worth: The liabilities of a looking glass self-orientation
among young adolescents. Journal of Research on Adolescence, 6, 285-3 08.

Harter, 8., Waters, P., & Whitesell, N.R. (l998a). Relational self-worth: Differences in
perceived worth as a person across interpersonal contexts among adolescents.
Child Development, 69, 756-766.

Harter, 8., Waters, P.L., Pettitt, L.M., Whitesell, N., Kofkin, J ., & Jordan, J. (1997b).
Autonomy and connectedness as dimensions of relationship styles in men and
women. Journal of Social and Personal Relationships, 14, 147-164.

Harter, 8., Waters, P.L., & Whitesell, N.R. (1997c). Lack of voice as a manifestation of
false self-behavior among adolescents: The school setting as a stage upon which
the drama of authenticity is enacted. Educational Psychologist, 32, 153-173.

154

Harter, 8., Waters, P.L., Whitesell, N.R., & Kastelic, D. (l998b). Level of voice among
female and male high school students: Relational context, support, and gender
orientation. Developmental Psychology, 34, 892-901.

Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment process.
Journal of Personality and Social Psychology, 52, 51 1-524.

Hill, J .P., & Holmbeck, G.N. (1986). Attachment and autonomy during adolescence.
Annals of Child Development, 3, 145-189.

Hill, J .P., & Lynch, ME. (1983). The intensification of gender-related role expectations
during early adolescence. In J. Brooks-Gunn & A.C. Petersen (Eds.), Girls at
puberty: Biological and psychological perspectives (pp. 201 -228). New York:
Plenum Press.

Hoffman, R.M., & Borders, L.D. (2001). Twenty-five years after the Bern Sex-Role
Inventory: A reassessment and new issues regarding classification variability.
Measurement and Evaluation in Counseling and Development, 34, 39-55.

Inderbitzen-Pisaruk, H., & Foster, S.L. (1990). Adolescent fiiendships and peer
acceptance: Implications for social skills training. Clinical Psychology Review,
10, 425-439.

Jack, DC. (1991). Silencing the self: Women and depression. Cambridge, MA: Harvard
University Press.

Jack, D.C., & Dill, D. (1992). The Silencing the Self Scale: Schemas of intimacy
associated with depression in women. Psychology of Women Quarterly, 1 6, 97-
106.

J ensen-Campbell, L.A., Adams, R., Perry, D.G., Workman, K.A., Furdella, J .Q., & Egan,
SK. (2002). Agreeableness, extraversion, and peer relations in early adolescence:
Winning fiiends and deflecting aggression. Journal of Research in Personality,
36, 224-251.

J ensen-Campbell, L.A., & Graziano, W.G. (2001). Agreeableness as a moderator of
interpersonal conflict. Journal of Personality, 69, 323-362.

J ezl, D.R., Molidor, C.E., & Wright, TL. (1996). Physical, sexual and psychological
abuse in high school dating relationships: Prevalence rates and self-esteem issues.
Child and Adolescent Social Work Journal, 13, 69-87.

Jones, D.J., & Costin, SE. (1995). Friendship quality during preadolescence and
adolescence: The contributions of relationship orientations, instrumentality, and
expressivity. Merrill-Palmer Quarterly, 41, 517-535.

155

Jordan, J .V. (1997). A relational perspective for understanding women's development. In
J.V. Jordan (Ed.), Women's growth in diversity: More writings from the Stone
Center (pp. 9-24). New York: The Guilford Press.

Jordan, J.V., Kaplan, A.G., Miller, J .B., Stiver, I.P., & Surrey, J .L. (Eds). (1991).
Women's growth in connection: Writings from the Stone Center. New York:
Guilford Press.

Kenny, M.E., Lomax, R., Brabeck, M., & Fife, J. (1998). Longitudinal pathways linking
adolescent reports of maternal and paternal attachments to psychological well-
being. Journal of Early Adolescence, 18, 221-243.

Kenny, M.E., Moilanen, D.L., Lomax, R., & Brabeck, M.M. (1993). Contributions of
parental attachments to view of self and depressive symptoms among early
adolescents. Journal of Early Adolescence, 13, 408-430.

Kerns, K.A. (1996). Individual differences in friendship quality: Links to child-mother
attachment. In W.M. Bukowski, A.F. Newcomb & W.W. Hartup (Eds.), The
company they keep: Friendship in childhood and adolescence (pp. 137-157). New
York: Cambridge University Press.

Kerns, K.A., Klepac, L., & Cole, A. (1996). Peer relationships and preadolescents'
perceptions of security in the child-mother relationship. Developmental
Psychology, 32, 457-466.

Kerns, K.A., Tomich, P.L., Aspelmeier, J.B., & Contreras, J .M. (2000). Attachment-
based assessments of parent-child relationships in middle childhood.
Developmental Psychology, 36, 614-626.

Kessler, R.C., & McLeod, J .D. (1985). Social support and mental health in community
samples. In S. Cohen & S.L. Syme (Eds.), Social support and health (pp. 219-
240). New York: Academic Press.

Kobak, R.R., Cole, H.E., Ferenz-Gillies, R., Fleming, W.S., & Gamble, W. (1993).
Attachment and emotion regulation during mother-teen problem solving: A
control theory analysis. Child Development, 64, 231-245.

Kobak, R.R., & Ferenz-Gillies, R. (1995). Emotion regulation and depressive symptoms
during adolescence: A functionalist perspective. Development and
Psychopathology, 7, 183-192.

Kobak, R.R., & Sceery, A. (1988). Attachment in late adolescence: Working models,
affect regulation, and representations of self and others. Child Development, 59,
1 3 5-146.

156

Kobak, R.R., Sudler, N., & Gamble, W. (1991). Attachment and depressive symptoms
during adolescence: A developmental pathways analysis. Development and
Psychopathology, 3, 461-474.

Koutrelakos, J ., Baranchik, A.J., & Darnato, N. (1999). Cultural and sex differences in
rating the self and a hypothetical well-adjusted person on Jack's divided self and
care as self-sacrifice subscales. Psychological Reports, 84, 67-83.

Kovacs, M. (1992). Children '3 Depression Inventory (CD1) Manual. North Tonawanda,
NY: Multi-Health Systems.

Lamb, M.E., Chuang, S.S., Wessels, H., Broberg, A.G., & Hwang, CR (2002).
Emergence and construct validation of the Big Five factors in early childhood: A
longitudinal analysis of their ontogeny in Sweden. Child Development, 73, 1517-
1 524.

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

Lamke, L.K. (1982). The impact of sex-role orientation on self-esteem in early
adolescence. Child Development, 53, 1530-1535.

Larson, R.W., Richards, M.H., Moneta, G., Holmbeck, G., & Duckett, E. (1996).
Changes in adolescents' daily interactions with their families from ages 10-18:
Disengagement and transformation. Developmental Psychology, 32, 744-754.

Levendosky, A.A., Huth-Bocks, A., & Semel, M.A. (2002). Adolescent peer relationships
and mental health functioning in families with domestic violence. Journal of
Clinical Child and Adolescent Psychology, 31, 206-218.

Levesque, R.J.R. (1993). The romantic experience of adolescents in satisfying love
relationships. Journal of Youth and Adolescence, 22, 219-251.

Lieberman, M., Doyle, A., & Markiewicz, D. (1999). Developmental patterns in security
of attachment to mother and father in late childhood and early adolescence:
Associations with peer relations. Child Development, 70, 202-213.

Lintunen, T., Leskinen, E., Oinonen, M., Salinto, M., & Rahkila, P. (1995). Change,
reliability, and stability in self-perceptions in early adolescence: A four-year
follow-up study. International Journal of Behavioral Development, 18, 351-364.

Lysne, M., & Levy, GD. (1997). Differences in ethnic identity in Native American
adolescents as a function of school context. Journal of Adolescent Research, 12,
372-3 88.

Markstrom-Adams, C. (1989). Androgyny and its relation to adolescent psychosocial
well-being: A review of the literature. Sex Roles, 21, 325-340.

157

Mendelson, M.J., & Aboud, RE. (1999). Measuring fiiendship quality in late adolescents
and young adults: McGill Friendship Questionnaires. Canadian Journal of
Behavioural Science, 31, 130-132.

Miller, J .B. (1991). The development of women's sense of self. In J .V. Jordan, A.G.
Kaplan, J .B. Miller, I.P. Stiver & J .L. Surrey (Eds.), Women's growth in
connection: Writings from the Stone Center. (pp. 11-26). New York: Guilford
Press.

Miller, J.B., Jordan, J .V., Kaplan, A.G., Stiver, I.P., & Surrey, J .L. (1997). Some
misconceptions and reconceptions of a relational approach. In J .V. Jordan (Ed.),
Women's growth in diversity: More writings fi'om the Stone Center (pp. 25-49).
New York: The Guilford Press.

Molidor, CE. (1995). Gender differences of psychological abuse in high school dating
relationships. Child and Adolescent Social Work Journal, 12, 1 19-134.

Molidor, C.E., & Tohnan, R. (1998). Gender and contextual factors in adolescent dating
violence. Violence Against Women, 4, 180-194.

Moran, P.B., & Eckenrode, J. (1991). Gender differences in the costs and benefits of peer
relationships during adolescence. Journal of Adolescent Research, 6, 396-409.

Nolen-Hoeksema, S. (1994). An interactive model for the emergence of gender
differences in depression in adolescence. Journal of Research on Adolescence, 4,
519-534.

Nolen-Hoeksema, S., & Girgus, J .S. (1994). The emergence of gender differences in
depression during adolescence. Psychological Bulletin, 115, 424-443.

Obeidallah, D.A., McHale, S.M., & Silbereisen, R.K. (1996). Gender role socialization
and adolescents' reports of depression: Why some girls and not others? Journal of
Youth and Adolescence, 25, 775-785.

Owens, G., Cromwell, J., Pan, H., Treboux, D., O'Connor, E., & Waters, E. (1996). The
prototype hypothesis and origins of attachment working models: Adult
relationships with parents and romantic relationships. In E. Waters, B. Vaughn, G.
Posada & K. Kondo-Ikemura (Eds.), New growing points of attachment.
Monographs of the Society for Research in Child Development (V 01. 60, pp. 216-
233)

Papini, D.R., & Roggrnan, L.A. (1992). Adolescent perceived attachment to parents in
relation to competence, depression, and anxiety: A longitudinal study. Journal of
Early Adolescence, 12, 420-440.

158

Parker, J .G., & Asher, SR. (1993). Friendship and fiiendship quality in middle
childhood: Links with peer group acceptance and feelings of loneliness and social
dissatisfaction. Developmental Psychology, 29, 611-621.

Petersen, A.C. (1988). Adolescent development. Annual Review of Psychology, 39, 583-
607.

Petersen, A.C., Crockett, L., Richards, M., & Boxer, A. (1988). A self-report measure of
pubertal status: Reliability, validity, and initial norms. Journal of Youth and
Adolescence, 1 7, 1 17-133.

Petersen, A.C., Leffert, N., & Graham, BL. (1995). Adolescent development and the
emergence of sexuality. Suicide and Life Threatening Behavior, 25, 4-17.

Petersen, A.C., Sarigiani, P.A., & Kennedy, RE. (1991). Adolescent depression: Why
more girls? Journal of Youth and Adolescence, 20, 247-271.

Phinney, J.S. (1990). Ethnic identity in adolescents and adults: Review of research.
Psychological Bulletin, 108, 499-514.

Phinney, J .S. (1996). When we talk about American ethnic groups, what do we mean?
American Psychologist, 51, 918-927.

Phinney, J.S., Cantu, C.L., & Kurtz, DA. (1997). Ethnic and American identity as
predictors of self-esteem among Afiican American, Latino, and White
adolescents. Journal of Youth and Adolescence, 26, 165-185.

Phinney, J .S., & Rosenthal, DA. (1992). Ethnic identity in adolescence: Process, context,
and outcome. In G.R. Adams & T.P. Gullotta (Eds.), Adolescent identity
formation. Advances in adolescent development, Vol. 4 (pp. 145-172). Thousand
Oaks, CA: Sage Publications, Inc.

Pipher, M. (1994). Reviving Ophelia: Saving the selves of adolescent girls. New York:
Ballantine Books.

Price, E.L., & Byers, ES. (1999). The attitudes toward dating violence scales:
Development and initial validation. Journal of Family Violence, 14, 351-375.

Procidano, M., & Heller, K. (1983). Measure of perceived social support from friends and
family. American Journal of Community Psychology, 11, l-24.

Remen, A.L., Charnbless, D.L., & Rodebaugh, TL. (2002). Gender differences in the
construct validity of the Silencing the Self Scale. Psychology of Women
Quarterly, 26, 151-159.

159

Renouf, A.G., & Harter, S. (1990). Low self-worth and anger as components of the
depressive experience in young adolescents. Development and Psychopathology,
2, 293-310.

Rice, K.G., Cole, D.A., & Lapsley, D.K. (1990). Separation-individuation, family
cohesion, and adjustment to college: Measurement validation and test of a
theoretical model. Journal of Counseling Psychology, 3 7, 195-202.

Roberts, L.R., Sarigiani, P.A., Petersen, A.C., & Newman, J.L. (1990). Gender
differences in the relationship between achievement and self-image during early
adolescence. Journal of Early Adolescence, 10, 159-175.

Roberts, R.E., Phinney, J.S., Masse, L.C., Chen, Y.R., Roberts, C.R., & Romero, A.
(1999). The structure of ethnic identity of young adolescents from diverse
ethnocultural groups. Journal of Early Adolescence, 19, 301-322.

Robinson, NS. (1995). Evaluating the nature of perceived support and its relation to
perceived self-worth in adolescents. Journal of Research on Adolescence, 5, 253-
280.

Rose, A.J., & Montemayor, R. (1994). The relationship between gender role orientation

and perceived self-competency in male and female adolescents. Sex Roles, 31,
579-595.

Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton
University Press.

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

Salzman, J .P. (1990). Save the world, save myself: Responses to problematic attachment.
In C. Gilligan, N.P. Lyons & T.J. Hammer (Eds.), Making connections: The
relational worlds of adolescent girls at Emma Willard School (pp. 110-146).
Cambridge, MA: Harvard University Press.

Salzrnan, JP. (1996). Primary attachment in female adolescents: Association with
depression, self-esteem, and maternal identification. Psychiatry: Interpersonal
and Biological Processes, 59, 20-33.

Salzrnan, J .P. (1997). Ambivalent attachment in female adolescents: Association with
affective instability and eating disorders. International Journal of Eating
Disorders, 21, 251-259.

Schneider, B.H., & Younger, AJ. (1996). Adolescent-parent attachment and adolescents'
relations with their peers: A closer look. Youth and Society, 28, 95-108.

160

Schumacker, R.E., & Lomax, R.G. (1992). A beginner’s guide to structural equation
modeling. Mahwah, NJ: Lawrence Erlbaum Associates.

Shuhnan, S., Elicker, J., & Sroufe, L.A. (1994). Stages of fiiendship growth in
preadolescence as related to attachment history. Journal of Social and Personal
Relationships, 1 I, 341-361.

Shulman, S., & Knafo, D. (1997). Balancing closeness and individuality in adolescent
close relationships. International Journal of Behavioral Development, 21, 687-
702.

Shulman, S., & Scharf, M. (2000). Adolescent romantic behaviors and perceptions: Age-
and gender-related differences, and links with family and peer relationships.
Journal of Research on Adolescence, 10, 99-1 18.

Simmons, R.G., & Rosenberg, F. (1975). Sex, sex roles, and self-image. Journal of Youth
and Adolescence, 4, 229-258.

Smith Slep, A.M., Cascardi, M., Avery-Leaf, S., & O'Leary, K.D. (2001). Two new
measures of attitudes about the acceptability of teen dating aggression.
Psychological Assessment, 13, 306-318.

Smolak, L., & Munstertieger, BR (2002). The relationship of gender and voice to
depression and eating disorders. Psycholog of Women Quarterly, 26, 234-241.

Sommers, OH. (2000, May). The war against boys. Atlantic Monthly, 285, 59-74.

Spangler, S. (1999). Ophelia speaks: Adolescent girls write about their search for self.
New York: HarperCollins.

Spence, J .T., Helrnreich, R., & Stapp, J. (1975). Ratings of self and peers on sex role
attributes and their relation to self-esteem and conceptions of masculinity and
femininity. Journal of Personality and Social Psychology, 32, 29-39.

Spratt, C.L., Sherman, M.F., & Gilroy, F .D. (1998). Silencing the self and sex as
predictors of achievement motivation. Psychological Reports, 82, 259-263.

Sroufe, L.A., Carlson, E., & Shulman, S. (1993). Individuals in relationships:
Development from infancy through adolescence. In DC. Funder, R. Parke, C.
Tomlinson-Keasey & K. Widarnan (Eds.), Studying lives through time:
Personality and development (pp. 315-342). Washington, DC: American
Psychological Association.

Sroufe, L.A., & Fleeson, J. (1986). Attachment and the construction of relationships. In
W.W. Hartup & Z. Rubin (Eds.), Relationships and development (pp. 51-67).
Hillsdale, NJ: Lawrence Erlbaum Associates.

161

Sroufe, L.A., & Rutter, M. (1985). The domain of child psychopathology. Child
Development, 55, 17-29.

Steinberg, L. (2001). We know some things: Parent-adolescent relationships in retrospect
and prospect. Journal of Research on Adolescence, 11, 1-19.

Steinberg, L. (2002). Adolescence (6th ed.). New York: McGraw-Hill.

Steinberg, L., & Morris, AS. (2001). Adolescent development. Annual Review of
Psychology, 52, 83-110.

Steinberg, L., & Silverberg, SB. (1986). The vicissitudes of autonomy in early
adolescence. Child Development, 5 7, 841-851.

Steiner-Adair, C. (1990). The body politic: Normal female adolescent development and
the development of eating disorders. In C. Gilligan, N.P. Lyons & T.J. Hammer
(Eds.), Making connections: The relational worlds of adolescent girls at Emma
Willard School (pp. 162-182). Cambridge, MA: Harvard University Press.

Straus, M.A. (1979). Measuring intrafarnily conflict and violence: The Conflict Tactics
(CT) Scales. Journal of Marriage and the Family, 41 , 75-88.

Sullivan, HS. (1953). The interpersonal theory of psychiatry. New York: Norton.

Surrey, J .L. (1991). The self-in-relation: A theory of women's development. In J.V.
Jordan, A.G. Kaplan, J .B. Miller, I.P. Stiver & J .L. Surrey (Eds.), Women's
growth in connection: Writings from the Stone Center. (pp. 51-66). New York:
Guilford Press.

Taylor, J .M., Gilligan, G., & Sullivan, A.M. (1995). Between voice and silence: Women
and girls, race and relationships. Cambridge, MA: Harvard University Press.

Thompson, J .M., & Hart, BL (1996, August). Attachment dimensions and patterns
associated with silencing the self. Poster presented at the APA Annual
Convention, Toronto, Ontario.

Thompson, R., & Zuroff, DC. (1999). Development of self-criticism in adolescent girls:
Roles of maternal dissatisfaction, maternal coldness, and insecure attachment.
Journal of Youth and Adolescence, 28, 197-210.

Tohnan, D.L. (1991). Adolescent girls, women and sexuality: Discerning dilemmas of
desire. Women and Therapy, 11, 55-69.

Tolman, D.L. (1994). Doing desire: Adolescent girls' struggles for/with sexuality. Gender
and Society, 8, 324-342.

162

Tolman, D.L., & Porche, M.V. (2000). The adolescent femininity ideology scale:
Development and validation of a new measure for girls. Psychology of Women
Quarterly, 24, 365-376.

Turner, P.J. (1991). Relations between attachment, gender, and behavior with peers in
preschool. Child Development, 62, 1475-1488.

Unger, R., & Crawford, M. (1992). Women and gender: A feminist psychology. New
York: McGraw-Hill, Inc.

Way, N. (1995). "Can't you see the courage, the strength that I have?": Listening to urban
adolescent girls speak about their relationships. Psychology of Women Quarterly,
19, 107-128.

Way, N., & Chen, L. (2000). Close and general fiiendships among African American,
Latino, and Asian American adolescents from low-income families. Journal of
Adolescent Research, 15, 274-301.

Way, N., Cowal, K., Gingold, R., Pahl, K., & Bissessar, N. (2001). Friendship patterns
among Afiican American, Asian American, and Latino adolescents from low-
income families. Journal of Social and Personal Relationships, 18, 29-53.

Way, N., & Pahl, K. (2001). Individual and contextual predictors of perceived fiiendship
quality among ethnic minority, low-income adolescents. Journal of Research on
Adolescence, 11, 325-349.

Wehner, E., & Furman, W. (2000). The Behavioral Systems Questionnaire. Unpublished
manuscript, University of Denver.

Weinfield, N.S., Ogawa, J .R., & Sroufe, L.A. (1997). Early attachment as a pathway to
adolescent peer competence. Journal of Research on Adolescence, 7, 241-265.

Wekerle, C., & Wolfe, DA. (1998). The role of child maltreatment and attachment style
in adolescent relationship violence. Development and Psychopathology, 10, 571-
586.

Wells, K. (1980). Gender-role identity and psychological adjustment in adolescence.
Journal of Youth and Adolescence, 9, 59-73.

Wenz-Gross, M., Siperstein, G.N., Untch, A.S., & Widaman, K.F. (1997). Stress, social
support, and adjustment of adolescents in middle school. Journal of Early
Adolescence, 1 7, 129-151.

Whitley, BE. (1983). Sex role orientation and self-esteem: A critical meta-analytic
review. Journal of Personality and Social Psychology, 44, 765-778.

163

Wichstrom, L. (1999). The emergence of gender difference in depressed mood during
adolescence: The role of intensified gender socialization. Developmental
Psychology, 35, 232-245.

Wolfe, D.A., & Feiring, C. (2000). Dating violence through the lens of adolescent
romantic relationships. Child Maltreatment, 5, 360-363.

Wolfe, D.A., Scott, K., Reitzel Jaffe, D., Wekerle, C., Grasley, C., & Straatrnan, AL.
(2001). Development and validation of the Conflict in Adolescent Dating
Relationships Inventory. Psychological Assessment, 13, 277-293.

164

APPENDICES

Appendix A: Parent Consent Form

165

February 25, 2002

Dear Parents,

During the next week, your child’s classroom will participate in a study being conducted
by researchers at Michigan State University. The principal and classroom teacher are in
agreement with the study. Only children whose parents sign this permission form will participate.
Staff (students) from MSU will visit the school for one day. Children will not leave the school
building and will make up any missed work. The study will involve filling out some
questionnaires; the goal of the study is to frnd out how empowerment, or the way that children
feel about themselves, affects children’s relationships, like with friends and anyone they may be
dating. Your child will be asked about her/his relationships with family and friends, and how
s/he has been feeling lately. Before your child participates, s/he will be asked to volunteer to
participate; if she declines, then s/he will not be required to participate. Your child may decline
or stop participation at any time without consequence. We will not be seeking any school records
on your child.

After your child completes the questions, which takes approximately 45 minutes, s/he will
receive a small gift, like a coupon to McDonald’s. In addition, each school will receive $100 for
school supplies. Aside fiom these gestures of appreciation by MSU, your child will not directly
benefit from being in the study. Attached is a copy of this consent form, which is yours to keep
for your personal records. Also, after your child completes the questionnaires, s/he will be asked
if s/he would like to be contacted in the future to talk to someone about how s/he feels about
herself/himself and her/his friends; this is completely optional, and does not affect her/his
participation in anything else. If your child does choose to become involved in this separate
study, they would be paid $10, and you will sign a separate form so that they can take part in the
separate study.

All information and responses your child gives will be confidential. Your identity and your
child’s identity as a participant in the study will be known only to the staff involved in this study.
No one will be identifiable in any reports written about this study. Information will be reported
only about groups rather than individuals. Your child’s privacy will be protected to the
maximum extent allowable by law.

If at any time now or later you have questions or concerns about the study, you can speak with
Sally Theran, M.A., (517) 432-1447 of the MSU Psychology Department, 129 Psychology
Research Bldg, East Lansing MI, 48824-1117. If you have questions about your child’s rights as
a participant in this research study you may contact Dr. Ashir Kumar of the Michigan State
University Committee for Research Involving Human Subjects (517) 355-2180. Ms. Theran is
the principal investigator and is responsible for the study.

I give permission for the child named below to participate in the study
I do not give permission for the child named below to participate in the study

Child’s first and last name (please print)
Parent first and last name (please print)
Parent Signature Date

 

 

PLEASE RETURN THIS FORM TO YOUR CHILD’S TEACHER

166

Appendix B: Child Consent Form

167

February 25, 2002

Dear teenager,

This study will involve filling out some questionnaires. The goal of this study is to find
out how the ways that you feel as a teenager about yourself may affect your relationships
with people in your life, such as fiiends and anyone you may be dating. You will be
asked about your relationships with family and fiiends, and how you have been feeling
lately. You are volunteering to participate. If you decline, then you will not be required
to participate. You may decline or stop participation at any time without consequence.
We will not be seeking any school records on you.

After you complete the questions, which takes approximately 45 minutes, you will
receive a small gift, like a coupon to McDonald’s. In addition, each school will receive
$100 for school supplies. Aside from these gestures of appreciation by MSU, you will
not directly benefit from being in the study. Attached is a copy of this consent form,
which is yours to keep for your personal records. Also, after you complete the
questionnaires, you will be asked if you would like to be contacted in the future to talk to
someone about how you feel about yourself and your fiiends; this is completely optional,
and does not affect your participation today.

All information and responses you give will be confidential. Your identity as a participant
in the study will be known only to the staff involved in this study. No one will be
identifiable in any reports written about this study. Information will be reported only
about groups rather than individuals. Your privacy will be protected to the maximum
extent allowable by law.

If at any time now or later you have questions or concerns about the study, you can speak
with Sally Theran, M.A., (517) 432-1447 of the MSU Psychology Department, 129
Psychology Research Bldg, East Lansing MI, 48824—1117. If you have questions about
your rights as a participant in this research study you may contact Dr. Ashir Kumar of the
Michigan State University Committee for Research Involving Human Subjects (517) 355-
2180. Ms. Theran is the principal investigator and is responsible for the study.

Your first and last name (please print)

 

Your Signature Date

168

Appendix C: Demographic Form

169

1. What grade are you in?

 

2. What is your birthdate? / /
month day year

3. What is your racial or ethnic background?

a) Black/Afiican-American

b) Caucasian/White

c) Latino/Latina

d) Asian

e) Native-American

f) Biracial (specify: )

g) Other (specify: )

 

4. What is your religious affiliation (for example, Baptist, Protestant, Catholic)

 

5. What is your gender (circle one)? Male Female

6. What is the highest level of education that your mother completed? (Circle one)
1 = grades 1, 2, 3, 4, 5, or 6 (circle specific grade)
2 = grades 7, 8, 9, 10, ll, 12, GED (circle specific grade)

 

 

 

 

 

3 = some college Where?
4 = AA degree Where?
5 = BA/BS Where?
6 = some grad school Where?
7 = graduate degree Where?

MA?

Ph.D.?

Law?

MD?

 

8 = other; Specify (e.g., Beauty School, nursing school)

 

7. What is the highest level of education that your father completed?
1 = grades 1, 2, 3, 4, 5, or 6 (circle specific grade)
2 = grades 7, 8, 9, 10, ll, 12, GED (circle specific grade)

 

 

 

 

 

3 = some college Where?
4 = AA degree Where?
5 = BA/BS Where?
6 = some grad school Where?
7 = graduate degree Where?

MA?

Ph.D.?

Law?

MD?

 

8 = other; Specify (e.g., Beauty School, nursing school)

 

170

8. Does your mother work? YES / NO
If YES, what does she do, and where does she work?

 

 

 

 

9. Does your father work? YES / NO
If YES, what does he do, and where does he work?

 

 

 

 

10. What is your grade point average?

1 1. What was your grade in English for the most recent marking term?
What was your grade in Social Studies for the most recent marking term?
What was your grade in Science for the most recent marking term?

What was your grade in Math for the most recent marking term?

171

Appendix D: Pubertal Development Scale

172

1. How tall are you, without shoes?
2. How much do you weigh, without shoes and heavy clothes?

To answer each question, please circle the number in fi'ont of the answer that best
describes what is happening to you. Please choose only gng for each question.

3. Would you say that your growth in height:

1. has not yet begun to spurt (spurt means more growth than usual)
2. has barely started

3. is definitely underway

4. seems completed

4. And how about the growth of body hair (“body hair” means underarm and pubic
hair)? Would you say that your body hair has:

not yet started growing
has barely started growing
is definitely underway
seems completed

99°19?

5. Have you noticed any skin changes, especially pimples?

not yet started showing changes
have barely started showing changes
skin changes are definitely underway
skin changes seem completed

99’1”?"

6. Have your breasts begun to grow?
1. not yet started growing
2. have barely started growing
3. breast growth is definitely underway
4. breast growth seems completed
To answer each question, fill in the blanks with the best answer you can give

7. Have you begun to menstruate?

1. No
2. Yes

If you answered “Yes”, how old were you when you first menstruated?
Age: I was year and months old when I began to menstruate.

173

Appendix E: Behavioral Systems Questionnaire

174

MY PARENTS
For this questionnaire we are interested in how you TYPICALLY feel and act in your
relationships with your 1parents. By parents, we mean all the people you consider to be
parental figures; these gures may include natural, adopted, or stepparents--whomever
you consi or to be parental figures. Of course, your answers may be more influenced by
the arent or parents that is/are more important to you. Some of these questions ma not

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

app y to all of your arental figures, but consider ow they TYPICAL Y apply. P case
use the following s e.
l 2 3 4 5
Strongly Disagree Neither Disagree Agree Strongly
Disagree Nor Agree Agree
1. “MY PARENTS” act as if I count on them too much. 1 2 3 4 5
2. I consistently turn to "MY PARENTS" when upset or 1 2 3 4 5
worried.
3. I am afraid that I turn to "MY PARENTS" more often 1 2 3 4 5
than they want me to.
4. I seek out "MY PARENTS" when something bad 1 2 3 4 5
happens.
5. I am r_r(_)t the kind of person who quickly turns to "MY 1 2 3 4 5
PARENTS" in times of need.
6. I do Egyfien ask "MY PARENTS" to comfort me. 1 2 3 4 5
7. I feel that “MY PARENTS” believe that I depend on 1 2 3 4 5
them too often
8.. I rely on "MY PARENTS" when I'm having troubles. l 2 3 4 5
9. I worry that "MY PARENTS" think I need to be 1 2 3 4 5
comforted too much
10. I rarely feel like I need help fiom "MY PARENTS." l 2 3 4 5
11. I rarely turn to "MY PARENTS" when upset. I 2 3 4 5
12. I seek out "MY PARENTS" for comfort and support. 1 2 3 4 5
13. It is easy for me to turn to "MY PARENTS" when I l 2 3 4 5
have a problem.
14. I do r_r_o_t like to turn to "MY PARENTS" when I'm 1 2 3 4 5
bothered about something.
15. I am afraid that "MY PARENTS" think I am too 1 2 3 4 5
dependent.

 

 

 

 

 

 

 

 

175

The followin statements refer. to carin for your parents. Again, we are interested in
what rs typrc of you. Please crrcle o y one response for each statement.

 

l 2 3 4 5
Strongly Disagree Neither Disagree Agree Strongly
Disagree Nor Agree Agree
1. I would rather "MY PARENTS" work out their problems 1 2 3 4 5
by themselves.

 

2. I am M comfortable dealing with "MY PARENTS" 1 2 3 4 5
when they are worried or bothered about a problem.
3. I enjoy being able to take care of "MY PARENTS." 5
4. I often help "MY PARENTS" more than they need or 1 2 3 4 5
want.
5. I do n_o_t like having to comfort or reassure "MY 1 2 3 4 5
PARENTS."
6. I find it easy to be understanding of "MY PARENTS" 1 2 3 4 5
and their needs.

 

N
b)
A

 

 

 

 

 

 

7. I get too wrapped up in my "MY PARENTS’" worries. l 2 3 4 5

8. I feel comfortable with "MY PARENTS" coming to me 1 2 3 4 5
for help.

9. I do 291 like "MY PARENTS" to depend on me for help. 1 2 3 4 5

 

10. I create difficulties by taking on "MY PARENTS'" l 2 3 4 5
problems as if they were mine.
11. I am comfortable with the responsibilities of caring for l 2 3 4 5

 

 

 

"MY PARENTS."

12. It is relatively easy to respond to "MY PARENTS'" l 2 3 4 5
needs.

13. Iwant "MY PARENTS" to be independent and go_t l 2 3 4 5
need me.

 

N
DJ
.5
M

14. I get over-involved in "MY PARENTS" problems.
15. Sometimes I try to comfort "MY PARENTS" more than 1 2 3
the situation calls for.

 

A
M

 

 

 

 

 

 

 

 

176

The following statements refer to other feelings in relationships with your parents.
Agrlrlrn, we are mterested in what is typical of you. Please circle only one response for
eac statement.

1. I contribute more to making our relationship work than 1 2 3 4 5
"MY PARENTS" do.
2. Both "MY PARENTS" and I make frequent efforts to see 1 2 3 4 5

 

 

 

 

or talk with each other.

3. Spending time together is more important to me than to 1 2 3 4 5
"MY PARENTS."

4. Truthfirlly, my relationships with "MY PARENTS" are l 2 3 4 5
just not that important to me.

 

5. I do g9_t want to put much energy into my relationship 1 2 3 4 5
with "MY PARENTS."
6. "MY PARENTS" and I jointly make the important 1 2 3 4 5
decisions in our relationship.
7. Iwant to do more things with "MY PARENTS" than they 1 2 3 4 5
want to.
8. I do r_19_t put much effort into trying to have good 1 2 3 4 5
relationships with "MY PARENTS."
9. "MY PARENTS" and I both contribute a lot to our 1 2 3 4 5
relationship.
10. Our relationship is valued by both "MY PARENTS" and 1 2 3 4 5
me.
11. I find that "MY PARENTS" are reluctant to get as close 1 2 3 4 5
as I would like.
12. I am _r_1_9_t that invested in my relationships with "MY 1 2 3 4 5
PARENTS."
13. I want to be closer to "MY PARENTS" than they want 1 2 3 4 5
to be with me.
14. lam ggt that interested in making my relationships with l 2 3 4 5
"MY PARENTS" the best they could be.
15. "MY PARENTS" and I really try to understand each 1 2 3 4 5
others' points of view.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

In this questionnaire we asked you to talk about your relationships with different parents.
Different people may have been thinking about different parental figures. You may have
thought mostly of one figure or several figures.

I was mostly thinking about: (check all that apply):

 

D A natural/adopted mother El A stepmother C] Other (fill in)
CI A natural/adopted father CI A stepfather D Other (fill in)

 

 

 

 

 

 

177

Appendix F: Children’s Sex Role Inventory

I78

Please read each statement and write in the ONE number that tells how true the statement
is for you.

 

Not at all true of me A little true of me Mostly true of me Very true of me

 

 

 

 

 

 

l 2 3 4

 

I think I’m better than most of the other people I know.

I control a lot of the kids in my class

I care about what happens to others

I have many fiiends.

When a decision has to be made, it’s easy for me to take a stand

When someone’s feelings have been hurt, I try to make them feel better.
It’s easy for me to fit into new places.

I am a leader among my fiiends.

PWSQE‘PP’N?‘

I am a warm person.

fl
9

I’m always losing things.

—s
H

. When I play games, I really like to win.

fl
1"

I am a kind and caring person.

p—s
DJ

. I like to do things that other people do.

y—s
:3;

I am sure of my abilities.
. I like babies and small children a lot.

I am a moody person.

I—iI-‘h—O
\rgxu.

. I stand up for what I believe in.

H
00

. I am a gentle person.

H
1°

I like acting in fiont of other people.

N
.9

I am good at sports.

N
.—s

. I am a cheerful person.

N
N

. I never know what I’m going to do from one moment to the next.

N
U)

It’s easy for me to tell people what I think, even when I know they will
probably disagree with me.

N
A

. When I like someone, I do nice things for them to show them how I feel.

N
U!

. I always do what I say I will do.

N
O\

. I make a strong impression on most people I meet.

N
\l

. I like to do things that girls and women do.

N
00

. I feel bad when other people have something that I don’t have.

N
\O

. I am good at taking charge of things.

lllllll lllllllllllllllllllllll

t»
O

. It makes me feel bad when someone else is feeling had.

179

Appendix G: Inauthentic Self in Relationships Scale

180

Please read each statement and write in the ONE number that tells how much you agree
with each statement.

 

 

 

 

Strongly Moderately Somewhat Somewhat Moderately Strongly
Disagree Disagree Disagree Agree Agree Agree
1 2 3 4 5 6

 

 

 

 

 

 

. I would tell a fiiend she looks nice, even if I think she shouldn’t go out of the

house dressed like that.

I express my opinions only if I can think of a nice way of doing it.

I worry that I make others feel bad if I am successful.

I would not change the way I do things in order to please someone else.

I tell my fiiends what I honestly think even when it is an unpopular idea.

Often I look happy on the outside in order to please others, even if I don’t feel
happy on the inside.

I wish I could say what I feel more often than I do.

I feel like it’s my fault when I have disagreements with my friends.

. When my fiiends ignore my feelings, I think that my feelings weren’t very

important anyway.

10. I usually tell my fiiends when they hurt my feelings.

181

Appendix H: Teenage Voice

182

Read each statement all the way across. Each statement describes two kinds of teenagers,
one on the left, and one on the right. Eir_st decide which kind of teenager YOU are most
Egg, the one on the left, or the one on the right. Pick one. Then for each kind of
teenager, check whether that description is really true for you or just sort of true for you.
So for each numbered item you will be checking only one box. Sometimes it will be on
the left, sometimes on the right. DO NOT check both sides. Just put a check on the side

that is most like you.

SAMPLE SENTENCE
Really Sort of Sort of Really
True True True True
For me for me for me for me
1. El E1 Some teenagers like to Other teenagers would E] [:1
go to the movies in BUT rather go to sports events.

their spare time

SAYING WHAT I THINK AROUND MY MOTHER

Really Sort of Sort of Really
True True True True
For me for me for me for me
1. [:1 [:1 Some teenagers share Other teenagers find it Cl 1:]
what they are really BUT hard to share what they are
thinking with their mother. thinking with their mother.
2. Cl [:1 Some teenagers usually Other teenagers do say Cl C]
don’t share what’s on BUT what’s on their mind.
their mind around their mother. around their mother
3. Cl C] Some teenagers are able to Other teenagers have Cl 13
express their opinions to BUT trouble expressing their
their mother. opinions to their mother.
4. E] [:1 Some teenagers are able to Other teenagers are not able E] El
let their mothers know BUT to let their mother know what’s
what’s important to them. important to them.
5. El E1 Some teenagers have a hard Other teenagers can [:1 El
time expressing their point BUT express their point
of view to their mother. of view to their mother.

183

SAYING WHAT I THINK AROUND MY CLASSMATES

Really Sort of
True True
for me for me

1. El E1 Some teenagers share
what they are really BUT
thinking with their classmates

Some teenagers usually
don’t share what’s on
classmates.

BUT

Some teenagers are able to
express their opinions to
their classmates.

BUT

Some teenagers are able to
let their classmates know BUT
what’s important to them.

Some teenagers have a hard
time expressing their point BUT
of view to their classmates.

Sort of Really
True
for me for me

True

SAYING WHAT I THINK AROUND MY TEACHERS

Really Sort of
True True
For me for me
1. El E1 Some teenagers share what
they are really thinking

with their teachers.

BUT

Some teenagers usually don’t
share what’s on their mind BUT
around their teachers.

Some teenagers are able to

express their opinions to BUT
their teachers.

4. Cl 1:] Some teenagers are able to
let their teachers know BUT

what’s important to them.

Some teenagers have a hard
time expressing their point BUT

184

Other teenagers find it Cl C]
hard to share what they are
thinking with their classmates
Other teenagers do say Cl Cl
what’s on their mind around their
their mind around their classmates.
Other teenagers have Cl [:1
trouble expressing
their opinions to their classmates.
Other teenagers are not E] El
able to let their classmates
know what’s important to them.
Other teenagers can Cl 13
express their point of view
to their classmates.

Sort of Really

True True

for me for me

Other teenagers find it hard C1
to share what they are thinking
with their teachers.

Other teenagers do say El
what’s on their mind
around their teachers.

Other teenagers have [:1
trouble expressing
their opinions to their teachers.

Other teenagers are not [3
able to let their teachers know
what’s important to them.

Other teenagers can express CI
their point of view to their teachers.
of view to their teachers.

El

SAYING WHAT I THINK AROUND MY FATHER

Really Sort of Sort of Really
True True True True
For me for me for me for me
1. El Cl Some teenagers share Other teenagers find it El E1
what they are really BUT hard to share what they
thinking with their father. are really thinking with their father.
2. El E1 Some teenagers usually Other teenagers do say C] 13
don’t share what’s on their BUT what’s on their mind
mind around their father. around their father.
3. El E1 Some teenagers are able to Other teenagers have Cl C]
express their opinions to BUT trouble expressing their opinions
their father. to their father.
4. Cl 13 Some teenagers are able to Other teenagers are not able E] El
let their fathers know BUT to let their father know what’s
what’s important to them. important to them.
5. E] [:1 Some teenagers have a Other teenagers can express [:1 Cl

hard time expressing their BUT their point of view to their father.
point of view to their father.

SAYING WHAT I THINK AROUND MY BEST FRIEND

Really Sort of Sort of Really
True True True True
For me for me for me for me
1. E] El Some teenagers share Other teenagers find it hard El Cl
what they are really BUT to share what they are thinking
thinking with their best fiiend. with their best friend.
2. Cl C] Some teenagers usually Other teenagers do say Cl 13

don’t share what’s on BUT what’s on their mind
their mind around their best fiiend. around their best fiiend.

3. El Cl Some teenagers are able to Other teenagers have trouble El E1
express their opinions to BUT expressing their opinions
their best fiiend. to their best friend.

4. El El Some teenagers are able to Other teenagers are not able Cl Cl
let their best fiiends know BUT to let their best fiiend know what’s
what’s important to them. important to them.

5. El D Some teenagers have 3 Other teenagers can express El El
hard time expressing their BUT their point of view to their best friend.
point of view to their best fiiend.

185

   

1"" mm ‘Wt’drz‘m; ..

a... ?-_-

.‘k

axilla'g'" (is: gran-gun} . 1".

7v..

. , {a q . . . a . .r u c o
_ . e... a... oases...» E sees...35?? ”neither.

. . F .. ..4..:. its. 0. . a. . r . a 4 .. o L . .« a. . . . .1. 5.1. a . V1. .. n. I .. T. l
...a....y....:x”.._.ra_.. big... -5 : . ,.. .. .._......,.,.,....:...._.;._m.iMitzi”:... IT....,e.h$m.w..,§.r.a_1.1}... .TH...:.3...:E?. _

a u .

 

ronnarre

riendship Qualities Quest
1 86

F

Appendix I

Put the initials of your very best fiiend here

 

We want to ask some questions just about you and the person you think of as your best
friend so we can know what your best fiiend is like. We have some sentences that we
would like you to read. Please tell us whether this sentence describes your fiiendship or
not. Some of the sentences might be really true for your fiiendship while other sentences
might be not very true for your fiiendship. We simply want you to read the sentence and
tell us how true the sentence is for your fiiendship. Remember, there are no right or
wrong ways to answer these questions, and you can use any of the numbers on the scale.

At the top of each page there is a scale that goes from 1 to 5.

" 1" means the sentence is probably not true for your friendship,
"2" means that it might be true,

"3" means that it is usually true,

"4" means that it is very true,

"5" means that it is really true for your friendship.

Write in the number on the scale that is best for you.
Be sure to read carefully and answer as honestly as possible.

 

Probably Not True Might be True Usually True Very true Really True

 

 

 

 

 

 

 

l 2 3 4 5

 

1. Even if my fiiend and I have an argument we would still be able to be fiiends
with each other.
2. I can trust and rely upon my fiiend.
3. If my fiiend had to move away I would miss her/him.
4. When I do a good job at something my friend is happy for me.
5. There is nothing that would stop my fiiend and I from being fiiends.
6. Sometimes my fiiend does things for me or makes me feel special.
7. When my friend and I have an argument, s/he can hurt my feelings.
8. When I have not been with my fiiend for a while I really miss being with
him/her.
9. I can get into fights with my fiiend.
10. If I have a problem at school or at home I can talk to my fiiend about it.
11. My friend can bug me or annoy me even though I ask her/him not to.
12. If I said I was sorry after I had a fight with my fiiend s/he would still stay mad
at me.
13. Even if other persons stopped liking me, my fiiend would still be my friend.
14. I know that I am important to my fiiend.
15. If there is something bothering me I can tell my fiiend about it even if it is
something I cannot tell to other people.
16. My fiiend puts our friendship ahead of other things.

187

 

Probably Not True Might be True Usually True Verytrue Really True

 

 

 

 

 

 

 

1 2 3 4 5

 

17. If my friend or I do something that bothers the other one of us we can make up
easily.

18. My friend and I can argue a lot.

19. My fiiend and I disagree about many things.

20. If my friend and I have a fight or argument we can say "I'm sorry" and
everything will be alright.

21. I feel happy when I am with my fiiend.

22. I think about my fiiend even when my friend is not around.

188

Appendix J: Conflict in Adolescent Dating Relationships

189

Please check the statement that best applies to you.
D I have not yet begun dating. (Skip to page 30)
1:1 I have begun dating and/or had a boyfriend/girlfriend.
Please check all the boxes below that describe the kinds of dating relationships you are
currently experiencing and those you have experienced in the past:
13 going out in male/ female groups
El dating different people
El dating one person without any definite commitment
El dating one person exclusively
El engaged

If you have ever been in a dating relationship or been going out with someone, please
answer the following questions: ‘
At what age did you start going out/ having boyfi'iends/girlfi'iends?
How many boyfiiends/girlfi'iends have you had (not including childhood crushes)?
How many boyfiiends/girlfriends did you have/have you had in:

 

 

 

Grade 7 [# of weeks] [# of weeks]
Number of boy/ girlfriends Longest relationship Shortest relationship
Grade 8 [# of weeks] [# of weeks]
Number of boy/girlfiiends Longest relationship Shortest relationship

 

The next few pages ask you to answer questions thinking about your current or recent ex-
boyfriend/girlfriend. Please check which person you will be thinking of when you answer
these questions:

[:1 I am thinking of somebody that is my boyfriend/girlfriend right now. (Go to A)

[J I am drinking of a recent ex-boyfriend/girlfiiend (within the past year). (Go to B)
:A) If this is your current boyfriend/girlfriend:
How long have you been dating/ going out?
EHow often do you see each other? Circle the best response.
:1? very day at school Every day at school 2-3 times per week Once per week or less

 

 

, and every day out of school

How much time do you spend alone together? hours per day OR hours per week
EHow old is s/he?

:How important is this relationship to you? (Circle one of the responses below).

:Not very important Somewhat important Important Very important
{99.1.0.tltenettpese ............................................................

B) If this is your ex boyfriend/girlfriend:

How long did you go out together?

How often did you see each other? Circle the best response below.

Every day at school Every day at school 2-3 times per week Once per week or less
and every day out of school

How much time did you spend alone together? hours per day OR hours per week

What kinds of things did you do together?

When did you stop going together/seeing each other?

How old was s/he?

How important was this relationship to you? (Circle one of the responses below).

Not very important Somewhat important Important Very important

 

 

 

 

190

The following questions ask you about things that may have happened to you with your
boyfiiend while you were having an argument. Check the box that is your best estimate of how
often these things have happened with your current or ex-boyfiiend/girlfiiend in the past year.
Please remember that all answers are confidential. As a guide use the following scale:

During a conflict or argument with my boyfriend/girlfriend in the past year:

 

Never: this has never happened in your relationship

Seldom: this has happened only 1-2 times in your relationship
Sometimes: this has happened about 3-5 times in your relationship
Often: this has happened 6 times or more in your relationship

 

 

 

Never Seldom Sometimes Often

5. I told him/her that I was partly to blame.
S/He told me that he was partly to blame.

I. I gave reasons for my side of the argument. E] E] 13 El
S/He gave reasons for his/her side of the argument. El Cl E] El
2. I tried to turn his/her friends against him/her. El El El E1
S/He tried to turn my fiiends against me. [:1 Cl El El
3. I did something to make him/her feel jealous. El [:1 El E1
S/He did something to make me feel jealous. El El E] El
4. I destroyed or threatened to destroy something

s/he valued. El E] El El

S/He destroyed or threatened to destroy something
I valued. El El Cl E]
El 13 El El
Cl El E1 El

During a conflict or argument with my boyfriend/girlfriend in the past year:
6. I brought up something bad that s/he had done

in the past. E] El Cl C]

S/He brought up something bad that I had done
in the past. El El El E1
7. I threw something at him/her. El E1 [3 Cl
S/I-Ie threw something at me. El E1 El Cl
8. I said things just to make him/her angry. 13 Cl El El
S/He said things just to make me angry. El El E1 El
9. I gave reasons why I thought s/he was wrong. El El El E1
S/He gave reasons why s/he thought I was wrong. El [:1 El E1
10. I agreed that s/he was partly right. El Cl [3 El
S/He agreed that I was partly right. El El E1 D

During a conflict or argument with my boyfriend/girlfriend in the past year:

11. I spoke to him/her in a hostile or mean tone of voice. El
S/He spoke to me in a hostile or mean tone of voice. El

12. I offered a solution that I thought would make us

both happy. El

S/He offered a solution that he thought would make us

both happy. E1

ElElElCl
DUDE!

DUDE]

191

 

 

Never: this has never happened in your relationship

Seldom: this has happened only 1-2 times in your relationship
Sometimes: this has happened about 3-5 times in your relationship
Often: this has happened 6 times or more in your relationship

 

 

During a conflict or argument with my boyfriend/girlfriend in the past year:
Never Seldom Sometimes Often

13. I put off talking until we calmed down.
S/He put off talking until we calmed down.
14. I insulted him/her with put downs.
He insulted me with put-downs.
15. I discussed the issue calmly.
S/He discussed the issue calmly.

DECIDED
ClElElElElEl
130131312113
121131313131]

During a conflict or argument with my boyfriend/girlfriend in the past year:
16. I said things to his/her fiiends about him/her

to turn them against him/her. Cl El E1 El
S/He said things to my friends about me to turn them

against me. El El El El

17. I ridiculed or made fun of him/her in front of others. El El El [:1
S/I-Ie ridiculed or made fun of me in fi'ont of others. 13 E] El El

18. I told him/her how upset I was. Cl E] El E1
S/He told me how upset s/he was. El El El El

19. I kept track of who s/he was with and where s/he was. E] El El 121
S/I-le kept track of who I was with and where I was. El 1:1 El E1

20. I blamed him/her for the problem. El [:1 El El
S/He blamed me for the problem. El El El E1

21. I kicked, hit or punched him/her. Cl Cl E] El
S/He kicked, hit or punched me. E] El El E1

During a conflict or argument with my boyfriend/girlfriend in the past year:
22. I left the room to cool down.
S/He left the room to cool down.
23. I gave in, just to avoid conflict.
S/He gave in, just to avoid conflict.
24. I accused him/her of flirting with another girl/ guy.
S/He accused me of flirting with another girl/ guy.
25. I deliberately tried to fiighten him/her.
s/He deliberately tried to fiighten me.
26. I slapped him/her or pulled his/her hair.
S/He slapped me or pulled my hair.

DDDDDDDDDD
DDDCICJEIEJDDEI
DDEIEIEIDCICIDEI
[1013000013130

192

 

 

Never: this has never happened in your relationship

Seldom: this has happened only 1-2 times in your relationship
Sometimes: this has happened about 3-5 times in your relationship
Often: this has happened 6 times or more in your relationship

 

 

During a conflict or argument with my boyfriend/girlfriend in the past year:
Never Seldom Sometimes Often

27. I threatened to hurt him/her/her.

S/He threatened to hurt me.
28. I threatened to end the relationship.

S/He threatened to end the relationship.
29. I threatened to hit him/her or throw something

at him/her.

S/He threatened to hit me or throw something at me.
30. I pushed, shoved or shook him/her.

S/He pushed, shoved, or shock me.

DUDE] DUDE]
DUDE] DUDE]
CIDCID DUDE]
DUDE] ElElElEl

193

Appendix K: Romantic Relationship Satisfaction

194

Please respond to these questions thinking about your current dating partner. If you do
not have a current dating partner, respond thinking about your most recent dating partner.
Please read each statement and write in the OE number that tells how much you agree
with each statement.

 

 

 

 

 

 

 

 

Strongly Moderately Somewhat Somewhat Moderately Strongly
Disagree Disagree Disagree Agree Agree Agree
1 2 3 4 5 6

 

. I really care for him/her.

. S/He acts thoughtfully.
10. S/he is a great companion.
l 1. I like the way I feel when I am with him/her.
12. I get upset when s/he shows interest in other girls/boys.

24. I never have to lie to him/her.

25. S/He listens to me when I need someone to talk to.
26. I find it easy to tell him/her how I feel.

195

. In general, I am satisfied with our relationship.

1
2
3. I often wish I hadn’t gotten into this relationship
4. Our relationship has met my best expectations.
5

6
7. I try to arrange my time so that I can be with him/her.
8
9

13. I like it when s/he pays attention to only me.
14. I watch other girls'/boys’ reactions to him.
15. S/He watches how I act with other guys/ girls.
16. Sometimes s/he doesn’t believe that I love only him/her.
17. S/He’s jealous of my relationships with other people.
18. I am happy when s/he succeeds.
19. I want him/her to be a success according to his/her own standards.
20. I like it when s/he does things on her/his own.
21. S/He makes me feel complete.
22. S/He helps me become what I want to be.
23. S/He helps me feel emotionally stronger.

. Compared to other people’s relationships, ours is pretty good.

. Our relationship is just about the best relationship I could hope to have with
anybody.
. I am happiest when we are together.

 

 

 

 

 

 

 

 

 

 

32.
33.
34.
35.
36.

37.
38.
39.

40.
41.
42.
43.

45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.

Strongly Moderately Somewhat Somewhat Moderately Strongly
Disagree Disagree Disagree Agree Agree Agree
1 2 3 4 5 6
__ 27. I really listen to what s/he has to say.
__ 28. S/He tells me about her/his weaknesses and strengths.
_ 29. S/He finds it easy to tell me how s/he feels.
_ 30. I am patient with him/her.
__ 31. I accept him/her for what he is.

I’m willing to forgive him/her for almost anything.

S/He recognizes and accepts faults in me.

S/He takes me for what I am.

I think s/he has good ideas.

I admire her/his persistence in getting after things that are important to
him/her.

I take pride in her/his accomplishments.

S/I-Ie thinks my ideas are important.

S/He respects my values and beliefs, although they don’t always agree with
hers/his.

S/He knows when something is bothering me.

I help him/her through difficult times.

I make him/her feel self-confident.

I am concerned with how s/he feels.

S/He helps me find solutions to my problems.

S/He comforts me when I need comforting.

S/He tries to get me in a good mood when I am angry.

S/He sometimes gets angry at me.

Dating can sometimes be painful for him/her.

Sometimes I really upset him/her.

I sometimes get upset because things don’t go well between us.

S/He can really hurt my feelings.

Sometimes I don’t know why I put up with the things s/he does or says.

I want to be special in his/her life.

No one could love him/her as much as I do.

I treat him/her as very special.

S/He is the most important person in my life.

196

 

 

 

 

 

 

 

 

 

 

 

 

 

Strongly Moderately Somewhat Somewhat Moderately Strongly
Disagree Disagree Disagree Agree Agree Agree
1 2 3 4 5 6
57. I feel that s/he was meant for me.
58. S/He is the person that best understands me.
During the last two questionnaires I was mostly thinking about:
1:] a boyfiiend C] an ex-boyfiiend 1:] other (fill in)
D a girlfiiend 1:] an ex-girlfiiend 1:] other (fill in)

 

 

 

197

 

 

Appendix L: Children’s Depression Inventory — Short Form

198

People sometimes have different feelings and ideas. This form lists the feelings and ideas
in groups. From each group of three sentences, pick one sentence that describes you best
for the past two weeks. After you pick a sentence from the first group, go on to the next

group.

Remember, pick out the sentences that describe you best in the past two weeks.
1. I am sad once in a while.
I am sad many times.
I am sad all the time.

Nothing will ever work out for me.
I am not sure if firings will work out for me.
Things will work out for me ok.

I do most things ok.
I do many things wrong.
I do everything wrong.

I hate myself.
I do not like myself.
I like myself.

I feel like crying everyday
I feel like crying many days.
I feel like crying once in a while

Things bother me all the time.
Things bother me many times.
Things bother me once in a while.

I look ok.
There are some bad things about my looks.
I look ugly.

I do not feel alone.
I feel alone many times.
I feel alone all the time.

I have plenty of fiiends.

I have some fiiends but I wish I had more.
I do not have any fiiends.

10. Nobody really loves me.

I am not sure if anybody loves me.
I am sure that somebody loves me.

9‘
[31:11:] ClElEl DUI—.1 DUE] DUE! DUE] ClElCl DUE! DUE] 1:11:11]

199

Appendix M: Rosenberg Self-Esteem Questionnaire

200

Please read each statement and write in the ONE number that tells how much you agree
with each statement.

 

Strongly Disagree Disagree Agree Strongly Agree

 

 

 

 

 

1 2 3 4

 

I feel that I am a person of worth, at least on an equal basis with others.

I feel that I have a number of good qualities.

All in all, I am inclined to feel that I am a failure.

I am able to do things as well as most other people.

I feel I do not have much to be proud of.

Pesewewwr

I take a positive attitude toward myself.
On the whole, I am satisfied with myself.

I wish I could have more respect for myself.

I

I

l

I certainly feel useless at time.
10. At times, I think I am no good at all.

I

201

 

Appendix N: The Self-Perception of Social Support - Approval subscale — Revised

202

Read each statement all the way across. Each statement describes two kinds of teenagers, one on
the left, and one on the right. m decide which kind of teenager YOU are most like the one
on the left, or the one on the right. Pick one. Then for each kind of teenager, check whether that
description is @111 true fomu or just sort of true for you. So for each numbered item you will
be checking only one box. Sometimes it will be on the left, sometimes on the right. DO NOT
check both sides. Just put a check on the side that is most like you.

 

What people in my life think about me as a person

Really Sort of Sort of Really
True True True True
For me for me for me for me
1. E] El Some teenagers have Other teenagers have E] El
mothers who do not BUT mothers who often praise
praise them or say nice them or say nice things
things about them about them

2. Cl [:1 Some teenagers have Other teenagers have Cl C]
fathers who are really BUT fathers who are n_ot very proud of
proud of their children their children.

3. El CI Some teenagers have Other teenagers have best El D

best friends who do n_ot BUT friends who do like most of
like many of the things they do the things they do

4. U [3 Some teenagers have Other teenagers have Cl C]
classmates who like BUT classmates who do not
the kind of person they are like the kind of person they are

5. [3 Cl Some teenagers have Other teenagers haves El El
teachers who praise them BUT teachers who do not praise
or say nice things about them them or say nice things about them

6. E] El Some teenagers have a person Other teenagers have a D E]
they are romantically BUT person they are romantically
interested in who does n_ot interested in who does like

like the kind of person they are the kind of person they are

7. C] I] Some teenagers have mothers Other teenagers have Cl C]
who are n_ot very proud of BUT mothers who are really proud of
their children their children.

8. E] El Some teenagers have fathers Other teenagers have Cl C]
who like most of the BUT fathers who do M like many of
things their children do the things their children do

9. E] El Some teenagers have best Other teenagers have best El E1
friends who do not like BUT friends who do like the kind
the person they are of person they are

203

Really Sort of

True

For me for me

10. C]

11.0

12. CI

13. CI

14. C]

15. E]

16.0

17.0

18. CI

L'J Some teenagers have

classmates who do not BUT
praise them or say nice

things about them.

Some teenagers have
teachers who are really BUT
proud of their students

Some teenagers have a

person they are BUT
romantically interested in

who does n_ot praise them

or say nice things about them.
Some teenagers have

mothers who do n_ot BUT
like many of the

things their children do

Some teenagers have fathers
who like the kind of BUT
person their children are.

Some teenagers have best

friends who do n_ot BUT
praise them or say nice

things about them

Some teenagers have
classmates who are really BUT
proud of them

Some teenagers have
teachers who do n_ot like BUT

many of the things their students do

Some teenagers have

a person they are BUT
romantically interested

in who is really proud of them.

204

Sort of Really
True True
for me for me

Other teenagers have E] El
classmates who often

praise them or say nice
things about them

Other teenagers have E] El
teachers who are mt very
proud of their students

Other teenagers have a Cl C]
person they are

romantically interested in

who does praise them or

say nice things about them.

Other teenagers have Cl C]
mothers who like most of
the things their children do

Other teenagers have fathersCl Cl

who do not_ like the kind

of person their children are

Other teenagers have best Cl C]
friends who do praise them

or say nice things about them

Other teenagers have E] El
classmates who are n_ot

very proud of them.

Other teenagers have D D
teachers who do like most of

the things their students do

Other teenagers have Cl C]
a person they are

romantically interested in
who is n_ot very proud of them.

Really Sort of
True True
For me for me

19. U C] Some teenagers have
mothers who do n_ot
like the kind of person
their children are

20. E] El Some teenagers have

fathers who do Lg;
praise them or say nice
things about them

21. Cl C] Some teenagers have best

friends who are mt very
proud of them.

22. Cl C] Some teenagers have

classmates who like most
of the things they do.
23. D D Some teenagers have
teachers who do not like
the kind of person
their students are.
24. El El Some teenagers have
a person they are
romantically interested
in who does n_ot like

most of the things they do.

BUT

BUT

BUT

BUT

BUT

BUT

205

Sort of Really
True True
for me for me

Other teenagers have D C]
who do like the kind

of person their
children are.

Other teenagers have E] El
fathers who do praise them

or say nice things

about them.

Other teenagers have best E] El
friends who are really
proud of them.

Other teenagers have Cl C]
classmates who do not like

many of the things they do.

Other teenagers have E] El
teachers who do like the

kind of person their students are.

Other teenagers have E] El
a person they are

romantically interested in
who does like most of the
things they do.

Appendix 0: The Perceived Social Support Scale

206

Directions: The statements that follow refer to feelings and experiences that occur to
most people at one time or another in their relationship with friends. For each statement
there are four possible answers:
Strongly Disagree Disagree Agree Strongly Agree
1 2 3 4

 

 

 

 

 

 

 

 

 

1. My friends give me the moral support I need.

2. Most other people are closer to their friends than I am.

3. My friends enjoy hearing about what I think.

4. Certain friends come to me when they have problems or need advice.
5. I rely on my friends for emotional support.

6. If I felt that one or more of my friends were upset with me, I’d just keep it to
myself.

7. I feel that I’m on the fringe of my circle of friends

8. There is a friend I could go to if I were just feeling down, without feeling funny
about it later.

9. My friends and I are very open about what we think about things.

10. My fi‘iends are sensitive to my personal needs.

11. My friends come to me for emotional support.

12. My friends are good at helping me solve problems.

13. I have a deep sharing relationship with a number of fiiends.

14. My fi'iends get good ideas about how to do things or make things from me.
15. When I confide in friends, it makes me feel uncomfortable.

16. My friends seek me out for companionship.

17. I think that my friends feel that I’m good at helping them solve problems.

18. I don’t have a relationship with a friend that is as intimate as other people’s
relationships with friends.

19. I’ve recently gotten a good idea about how to do something from a friend.

20. I wish my friends were much different.

207

 

Directions: The statements that follow refer to feelings and experiences that occur to
most people at one time or another in their relationships with their families. For each
statement there are four possible answers:

 

 

 

 

 

 

 

 

Strongly Disagree Disagree Agree Strongly Agree
1 2 3 4
1. My family gives me the moral support I need.

2 I get good ideas about how to do things or make things from my family.

3 Most other people are closer to their family than I am.

4. When I confide in the members of my family who are closest to me, I get the

idea that it makes them uncomfortable.

5 My family enjoys hearing about what I think.

6 Members of my family share many of my interests.

7 Certain members of my family come to me when they have problems or need
advice.

8 I rely on my family for emotional support.

9. There is a member of my family that I could go to if I were just feeling down,
without feeling fimny about it later.

10. My family and I are very open about what we think about things.

1 1. My family is sensitive to my personal needs.

12. Members of my family come to me for emotional support.

13. Members of my family are good at helping me solve problems.

14. I have a deep sharing relationship with a number of members of my family.

15. Members of my family get good ideas about how to do things or make thing
from me.

16. When I confide in members of my family, it makes me uncomfortable.

17. Members of my family seek me out for companionship.

18. I think that my family feels that I’m good at helping them solve problems.

19. I don’t have a relationship with a member of my family that is as close as other
people’s relationships with family members.

20. I wish my family were much different.

208

   

uttttttjnjtgtgjtitty: