. .V .. . . H .... .‘,..‘.,.. .. . , .A . .... . ‘ ., .‘ “V M. .. .k. -. u... ”a...“
, _ , _ ,
‘ ‘ ‘ . , ‘
‘ , . , 4 ,
_ ‘V A
f ‘ * ‘ ‘ »- . , w
‘ ’ ‘ . . ‘ ‘ . . . .‘ _ . ~ ‘ ‘

 

THESIS

WWW/WW

This is to certify that the

dissertation entitled
ATTACHMENT STYLES AND CLOSE RELATIONSHIPS
presented by
Carole Nhu'y Hodge

has been accepted towards fulfillment
of the requirements for

Ph . D. degreeinPSVChOlOQSL

Wm a, am

Major pfgfissor

Date ”CV. 15% IqQ‘

MSUt’: an Affirmative Action/Equal Opportunity Institution 0-12771

 

 

LIBRARY

Michlgan State
University

 

 

 

record.
PLACE IN RETURN BOX to remove thb stroked from your
To AVOID FINES return on or before data no.

DATE DUE DATE DUE DATE DUE

 

I i r l

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ATTACHMENT STYLES AND CLOSE RELATIONSHIPS
By

Carole Nhu'y Hodge

A DISSERTATION
Submitted to
Michigan State University

in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Psychology

1 996

ABSTRACT
ATTACHMENT STYLES AND CLOSE RELATIONSHIPS
By

Carole Nhu’y Hodge

The aim of this study was to fiirther our understanding of the links between adult
attachment styles and close interpersonal relationships. In its original form, attachment
theory addressed the processes by which afi‘ectional bonds are forged and broken between
infant and caregiver (Bowlby, 1969). According to Bowlby, an attachment is an
emotional bond involving comfort, security, trust, and support. Research suggests that
adult attachment style is an orientation that evolves primarily from people’s childhood
relationships with their parents (e.g., Ainsworth et al., l978). Studies also have found
that adult attachment styles affect people’s experience of romantic love (e. g., Hazan &
Shaver, 1987). However, close friendship, another important adult relationship, has yet
to be examined in terms of its potential links to attachment styles. The present study
attempted to replicate past work on the association between adult attachment styles and
romantic relationships and to extend understanding of this important psychological
phenomenon by examining its links to close friendships.

Psychology undergraduates (583 females, 213 males) completed two adult
attachment scales (Bartholomew & Horowitz, 1991; Hazan & Shaver, 1987) and rated

their romantic partner, other-sex best friend, and same-sex best friend on important

relationship components using the Davis and Todd’s Relationship Rating Form.

As expected, results for the romantic relationship hypotheses replicated those of
previous research. Secure individuals rated their romantic partner higher on positive
relationship components compared to both anxious-ambivalent and avoidant individuals.
As predicted for the friendship hypotheses, attachment style also affected ratings of
certain relationship components. In general, secure individuals rated their other-sex and
same-sex best friends higher on relationship components such as viability, intimacy, and
care than did either anxious-ambivalent or avoidant individuals.

The importance of this research is that it extends our understanding of attachment
style beyond romantic relationships to other close relationships by showing that
regardless of relationship type, an individual’s attachment style predicts their perceptions
of their close bonds. This puts adult attachment research into a broader interpersonal
relationship matrix by expanding the generalizability of attachment style to different

types of relationships.

Copyright by
CAROLE NHU’Y HODGE
1996

This dissertation is dedicated to M.L.O.

ACKNOWLEDGMENTS

I am grateful to my two chairs, Dr. Joel Aronofi‘ and Dr. Ellen Strommen for all
their help and guidance throughout my dissertation process. Their knowledge and
expertise in the areas of attachment style and relationships have been of great benefit to
me. I have enjoyed working with them and am forever grateful to them for all that they
have done. I would also like to express my deepest appreciation to Dr. Larry Messé for
all his help in the design of my dissertation, in the statistical analyses, and all his very
helpful comments and suggestions. I would also like to thank Dr. Galen Bodenhausen for
his careful consideration and helpful suggestions of this manuscript.

I would like to give a big thanks to Dr. Linda Sullivan for her friendship, help,
social support, understanding, encouragement, and graduate school wisdom to help me
see that there is a light at the end of the tunnel--and it’s not a train coming towards me.
This dissertation could not have been completed without her help.

I would like to thank my father (Harold) and my family for all their moral and
financial support. See, it was all worth it in the end.

I would like to give a big thanks to my best buddy Geri for all her help in this
whole process and for being there for me during all my years in graduate school. Thanks

for helping me clean data, collate questionnaires, and being so good about joining in

vi

many other fun research things that I volunteered you for.

Thanks to my other family/friends for all their encouragement. Brad & Barb,
thanks for all your advice about graduate school and for supporting and believing in me
all along the way. Akemi, thanks for being my vacation buddy. Joan & Fergus, thanks
for all your encouragement. Anne & Dan, thanks for your support. Jo & Julia, thanks for
your encouragement. Julia, thanks for being so enthusiastic about my topic.

There are thanks I would like to give to my friends at MSU for all their support
and help throughout my years in graduate school. Suzy, thanks for answering all my
questions, telling me all the important dates to look out for, being a constant source of
social support, and being my fi'iend. Seka, I could not have gotten through my
dissertation without your help and moral support. Thanks for being such a nice person
and for the angels that you have sent my way. They have been a great help. Thanks to all
the spigkids for all their humor and understanding.

Last but not least, I would like to thank my new but true friends in my Marketing
Department at CRM Films (Lyndi, Andrea, Danielle, & Liad). Lyndi, thanks for being a
wonderful boss, a good friend, a constant source of support and encouragement, and a
person I can laugh with and talk to on a daily basis. Andrea, thanks for letting me drive
you mad and being my nerf ball buddy. Danielle, thanks for giving me real world
insights into relationships. Liad, thanks for being so nice and supportive. You all make
it fun to come to work everyday. I would also like to say thanks to everyone else at CRM
Films for their support. Lisa, thanks for being a friend and the coolest HR person. Now

you can all call me Dr. Carole!

vii

TABLE OF CONTENTS

LIST OF TABLES .......................................................................................................... xiii
INTRODUCTION .............................................................................................................. 1
Overview ............................................................................................................................. 1
Attachment Theory ........................................................................................................... 3
Brief History of Attachment Theory ............................................................................. 3
Infant Attachment ......................................................................................................... 5
Attachment Style across the Lifespan ............................................................................... 7
Development of Mental Models ................................................................................... 7
Multiple Attachments ................................................................................................... 8
Continuity of Attachment Style .................................................................................... 8
Adult Attachment ............................................................................................................ 11
Mental Models in Adults ............................................................................................ 11
Measurement of Adult Attachment ............................................................................. 13
Attachment Style and Adult Romantic Relationships .................................................... 18
Relevant Research on Adult Attachment Styles and Romantic Relationships ........... 18

Summary of Relevant Research Results on Sex Differences in Attachment Styles....21

Summary of Results on Attachment Styles and Adult Love Relationships ................ 25

viii

Friendship ....................................................................................................................... 28

Brief History of Research on Friendship .................................................................... 28
Characteristics of a Friend .......................................................................................... 30
Types of Friendship ................................................................................................... 31
Benefits of Friendship ................................................................................................. 31
Components of Friendship .......................................................................................... 32
Female-Female Friendships ........................................................................................ 34
Male-Male Friendships ............................................................................................... 34
Summary of Friendship Research ............................................................................... 35
Comparing Romantic Relationships and Best Friendships ............................................. 37
Similarities between Romantic Relationships and Best Friendships .......................... 39
Differences between Romantic Relationships and Best Friendships .......................... 40
Close Relationship Components ................................................................................. 42
Pertinent Research Studies for the Present Study ........................................................... 44

Why would different Attachment Styles have different ratings for Best Friend

and Romantic Partner .................................................................................................. 46
Overall Summary ............................................................................................................ 50
Hypotheses ...................................................................................................................... 54

Replication Hypotheses .............................................................................................. 54
Extension Hypotheses ................................................................................................. 57
Differential Hypotheses .............................................................................................. 59
Additional Hypotheses ................................................................................................ 62

ix

CHAPTER 11: METHOD ................................................................................................ 64

Participants ...................................................................................................................... 64
Attachment Style Classification .................................................................................. 65
The Analysis Sample .................................................................................................. 66
The Between-Subjects Sample ................................................................................... 68
The Within-Subjects Sample ...................................................................................... 69

Materials ......................................................................................................................... 73

Procedure ........................................................................................................................ 76

CHAPTER III: RESULTS .............................................................................................. 79

Overview ......................................................................................................................... 79

Reliability of the RRF ..................................................................................................... 81

Tests of the Hypotheses .................................................................................................. 83

Replication Hypotheses: Effects of Attachment Style on Viability, Intimacy,

Global Satisfaction, Care, Passion, and Conflict-Ambivalence ratings of

Romantic Partner .......................................................................................................... 83
Replication Hypotheses: Summary ................................................................................ 86
Extension Hypotheses: Effects of Sex, Attachment Style, and Relationship

Type on Viability, Intimacy, Global Satisfaction, and Care ratings of

Best Friends .................................................................................................................. 89
Extension Hypotheses: Summary ................................................................................... 92
Differential Hypotheses: Effects of Sex and Attachment Style on Intimacy,

Global Satisfaction, and Care ratings of Relationship Types ....................................... 96

Differential Hypotheses: Summary ............................................................................... 101

Additional Hypotheses .................................................................................................. 106
Additional Hypotheses: Summary ................................................................................ 108
Overall Analyses .......................................................................................................... 108
Overall Analyses: Summary ......................................................................................... 116
Supplemental Analyses ................................................................................................ 130
CHAPTER IV: DISCUSSION ...................................................................................... 137
Replication .................................................................................................................... 138
Extension ...................................................................................................................... 144

The Effects of Attachment Style, Sex, and Relationship Types on Ratings of

Relationship Components .......................................................................................... 146
Sex Differences ............................................................................................................ 154
Best Friends as Attachment Figures ............................................................................. 156
Limitations .................................................................................................................... 158
Future Directions .......................................................................................................... 159
Conclusion .................................................................................................................... 161

REFERENCES ............................................................................................................... 163
APPENDIX A-l: CLOSE RELATIONSHIPS QUESTIONNAIRE Part 1 .................. 174
APPENDIX A-2: CLOSE RELATIONSHIPS QUESTIONNAIRE Part 2 .................. 175
APPENDIX B-l: The Short Form of the DSHQ ........................................................... 176
APPENDIX B-2: Demographic and Social History Questionnaire ............................... 177
APPENDIX C: Departmental Research Consent Form #1 ......................................... 180

APPENDIX D: Future Participation Sheet .................................................................. 181

APPENDIX E]: Davis & Todd’s Relationship Rating Form
Same-Sex Best Friend Version .......................................................... 182

APPENDIX E-2: Davis & Todd’s Relationship Rating Form

Other-Sex Best Friend Version .......................................................... 184
APPENDIX E-3: Davis & Todd’s Relationship Rating Form

Romantic Partner Version .................................................................. 186
APPENDIX F: Departmental Research Consent Form #2 .......................................... 188
APPENDIX G: Participant Information Sheet ............................................................. 189
APPENDIX H-l: Researcher Script (Mass Testing Version) ....................................... 190
APPENDIX H-2: Researcher Script (Social Relationship Study Version) ................... 191
FOOTNOTE .................................................................................................................... 192

xii

7a.

7b.

10.

11.

12.

LIST OF TABLES

Six Components of Close Relationships from Davis & Todd’s Relationship

Rating Form ............................................................................................................... 43
Sample Size by Attachment Style and Sex ............................................................... 70
Proportional Counts by Number of Relationships and Relationship Type

for the Analysis Sample ............................................................................................. 71
Sex breakdown for each incidence of a Relationship Type ....................................... 72
Reliabilities (coefficient alphas) for the RRF ........................................................... 82

Mean ratings on relationship variables for Romantic Partner for each
Attachment Style for the Analysis Sample ............................................................... 88

Mean ratings on relationship variables for Other-Sex Best Friend for each
Attachment Style for the Between-Subjects Sample and the Within-
Subjects Sample ........................................................................................................ 94

Mean ratings on relationship variables for Same-Sex Best Friend for each
Attachment Style for the Between-Subjects Sample and the Within-

Subjects Sample ........................................................................................................ 95
Means for Paired-Sample T-Tests for the Analysis Sample ................................... 104
Means for Paired-Sample T-Tests for the Within-Subjects Sample ........................ 105
Means for the 2 X 3 X 3 AN OVAs for the Between-Subjects Sample ................... 120
Means for the 2 X 3 X 3 AN OVAS for the Within-Subjects Sample ...................... 121

2 (Sex) X 3(Attachment Style) X 3(Relationship Type) AN OVA on:
Viability, Intimacy, Global Satisfaction, Care, Passion, and
Conflict-Ambivalence for the Between-Subjects Sample ....................................... 122

xiii

13.

14.

15.

16.

17.

l8.

19.

20.

21.

2 (Sex) X 3(Attachment Style) X 3(Relationship Type) ANOVA on:
Viability, Intimacy, Global Satisfaction, Care, Passion, and

Conflict-Ambivalence for the Within-Subjects Sample .......................................... 124

Attachment Style Means and Main Effects for the Between-Subjects and

the Within-Subjects Samples ................................................................................... 126

Relationship Type Means and Main Effects for the Between-Subjects

and the Within-Subjects Samples ............................................................................ 127

Simple effects for Sex By Relationship Type Interactions for the Within-

Subjects Sample ...................................................................................................... 128

Simple effects for Attachment Style By Relationship Type Interaction

for the Between-Subjects Sample ........................................................................... 129

Mean ratings on Length of Time with Romantic Partner, Other-Sex Best

Friend, and Same-Sex Best Friend by Attachment Style and Sex .......................... 133

Mean ratings on Closeness with Romantic Partner, Other-Sex Best Friend,

and Same-Sex Best Friend by Attachment Style and Sex ...................................... 134

Mean ratings on Expectation of Future Togethemess with Romantic Partner,
Other-Sex Best Friend, and Same-Sex Best Friend by Attachment Style and

Sex .......................................................................................................................... 135

Mean ratings on Past Numbers of Romantic Partner, Other-Sex Best Friend,

and Same-Sex Best Friend by Attachment Style and Sex ...................................... 136

xiv

INTRODUCTION

Mm

The overall aim of the present study was to further our understanding of the
relationship between attachment styles and close interpersonal relationships. This study
examined how individuals within each of the three attachment styles rated their other-sex
and same-sex best friends, as well as their romantic partner, on important relationship
components.

There were three purposes to this research. The first was to replicate previous
findings by examining the relationship between adult attachment style and romantic
relationships. The second purpose was to extend the previous findings of attachment
style correlates of relationship components within romantic relationships to the area of
friendship by examining the links between adult attachment style and relationship
components in both same-sex and other-sex best friends. The third purpose was to
examine the effects of attachment style, sex, and different types of relationships (i.e.,
romantic relationship, other-sex and same-sex best friendships) on ratings of relationship
components within a more comprehensive framework. These close relationships are
similar on several factors, such as closeness and intimacy, but different on others, such as

sexual intimacy. It is hypothesized that the attachment style and sex of the participant

play a role in the ratings of relationship characteristics.

Friendships play an important role in adult life along with romantic relationships.
Best friends have been found to provide each other with social support, care, comfort and
to keep each other mentally and physically healthy. While there is a great deal of
research on adult attachment style and adult romantic or love relationships, there is little
research on adult fiiendships and adult attachment. People with different attachment
styles have been found to differ in their ratings of certain components of relationships
with their romantic partner, such as trust and self-disclosure. It is predicted that in
general the findings linking attachment styles and best friends will replicate those of
attachment styles and romantic relationships. However, because close friendships are
different from romantic relationships, different ratings are predicted to occur for certain
relationship components among the three attachment styles. Extending attachment theory
to adult best friendship is a natural and necessary extension to further our understanding
of adult attachment and close adult relationships.

In the pages that follow, attachment theory is discussed first by reviewing the
history of attachment theory, infant attachment, adult attachment, and how attachment
theory has been extended to the close adult relationship area, specifically to adult
romantic and love relationships. The friendship literature is discussed next, starting with
a brief history of friendship, the characteristics of friendship, types of friendship,
components of friendship, and why the adult attachment literature should be extended to
adult best friendships. A discussion of the similarities and differences between

friendships and romantic relationships is discussed next. Lastly, hypotheses are

developed about the relationship between attachment style, sex, and relationship type on

various important relationship components.

Attachmenuhecni
E . [H E ! l I]

Attachment theory addresses the processes by which affectional bonds are forged
and broken between infant and caregiver (Bowlby, 1969, 1973, 1979, 1982, 1988).
According to Bowlby, an attachment is an emotional bond. This bond involves comfort,
security, trust, and support. It describes and explains how infants become emotionally
distressed when separated from their primary caregiver. The basic assumption of
attachment theory is that humans are immature at birth and can only survive if an adult is
willing to provide protection and care. Because of this, behaviors have evolved to ensure
that the infant maintains in close relation to the caregiver/protector. Bowlby stated that
there are three defining features of attachment and functions of an attachment
relationship. Proximity maintenance or staying near the caregiver is one. A second
feature is using the caregiver as a secure base, meaning that the infant can feel
comfortable enough to engage in other behaviors such as exploring the environment. The
third feature is using the caregiver as a safe haven, meaning that the infant can rely on
the caregiver for comfort, support, and reassurance. When an individual perceives a
threat to proximity maintenance, attachment behaviors are activated resulting in a search
and re-establishment of proximity to the attachment figure. The goal of the attachment
system is to keep the infant near the caregiver for protection. Ainsworth (1982)

broadened this goal to define the attachment system as a secure base from where infants

4

can feel safe to confidently explore their environment. The three criteria for
infant-caregiver attachment according to Bowlby are the association of the attachment
figure with feelings of security, the greater likelihood of attachment behavior when the
infant is in a situation of apparent threat, and the tendency of infants to attempt to ward
off separation from an attachment figure by calling or crying, and when able, by
following.

While working for the World Health Organization, Bowlby observed and reported
that inadequate maternal care during early childhood had negative influences on
personality development. Separation from their caregiver(s) had profound effects on
children’s relations to their caregivers when reunion took place. With this important
finding, Bowlby wanted to develop a theory of personality development that started at the
core of personality development and followed it through its natural course. Attachment
theory was developed out of concepts and ideas from the obj cot-relations tradition in
psychoanalysis, evolution theory, ethology, control theory, and cognitive psychology.

Attachment theory can be Stunmarized in terms of three propositions. First, when
an individual is confident that an attachment figure will be available whenever desired,
that person will be much less prone to chronic fear than will an individual who, for
whatever reason, has no such confidence. The second proposition concerns the sensitive
period during which such confidence develops. Attachment theory postulates that
confidence in the availability of attachment figures is built up slowly during the early
years of childhood---and that whatever expectations are developed during those years

tend to persist relatively unchanged throughout the rest of life. However, some current

researchers such as Kobak and Hazan (1991) challenge this notion. They suggest that
working models of others can be assimilated and accommodated continuously throughout
life. The third proposition concerns the role of actual experience. It postulates that the
varied expectations of the accessibility and responsiveness of attachment figures, which
individuals develop during the years of immaturity, are reflections of the experiences
those individuals have actually had (Bowlby, 1973, p. 235).

Attachment is one of several different and important behavioral systems that
facilitate human survival. The attachment system dominates in infancy and childhood.
When this system is in a crisis, all other behavioral systems such as exploration are
postponed until the crisis is over. An attachment behavior can be any of various forms of
behavior that a person engages in from time to time to obtain and/or maintain a desired
proximity to the caregiver. Attachment behavior is most obvious in early childhood but
can be observed throughout the life cycle, especially in emergencies. Researchers (e. g.,
Shaver, Hazan, & Bradshaw, 1988) have pointed out substantial Similarities between
attachment in childhood and adult romantic love relationships.

InfantAttachmem

Ainsworth, Blehar, Waters, and Walls (1978) made a major contribution to
attachment theory by identifying three different types of attachment pattems/styles that an
infant can have with his or her primary caregiver. The three attachment styles identified
were secure, WW, and Wright. “Secure attachment develops
when infants are confident that their parent (caregiver) will be available, responsive, and

helpful should they encounter adverse or frightening situations. With this assurance, they

feel bold in their exploration of the world. AnximAmhiyalem attachment develops
when infants are uncertain whether their parent will be available, responsive, or helpful
when called upon. Because of this uncertainty, they are always prone to separation
anxiety and tend to be clinging and are anxious about exploring the world. Anxieus;
Ayeidam attachment develops when individuals lack confidence that when they seek
care, they will receive a helpful response. On the contrary, they expect to be rebuffed.
These individuals are compulsively self-reliant, independent of affectional ties, and
emotionally self-sufficient. When this happens to a marked degree, these individuals
attempt to live life without the love and support of others; they try to become emotionally
self-sufficient and may later be diagnosed as narcissistic or as having a false self ’
(Bowlby, 1988, p. 124). In this study, anxious-avoidant will be referred to as avoidant.

Ainsworth and her colleagues (1978) and Egeland and Farber (1984) have
remarked that the three styles stated above seem closely associated with differences in
caregiver warmth and responsiveness. The Seeme pattern of attachment is promoted by a
caregiver who is readily available, sensitive to the child's signals, and lovingly responsive
when the child seeks protection and comfort. The W pattern of
attachment is promoted by a caregiver who is available only on some occasions. The
caregiver uses threats of abandonment as a means of control. The Ayeidam pattern of
attachment results fi'om constant rejections by the caregiver to approaches for comfort or
protection. Ainsworth et a1. (1978) found that 56% of infants had a secure attachment
style, 19% had an anxious-ambivalent attachment style, and 21% had an avoidant

attachment style. However, 4% of the infants were classified as disorganized/disoriented.

Attachmentflxleamssjhmm

DeedepmemeflMemalMQdels. Bowlby stated that based on an individual's
relationship with the caregiver, he or she creates working models of self and others which
are beliefs and expectations about self, others, and relationships (Bowlby, 1973). These
internal representations of others and self provide the groundwork for personality
development and conceptualization about relationships and others (Ainsworth, et a1.
1978; Bowlby, 1973; Main, Kaplan, & Cassidy, 1985). These mental models become the
basis for attachment patterns or styles that individuals develop at a very young age. Once
an individual develops an attachment style or pattern, the style tends to persist and
become more difficult to change with time because of the internalization of the mental
models, although change through subsequent relationship experience remains possible.
Most of us develop some style of attachment by the end of our first year of life. These
mental models carry over to new relationships where they guide and organize
expectations, perceptions, and behavior (Bowlby, 1973). As time passes, these models
become more important for children's general style of interaction with others. Belsky and
Cassidy (1994) have suggested that mental models are constructed from real relationship
experiences. The people with whom individuals interact play a role in their developing
mental models. These models play a major role during interaction with others because
they not only help the child interpret the situation but they also integrate new experiences
into the existing perceptions of self and others. In addition, they also provide predictions

for future interactions and relationships with others.

MnltipleAnaehmems. It is suggested that individuals can have multiple
attachment figures throughout life. Bowlby (1982) stated that there is a plurality of
attachment figures. However, all figures are not equal in that there is a hierarchical order
to attachment figures. Even in infancy where individuals usually have a primary
attachment figure, the mother, the father is also another attachment figure. Ainsworth
(1982) stated that “attachment patterns in infancy and childhood usually refer to how
infants and children organize their attachment behavior in terms of a particular
caregiver. " However, Bowlby (1988) stated that "As yet little is known about how the
influence on personality development of interactions with the mother compares with the
influence of those with the father. It would hardly be surprising were different facets of
’ personality, manifest in different situations, to be influenced differently. In addition,
their respective influences on males may be expected to differ from their respective
influences on females" (p. 129). And in fact, research with infants shows that attachment
is not always the same to mothers as to fathers. These styles may change with the
appropriate environment and interactant. Extending this, a person can have a different
attachment style towards different attachment figures or significant others. For adults,
these different attachment figures can be a best friend or a romantic partner.

Cominn'mmfAmmmenLStyle. Previous research on attachment lends some
support to the continuation of patterns of attachment (i.e., attachment style) across the life
span. Bowlby (1979, p. 129) maintained that "attachment behavior characterizes human
beings from cradle to grave." However, this is different from saying the same patterns

characterize the same individual from cradle to grave. A person's attachment style may

be different at infancy and at adulthood. However, researchers have found that these
styles are relatively stable (e.g., Main & Cassidy, 1988; Wartner, 1986). Nevertheless, as
a person interacts with others, he or she accumulates information about self and
relationships. She or he can either add this information to her or his mental model or
discard it. Continuity of attachment styles have been examined by various researchers

(e. g., Elicker, Englund, & Sroufe, 1992; Grossman & Grossman, 1991) and they have
found attachment styles to be relatively stable over the first 10 years of life, given a stable
home environment. Hazan and Hutt (1993) found that stability was more common that
change (78% vs. 22%) in attachment styles.

Regardless of continuity or discontinuity, attachment style can now be measured
in adults. Main and her colleagues (1985) have developed several intensive interview
procedures to measure attachment in both children and adults, with the goal being to
extend attachment findings across the lifespan. Their research and several other studies,
using paper and pencil self-report measures (e.g., Hazan & Shaver, 1987; Collins & Read,
1990), indicate that attachment style in infancy and adulthood are similar. For instance,
Hazan and Shaver (1987) found that attachment styles Similar to those found in children
also can be identified in adults and that the distribution of adults across styles was very
similar to those found in infants. F ifty-six percent of adults were secure, 19% were
anxious-ambivalent, and 25% were avoidant. Other researchers have found many strong
similarities between infant and adult attachment (e.g., Collins & Read, 1990). According
to Hazan and Shaver (1987) adult attachment bonds are primarily manifested in romantic

love.

10

Even though there seems to be the understanding that attachment styles are very
similar for infants and adults, there are definite differences that exist between infants and
adults for the behaviors displayed, roles played, and thinking mode about attachment
(Weiss, 1982, 1991). First, the attachment relationship in infancy and childhood is
complementary where the infant/child seeks security and care but does not provide these,
while the caregiver provides security and care but does not seek them. In contrast, the
attachment relationship in adulthood is reciprocal; each person in the relationship both
provides and receives care. Second, in infancy and childhood, the attachment
relationship is manifested in physical behaviors and can be observed. In adulthood, the
attachment relationship is formed into an internal belief system about the partner (Main et
al., 1985). Third, the attachment system in adulthood is not manifested to the same
degree as it was in infancy. During infancy, separation from the attachment figure leads
to an inability to concentrate on anything else. During adulthood, separation does not
typically have as intense an effect. Adults can continue to perform their everyday
functions when separated from their attachment figure, though they may Show some
distraction. However, the emotions can be as intense but adults psychologically know
that there is security. Adults can perform functions with stress fiom separation. And
finally, an infant or child's attachment figure is usually a parent whereas an adult's
attachment figure is most likely to be a peer and usually is the sexual partner (at least as

conceptualized and studied to date).

11

AdulLAttachment

Through adolescence and young adulthood, individuals begin to form attachment
to new attachment figures, such as friends and romantic partners (e. g., Hazan & Shaver,
1987; Bennan, Marcus, & Berman, 1994; Levitt, 1991). The new attachment figure is
perceived to be one who can be trusted to provide comfort, support, and security when
necessary. In this new relationship, all criteria of attachment bonding can be observed.
Around this time, attachment functions (i.e., proximity seeking, safe haven, and secure
base) are transferred from parent to other attachment figures (e. g., peer). However,
attachment to parents does persist and is not replaced, though it may become
overshadowed by new attachments. Even as adults, when anxious or distressed, people's
attachment system will be activated and they will seek and desire protection and comfort
from an attachment figure, whomever this person may be. F eeney and Noller (1990)
described attachment relationships as characterized by a seeking of comfort and security
within relationships. According to Ainsworth (1989) dyadic attachment relationships are
affectional bonds and are enduring.

’ WW. Mental models are beliefs and expectations about self,
others, and relationships that were created in infancy from the relationship with the
caregiver (Bowlby, 1973). Mental models are relatively stable in regulating a person's
affect and behaviors throughout life (Mikulincer & Nachshon, 1991) and they play an
important role in understanding how early relationships affect adult relationships. The
early relationship becomes a model for later relationships and Hazan and Shaver (1987)

stated that variations in early social experiences produce relatively enduring differences

12

in relationship styles. These models have been suggested to persist throughout the
lifespan. According to these different models, individuals with secure, anxious-
ambivalent, and avoidant attachment styles should be involved in different kinds of
romantic relationships (Bowlby, 1973, 1980). Researchers have found that individuals
within each attachment style report different attributes about themselves, others, and
relationships.

Recent research has been conducted specifically examining what mental models
were developed by individuals possessing different attachment styles. Researchers have
found the following: people who possess a secure attachment style tend to develop mental
models of themselves as being fiiendly, good-natured, and likable and of significant
others as being generally well-intentioned, reliable, and trustworthy (Collins & Read,
1990; Feeney & Noller, 1990; Hazan & Shaver, 1987; Levy & Davis, 1988; Simpson,
1990); people who possess an anxious-ambivalent attachment style tend to develop
models of themselves as being misunderstood, unconfident, and under-appreciated and of
significant others as being typically unreliable and either unwilling or unable to commit
themselves to permanent relationships (Collins & Read, 1990; Feeney & Noller, 1990;
Hazan & Shaver, 1987; Mikulincer & Nachshon, 1991; Simpson, 1990), and they also
have a more negative view of themselves; people who possess an avoidant attachment
Style tend to develop models of themselves as being suspicious, aloof, and skeptical and
of significant others as being unreliable and overly eager to commit themselves to

relationships (Collins & Read, 1990; Feeney & Noller, 1990; Hazan & Shaver, 1987).

13

MeasurememeflAdmLAttaehmem. Several measures have been developed for
adult attachment. The first measure was developed by George, Kaplan, and Main (1984)
and is called the Adult Attachment Interview. This measure uses an interview format to
assess the quality of an adult's current internal representation of the childhood attachment
relationships with his or her parents. From this measure, three patterns of adult
attachment can be derived: secure, preoccupied, and dismissing. These three patterns
correspond to infant attachment patterns of secure, anxious-ambivalent, and avoidant,
respectively, as described by Ainsworth and her colleagues (197 8). While these
interviews produced very detailed and informative findings, they were also very time
consuming.

Because of the problems presented by interview measures, self-administered
paper and pencil scales have been developed to measure adult attachment. The first of
these was created by Hazan and Shaver in 1987. This scale has been used as a basis for
the development of subsequent adult attachment scales (e.g., Collins & Read, 1990;
Feeney & Noller, 1990; Simpson, 1990). Hazan and Shaver’s adult attachment scale is
both simple and self-administered. The measure consists of two parts. The first part is
categorical and consists of three paragraphs, each describing a style of attachment. The
participant is requested to select the paragraph that best describes them. During the
second part, the participant is requested to rate each of the above paragraphs on a 7-point
scale (1 = not at all like me to 7 = very much like me). Hazan and Shaver based their
measure on the attachment patterns of Ainsworth and her colleagues' classification of the

three attachment patterns of infancy/childhood (Ainsworth et al., 197 8). They found that

14

the three attachment styles found in infants exist in adults, and that these styles can have
important significance for adult romantic relationships. In their scale, a person selected
the one style of the three presented, that best fit them in a relationship context. The styles
were named secure, anxious-ambivalent, and avoidant complying with Ainsworth et al.'s
three patterns of infant attachment. In their study, Hazan and Shaver (1987) found 51-
56% participants selected the secure style, 19-21% selected the anxious-ambivalent style,
and 23-28% selected the avoidant style. Test-retest reliabilities of the three
attachment-style ratings were as follows: secure = .56; avoidant = .68; and anxious-
ambivalent = .56. Pistole (1989) also found adequate test-retest reliability (.59) for this
attachment scale over a one week period. The advantages to using this scale are that it
follows directly from the work of Ainsworth and her colleagues on the three attachment
patterns and it has been used by a variety of researchers. Hazan and Shaver’s scale can be
found in Appendix A-l.

Simpson's (1990) adult attachment scale, like Hazan and Shaver's, tests for
Ainsworth et al.'s three patterns of attachment. Simpson's scale is rooted in Hazan and
Shaver's scale in that Simpson broke Hazan & Shaver's three attachment style paragraphs
into 13 sentences. Examples of these statements are: "I find it relatively easy to get close
to others," "I Ofien worry that my partner(s) don’t really love me." Each statement relates
to some aspect of one of the attachment styles. Participants rate each statement on a 7-
point Likert-type scale (1 = strongly disagree to 7 = strongly agree). Reliabilities for the
styles in the scale were: secure, alpha = .51; Anxious-ambivalent, alpha = .79; and

avoidant, alpha = .59. For more information on reliabilities and validities see Simpson

15

(1990). The disadvantage of this scale is that it lacks confirming reliability and validity
testing.

Collins and Reads (1990) adult attachment scale was also based on Hazan and
Shaver's (1987) adult attachment measure. Collins and Read's scale consists of 18 items
and three underlying factors. The first factor is called Depend because it contains items
relating to trust and dependency on others (e.g., "I find it difficult to allow myself to
depend on others"). The second factor is called Anxiety because it contains items relating
to anxiety in relationships (e.g., "I do not often worry about being abandoned"). The
third factor is called Close because it contains items relating to being comfortable with
closeness and intimacy (e.g., “I find it relatively easy to get close to others"). Participants
are requested to rate their feelings on each statement on a 5-point scale (1 = not at all
characteristic to 5 = very characteristic). The internal consistencies were as follows:
Depend, alpha = .75; Anxiety, alpha = .72; Close, alpha = .69.

Collins and Read claimed that their scale is a more sensitive measure of adult
attachment and more precise than Hazan and Shaver's three styles. The disadvantage of
this scale is that the items do not correspond to the original three styles of attachment
(i.e., avoidant, anxious-ambivalent, secure) but instead to other dimensions depend,
anxiety, and close respectively. Also, it lacks confirming reliability and validity testing
by other researchers.

A four style attachment model to measure adult attachment was preposed by
Bartholomew (1990) and tested by Bartholomew and Horowitz (1991). This scale

measures adult attachment style but it differs fi'om Hazan and Shaver in that

16

Bartholomew differentiated the avoidant style into two separate styles: dismissing and
fearful. The scale discussed here is the simple and self-administered version of the more
complex self-administered scale from Bartholomew and Horowitz. The scale is very
similar to Hazan and Shaver's scale, where subjects selected one style that best fit them
and consists of two parts. The first part is categorical and consists of three paragraphs,
each describing a style of attachment. The participant is requested to select the paragraph
that best describes them. During the second part, the participant is requested to rate each
of the above paragraphs on a 7-point scale (1 = not at all like me to 7 = very much like
me). The four styles are secure, preoccupied, fearful, and dismissing. Bartholomew
and Horowitz (1991) found in their sample 47% selected the secure style, 18% selected
the dismissing style, 14% selected the preoccupied style, and 21% selected the fearful
style. Brennan, Shaver, and Tobey (1991) found that this measure was highly related to
Hazan and Shaver's (1987) scale, x2 (6) = 370.31, p<.0001. The advantage of this scale is
that it measures the four attachment styles defined by positive and negative mental
models of self and others. The disadvantage of this scale is that because of its recent
development, little testing has been conducted. Bartholomew and Horowitz’s scale can
be found in Appendix A-2.

For the purposes of this study, both Hazan and Shaver‘s Adult Attachment Scale
and Bartholomew and Horowitz’s Adult Attachment Measure were utilized. The two
scales are similar to each other except for some wording changes within each style

description. This was to ensure that an individual’s chosen attachment style would be

17

more valid and reliable than each scale by itself. The two scales were combined and
Hazan and Shaver’s (1987) three styles were selected over Bartholomew’s (1990) four
styles. Thus, individuals in Bartholomew’s fearful and dismissing styles were combined
into Hazan and Shaver’s avoidant style. The three styles were selected for the following
reasons. One, it followed directly from the work of Ainsworth and her colleagues on the
three attachment patterns. Ainsworth's work with infancy attachment patterns were
revised to measure adult attachment patterns. Two, many researchers who have used
Hazan and Shaver's Adult Attachment Scale have found their results to correspond
closely with other researchers. The scale has been used by a variety of researchers, in
various populations, and has been found to have high reliabilities and validities (e.g.,
Feeney & Noller, 1990; Mikulincer & Nachshon, 1991). Three, results from Hazan and
Shaver's scale have also found frequencies for the three adult attachment styles very
similar to that of the infant's three attachment styles. Four, the scale focused only on the
three pattems/ styles of attachment and not on other factors discussed by other researchers
(e.g., Collins and Read, 1990) on closeness and dependency. Five, the scale has been
used as a basis for many other adult attachment measures (e.g., Collins & Read, 1990;
Feeney & Noller, 1990; Simpson, 1990). And finally, both the categorical and
continuous measures yield fairly comparable results for both scales.

In a recent article, Hazan and Shaver (1994) proposed that attachment can be an
organization for research on close relationships, specifically adult love relationships.

Thus far, research on adult attachment has focused on adult romantic or love

18

relationships. The result of this can be seen in the adult attachment literature that has

focused on attachment style and romantic relationships, and is discussed below.

9119] 81 1B .31.].

0-. 0-. _ _ o _ ' o. ’ '
\ 4| \ -. I035. ;I_- our a euro”... ‘ g OIII

Hazan and Shaver (1987) extended attachment theory into the area of adult love/romantic
relationships. They found that individuals in the three attachment styles differed in how
they perceived and experienced romantic love. Love relationships of secure people were
characterized by trust, friendship, and positive emotions. Love relationships of anxious-
ambivalent people were characterized by obsession, desire for reciprocity and union,
emotional highs and lows, and extreme sexual attraction and jealousy. In addition,
anxious-ambivalent people had the shortest relationship duration while secure people had
the longest relationship duration. Love relationships of avoidant people were
characterized by fear of intimacy, emotional highs and lows, and jealousy. Their
relationships were also the least satisfying compared to people in the other two styles.
Collins and Read (1990) examined adult attachment, relationship quality, and
working models. They found that individuals tended to be in relationships with partners
who shared similar beliefs and feelings about becoming close and intimate with others
and about the dependability of others. However, they found that anxious-ambivalent
people, who desire to get close and commit to another person, were attracted to others
who wanted distance and were uncomfortable with getting close. These individuals were
seeking others who did not share their beliefs about relationships but confirmed their

expectations of what others think about relationships. Weiss (1982) suggested that

19

people may seek partners for whom their attachment system is already prepared to
respond; people who would fulfill their expectations. Sroufe and F leeson (1986)
suggested that people need to be consistent within their own mental models of self and
others; thus they may seek to establish relationships that are similar to past relationships.

F eeney and Noller (1990), like Hazan and Shaver (1987), examined how
attachment style can function as a predictor for adult romantic love. They found that
people with each attachment style experience and view love differently. Secure people
were more successful in relationships. They tended to emphasize the importance of
openness and closeness in their relationships, while at the same time seeking to retain
their individual identity. They also had the longest lasting love relationships when
compared to the other two attachment styles. Anxious-ambivalent people reflected a
demanding relational style characterized by overinvolvement and idealization of their
partner. They also had the least enduring love relationships. Avoidant people were more
likely to report never having been in love or not being in love at the time of the study; to
indicate low intensity of love experiences; to mistrust others; and were most likely to try
to maintain distance in their relationships.

Simpson (1990) also found differences in people's romantic relationships
according to their attachment style. People with secure attachment styles had more
positive emotions about their relationships and their relationships contained higher levels
of interdependence, trust, commitment, and satisfaction. People with the insecure

attachment styles (anxious-ambivalent and avoidant) had more negative emotions about

20

their relationships. While avoidant people also maintained relationships, they tended to
avoid intimacy. Anxious-ambivalent people reported less trust in their relationships.

Studies examining marital/relationship adjustment and attachment styles were
conducted by Kobak and Hazan (1991) and Senchak and Leonard (1992). There were
similar findings in these two studies. Both studies found a higher proportion of secure
people in their study than was found in other studies. In Senchak and Leonard's study,
over 80% of both husbands and wives were secure. Along this line, Kirkpatrick and
Davis (1994) found that in their study of dating couples that over 74% of men and
women classified themselves as secure. These researchers have suggested that secure
individuals may simply be more likely to marry or become committed to a long-term
relationship than insecure individuals. Kobak and Hazan (1991) suggest an additional
mechanism -- people who were insecure may become secure once they are in a
relationship. It could be that marriage attracts secure people or that marriages transform
people to the secure style.

Kobak and Hazan (1991) examined the issue of attachment in marriage and its
effects on security and accuracy of working models. They found that when husbands and
wives reported relying on their partner, both reported higher levels of relationship
adjustment. Senchak and Leonard (1992) examined attachment styles and marital
adjustment among 322 newlywed couples. They found that secure-secure couples (i.e.,
both wife and husband have a secure attachment style) perceived more intimacy and

evaluated each other more favorably than did any other combination of couples. The

proportions of their couples in the different attachment style combinations were: for same

21

attachment style couples, 70% secure-secure, 0.5% Anxious-ambivalent-Anxious-
ambivalent, and 1.5% avoidant-avoidant; for mixed couples, husband secure and wife
anxious-ambivalent 4%, wife avoidant 7%; for mixed couples, wife secure and husband
anxious-ambivalent 3%, husband avoidant 9%; for mixed couples, husband avoidant and
wife anxious-ambivalent 2%, wife avoidant and husband anxious-ambivalent 1.5%.
Senchak and Leonard stated that attachment style is used as a basis for pairing in marital
relationships because a greater proportion of insecure people (anxious-ambivalent and
avoidant) were paired with secure people. However, because of the greater proportion of
secure people in the population compared to insecure people, both insecure and secure
people would be more likely to marry secure people.

.uu-J o i" -_u i: [at n t- .u , -. Ii‘ru .- ' “A mu 5‘
Overall, the majority of studies on adult attachment (e.g., Brennan & Shaver, 1995; Collins
& Read, 1990; Feeney & Noller, 1990; Hazan & Shaver, 1987, Kirkpatrick & Davis, 1994,
Mikulincer & Nachshon, 1991) have found no sex differences on the measure of adult
attachment styles, in that there were similar proportions of men and women within each
attachment style. Correspondingly, there have been no sex differences found within the
infant attachment literature (Campos, Barrett, Lamb, Goldsmith, & Stemberg, 1983). These
studies on adult and infant have used the three styles of attachment.
On the other hand, sex differences were found with Bartholomew’s four styles of

adult attachment. Bartholomew & Horowitz (1991), using an interview-based attachment
rating, found that women were rated by the interviewer as being more preoccupied than

were men. Men were rated by the interviewer as more dismissing than were women.

22

However, no sex differences were found on any of the four styles of adult attachment when
subjects rated themselves on a 1-7 rating scales. In a comparison of Hazan and Shaver’s
(1987) three styles of adult attachment and Bartholomew’s (1990) four styles of adult
attachment, Brennan and colleagues (1991) found sex differences with Bartholomew's four
styles but not with Hazan and Shaver’s three styles. There were similar proportions of
males and females within each of Hazan and Shaver’s three attachment categories. In
contrast, there were sex differences in all four styles of Bartholomew’s scale. Because of
previous findings (Bartholomew, 1990), Brennan et al. predicted and found a higher
proportion of males in the dismissing style and a higher proportion of females in the fearful
style. However, sex differences were also found in the other two styles, secure and
preoccupied. There were more secure women than men and there were more preoccupied
men than women. These findings were not predicted and Brennan and colleagues stated
that they were unexpected because they were incongruent with gender-role stereotypes.
There have been no findings of sex differences within each attachment style (i.e., the
percentage of men and women are the same) for the three styles of adult attachment and few
sex differences findings with the four styles. Sex differences within each attachment style
is a different issue than is the issue of sex differences interacting with attachment style on
ratings of relationship variables. First of all, when examining relationship variables within
the area of adult attachment, sex difference main effects are expected to occur. From the
gender-role literature, women have been designated as the relationship-oriented sex and
have been called “relationship specialists” (Chodorow, 1978). Research has found that

women are more oriented toward connecting with others and being concerned about others.

23

Relationships tend to be more important to women compared to men (Gilligan, 1982) and
thus they tend to devote more effort to relationship maintenance than men. Because of the
importance of relationships for women, they are more concerned about how decisions
impact their relationships (Gilligan, 1982; Huston, Surra, Fitzgerald, & Cate, 1981). Men,
on the other hand, tend to be more oriented toward individuality (Gilligan, 1982). A similar
distinction is made by Bakan (1966) labeling the female personality cluster as communion
and the male personality cluster as agency. It would be expected that across all attachment
styles, women would have higher ratings for relationship components compared to men.
Attachment researchers would concur that gender-role differences exist in
perceptions of others and of relationships within the framework of attachment. According
to Pietromonaco and Camelley (1994), "Given differences in gender role socialization,
particular attachment experiences will lead to different meaning and differential
psychological consequences for women versus men"(p.64). Pietromonaco and Camelley
(1994) have suggested that gender may be a moderator of individuals’ working models and
their perceptions and behaviors within relationships. Previous research on couples (Collins
& Read, 1990; Kirkpatrick & Davis, 1994; Simpson, 1990) would support this framework.
Pietromonaco and Camelley (1994), along with previous researchers (Collins & Read,
1990; Kirkpatrick & Davis, 1994), have found that feelings about relationships are linked to
both individuals' working models and their gender. The consequences for the insecure
styles (i.e., anxious-ambivalent and avoidant) are different for men and women.
Pietromonaco and Camelley (1994) found that avoidant men and anxious-ambivalent

women had more negative feelings after imagining being in a relationship than their

24

counterparts (i.e., insecure individuals had negative feelings and secure individuals did not).
These two groups have been found to have lower satisfaction in their relationships (Collins
& Read, 1990; Kirkpatrick & Davis, 1994) and have negative perceptions of relationships
(Pietromonaco & Camelley, 1994) compared to women and men in the other styles.

Within the attachment literature, when focusing on relationship variables, anxious-
ambivalent or preoccupied women tend to stand out more than secure or avoidant women in
their perceptions of relationships. Among women, anxious-ambivalent types tend to have
the most negative view about relationships compared to other women (Pietromonaco &
Camelley, 1994). These women have been found to be less satisfied with the relationship
and viewed it more negatively (Collins & Read, 1990). They also perceived their partners
as less trusting and less dependable and were less close to their partner (Collins & Read,
1990). Kirkpatrick and Davis (1994) found that anxious women had lower scores on
satisfaction, viability, and care and higher scores on conflict-ambivalence. They are more
likely to be in relationships with less commitment and trust (Simpson, 1990). Anxious-
arnbivalent women may be less satisfied because they become anxious when thinking about
relationships and this feeling may be exaggerated by their fear and anxiety over not finding
the right partner. Their fear of abandonment adds to their low level of satisfaction. A trait
of being anxious is the fear of being abandoned. Such a trait would make these individuals
uncomfortable thinking about and being in a relationship. Secure women were more
comfortable with closeness and perceived less conflict in their relationship. They also felt
that their partner was more dependable and had a higher level of self-disclosure (Collins &

Read, 1990).

25

Among men, avoidant types tend to have the most negative view about relationships
compared to other men (Pietromonaco & Camelley, 1994). They have been found to have
lower scores on satisfaction, viability, intimacy, care, and passion (Kirkpatrick & Davis,
1994). Kirkpatrick and Davis define satisfaction as involving success, enjoyment,
reciprocity, and esteem. Viability involves trust, respect, and acceptance\tolerance.
Intimacy involves confiding and understanding. Care involves giving the utmost,
championing, and assistance. And passion as involving fascination, sexual intimacy, and
exclusiveness. Avoidant individuals prefer to avoid intimacy and maintain emotional
distance in their relationships. Because of their low levels of intimacy and emotionality
within the relationship, they tend to experience less distress following a break-up (Simpson,
1990). Avoidant men are not relationship-oriented. Anxious-ambivalent men also view
relationships negatively. They tend to be involved in relationships with low trust and low
satisfaction (Simpson, 1990) and feel that their partners are not dependable (Collins &
Read, 1990). Secure men were comfortable with closeness and viewed relationships as
more positive and satisfying. They felt closer to their partner, trusted her more, liked her,
and felt that she was dependable (Collins & Read, 1990).

uni-1 . t- .u in... mm a- :u u. . -:- .. u u- . Secure
lovers describe their most important love experience as especially happy, friendly, and
trusting (Hazan & Shaver, 1987). They can accept and support their partner, which leads
them to be confident in their emotional attachment. Their relationship is stable and
supportive and there is a high level of trust, interdependence, commitment, satisfaction

and positive feeling about the relationship (Collins & Read, 1990; Feeney & Noller,

26

1990; Hazan & Shaver, 1987; Levy & Davis, 1988; Simpson, 1990). There is caring,
intimacy, support, and understanding in the relationship (Mikulincer & Nachshon, 1991).
They describe themselves as easy to get to know, liked by most people, comfortable with
closeness, able to depend on others, and not worried about being abandoned (Hazan &
Shaver, 1987). They also had higher self-esteem, higher self-worth, greater
expressiveness, greater social confidence, lower ratings on neuroticism, higher ratings on
extroversion than insecure people and higher levels of appropriate self-disclosure to
others (Collins & Read, 1990; Feeney & Noller, 1990; Hazan ‘& Shaver, 1987; Kobak &
Sceery, 1988; Mikulincer & Nachshon, 1991; Simpson, 1990; Shaver & Brennan, 1992).
Secure people had the longest relationship durations compared to avoidant and anxious-
arnbivalent individuals (Feeney & Noller, 1990; Hazan & Shaver, 1987).

Anxious-ambivalent lovers experience love as involving obsessions, desire for
reciprocation and union, emotional highs and lows, extreme sexual attraction and
jealousy, and a high level of anxiety (Collins & Read, 1990; Feeney & Noller, 1990;
Hazan & Shaver, 1987; Kobak & Sceery, 1988). There is clinging, neediness, emotional
instability, and ambivalent feelings within the relationship (Hazan & Shaver, 1987;
Mikulincer & Nachshon, 1991). There is considerable ambivalence toward their
romantic partners (Simpson, 1990). Within relationships, they reported less satisfaction
and more conflict and ambivalence and having difficulty maintaining a relationship,
which leads to a high rate of relationship dissolution (Collins & Read, 1990; Kirkpatrick
& Davis, 1994; Feeney & Noller, 1990; Levy & Davis, 1988; Simpson, 1990). They

need to be in a relationship and they do not care with whom because their primary goal is

27

emotional security (N ewcomb, 1981). Anxious-ambivalent people, compared to other
groups, had lower self-esteem, scored higher on neuroticism, and self-disclosed
inappropriately compared to secure people (F eeney & Noller, 1990; Mikulincer &
Nachshon, 1991; Shaver & Brennan, 1992). They were subject to anxiety and fear and
worried about being abandoned and unloved but they were also comfortable with being
close to others (Hazan & Shaver, 1987). Anxious-ambivalent people had the shortest
relationship durations compared to secure and avoidant individuals (Feeney & Noller,
1990; Hazan & Shaver, 1987).

Avoidant lovers were characterized by fear of intimacy, emotional highs and
lows, jealousy, self-reliance, a tendency to hide feelings of insecurity, and emotional
distancing between themselves and their partner (Collins & Read, 1990; Feeney &
Noller, 1990; Hazan & Shaver, 1987; Mikulincer & Nachshon, 1991). They reported a
lower level of trust, interdependence, commitment, intimacy, and satisfaction in their
relationships (Feeney & Noller, 1990; Levy & Davis, 1988). They reported being
sensitive to rejection (Simpson, 1990). They scored lower in self-esteem (F eeney &
Noller, 1990), high on neuroticism (Shaver & Brennan, 1992) when compared to secure
individuals, and showed a fear of intimacy and closeness (Hazan & Shaver, 1987). They
avoided self-disclosing to others and were uncomfortable when others self-disclosed
(Mikulincer & Nachshon, 1991).

So far, discussion has been based on attachment style and romantic relationships.
However, romantic relationships are only a subset of adult close relationships. Other

important subsets include friendships, exist. Friendship is an important adult close

28

relationship which should be examined within attachment style to further our
understanding of attachment style and close adult relationships. Although there has been
a plethora of research on the relationships between attachment styles and romantic
relationships, there is a lack of research on friendship and attachment styles. Friendship
is an important adult relationship and can be as impactfirl as romantic relationships on
mental and physical well-being. According to O' Connor (1992) "...theorists who have
been concerned with needs, love, and attachment styles have shown very little interest
indeed in the friendship relationship per se" (p.4).

E . l l .

Brief Hi stem Qf Researeh en En'endship. According to Blieszner and Adams
(1992), research on adult friendship is relatively new, extending back only 25 years.
Research on childhood and adolescent friendship had been well established by the time
research on adult friendship was initiated. Thus, it is understandable that research
theories and methodologies in the adult fiiendship area were established by research
within the childhood and adolescent fiiendship areas (Blieszner & Adams, 1992).

In this society, fiiendships are considered to be voluntary relationships because
there are no formal structures to them and they are not formally celebrated (Blieszner &
Adams, 1992). However, norms do exist within fiiendship. Norms predict whom we
choose as our friends, how we treat our friends, and what is acceptable to expect of our
fiiends. Research has found that people choose friends that are of the same age, sex, race,
religion, geographic area, and social status levels (Blau, 1973; Booth & Hess, 1974) as

themselves. Studies have found that similarity of interests, attitudes, skills, and economic

29

status seem to be the basis for interpersonal attraction and they influence fiiendship
relations and functions (e.g., Berscheid & Walster, 1969; Byme, Clore, & Worche, 1966;
Weiss & Lowenthal, 1975). Usually friendships are viewed as intimate, at least to some
degree, and egalitarian (Allan & Adams, 1989; Reohr, 1991; Thomas, 1987).

There are different processes that exist within adult fiiendship. Two important
processes are relevant to this study which are cognitive processes and affective processes.
According to Blieszner and Adams (1992), cognitive processes are internal thoughts that
a person has about herself or himself, the friend, and the fiiendship. Some of these would
be very similar to attachment theory's mental models. These thoughts also concern the
evaluations of many things: the performance of the friend role from the perspective of
both the self and the partner; the stability of the friendship; the behaviors and intentions,
or needs of self and the partner; and other important qualities such as values, needs, and
personality traits. Affective processes are emotional reactions that individuals have
within a fiiendship (Blieszner & Adams, 1992). They encompass both positive and
negative emotional reactions to friends and fiiendships. Positive emotions consist of
empathy, trust, loyalty, satisfaction, and commitment while negative emotions include
indifference, anger, hostility, and jealousy.

Research on sex differences in friendship has found that women choose to affiliate
in small groups or dyads whereas men tend to affiliate in large groups (Bell, 1981; Rubin,
1985). Women spend much of their time with each other talking and sharing personal
problems, feelings, and ideas while men have more activity-oriented friendship (Rubin,

1985). Women tend to be closer to their friends, spend more time with them, have more

30

frequent thoughts about friends, and have a high intimacy level within these friendships
(Jones, 1991; Wong & Czikszentmihalyi, 1991; Wright, 1982). Women value self-
disclosure and understanding in friendship more than do men (Aries & Johnson, 1983;
Helgeson, Shaver, & Dyer, 1987; Monsour, 1992). For women, friendship includes a
loving, sharing, understanding relationship. On the other hand, men tend to have more
friends in general, and have a low intimacy level with these fiiendships. Men also define
friendship in terms of trust and intimacy, but emphasize instrumental activity more than
do women (Swain, 1989). Buhrke and Fuqua (1987) comparing same-sex fiiendship of
women and men found that women had more contact under stress; were closer and more
satisfied; and perceived that they knew their friends better and were better known by their
fiiends.

WW. Research shows that there are general characteristics
of fiiends upon which there is a good deal of agreement. A friend is someone with whom
one is intimate; to whom one self-discloses, trusts, gives and receives social support;
shares activities, interests, and information; accepts who we are; and values, respects, and
shares time with us (Davis & Todd, 1985; Duck, 1991; Jones, 1991; Rawlins, 1992).
Friends keep us emotionally stable and provide assistance and physical and psychological
support (e. g., Duck, 1983). Friends share activities and interests and are people with
whom an individual spends time (Howes, 1981). Friends help each other (Wright, 1984)
and enhance each other's self-esteem, self-confidence, and self-competence. Among
women fiiends, several functions (e. g., intimacy, assistance, power) have been found to

exist in their friendships (Candy, Troll, & Levy, 1981).

31

We. There exist a variety of friendship types: acquaintances,

casual friends, long-distance friends, cross-generational friends, cross-sex fiiends, same-
sex fiiends, close fiiends, and best fiiends (Hays, 1988). The types of fiiendships can be
differentiated by the level of closeness between the two participants. Acquaintances may
have the lowest level of closeness while best fiiends will likely have the highest level of
closeness. La Gaipa (1977) identified five different levels of fiiendship: casual
acquaintance, social acquaintance, close friend, very close fiiend, and best fiiend. These
levels extend from having great distance between the two participants to having the two
people be very close. Davis and Todd (1982) found that differences exist between
friendship in general and best friendship. They found that there were higher ratings for
intimacy, support, enjoyment, understanding, and confiding for best fiiends compared to
same-sex close friends.

Wendship. Friends have important roles or functions in people's lives
(Ginsberg, Gottrnan, & Parker, 1986). Friendships are found to occur throughout the
lifespan and are considered by most people to have a central role in adapting to life (Bell,
1981; Pogrebin, 1987; Rubin, 1985; Shulrnan, Elicker, & Sroufe, 1994). Candy and
colleagues (1981) stated that fiiends serve as a source of continuity for people as they go
through adulthood. Friendship is one relationship that offers companionship as well as
many other valuable functions such as emotional and physical support (Duck, 1983;
Rubin, 1985). It is believed that friends provide a close, secure, and accepting
environment where a person can truly be himself or herself (Duck, 1984). The behaviors

(e. g., closeness, caring) that best fiiends provide each other are similar to attachment

32

behaviors in an attachment bond and thus best friends may be attachment figures for each
other. They provide a secure base and a safe haven where each partner can feel safe and
protected.

Cempenmtsflrismdships. The components of fiiendship thought to be
important by close relationships researchers are self-disclosure, intimacy, trust,
satisfaction, and social support as evidenced by their examination in many close
relationship studies (e. g., Davis & Todd, 1982; Hazan & Shaver, 1987; Helgeson et al.,
1987; Kirkpatrick & Davis, 1994; Jones, 1991; Jourard, 1974; Rawlins, 1992; Shea &
Adams, 1984). It is important to point out that these components are not distinct
concepts. Some researchers have used some of the concepts interchangeably and some
researchers view some of the concepts as sub-components of others. According to
Wright (1988), "Virtually all close friendships involve shared interests and activities,
various kinds of intimacy including self-disclosures and the sharing of confidences,
emotional support, small talk, shop-talk and exchanges of tangible favours..." (p. 370).

Self-disclosure is a component of intimacy (Wright, 1988) and is an important
aspect of friendship, especially intimate fiiendship (e.g., Monsour, 1992; Reohr, 1991).
Self-disclosing has been found to contribute to the development and maintenance of close
relationships. Jourard (1974) stated that "mutual self-disclosure is what differentiates
personal relationships of love and fiiendship from formal role relations because
participants seek to make their subjective worlds known to one another, in an ongoing
dialogue" (p. 222). Intimacy has been widely noted as a component of friendship (Hays,

1988). According to Hatfield (1984), intimacy is “A process by which people attempt to

33

get close to another; to explore similarities and differences in the ways they both think,
feel, and behave” (p. 208). Reis and Shaver (1988) stated that intimacy is important to
any close interpersonal relationships, “Intimacy has been shown to be important to
people’s health and well-being, and to be a worthy goal in its own right. Its repeated
appearance in the literature on friendships, marriage, lead us to believe that it will
remain an essential focus of theories in interpersonal relations” (p. 389).

Trust has been considered to be an important quality in friendship (e.g., Sullivan,
1953). Holmes and Rempel (1989) suggest that one’s most important and satisfying
relationships are characterized by trust. Trust is the ability to rely on the relationship
partner and to be oneself without fear of rejection or betrayal (Jones & Burdette, 1994).
Trust is important in friendship because it means that the participants are able to keep
confidences, are trustworthy, and able to trust others (Argyle & Henderson, 1984). It is
considered to be an essential element underlying the development of self-disclosure and
intimacy in friendships (Buhrrnester & F urman, 1986). Satisfaction can be defined as a
feeling that the relationship is rewarding and satisfying (Parker & De Vries, 1993).
Satisfaction has been found to have an influence on the stability of the friendship (Berg,
1984) and to increase the overt enjoyment of the relationship. Social support can be
defined as various ways of interacting within a relationship to improve coping, esteem,
belonging, and competence (Gottlieb, 1994). Social support can lead individuals to
believe that they are cared for and loved and held in esteem and value (Cobb, 1976).
Hays (1989) found that close fiiends provided more emotional support and information

than did casual fiiends.

34

Jones (1991) examined sex differences in fiiendship satisfaction. She was
interested in investigating sex differences in the conceptions of fiiendship and
relationships. Three core provisions of friendships were measured: trust in friends,
orientation to assistance, and attributes of friends associated with companionship. It was
proposed that these concepts are similar to attachment theory's working models of self
and others in relationships and that these models contribute to satisfaction in a
relationship. It was found that friendship satisfaction was associated with self-disclosure,
trust and fiiendship enjoyment, for both males and females.

Eemflgfiemflefinendshrps. Women throughout history have always had closer
and more intimate fiiendships with each other than have men, and today it is still
acceptable for women to be emotionally and physical expressive to one another (Bell,
1981). Women see their friendship with other women as a close, supportive, highly
secure, highly committed, revealing, and meaningfully intimate, compared to how men
view their friendships with other men (Caldwell & Peplau, 1982; Dickens & Perlrnan,
1981; Reis, 1984; Reisman, 1981, 1990). Women's fiiendship research reveals emotional
sharing (Caldwell & Peplau, 1982), affective generosity (Fox, Gibbs, & Auerbach, 1985),
positive affect (Ginsberg, et a1. 1986), intimacy (Fischer & Narus, 1981), acceptance
(Rose, 1985), interdependence and satisfaction (Wright, 1982). Women tend to find more
comfort and support with a same-sex friend even while in a romantic relationship (Rubin,
1985). Limb (1989) in her research found that some female fiiendships are life long.

MaleMaleEnendships. Bell (1981) stated that men spend less time with women

than with other men and their fiiendships are outside of the family. Men report doing

35

more activities with each other men than with women (Caldwell & Peplau, 1982; Fox et
al., 1985; Rubin, 1985; Wright, 1982) which is very similar to women’s friendships. In
that there are more activities within their same-sex than with the other-sex. However, the
difference in time spent with same-sex friends is that men usually participate in some sort
of physical activity while women usually participate in nonphysical activity (e. g.,
talking). Researchers have found that males confide more in their best female fiiends
than their best male fiiends (Komarovsky, 1974). Wheeler, Reis, and Nezlek (1983)
observed that males experienced less closeness with their friends than females did.
Wheeler and colleagues (1983) and Hays (1989) found that men and women prefer
intimate partners who are women and Rubin (1985) found that men's closest friends are
women. Both sexes find it easier to disclose to their female friends than to their male
friends (Reis, Senchak, & Solomon, 1985). Men tend not to be very emotionally close or
intimate with their male friends because they cannot let their friends perceive them to be
vulnerable. Men also exhibit less intimacy and affection in their fiiendships than women
do (Fischer & Narus, 1981; Williams, 1985) and men rarely tell their male fiiend that
they like him.

SemmamngnendshmReseareh. To summarize, fiiendships play an important
role in people's lives. Many people find their friendships very enjoyable and satisfying.
Friends offer each other both psychological and physical support. People trust and
disclose personal information to their friends. Most people, especially women, have very

intimate and close fiiendships even while in a romantic relationship.

36

Despite the importance of friendship for adults, friendship has been neglected by
adult attachment researchers. Why has friendship been neglected in research on adult
attachment? Hazan and Shaver's (1987) research on adult attachment and adult romantic
love was the flagship for the extension of attachment into the adult relationship area.
Since this research, most adult attachment researchers have replicated and extended their
work and thus concentrated in the area of romantic love. This is understandable because
adult love relationships share many variables with infant-caregiver relationships. Many
individuals would consider the adult romantic relationship between two peOpIe to be
exclusive, unique, intimate, and close, which is very similar to the infant-caregiver
relationship. However, because of this narrow focus on romantic relationships,
researchers have neglected to explore other close adult relationship that may be
considered by individuals to be just as important, close, and intimate as adult romantic
love specifically, here, adult best friendship. According to Davis and Todd (1982) “To
not have a friend is to miss something vital for full-fledged human development...The
personal relationship of fiiendship is seen as providing the context within a number of
basic human needs can be met” (p.21).

Although research in adult attachment has been concentrated in the area of adult
love, Hazan and Shaver (1994), when proposing that attachment can be an organizational
framework for research on close relationships that are focused on romantic love, stated
that other close relationships can function within this new framework of attachment.
Bowlby's attachment theory is not specific to sexual or love relationships which means

that it can be extended to other close relationships such as friendship. Attachment theory

37

clearly has the potential to increase our understanding of personal relationships of many
kinds (N oller & F eeney, 1994). A logical next step would be to examine how attachment
style affects people's fiiendship relationships. Extending adult attachment research to
friendship will increase our understanding of the relationships between attachment styles
and close adult relationships. In addition to examining attachment styles and fi'iendships,
it is important to compare romantic partner and best friend as this approach can further
our understanding of the relationship between attachment styles and close adult
relationships.

Before comparing these two relationships, it is best to define what is a best friend
and what is a romantic partner. A best fiiend is someone who provides more emotional
and informal support than do casual friends (Hays, 1989). Usually the label 'best friend'
signifies a level of attachment, of intimacy, of commitment, of sharing, that transcends all
other friendships (e.g., Davis & Todd, 1982; Rubin, 1985). Best fiiendship is "... a
relationship not bounded by the usual limits that constrain friends" (Rubin, 1985, p. 178).
The best friend can be of the same-sex or of the other-sex. On the other hand, a romantic
partner is someone, usually of the other-sex, with whom someone has a relationship that
includes sexual dimensions. This person can be a significant other, a boy/girlfiiend, or a
husband/wife.

Earlier research examining the relationship between romantic relationships and
friendship has been within the areas of liking and loving (e.g., Rubin, 1973). Rubin

(1970, 1973) was one of the first researchers to conceptualize liking and loving as

38

different entities with overlapping attributes. To test this conceptualization, Rubin (1970)
created two scales, the Liking Scale and the Love Scale. Four main components were
derived from the Love Scale -- Needing, Caring, Trust, and Tolerance. The Liking Scale
measured predominantly the components of Respect. Rubin found that men and women
liked and loved their romantic partner more than their fiiends. However, only loving
scores were significantly higher for romantic partner than friends. Rubin (1970) and
other researchers (Stemberg & Barnes, 1985; Stemberg & Grajek, 1984) found that liking
for one's partner was positively correlated with love for the partner.

In 1982, Davis and Todd set out to compare fiiendship and love relationships.
They differentiated between liking and loving and they also conceptualized other
important relationship characteristics which exist for love relationships and fiiendships.
They agreed with Rubin (1973) and others that love relationships and fiiendships were
conceptually distinct relationships but Shared many attributes. In their work, they created
the Relationship Rating Form (RRF) which was a scale to measure both love
relationships and fiiendships. The RRF was created to measure essential attributes
important to close and intimate adult relationships. They began their series of studies
with eighteen conceptually distinct aspects of personal relationships that were important
both to fiiendship and love relationships. After conducting confirmatory factor analyses,
the RRF ended up with six general scales of relationship characteristics which are
described and discussed below. Davis and colleagues (1982, 1985) concluded that the
RF was a reliable and valid scale of close relationships. According to Davis and Todd

(1982), the RF was constructed with the cultural expectations definitions of fiiendships

39

and love relationships. In summary, Davis and colleagues have presented empirical
evidence that friendships and love relationships are conceptually and empirically distinct
relationships which share several important relationship characteristics.

. Romantic

 

relationships and friendships are considered to be both important close relationships in
adulthood for many individuals. As discussed above, they tend to share many important
relationship variables. First, besides the fact that these relationships play important roles
in life, there are a number of characteristics that exist in any good relationship (romantic
or friendship). The parties involved share experiences, interests, activities, and values
and describe the other as someone whom they can relax with, receive support from, be
intimate with, self-disclose to, trust, and share with (Bell, 1981; Jones, 1991). Second,
both persons state that there are feelings of love, affection, warmth, acceptance,
commitment, loyalty, and comfortableness (Rubin, 1985). Third, both types of
relationships have high levels of viability, spontaneity, and intimacy (Davis & Todd,
1982) and the individuals involved enjoy the other person's company. Fourth, both types
of relationships are perceived to be voluntary, enjoyable, meaningful, and to involve
caring and mutual liking. Fifth, most close relationships are more likely to be of a
communal nature where the two people involved do not exploit each other (Clark &
Mills, 1979). Sixth, although most studies have found the best friend to be a different
person than a romantic other for many people, some studies have found that the close or
best friend is also the romantic partner (Bell, 1981). This tends to be more the case for

married individuals compared to single individuals. Seventh, these two relationships are

40

considered to be very close and intimate and thus would be considered to be socially
supportive relationships and would be characterized by feelings of attachment
(Henderson, Byme, & Duncan-Jones, 1981). Hazan and Shaver (1994) stated that close-
bond relationships, which would include romantic relationships and best fiiendships,
would involve an attachment bond between the two participants. And finally, both types
of relationships usually occur within the same socioeconomic level, culture, and race, but
of course, there are always exceptions to this general pattern.

Many similarities do exist for these two relationships because they are close,
intimate, and voluntary. However, because of the different and sometimes separate roles
and functions that each relationship has, different expectations and perceptions exist

between the two relationships.

 

romantic relationships and friendship are both close, primary relationships, each
relationship satisfies distinct needs and they are not alternatives for each other (Bulrner,
1987). First, the most distinct difference between these two relationships is the sexual
aspect. This sexual aspect distinguishes what is a romantic relationship from what is a
friendship. This sexual aspect has been named "Passion" by Davis and Todd (1982) in
their RRF scale. Passion contains dimensions of fascination, sexual intimacy, and
exclusiveness. Davis and Todd (1982) found that between lovers, there were higher
levels of passionate aspects of love such as fascination, exclusiveness, and sexual desire
compared to best friendships. Even though best friendship is a close and intimate

relationship, it is usually nonsexual and does not receive the emotional intensities and

41

emotional expressions that accompanies sexual desire. Second, while love relationships
tend to have higher ratings of passion, mutual love, and support, friendships tend to have
higher ratings of stability (Davis & Todd, 1982). Because most individuals have had
friends from a very young age, they have more experience with friendships than with
romantic relationships, which may explain the stability aspect. In addition, the
individuals most often studied are young adults (i.e. college students) and this group
tends to not have had many romantic relationships. Third, according to some researchers
of romantic love (e.g., Davis & Todd, 1982), the lover will or would do almost anything
for the other such as "giving the utmost" or "being a champion for the other." Davis and
Todd (1982) named this aspect of romantic support "Care." Fourth, some researchers

(e. g., Davis & Todd, 1982) have found that fiiendship is based more on reciprocal trust,
respect, and confiding while Other characteristics such as caring have a more important
role in romantic love (Steck, Levitan, McLane, & Kelley, 1979).

And finally, because of the early relationship with the mother, intimacy and
relationship issues may be perceived to be more related to the female figure in the
relationship (Rubin, 1985). Both boys and girls learn to derive intimacy from their
mother before any other person. As adults, both men and women turn to a woman for
intimacy in their social milieu (Wheeler et al., 1983). Thus, individuals may relate
intimacy with women. Some researchers have stated that men can derive all their
intimacy needs from their romantic partner, who most likely would be a woman (e.g.,
Rubin, 1985). Some women, however, may not receive all their intimacy needs from

their romantic partner, who is most likely a man, and must attain intimacy through a

42

different source, perhaps a same-sex best friend or a female relative. Women firlfill
certain needs such as understanding, acceptance, and reciprocal sharing of the self and
experiences with their same-sex friends which their romantic partners may not or cannot
provide or satisfy (O'Connor, 1992; Rubin, 1985).

WWW. Thus far, many important relationship variables
for romantic relationships and best fiiendships have been discussed. Most of these
overlap greatly in their importance with these two relationship types. Others, such as
sexual intimacy, distinguish one type of relationship from the other. Many of the
variables are measured in Davis and Todd's (1982) RF scale. The RF scale measures
six general scales of relationship characteristics important to both romantic relationships
and friendships. The six scales are Viability, Intimacy, Glebalfiatisfaetien, Care,
Bassien, and W. The first four components were chosen because of
their generality and importance to both friendships and romantic relationships. The last
two components were chosen because they play a major role in romantic relationships
and thus ratings for them should be different between romantic partners and best friends.
This scale was used in the present study because it has been found to be a reliable and

valid measure of components of close adult relationships. (See Table l)

43

 

 

Iahlel
ouorn‘u, o o ‘ i‘ you is non D: ' 5: mo 1' .1. in {.u- om
Components Description
(Characteristics included in each component)
Viability This component deals with trust, respect, and acceptance of the
other’s fault.
(Trust, Respect, and Acceptance\Tolerance)
Intimacy This component deals with confiding and mutual understanding.
(Confiding and Understanding)
Global Satisfaction This component reflects the degree to which the relationship
has been a success, meets one’s needs, is a source of enjoyment,
enhances one’s sense of worth, and is reciprocated.
(Success, Enjoyment, Reciprocity, and Esteem)
Care This component deals with a willingness to give one’s utmost
for the partner, to champion the partner’s interests, and to provide
practical and emotional support.
(Giving the Utmost, Championing, and Assistance)
Passion This component deals with fascination, exclusiveness, and
sexual desire.
(Fascination, Sexual Intimacy, and Exclusiveness)
Conflict- This component deals with the degree to which the relationship is
Ambivalence marked by tension, conflict, feelings of uncertainty and feelings of

being trapped.
(Conflict and Ambivalence)

 

Note: Description taken from Kirkpatrick and Davis (1994) and Levy and Davis (1988)

44

E . B l S l' E l E S 1

Many research studies have examined the relationship between attachment styles
and romantic relationships on various relationship characteristics (e. g., Bartholomew &
Horowitz, 1991; Brennan & Shaver, 1990, 1995; Collins & Read, 1990; Feeney & Noller,
1990; Hazan & Shaver, 1987 ; Simpson, 1990). Of all these studies, two stand out in their
relevance to the present study -- Levy and Davis (1988) and Kirkpatrick and Davis
(1994). The present study examined the relationships between attachment styles and
close relationships on various relationship components, specifically those measured in the
RRF (Davis & Todd, 1982, 1985). Levy and Davis (1988) and Kirkpatrick and Davis
(1994) examined attachment styles and romantic relationships on similar relationship
components. The present study examined the association between three different types of
close relationships (e.g., romantic relationships, best fiiendships: same-sex and other-sex)
and attachment styles on various relationship characteristics. The RRF was chosen
because of its ability to assess relationship characteristics that were important to both
romantic relationships and friendships. As discussed earlier, the RRF contains six
relationship clusters: Viability, Intimacy, filehaLSatisfaetign, Care, Passion, and
W. Davis and Todd (1982) have used the RF to examine and
compare how romantic partners and best friends are similar to and different from each
other on various relationship components. Davis and Todd's (1982) study was discussed
in detailed earlier.

Levy and Davis (1988) compared attachment styles and love styles on various

relationship characteristics. They found that certain love styles (e.g., eros and agape)

45

were related to certain relationship characteristics (e. g., intimacy, viability). They found
that the secure style was positively correlated with intimacy, passion, and satisfaction.
The avoidant style was negatively correlated with viability, intimacy, care, commitment,
and satisfaction and positively correlated with conflict. The anxious-ambivalent style
was negatively correlated with viability, intimacy, care, and satisfaction and positively
correlated with conflict. Note that avoidant and anxious-ambivalent styles have very
similar findings. Levy and Davis reported that their findings are similar to Hazan and
Shaver's (1987) study in that each attachment style was associated as expected to certain
relationship components (e. g., the secure style was associated positively with viability,
intimacy, and satisfaction).

Kirkpatrick and Davis (1994) examined attachment styles, sex, and relationship
stability over a three year period. They used the RRF to assess relationship
characteristics. They found that secure men had the highest ratings on satisfaction,
viability, and intimacy and the lowest rating on conflict-ambivalence. Secure women had
the highest ratings on commitment, satisfaction, viability, intimacy, and care and the
lowest rating on conflict-ambivalence. Avoidant men had the lowest ratings on
commitment, satisfaction, viability, intimacy, passion, care, and the highest rating on
conflict-ambivalence. Avoidant women had the lowest ratings on intimacy, and passion
and the highest rating on conflict-ambivalence. Anxious-ambivalent men had the highest
ratings on passion, care, and commitment. Anxious-ambivalent women had the highest
rating on passion and the lowest ratings on care, commitment, satisfaction, and viability.

Kirkpatrick and Davis (1994) reported that their findings are similar to other adult

46

attachment research (e. g., Collins & Read, 1990; Simpson, 1990) in that individuals in
different attachment styles differ in how they rate certain relationship components.

.1: .0 .0 ci‘ 4 él.-,. urn \‘ i. 'ci‘ "O raw 0 =‘ ru an
RomantieEartner‘Z

Ainsworth and other attachment researchers have found that relationships with
caregivers during the early years have effects on how individuals perceive themselves,
others, and relationships. These perceptions become mental models which will influence
future relationships. This early relationship also predisposes individuals to a specific
attachment style. These attachment styles were discussed at length earlier. Research
suggests that adult attachment styles affect the experience of romantic love in adulthood
(Hazan & Shaver, 1987; Shaver et al., 1988).

From the above discussion, it should be clear that because of their close and
intimate nature, romantic relationships and best fiiendships should share many positive
characteristics because they play important roles in an individual's life. It will be
predicted that each style's mental models of others and relationships will be similar for
romantic partner and best friend because mental models should be similar across close
relationships. However, because of their different functions, romantic relationships and
fiiendships should differ from each other in specific ways, such as sexual intimacy, and
thus are predicted to be rated differently on these relationship components.

Secure individuals have positive mental models of others and relationships.
Research (e.g., Collins & Read, 1990; Hazan & Shaver, 1987) has found that secure

individuals report their love relationships to be happy, trusting, caring, supportive,

47

satisfying, friendly, and intimate. Thus, on general positive relationship components of
viability and global satisfaction, it will be predicted that secure individuals will have
similar ratings for their romantic partner and their best fiiend. However, on relationship
components more related to love relationships (i.e., intimacy, passion, and conflict-
ambivalence), it will be predicted that secure individuals will have different ratings for
romantic partner and their best friend. The variable Care would normally be predicted to
be rated similar for the two relationship types. However, Davis and Todd's
conceptualization of Care is more within the realm of romantic relationships. Secure
individuals will be predicted to have a higher rating for their romantic partner compared
to their best fiiend.

Anxious-ambivalent individuals have been found to describe their love experience
as filled with obsession, emotional highs and lows, and extreme sexual attraction (e. g.,
Collins & Read, 1990; Feeney & Noller, 1990). They also desire union with another
while behaving in a clingy and needy manner. They report less satisfaction and more
conflict within their relationships (e.g., Kirkpatrick & Davis, 1994; Levy & Davis, 1988).
Because of their obsession and desire to be in a relationship, most likely a romantic
relationship, they will be predicted to rate their romantic partner higher than their best
fiiend on a variety of relationship components -- viability, intimacy, care, passion, and
conflict-ambivalence. Because of their obsessive nature, emotional extremes, and lower
satisfaction level, it will be predicted that they will rate their best friend higher on

satisfaction compared to their romantic partner.

48

Avoidant individuals have been found to fear intimacy and closeness within their
love relationships (e.g., F eeney & Noller, 1990). They report lower levels of satisfaction,
trust, and intimacy in their relationship compared to both secure and anxious-ambivalent
individuals. It is proposed that these negative relationship ratings are more related to
romantic partners than to best friends. Best friends will be predicted to have higher
ratings on viability, intimacy, and satisfaction. However, because of their relevancy to
love relationships, passion and conflict-ambivalence will be predicted to have higher
ratings for romantic partner compared to best friend.

Each style may have different definitions for their best friendship and their
romantic relationship which may cause the differences in their ratings of certain
relationship components. As discussed earlier, previous research has found that men and
women do define a best fiiend as being different from a romantic partner. And the sex of
the best fiiend also plays a role in what individuals receive and expect from that
relationship. Same-sex friendships for women are more face-to-face (women talk more to
their friends) while same-sex friends for men are more side-by-side (men do more things
with their fiiends) (Wright, 1982). Other-sex friends bring different things to men and
women, however both men and women stated that they are more intimate, more self-
disclosing, and feel closer with a female best friend compared to a male best friend
(Wheeler et al., 1983). This also may be derived from the early attachment relationships
that both males and females have with a female caregiver.

In adulthood, representations of fiiendships and love relationships would be

expected to be at least as important as representations of family relationships in defining a

49

current "attachment style" (Bartholomew, 1990). Bowlby stated that attachment implies
an enduring "affectional bond". This bond can be with friends, especially best fiiends.
Hazan and Shaver (1987) suggested that romantic love is an attachment process,
experienced somewhat differently by different people because of variations in their
attachment histories.

It was proposed that best friendship can also be considered to be an attachment
process. According to Hazan and Shaver (1994), “If adult peers begin to serve similar
functions and satisfy the same needs for emotional support and security for which parents
are primarily responsible during infancy and childhood, then at some point attachment
will be transferred from parents to peers” (p.8). Friendship may be experienced
differently by individuals with different attachment styles. Thus far, most researchers of
adult attachment have limited their expansion of attachment theory to adult romantic
relationships and most have not examined other significant figures in people's lives such
as best friends or close friends.

Most studies on attachment have examined the relationships between individuals
and their romantic partners, or their parents, or both (e.g., Hazan & Shaver, 1987; F eeney
& Noller, 1990; Keelan, Dion, & Dion, 1994; Simpson, 1990). Attachment theory clearly
has the potential to increase our understanding of personal relationships of many kinds
(Noller & F eeney, 1994). Bowlby's attachment theory is not specific to sexual or love
relationships which means that it can be extended to other close relationships such as
friendship. This study attempts to explore the possibility that the different attachment

styles identified by Ainsworth and her colleagues, resulting from the parent-child

50

relationship, may play a role in the determinants of adult close relationships which
include best fiiendship and romantic relationships. In addition, individuals within
different attachment styles may differ in their perceptions of same-sex and other-sex best
friends.
mum

In adulthood, friendships and romantic relationships, along with family
relationships, were expected to play a major role in individuals' adult attachment
experience (Bartholomew, 1990). Hazan and Shaver (1987) were the first researchers to
extend attachment theory to adult relationships. Research suggests that adult attachment
style, an orientation to relationships determined primarily in childhood relationships with
parents (Ainsworth et al., 1978; Bowlby, 1973, 1980, 1982), affects the experience of
romantic love in adulthood (Hazan & Shaver, 1987; Shaver et al., 1988). It has been
found that individuals possessing different attachment styles have different perceptions
concerning romantic relationships (e.g., Hazan & Shaver, 1987). To better understand
how attachment styles can affect other important close adult relationships, it would be
best to compare a relationship type that is very similar in terms of closeness and share
many other characteristics of romantic relationship yet be different on a few important
characteristics. Best friendship was chosen as an adult relationship that met these criteria.
This friendship type is believed to be as important and as close to that of romantic
relationships to many adults. The best friendship is likely to have the same levels of
intimacy, understanding, and commitment as a romantic relationship but differ on traits

such as sexual attraction and desire. Research comparing best friend and romantic

51

partner within the area of adult attachment would further our understanding of adult
attachment and close relationships.

There were three purposes to the present study. The first purpose was to examine
the relationship between attachment style and romantic relationship and to replicate
previous findings in this area. The second purpose was to examine the relationship
between adult attachment style and best fiiends (i.e., same-sex and other-sex). This
would extend the previous findings of attachment style and romantic relationships to the
area of friendship. The third purpose was to examine the effects of attachment style, sex,
and different types of relationships (i.e., romantic relationship, other-sex and same-sex
best fiiendships) on ratings of relationship components. Because of the commonalities
between romantic partners and best friends, previous findings on attachment styles and
romantic relationships are expected to be replicated. Bowlby stated that attachment
implies an enduring affectional bond. If so, this could be interpreted to mean that friends,
especially best friends, can have attachment bonds with each other. Hazan and Shaver
(1987) suggested that romantic love is an attachment process, experienced somewhat
differently by different people because of variations in their attachment histories. Best
friendship can also be considered as an attachment process and should also be
experienced differently by individuals with different attachment styles. So far, most
researchers of adult attachment have focused mostly on adult romantic relationships (e. g.,
Hazan & Shaver, 1987; F eeney & Noller, 1990; Keelan et al., 1994; Simpson, 1990).

Research has not examined other significant figures in people's lives, such as best friends

52

or close friends, and thus our understanding of attachment and close relationships in
adulthood has been limited.

The relationship components of interest were Viability, Intimacy, Global
Satisfaction, and Care. These variables were chosen because they were components most
identified as important variables in close relationships (e.g., Davis & Todd, 1982; Reis &
Shaver, 1988). Other relationship components, Passion and Conflict-Ambivalence, will
also be examined. All these components are described in Table 1. Most adult romantic
attachment research has examined similar relationship components (e.g., Hazan &
Shaver, 1987; Collins & Read, 1990; F eeney & Noller, 1990). Research in fiiendship has
also examined similar relationship components and has confirmed their important status
in close relationships (e.g., Cohen & McKay, 1984; Morgan, 1976).

Most of the hypotheses predict that secure individuals will have higher ratings on
positive relationship components compared with anxious-ambivalent and avoidant
individuals. Individuals in the secure attachment style are psychologically healthier and
have a more positive perception of relationships and others. They have also been found
to have higher levels of certain relationship components such as trust, intimacy, and
satisfaction in close relationships compared to their less healthy counterparts, anxious-
ambivalent and avoidant individuals. Feeney, Noller, and Hanrahan (1994) found in their
research that there were two major clusters of individuals, secure and insecure. However,
the insecure cluster could be broken down into smaller clusters. Noller and Feeney
(1994) commented that there may be only one way to be secure and a few ways to be

insecure. Thus these hypotheses contained an element of the obvious -- that secure

53

people score higher on positive relationship traits and lower on negative relationship
traits than do the other two groups. Not so obvious is the distinction between the two
insecure groups and that is what some of these hypotheses were developed to test.

The hypotheses in this study were based on the above research findings on
attachment styles and romantic relationships. The present study seeks not only to
replicate the studies discussed above but also to extend the area of attachment theory into
the friendship arena. So far, research on attachment styles has only examined romantic
relationships and has neglected to assess other close adult relationships such as best
friendships. The present study aims to replicate previous findings on attachment and
romantic relationships; to extend the research area by examining attachment styles and
best fiiendships (both same-sex and other-sex); and to examine the relationship between
attachment style, sex , and the three relationship types (i.e., romantic relationship, same-
sex best fiiendship, and other-sex best fiiendship) on various relationship characteristics.
The findings fi'om this study will extend our understanding of attachment theory and
close relationships.

The hypotheses to be examined in this study include renlieatienhypetheses,
hypetheses are hypotheses that replicate previous research on attachment styles and
romantic relationships. Extensienhymtheses are hypotheses that extend the findings on
romantic relationship to those found on attachment styles and friendship types (i.e., other-
sex, same-sex). Research has been conducted only on attachment styles and romantic

relationships. By extending this framework to include friendships, it was expected that

54

findings would expand the area of adult attachment and further understanding on how
attachment styles affect different types of close relationships. Because of similarities
between friends and lovers, it was expected that findings for attachment styles and best
friends would be similar to those of romantic relationships. DiflerentiaLhythheses are
hypotheses proposing the ways in which attachment styles may create differences in the
ratings for certain relationship components between romantic partners and best fiiends.
AdditignaLhymtheses are hypotheses that examine other relationship components and
how they are affected by attachment styles between the different types of relationships.

Overall Analyses were conducted to examine possible relationships among
attachment style, sex, and relationship types. Main effects were expected for attachment
style and sex. Females were expected to have higher ratings on positive relationship
components (e.g., viability) than males. Secure people were expected to have higher
ratings compared to anxious-ambivalent and avoidant people on positive relationship
components (e.g., care).
W
B l' . H l

Replication hypotheses were hypotheses for attachment styles and romantic
partner. They were expected to replicate previous findings on attachment styles and
romantic relationships.

Based on their attachment pattern, secure individuals were expected to have
positive perceptions of others and relationships. Thus, they would have higher ratings on

many positive relationship characteristics compared to anxious-ambivalent and avoidant

55

individuals. Recall fiom research discussed earlier (e.g., Collins & Read, 1990; Hazan &
Shaver, 1987; Mikulincer & Nachshon, 1991; Simpson, 1990) that secure individuals had
higher levels of trust, acceptance, self-disclosure, closeness, intimacy, satisfaction,
accepting and supportive of partner compared to avoidant and anxious-ambivalent
individuals. Levy and Davis (1988) found that the secure attachment style was positively
correlated with Viability, Intimacy, and Global Satisfaction, while both the anxious-
ambivalent and avoidant attachment styles were negatively correlated with all three
relationship components.

Kirkpatrick and Davis (1994) found that secure individuals had higher ratings on
Viability and Intimacy compared to anxious-ambivalent and avoidant individuals.
Kirkpatrick and Davis (1994) found that secure men had higher ratings for Global
Satisfaction compared to avoidant men. Secure women had higher ratings for Global
Satisfaction compared to anxious-ambivalent women.

Levy and Davis (1988) found that the secure attachment style was positively
correlated with Care while both the anxious-ambivalent and avoidant attachment styles
were negatively correlated with this component. The reverse was found for Conflict-
Ambivalence. The secure style was positively correlated with Passion while the avoidant
style was negatively correlated with it. Kirkpatrick and Davis (1994) found that secure
men had higher ratings for Care compared to avoidant men. Secure women had higher
ratings for Care compared to anxious-ambivalent women. Kirkpatrick and Davis (1994)
defined Care as being more related to romantic relationships than friendships thus sex

differences are examined for this variable for the romantic partner only. Anxious-

56

ambivalence individuals had higher ratings on Passion compared to both secure and

avoidant individuals. Avoidant individuals had higher ratings on Conflict-Ambivalence

compared to both secure and anxious-ambivalent individuals. From the research studies

on attachment styles and conflict and ambivalence (e.g., Brennan & Shaver, 1995;

Collins & Read, 1990), it has been found that secure individuals have the lowest ratings

on Conflict/Ambivalence compared to both anxious-ambivalent and avoidant individuals.

The following hypotheses are proposed based on findings by Kirkpatrick and Davis

(1994), Levy and Davis (1988), and other attachment researchers (e.g., Collins & Read,

1990; Hazan & Shaver, 1987).

R1. It was hypothesized that secure individuals would have higher ratings on liability
for their romantic partner compared to both anxious-ambivalent and avoidant
individuals.

R2. It was hypothesized that secure individuals would have higher ratings on Intimacy
for their romantic partner, followed by anxious-ambivalent individuals then by
avoidant individuals.

R3. It was hypothesized that secure individuals would have higher ratings on Glehal
Satisfacticn for their romantic partner compared to avoidant and anxious-ambivalent
individuals.

R4. It was hypothesized that secure individuals would have higher ratings on Care for
their romantic partner compared to avoidant and anxious-ambivalent individuals.

R4a. It was hypothesized that secure and anxious-ambivalent men would have higher

ratings on Care for their romantic partner compared to avoidant men.

57

R4b. It was hypothesized that secure and avoidant women would have higher ratings on
Care for their romantic partner compared to anxious-ambivalent women.
R5. It was predicted that anxious-ambivalent individuals would have higher ratings on
Bassign for their romantic partner compared to both secure and avoidant individuals.
R6. It was predicted that avoidant individuals would have higher ratings on Conflict:
Amhiyalenee for their romantic partner compared to secure and anxious-ambivalent
individuals.
Extensionfixmthm
Extension hypotheses were hypotheses extending the previous findings of
attachment styles and romantic relationships to attachment styles and best friendships.
As discussed above, most research studies within attachment styles have not
examined attachment styles and best fiiendships. Recall that mental models of others and
relationships are different for each attachment style. These mental models have been
found to have an impact on behaviors within love relationships. Individuals who are
secure have positive mental models of others and relationships. Anxious-ambivalent and
avoidant individuals have negative mental models of others and relationships. Research
on adult attachment has found support for these models. It was proposed that these
mental models should have similar effects for best friendships compared to romantic
relationships because for most individuals, a best friend is considered to be as important
as a romantic partner. Because of the Similarities between best friend and romantic
partner and their relationships, it is predicted that the findings for best friends and their

relationships will correspond highly to those of romantic partners and their relationships

58

within attachment styles because mental models should be constant across different close

relationships. Due to lack of empirical research, predictions for attachment style and best

friendships on Viability, Intimacy, Global Satisfaction, and Care were made based on
findings from attachment styles and romantic relationships discussed earlier. Thus far,
findings for avoidant individuals, in that they mistrust others and fear intimacy and
closeness with others, have been for romantic partner. Thus it was proposed that the
mistrust and fear of intimacy and closeness that avoidant individuals have are specific to
romantic relationships and not fiiendships. Avoidant individuals have been found to have

a high level of enjoyment in their non-romantic relationships (F eeney & Noller, 1990).

This enjoyment should lead to a higher level of satisfaction with these non-romantic

relationships because enjoyment is a characteristic of the component global satisfaction.

Best friends were analyzed separately for other-sex and same-sex.

El. It was predicted that secure individuals would have higher ratings on Viability for
their best fiiend compared to anxious-ambivalent and avoidant individuals.

E2. It was predicted that secure individuals would have higher ratings on Intimacy for
their best friends compared to anxious-ambivalent and avoidant individuals.

E3. It was predicted that secure and avoidant individuals would have higher ratings on
filebaLSatisfaeticn for their best friends compared to anxious-ambivalent
individuals.

E4. It was predicted that secure individuals would have higher ratings on Care for their

best friends compared to anxious-ambivalent and avoidant individuals.

59
Diffctcnfialflxpothescs

Differential hypotheses were hypotheses comparing romantic partner and best
friends on various relationship components within each attachment style.

In the extension hypotheses section, it was predicted that many similarities
between best friend and romantic partner would occur for certain relationship components
across the three attachment styles. However, because of the different roles and functions
that romantic partners and best friends have, individuals should differ in how they rate
certain relationship components and this would all vary according to an individual's
attachment style. Davis and Todd (1982) found that when comparing best friend and
lover/spouse on characteristics of Global Satisfaction, best friend received higher ratings
compared to spouse/lover. Secure individuals had a higher level of satisfaction within
their love relationships compared to anxious-ambivalent and avoidant individuals. Davis
and Todd (1982) designed the factor Care to contain variables pertaining more to
romantic relationships than to friendship. Giving the utmost and championing are done
for romantic partners more often than for fiiends. Caring is a relationship characteristic
that has been found to be more important in romantic love relationships compared to
friendships (Steck et al., 1979). With this in mind, it is predicted that secure and anxious-
arnbivalent individuals will discriminate between romantic partner and best friends on
Care. Avoidant individuals will not differentiate between their ratings of Care for
romantic partner and best fiiendships. Davis and Todd (1982) found that spouse/lover
received a significantly higher rating on Passion and Conflict/Ambivalence compared to

best fiiend. It is predicted that regardless of attachment style, individuals will

60

discriminate between romantic partner and best fiiends on Passion and
Conflict/Ambivalence.

Research studies (e.g., Davis & Todd, 1982) have compared romantic partner and
best friends on certain relationship components. Davis and Todd (1982) found that best
friends received higher ratings on Viability and Intimacy than spouse/lover. However,
these ratings were not significantly different. Previously it was shown that secure
individuals perceive others to be good and well intended and relationships to be
satisfying. These individuals should not discriminate between romantic partner and best
friends on these two components. Anxious-ambivalent individuals desire union with
others and are preoccupied and obsessed with relationships. However, their desire for
union may be specific to romantic partners and not best fiiends. They should
discriminate their ratings on Viability and Intimacy for romantic partners and best
fiiends. Avoidant individuals mistrust others and fear intimacy and closeness with
others. So far, all findings have been for romantic partner. Thus it will be proposed that
the fear of intimacy and closeness that avoidant individuals have is specific to romantic
relationships and not fiiendships and they will rate their best fiiends higher on Viability
compared to their romantic partner.

Sex differences were predicted to exist for secure and avoidant individuals on
Intimacy because it has been found that men tend to find intimacy needs fulfilled in love
relationships while women seem to find intimacy needs fulfilled within same-sex
fiiendships (Rubin, 1985). Researchers have found that males confide more in their

female best friends compared to their male best friends (Komarovsky, 1974). Wheeler

61

and colleagues (1983) and Hays (1989) found that men and women prefer intimate

friends who are women and they find it easier to disclose to them (Reis et al., 1985).

D1. It was hypothesized that secure and avoidant males would have higher ratings for
Intimacy for romantic partners than best friends.

D2. It was predicted that avoidant and anxious-ambivalent individuals would have
higher ratings on Glebalfiatisfacticn for their best fiiends compared to their
romantic partner.

D3. It was hypothesized that secure and anxious-ambivalent individuals would have
higher ratings on_Car_e for romantic partners compared to best friends.

D4. It was predicted that across all three styles of attachment, individuals would have
higher ratings on Passicn for their romantic partners compared to their best friends.

D5. It was predicted that across the three attachment styles, individuals would have
higher ratings on Cenflictfimbiyalencc for their romantic partners compared to
their best friends.

D6. It was hypothesized that anxious-ambivalent individuals would have higher ratings
on Viability for romantic partner and other-sex best fiiend compared to same-sex
best fiiend.

D7. It was hypothesized that avoidant individuals would have higher ratings on
Viability for same-sex best friend, followed by other-sex best friend, and finally by
romantic partner.

D8. It was hypothesized that secure and avoidant females would have higher ratings on

Intimacy for same-sex best friend, then romantic partner, then other-sex best friend.

62

D9. It was hypothesized that anxious-ambivalent individuals would have higher ratings
on Intimacy for romantic partner and other-sex best friend compared to same-sex
best fiiend.

! 1 1° . l H 1

Additional hypotheses: these hypotheses examined attachment styles and close
relationships on other important relationship variables. Some differences were predicted
to occur on certain relationship components across the three attachment styles. Due to
lack of empirical findings, some hypotheses below will be more vague and could be
perceived as interesting research questions instead of specific hypotheses.

As stated earlier, most research studies investigating attachment styles have not
examined best friendship. However, research within the friendship literature has found
that sex differences exist between men and women on intimacy for same-sex and other-
sex fiiendships. Overall, women tend to have more intimacy and value self-disclosure
and understanding within their same-sex friendships more compared to men (e. g., Aries
& Johnson, 1983; Bell, 1981; Caldwell & Peplau, 1982; Helgeson et al., 1987; Monsour,
1992). Sex differences may exist in that women have been found to find support in same-
sex best fiiendships while men find it in love relationships (e.g., Caldwell & Peplau,
1982; Reis, 1984; Rubin, 1985). Women have been found to be more satisfied with their
same-sex friendships compared to men (e.g., Buhrke & Fuqua, 1987).

Research on attachment styles has found that secure individuals have more
positive perceptions of others and relationships, believe that relationships are more

satisfying and worth maintaining, and thus would be more likely to be in a relationship

63

compared to avoidant and anxious-ambivalent individuals (i.e., more likely to have a

romantic partner, other-sex best friend, and same-sex best friend) (e. g., Hazan & Shaver,

1987). Feeney, Noller, and Patty (1993) found that avoidant individuals had the fewest

mixed-fiiend interactions when compared to secure and anxious-ambivalent individuals.

Al. It was hypothesized that across all attachment styles, women would have higher
ratings on Intimacy for their same-sex best friend compared to men.

A2. It was hypothesized that overall, across the three attachment styles, women would
have higher ratings on GlebaLSatisfacticn for their same-sex best friend compared
to men.

A3. It was hypothesized that overall, across the three attachment styles, women would
have higher ratings on Care for their same-sex best friend compared to men.

A4. It was hypothesized that secure individuals would be more likely to have a romantic
partner compared to anxious-ambivalent and avoidant individuals.

A5. It was hypothesized that secure individuals would be more likely to have a best
fiiend (other-sex, same-sex) compared to both avoidant and anxious-ambivalent

individuals.

CHAPTER 11

METHOD

E . .

During the first few days of classes of the Fall semester, I went to four
Introductory Psychology classes and four 200 level psychology classes (Social
Psychology, Personality Psychology, Developmental Psychology, and Psychology of
Women) to administer a short packet of questionnaires for the mass testing wave of this
study. This initial testing included 1,619 male and female undergraduate psychology
students. Of these, 46 participants were dropped from the study due to failure to follow
directions or because their responses indicated that they were not in any of the
relationships required for participation in this study. Thus, there were 1,573 participants
in the first wave of data collection. This was called the Mass Testing Sample.

Of this sample, 1,217 students volunteered to participate further in a Social
Relationship Study to earn three extra credit points toward their psychology class grades.
Of this sample, 118 participants did not complete their questionnaire packet correctly or
were not in any of the relationships required for the study, so their data had to be
discarded. Thus, the total number of usable data sets across the two waves of data

collection was 1,099. This was called the Social Relationship Study Sample. See Table

64

65

2 for number of participants for each sample of the study.
AttachmenLStyleLlassiflcaticn. To be eligible for the final sample used in the
data analyses reported below, participants had to indicate the same attachment style on
both Hazan and Shaver’s (1987) and Bartholomew and Horowitz’s (1991) measures.
These measures can be found in Appendices A-1 and A-2. Although this procedure had
the disadvantage of decreasing the number of classifiable participants, the identification
of the attachment styles for the selected participants has been found to be more reliable
and valid compared to the unselected group of participants (e. g., Brennan et al., 1991).
An analysis was conducted on these participants to assure that the attachment style they
had selected, on both adult attachment scales, was in fact rated higher on the parallel 1-7
rating scale compared to the other two attachment styles. One-way AN OVAs were
conducted on combined Hazan & Shaver’s and Bartholomew’s attachment style category
as the independent variable and the 1-7 rating scales as the dependent variables. On the
secure 1-7 rating scale, secure participants rated it significantly higher (M=6.l3)
compared to either anxious-ambivalent (M=3.48) or avoidant (M=2.75) participants
[E(2,790) = 1,288.79, p<.001]. On the anxious-ambivalent 1-7 rating scale, anxious-
ambivalent participants rated it significantly higher (M=6. 14) compared to either secure
(M=2.41) or avoidant (M=2.62) participants [E(2,787) = 442.22, p<.001]. On the
avoidant 1-7 scale, avoidant participants rated it significantly higher (M=5.26) compared
to either anxious-ambivalent (M=2.76) or secure (M=2.34) participants [E(2,786) =
864.20, p<.001]. Separate analyses were conducted for males and females and they gave

results Similar to the above.

66

IheAnalysiiSample. Out of the 1,099 participants, 796 participants could be
classified into an attachment style and thus constituted the Analysis Sample. This data
set contained the following percentages for each attachment style : 25.8% were Avoidant,
11.4% were Anxious-Ambivalent, and 62.8% were Secure. From previous research on
attachment style (e.g., F eeney & Noller, 1990; Hazan & Shaver, 1987; Mikulincer &
Nachshon, 1991), the expected ranges for each attachment styles are: Avoidant (14%-
30%), Anxious-Ambivalent (8%-19%), and Secure (55%-65%). The percentages of
participants of each sex and attachment style can be found in Table 2.

Besides meeting the attachment style criterion, participants also had to have either
a best friend or a romantic partner to be eligible for participation. This criterion was
established by the sign-up sheets and the researcher script; thus, all participants were in at
least one type of relationship. Some individuals had overlapping relationships in that
their romantic partner was also their best fiiend. This issue could confound the findings
for ratings of romantic partner and best fiiends because these two ratings were of the
same person. To reduce this potential problem, when selecting the rating for romantic
partner and best friends, only romantic partner ratings were used for analyses from the
individuals who nominated their romantic partner as their best fiiend. Thus, responses
about romantic partners in this study were not also those of the best fiiend.

Table 3 shows the number and type of relationships that participants in the study
bad. This table revealed that best friends tend to play a large role in individuals’
relationship type. When individuals only have one relationship, it is most likely to be a

same-sex best fiiend. Same-sex best friend still plays a large role when the focus is on

67

two relationships. Over 50% of individuals in the Analysis Sample stated that their two
relationships were other-sex best friend and same-sex best friend. Over 50% of all
individuals in the Analysis Sample had all three relationships, and very few individual
had only one relationship. More than half of all individuals had all three relationships.
There was a great overlap between best friend and romantic partner, thus the numbers for
other-sex best friend were inflated. When only independent relationships were taken into
account such that best fiiends were not also romantic partners, the actual total of
individuals with three relationships decreased from 470 (59% of the sample) to 227 (29%
of the sample). The total number having two relationships then became the most frequent
response. Regardless of the number of relationships, most participants had a same-sex
best friend.

To return to the Analysis Sample, almost half (45.6%) of all participants were
freshmen and the mean age was 19 years. Of the participants in the sample 86.5% were
White/Caucasian, 5.5% were Black/African-American, 5.2% were Asians/Pacific
Islander, 1.8% were Hispanic/Mexican-American, 0.5% were Native American, and 0.5%
were classified as other.

Within the Analysis Sample, two other samples were created; a Between-Subject
Relationship Type Sample and a Within-Subject Relationship Type Sample. The
Between-Subject Sample is comprised of individuals selected to be analyzed in only one
type of relationship. Thus, the sample consisted of three independent groups of
individuals who were analyzed for only one of the three relationship types. The Within-

Subject Sample is comprised of individuals who had all three relationships (i.e., romantic

68

partner, other-sex best friend, and same-sex best friend), with the qualification (noted
above) that their romantic partner could not be their best fiiend.

Table 4 shows proportion representative of types of relationships for each analysis
sample. Within the Analysis Sample, 94% of participants had a same-sex best friend,
81% of participants had an other-sex best fiiend, and 76% of individuals had a romantic
partner. However, the numbers for best fiiends were reduced when individuals who had a
romantic partner who was also their best friend were removed from the best friend counts.
In the Analysis Sample, 45% of participants had an other-sex best fiiend and 93% of
participants had a same-sex best friend.

W This sample contained 735 participants. This
group was within the expected range of percentages for each attachment style: 27.8%
were Avoidant, 12.2% were Anxious-Ambivalent, and 60.0% were Secure. See Table 2
for the percentages of participants within each sex and attachment style. There were 245
individuals from each relationship type (i.e., romantic partner, other-sex best friend, and
same-sex best fiiend). To ensure that the attachment style breakdown of individuals
(within each relationship type) would be representative of other samples, the proportion
of individuals within each attachment style was held constant within each relationship
type group. This meant that regardless of the percentage of individuals in each
attachment style and relationship type, appropriate percentages were selected from each
group such that approximately 28% (68) of these individuals were avoidant, 12% (30) of
these individuals were anxious-ambivalent, and 60% (147) of these individuals were

secure.

69

In the Between-Subject Sample, almost half (45.6%) of all participants were
freshmen and their mean age was 19 years. Of the participants in the sample, 86.1% were
White/Caucasian, 5.9% were Black/African-American, 5.2% were Asians/Pacific
Islander, 1.8% were HisparricMexican—American, 0.5% were Native American, and 0.5%
were classified as other. Table 3 shows the number of relationship that participants in the
study had. Table 4 shows the types of relationships that participants in the study had.

W In the Within-Subject Sample, there were 227
participants. All participants had all three relationships. This sample was within the
expected range of percentages for each attachment style: 24.2% were Avoidant, 11.5%
were Anxious-Ambivalent, and 64.3% were Secure. See Table 2 for percentage of
individuals within each attachment style. Almost half (41 .4%) of all participants were
freshmen and their mean age was 19 years. Of the participants in the sample 87.7% were
White/Caucasian, 4.0% were Black/African-American, 5.3% were Asians/Pacific

Islander, 2.2% were Hispanic/Mexican-American, and 0.9% were Native American.

70

 

 

Table 2
SampleSizchLAttachmanStxleanRScx
Females Males Total
(Row %) (Row %) (Row %) (Column %)
Massiesting
N 1,110 (70.6%) 463 (29.4%) 1,573 (100%)
Attachment Stylel
Avoidant 218 (73.9%) 77 (26.1%) 295 (100%) (18.7%)
Anxious-Ambivalent 125 (59.8%) 84 (40.2%) 209 (100%) (13.3%)
Secure 767 (71.7%) 302 (28.3%) 1,069 (100%) (68.0%)
S . l E l . l . S l S l
N 781 (71.1%) 318 (28.9%) 1,099 (100%)
Attachment Stylel
Avoidant 171 (73.1%) 63 (26.9%) 234 (100%) (21.3%)
Anxious-Ambivalent 95 (58.6%) 67 (41.4%) 162 (100%) (14.7%)
Secure 515 (73.3%) 188 (26.7%) 703 (100%) (64.0%)
W
N 583 (73.2%) 213 (26.8%) 796 (100%)
Attachment Style2
Avoidant 153 (74.6%) 52 (25.4%) 205 ( 100%) (25.8%)
Anxious-Ambivalent 57 (62.6%) 34 (37.4%) 91 (100%) (11.4%)
Secure 373 (74.6%) 127 (25.4%) 500 (100%) (62.8%)
W121:
N 533 (72.5%) 202 (27.5%) 735 (100%)
Attachment Style2
Avoidant 152 (74.5%) 52 (25.5%) 204 ( 100%) (27.8%)
Arixious-Ambivalent 56 (62.2%) 34 (37.8%) 90 (100%) (12.2%)
Secure 325 (73.7%) 116 (26.3%) 441 (100%) (60.0%)
ll? 1 . -S l . S l
N 161 (70.9%) 66 (29.1%) 227 (100%)
Attachment Style2
Avoidant 40 (72.7%) 15 (27.3%) 55 (100%) (24.2%)
Anxious-Ambivalent 16 (61.5%) 10 (38.5%) 26 (100%) (11.5%)
Secure 105 (71.9%) 41 (28.1%) 146 (100%) (64.3%)

 

Note. 1 Only Hazan & Shaver’s adult attachment scale was used.

2 Both Hazan & Shaver’s and Bartholomew’s adult attachment measures were used.

71

 

 

 

Table 3
’ooioith OJ. 0 k "01' O (‘9... '0 1":'1.'U l0 l'r' 5.
Sample
% in % in Total
Relationship Count Analysis
Category Sample
Relationship Type: Count showing that pattern of Relationships Only
100% 796
One Relationship Only 100% 7.7% 61
Romantic Partner Only 3.3% 0.3% 2
Other-Sex Best Friend Only 8.2% 0.6% 5
Same-Sex Best Friend Only 88.5% 6.8% 54
Two Relationships Only 100% 33.2% 264
Other-Sex Best Friend and Romantic Partner 14.4% 4.8% 38(8)
Same-Sex Best Friend and Romantic Partner 35.6% 11.8% 94(87)
Other-Sex and Same-Sex Best Friend 50.0% 16.6% 133
All Three Relationships 100% 59.0% 470(227)
Relationship Type: Overall Type and Number of Relationships
100% 796
One or More Relationships
Romantic Partner 75.9% 604
Other-Sex Best Friend 81.0% 645(378)
Same-Sex Best Friend 94.3% 751(73 8)
Two or Three Relationships
Other-Sex Best Friend and Romantic Partner 63.8% 508(235)
Same-Sex Best Friend and Romantic Partner 70.9% 564(314)
Other-Sex and Same-Sex Best Friend 63.2% 503

 

Nete, Numbers in parentheses display the actual number of best fiiends who are not also

a romantic partner.

72
Table 4

5111 E 1.1 EBl'l'I

 

 

Females Males Total
AnalysisSampl: 583 213 796
Romantic Partner 444 (73.5%) 160 (26.5%) 604 (76%)
Other-Sex Best Friend (OS) 478 (74.1%) 167 (25.9%) 645 (81%)
Same-Sex Best Friend (SS) 552 (73.5%) 199 (26.5%) 751 (94%)
OS is the RP 208 (36%) 59 (28%) 267 (34%)
SS is the RP 11 (2%) 2 (1%) 13 (2%)
B_:M::rtSubi::LSamul: 533 202 735
Romantic Partner 182 (74.3%) 63 (25.7%) 245
Other-Sex Best Friend 170 (69.4%) 75 (30.6%) 245
Same-Sex Best Friend 181 (73.9%) 64 (26.1%) 245
Best Friend 179 (73.1%) 66 (26.9%) 245
WSW: 161 66 227
Romantic Partner 161 (70.9%) 66 (29.1%) 227
Other-Sex Best Friend 161 (70.9%) 66 (29.1%) 227
Same-Sex Best Friend 161 (70.9%) 66 (29.1%) 227

 

73

Manuals

Participants completed the following questionnaires during the mass testing in
their psychology class; i.e., the screening wave of data collection.

CleaeRelaticnshmsflmstionnaireBaflL This is a measure of adult attachment
developed by Hazan and Shaver (1987), who derived this adult attachment scale from
Ainsworth and her colleagues’ (1978) description of infants' style of attachment. Each
attachment style (secure, anxious-ambivalent, avoidant) is described in a single
paragraph. Participants indicated which paragraph of the three attachment styles best
described them and rated each attachment style on a 7-point Likert-typed scale (l-not at
all like me to 7-very much like me). Brennan and Shaver (1990) have found 8-12 month
stability for this measure for 70-75% of their students. Test-retest reliabilities are in the
range of 0.60. Reliabilities and validity were described earlier in the Measurement of
Adult Attachment section. (See Appendix A—l for this scale).

WW2. This scale assesses attachment style
according to Bartholomew's (1990) four category model: secure, fearful, preoccupied,
dismissing. This scale contains two parts. In the first part, descriptions of each of the
four styles were provided and participants indicated which style best described their
general approach to relationships. In the second part, individuals indicated on a 1-7
(1=not very much like me to 7=very much like me) Likert-type scale their degree of
agreement with each of the four descriptions. Reliabilities and validity were described
earlier in the Measurement of Adult Attachment section, also see Bartholomew and

Horowitz (1991). (See Appendix A—2 for this scale).

74

'0‘ in om u .- I-uounu .u oi, .' on O : on. - ll .0 .
Participants responded to a few questions concerning their age and five types of personal
relationships. Participants responded to questions pertaining to their romantic partner, if
they have one; their same-sex best friend and their other-sex best friend; and their close
friendships. This scale is a shorter form of the actual DSHQ scale. This short form was
administered during the mass testing sessions. (See Appendix B-l for this scale).

Eunnefianicinaticnfiheet. Participants were requested to complete this sheet if
they wished to participate in the Social Relationship Study. Students wrote down their
name, phone number, best time to be reached by phone, and their psychology instructor's
name. Participants’ student ID numbers were used as subject ID numbers so that
students' names would not be used, in order to protect their anonymity. Students were
told that if they chose to participate in the Social Relationship Study, they could earn
three extra credits toward their psychology course. (See Appendix D for this sheet).

Participants completed the following questionnaires in the Social Relationship
Study.

W This version of
the RF (Davis & Latty-Mann, 1987) is a modified version of the original presentation of
Davis & Todd (1982, 1985). This version is a 68-item relationship questionnaire that
includes Six general scales and two subscales of relationship factors for both fiiendships
and love relationships and has been used in recent studies (e.g., Kirkpatrick & Davis,
1994). Participants rated each item on a 9-point scale (1=not at all to 9=completely or

extremely). The alphas and test-retest reliabilities are as follows: Viability (alphas: .80;

75

.85; test-retest = .74), Intimacy (alphas: .76; .79; test-retest = .78), Passion (alphas: .82;
.80; test-retest = .82), Care (alphas: .89; .87; test-retest = .79), Global Satisfaction
(alphas: .90; .93; test-retest = .73), and Conflict/Ambivalence (alphas: .73; .83; test-retest
= .68). The test-retest correlations were calculated over a three-week span. Sample items
are as follow: "Do you accept this person as she or he is?" "Do you confide in this
person?" and "Can you count on this person to support you in argument or dispute with
others?" This scale was revised to measure best friendships and contains 65 items. Three
items (28-30) were lefi out of the scale because they measured sexual intimacy.

However, participants still had to fill in a rating for each item of their scantron sheet so
they were told to give these items ratings of 10 (on their scantron sheets) for their other-
sex and same-sex best friend. A rating of 10 was equivalent to a missing rating. Some
participants completed these items if either their other-sex or their same-sex best fiiend
was also their romantic partner but these data were disregarded (see above). (See
Appendices E-l , E-2, E-3 which presents the RF for Same-Sex Best Friend, Other-Sex

Best Friend, and Romantic Partner, respectively).

 

. Participants responded
to questions concerning their age, race, class level, and five types of personal
relationships. They were also questions pertaining to their romantic partner, if they had
one; their same-sex best fiiend and their other-sex best friend; and their close friendships.
This scale included questions about best fiiends and romantic partner, on length of these

close relationships, closeness level with each person, and relations with each person.

76

This scale was constructed by the author to measure the history and status of these
relationships. (See Appendix B-2 which presents the DSHQ).
Procedure

A mass testing took place during the first few days of classes, when psychology
students completed the Close Relationships Questionnaire Parts 1 & 2, the Short Form of
the Demographic and Social History Questionnaire, the Future Participation Sheet, and
consent form #1 (See Appendices A-l , A-2, B-l, D, and C for copies of these materials).

' The researcher read the mass testing portion of the research script to the students (See
Appendix H-l for copy of researcher script). Those students who wished to participate in
the Social Relationship Study and earn three extra credit points for their psychology class
completed the Future Participation Sheet by writing their name, phone number, the times
when they could be reached by phone, and the name of their Psychology professor. After
the mass testing sessions, sign-up sheets for the Social Relationship Study were posted in
all introductory psychology classes and the four 200 level psychology classes. Students
signed up if they were interested in participating in the study.

At their session, participants were told that their participation would require about
one hour and thirty minutes and the completion of a packet of questionnaires about
friendships and romantic relationships; and that these questionnaires were standard
measures commonly used to assess personal relationships.

At the beginning of the session, the researcher began by explaining the purpose of
the study; then, she administered consent form #2 (See Appendices H-2 and F for the

research script for the Social Relationship Study and consent form #2, respectively).

77

These forms were collected after the participants read and signed them. Next, the
researcher administered the questionnaire packets and scantron Sheets. Each
questionnaire packet contained the following scales: three versions of the Davis-Todd’s
Relationship Rating Form (same-sex best friend, other-sex best friend, and romantic
partner) and the Demographic and Social History Questionnaire. All scales were counter-
balanced. The researcher gave instructions on questionnaire completion and answered
any questions that were raised. (See Appendices E-l, E-2, E-3 for the three versions of
the RRF: Same-Sex Best Friend, Other-Sex Best Friend, and Romantic Partner. See
Appendix B-2 for the Demographic and Social History Questionnaire).

When the participants were done completing their questionnaires, they returned
the questionnaire packet and the scantron sheets to the experimenter. Their psychology
credit cards were stamped with three points for their participation. The researcher then
read aloud the Participant Information Sheet and answered any questions that were raised.
All participants received a Participant Information Sheet which included information
about the study to take home with them; this form also listed the phone number of the
researcher in case the participants had any later questions about the study (See Appendix
G for the Participation Information Sheet).

Although many students signed up for the study, more participants were needed,
especially anxious-ambivalent males. For this reason, eligible students - those who had
completed the Future Participation Sheet at the mass testing and were in at least one of
the necessary relationships -- were contacted by telephone and asked if they would like to

participate in the Social Relationship Study. They were told that they had completed a

78

short questionnaire on either the first or second day of class and now they were eligible to
participate in this study to earn three extra credit points toward their psychology class.
They were told that their participation would consist of completing questionnaires about
friendship and romantic relationships and would take about one hour and thirty minutes.
If they agreed to participate, they were asked to sign up for a time session that best fit
their schedule. All students were thanked for taking the time to talk to the experimenter
on the phone. From this calling procedure, a total of 42 students Signed up to participate
in the study. Thirty-two students (12 females, 20 males) actually participated in the

study.

CHAPTER III

RESULTS

Qverxim

The goal of the present study was to further our understanding of attachment
styles and important components of close interpersonal relationships. More specifically,
it addressed the question of how attachment styles and different types of interpersonal
relationships -- particularly romantic/love relationships and best friendships -- are related
to feelings and beliefs that people have about these relationships. Within this framework,
the research examined how individuals with one of three attachment styles rate their best
fiiends and their romantic partner on important relationship components. Because of the
perceived differences in the roles and functions that different relationships may have for
individuals with different attachment styles, these people may rate their romantic partner
and their best fiiends differently on some of the components that characterize their
feelings about these “significant others.”

This study was conducted using a 2 (Sex of subject: Female, Male) X 3
(Attachment style of subject: Secure, Anxious-Ambivalent, Avoidant) X 3 (Relationship
Type: Romantic Partner, Same-Sex Best Friend, Other-Sex Best Friend) fully crossed

research design (with Relationship Type as either a between-subj ects or a within-subj ects

79

80

factor). The primary dependent variables examined were Viability, Intimacy, Global
Satisfaction, and Care. Passion and Conflict-Ambivalence were also examined. As
discussed in the introduction, these relationship components were selected for study
because they were thought to be important to both romantic relationships and best
friendships.

As noted, the 2 X 3 X 3 design was used to examine two separate data sets
(Within-Subjects Sample and Between-Subjects Sample) described in the Method
Chapter. With the Within-Subjects Sample, I was able to cast the 2 X 3 X 3 design with
relationship type as a repeated measure. With the Between-Subjects Sample, I was able
to cast the 2 X 3 X 3 design with relationship type as an independent-sample variable.

The data were examined in a number of ways to maximize the number of relevant
responses that would bear on a hypothesis or question. The Analysis Sample contained
all of the participants, most of whom had a romantic partner. This sample was utilized
for the Replication Hypotheses involving romantic partners to explore the extent to which
the present findings mirrored those of previous research that had focused on that
relationship-type exclusively. The Analysis Sample was also used to examine some of
the Differential Hypotheses and Additional Hypotheses. Both versions of the 2 X 3 X 3
design were utilized for examining the Extension Hypotheses. The Between-Subj ects
Sample was also used to examine some of the Differential Hypotheses.

In this chapter, reliabilities of the RF are presented before the results are
examined. Findings relevant to the replication hypotheses are discussed first, followed

by those pertaining to the extension hypotheses, then those related to the differential

81

hypotheses, then those relevant to the additional hypotheses and finally, those pertaining
to the overall analyses. Each section contains a brief introduction before the results are
presented and a brief summary is offered after them.
31.11.. [1 EEE

The main dependent variables were derived from the RF (Davis and colleagues,
1982, 1985, 1987). Subscales were formed based on Davis & Todd’s findings using the
RRF. Reliabilities (coefficient alphas) were calculated on the six subscales used in the
present study. The alphas found in the present study were similar to Davis and
colleagues’ findings, which were summarized in the Method Chapter. Reliability
coefficients were computed separately for the three relationship types (i.e., romantic
partner, other-sex best fiiend, same-sex best friend). For ease of comparisons, alphas

fiom the present study and those of Davis and colleagues are presented in Table 5.

82

 

 

 

Iah1:_i
31.11.. C m. 11 IE lEBE
Present Study
Relationship Type
Romantic Other-Sex Same-Sex Davis & Kirkpatrick
Partner Best Friend Best Friend Latty-Mann & Davis
(1987) (1994)
Variables
Viability .82 .83 .84 .80 .85
Intimacy .75 .75 .69 .76 .79
Global .87 .86 .85 .90 .93
Satisfaction
Care .87 .87 .86 .89 .87
Passion .76 .88 .75 .82 .80
Conflict- .68 .64 .68 .73 .83

Ambivalence

 

83

Rstsnftheflxpchses

1'. Au - urn: ‘ ‘ i' . o it... urn .\ ‘n in... on
.H ,- H. .H {H ni' -;m' 1... _fi '.- . in... . ”m.

These hypotheses addressed the issue of the extent to which the present data set
replicated previous research findings on romantic relationships and attachment styles.
Thus, analyses were conducted on all individuals who had a romantic relationship. Levy
and Davis (1988) and Kirkpatrick and Davis (1994) conducted research examining the
relationship variables from the RF within the areas of romantic relationship and
attachment styles. Based on this work, the hypotheses were straightforward: for each
relationship component (i.e., viability, intimacy, global satisfaction, and care), secure
individuals were predicted to have higher ratings for their romantic partner compared to
anxious-ambivalent and avoidant individuals; for passion, anxious-ambivalent individuals
were predicted to have higher ratings compared to secure and avoidant individuals; and,
for conflict-ambivalence, avoidant individuals were predicted to have higher ratings
compared to secure and anxious-ambivalent individuals.

Planned comparisons were conducted to examine the relationship between
attachment styles and relationship components. The Analysis Sample was employed for
these comparisons because it contained the greatest number of individuals who reported

having a romantic partner. See Table 6 for means on ratings of relationship variables for

romantic partner by each attachment style for the Analysis Sample.

84
R1. Viability

It was hypothesized that secure individuals would have higher ratings on Viability
for their romantic partner compared to both anxious-ambivalent and avoidant
individuals.

This hypothesis was supported in that mean differences were as predicted and a
planned comparison Showed that secure participants rated their romantic partner higher
on Viability than did anxious-ambivalent and avoidant participants [t (601) = 4.53,
p<.001].

R2. Intimacy
It was hypothesized that secure individuals would have higher ratings on Intimacy
for their romantic partner followed by anxious-ambivalent individuals then by
avoidant individuals.

Mean differences were as predicted, so two planned comparisons were conducted
to test this hypothesis. The first planned comparison was between secure and anxious-
ambivalent individuals. The second planned comparison was between anxious-
ambivalent and avoidant individuals.

For the first planned comparison, this hypothesis was supported. Secure
participants rated their romantic partner significantly higher on Intimacy than did the
anxious-ambivalent [t (601) = 2.34, p<.01]. For the second planned comparison, there
was marginal support for this hypothesis. Anxious-ambivalent participants rated their

romantic partner marginally higher than did avoidant participants [t_(601) = 1.49, p<.10].

85

R3. Global Satisfaction
It was hypothesized that secure individuals would have higher ratings on Global
Satisfactien for their romantic partner compared to avoidant and anxious-ambivalent
individuals.

This hypothesis was supported in that mean differences were as predicted and a
planned comparison Showed that secure participants rated their romantic partner
significantly higher on GlebaLSatisfacticn than did anxious-ambivalent and avoidant
participants [1 (601) = 5.80, p<.001].

R4. Care
It was hypothesized that secure individuals would have higher ratings on Care for
their romantic partner compared to avoidant and anxious-ambivalent individuals.

This hypothesis was supported in that mean differences were as predicted and a
planned comparison showed that secure participants rated their romantic partner higher
on Care than did anxious-ambivalent and avoidant participants [t (601) = 4.44, p<.001].
R4a. Care for Males

It was hypothesized that secure and anxious-ambivalent men would have higher
ratings on Care for their romantic partner compared to avoidant men.

This hypothesis was not supported. Inspection of the means showed that they
were not in the predicted direction. (See Table 6)

R4b. Care for Females
It was hypothesized that secure and avoidant women would have higher ratings on

Cane for their romantic partner compared to anxious-ambivalent women.

86

This hypothesis was not supported. Inspection of the means showed that they

were not in the predicted direction. (See Table 6)

R5. Passion
It was predicted that anxious-ambivalent individuals would have higher ratings
on Bassicn for their romantic partner compared to both secure and avoidant
individuals.

This hypothesis was supported in that mean differences were as predicted and a
planned comparison showed that anxious-ambivalent participants rated their romantic
partner significantly higher on Bassien than did secure and avoidant participants [1 (601)
= 2.21, p<.05].

R6. Conflict-Ambivalence
It was predicted that avoidant individuals would have higher ratings on
Ccnflictfimbiyahnce for their romantic partner compared to secure and
anxious—ambivalent individuals.

This hypothesis was not supported. Inspection of the means showed that they
were not in the predicted direction. (See Table 6)

As expected, the overall results tended to replicate previous findings of Levy and
Davis (1988) and Kirkpatrick and Davis (1994), not only in that generally, individuals
with different attachment styles differ in how they rate certain relationship characteristics
but also in how the styles differ. Secure individuals were found to have higher ratings on

relationship characteristics of viability, intimacy, global satisfaction, and care for their

87

romantic partner compared to avoidant and anxious-ambivalent individuals. Anxious-
arnbivalent participants had higher ratings for Intimacy and Passion compared to avoidant
participants. Anxious-ambivalent individuals had higher ratings on passion and conflict-
ambivalence than secure individuals. Avoidant individuals had higher ratings on
Conflict-Ambivalence compared to secure individuals. Overall, secure individuals had
higher ratings for positive relationship variables compared to individuals in the two

. l
insecure styles.

88

 

 

 

Iah1:_6
U'qle.1' H ‘quuoo euqo‘ 0:01:41! ”IV-J 0 ‘.-. téL- Ht‘u
fQLtheAndysiSSampl:
Attachment Style
Relationship Variables Avoidant Anxious-Ambivalent Secure
Viability 6.97 7.04 7.45
Intimacy 6.56 6.87 7.30
Global Satisfaction 7.21 7.14 7.87
Care 7.24 7.33 7.76
(Males) 7.12 7.07 7.33
(Females) 7.29 7.48 7.90
Passion 7.18 7.74 7.61
Conflict-Ambivalence 3.09 3.20 2.66

 

Nete: n = 604 (444 females; 160 males)

o
C
n
I
b
.
O
-
I
-l
O
O
.-
‘-
n
-
c
'
c
-
f
O
O
u
c
I
.
b
4'.
O
c

Having successfully replicated past results that demonstrated a link between
attachment style and certain relationship variables within romantic relationships, the next
question is whether these patterns extend to best fiiendships as well. The purpose of
these extension hypotheses was to broaden research on relationship variables and
attachment styles into the domain of fiiendship. Many researchers have conducted
investigations within domain of the romantic relationship, but few have conducted
research within other relationships such as close friendships. Best friends were separated
into two groups -- other-sex and same-sex -- because it was possible that people would
tend to vary their ratings on the relationship variables depending on the sex of their
friends. Previous research had not examined best friends of both sexes; thus, no specific
differences were predicted between other-sex and same-sex best friends.

These hypotheses were straightforward and similar to the romantic relationship
hypotheses; for viability, intimacy, and care, secure individuals were predicted to have
higher ratings for their other-sex and same-sex best friend compared to anxious-
ambivalent and avoidant individuals. For Global Satisfaction, secure and avoidant
individuals were predicted to have higher ratings for their other-sex and same-sex best
friends compared to anxious-ambivalent individuals. Participants who had an other-sex
best friend were included in the other-sex best fiiend analyses and participants who had a

same-sex best fiiend were included in the same-sex best fiiend analyses.

90

Planned comparisons were conducted to examine the relationship between
attachment styles and relationship components. Both the Between-Subj ects Sample and
the Within-Subjects Sample were employed for these comparisons. See Table 7a for
means on ratings of relationship variables for other-sex best friend by each attachment
style. See Table 7b for means on ratings of relationship variables for same-sex best
friend by each attachment style.

E1. Viability
It was predicted that secure individuals would have higher ratings on Viability for
their best fiiend compared to anxious-ambivalent and avoidant individuals.

For both samples, means were as predicted for other-sex best friends: secure
participants rated their other-sex best fiiend higher on Viability than did anxious-
ambivalent and avoidant participants. Moreover, planned comparisons indicated that the
hypothesis was supported by the Between-Subjects Sample [t (717) = 2.13, p<.05] and
marginally supported for the Within-Subjects Sample [1 (221) = 1.47, p<.10].

For both samples, means were as predicted for same-sex best friends: secure
participants rated their same-sex best fiiends higher on Viability than did anxious-
ambivalent and avoidant participants. And planned comparisons revealed that the
hypothesis for same;scx_bestjiend was supported for both the Between-Subj ects Sample
[t (717) = 2.41, p<.05] and the Within-Subjects Sample [1(221) = 1.67, p<.05].

E2. Intimacy
It was predicted that secure individuals would have higher ratings on Intimacy for

their best fiiends compared to anxious-ambivalent and avoidant individuals.

91

The hypothesis for W was supported by the Between-Subjects
Sample. Mean differences were as predicted and a planned comparison showed that
secure participants rated their other-sex best fiiend higher on Intimacy than did anxious-
ambivalent and avoidant participants [t (717) = 2.09, p<.05]. However, inspection of the
means for the Within-Subj ects Sample showed that they were not in the predicted
direction. (See Table 7a)

The hypothesis for mm was supported by the Between-Subjects
Sample and marginally supported for Within-Subjects Sample. Mean differences were as
predicted: secure participants rated their same-sex best friend higher on Intimacy than did
anxious-ambivalent and avoidant participants. And planned comparisons yielded a
significant effect for the Between-Subjects Sample [t (717) = 2.04, p<.05] and a
marginally significant effect for the Within-Subjects Sample [t (221) = 1.48, p<.10].

E3. Global Satisfaction
It was predicted that secure and avoidant individuals would have higher ratings on
GlebaLSatisfacticn for their best friends compared to anxious-ambivalent
individuals.

Inspection of the means for ethensezcbestfiend Showed that they were not in the
predicted direction in both the Between-Subj ects Sample and the Within-Subjects
Sample; therefore, the hypothesis was not supported. (See Table 7a)

The hypothesis for samtsexmstfiiend was supported by the Between-Subj ects
Sample. Mean differences were as predicted: secure and avoidant participants rated their

same-sex best fiiend higher on Glebalfiatisfacticn than did anxious-ambivalent

92

participants. A planned comparisons yielded a marginal effect for the Between-Subjects
Sample [t (717) = 1.41, p<.10]. However, inspection of the means for the Within-
Subjects Sample revealed that they were not in the predicted directions. (See Table 7b)
E4. Care
It was predicted that secure individuals would have higher ratings on Care for
their best friends compared to anxious-ambivalent and avoidant individuals.

For both samples, means were as predicted for other-sex best friends: secure
participants rated their other-sex best friend higher on Care than did anxious-ambivalent
and avoidant participants. Moreover, planned comparisons indicated that the hypothesis
was supported by the Between-Subjects Sample [t (717) = 2.26, p<.05] and marginally
supported for the Within-Subjects Sample [1 (221) = 1.47 , p<.10].

For both samples, means were as predicted for same-sex best friends: secure
participants rated their same-sex best friends higher on Care than did anxious-ambivalent
and avoidant participants. And planned comparisons revealed that the hypothesis for
mm was supported for the Between-Subjects Sample [t (717) = 3.69,
p<.01] and marginally supported for the Within-Subjects Sample [t (221) = 1.35, p<.10].
ExtensiQnHypethesesLSummam

The overall result of the extension hypotheses is that they extend the previous
findings of romantic relationships to other-sex and same-sex best fiiendships. Within the
friendship domain, as in romantic relationships, individuals with certain attachment styles
rated certain relationship characteristics higher than did those with other attachment

styles. For the Between-Subjects Sample, secure individuals rated their other-sex best

93

friend higher on viability, intimacy, and care than did either avoidant and anxious-
ambivalent individuals. Secure individuals rated their same-sex best fiiend higher on
viability, intimacy, global satisfaction, and care than did either avoidant and anxious-
ambivalent individuals. For the Within-Subjects Sample, secure individuals rated their
other-sex best fiiend higher on viability and care compared to anxious-ambivalent and
avoidant individuals. Secure individuals rated their same-sex best friend higher on

viability, intimacy, and care compared to anxious-ambivalent and avoidant participants.

94

 

 

 

it It i H n U0 «1.0 0 .JI'- u‘r run 0 auburn
\ 0 I 3‘s. i- I no in u‘ A '- in
Attachment Style
Relationship Samples Avoidant Anxious-Ambivalent Secure
Variables
Viability
Between-SS 7.18 7.27 7.50
Within-SS 7.32 7.31 7.48
Intimacy
Between-SS 6.56 6.97 7.08
Within-SS 6.60 7.17 7.10
Global Satisfaction
Between-SS 6.71 7.11 7.33
Within-SS 6.76 7.37 7.31
Care
Between-SS 6.81 7.1 1 7.29
Within-SS 6.84 7.23 7.29

 

Nate: Between-SS Sample (11 = 735); Within-SS Sample (11 = 227)

95

 

 

 

Iahleflz
U‘alfi.l° OIC<OIIO qun‘ 0 m = I‘M I .i III‘I
.\‘ 0 I'g‘IAHI'..0.'. 10.0.0... I' .O" "0
Attachment Style
Relationship Samples Avoidant Anxious-Ambivalent Secure
Variables
Viability
Between-SS 7.38 7.03 7.61
Within-SS 7.34 7.35 7.62
Intimacy
Between-SS 7.20 6.99 7.48
Within-SS 7.22 7.27 7.51
Global Satisfaction
Between-SS 7.38 7.10 7.61
Within-SS 7.27 7.50 7.53
Care
Between-SS 7.24 6.66 7.60
Within-SS 7.24 7.41 7.58

 

Nate: Between-SS Sample (11 = 735); Within-SS Sample (n = 227)

96

Ii- -. . i... .- -- i- , . -. ... it-.. ...-. . -.. .... u.
S . E . l C . [B l . l . I

Hypotheses D1-D9 were differential hypotheses. Differential hypotheses
concerned attachment style differences in the ratings of certain relationship components
between romantic partner and best fi'iends. Only individuals who had both of the selected
relationships were utilized in the analyses. So for example, only individuals who had
both a romantic partner and an other-sex best friend were selected for analysis in the
romantic partner versus other-sex best friend analyses; thus the number of individuals for
each comparison varied depending on the relationships and the groups being compared.
The Analysis Sample contained the most participants and thus was utilized for the
Differential Hypotheses Dl-D5.

Paired t-tests were conducted to test hypotheses D1-D5. Comparisons were made
between romantic partner and other-sex best fiiend and romantic partner and same-sex
best friend for Intimacy, Global Satisfaction, Care, Passion, and Conflict-Ambivalence
ratings. Means, SDs, and NS for these hypotheses can be foirnd in Table 8.

D1. It was hypothesized that secure and avoidant males would have higher ratings
for Intimacy for romantic partners than best fiiends.

The hypothesis for secure males’ ratings on Intimacy was not confirmed for either
comparison. Secure males did not rate their romantic partner significantly higher
(M=7.00) than they rated their 9111mm (M=6.93) [t (42) = 0.35, p=ns].

Also, the mean difference for same-sex best friend and romantic partner was not as

predicted. (See Table 8)

97

The hypothesis for avoidant males’ ratings on Intimacy also was not confirmed
for either comparison. Avoidant males did not rate their romantic partner significantly
different (M=6.30) fi'om their MW (M=5.87) [t (15) = 1.04, p=ns]. Also
the mean difference for same-sex best friend and romantic partner was not as predicted.
(See Table 8)

D2. It was predicted that avoidant and anxious-ambivalent individuals would have
higher ratings on GlebaLSatisfacticn for their best friends compared to their
romantic partner.-

The hypothesis for avoidant individuals on Global Satisfaction was not supported
for either comparison. Contrary to prediction, avoidant participants rated their ctbensex
bestfriend (M=6.68) lower than their romantic partner (M=6.82). And their rating for
their samefiexbestfijiend on Global Satisfaction (M=7.24) was not significantly higher
from that for their romantic partner (M=7.21) [t (120) = -0.23, p=ns].

The hypothesis for anxious-ambivalent participants’ ratings on Global
Satisfaction was not confirmed for either comparison. Although anxious-ambivalent
participants rated their cthezsexbestfieiid (M=7.37) higher than their romantic partner
(M=7.08), it was not significant. Moreover, their rating for their samegeacbestfiiend
(M=7 .3 6) was also not significantly higher from that for their romantic partner (M=7.10)
[t (61) = 1.00, p=ns].

D3. It was hypothesized that secure and anxious-ambivalent individuals would

have higher ratings on_C_are for romantic partners compared to best fiiends.

98

This prediction for secure participants’ ratings on Care was supported for both
other-sex best friend and same-sex best friend. Secure participants rated their romantic
partner significantly higher (M=7.54) compared to their cthenscxbestfiiend (M=7 .28) [t
(151) = 2.39, p<.05]. Their rating for their romantic partner on Care (M=7.76) was also
significantly higher than for their same;sex_t&st_fricnd (M=7.47) [t (367) = 4.17, p<.001].

The hypothesis for anxious-ambivalent participants’ ratings on Care was not
confirmed for either comparison. Contrary to prediction, anxious-ambivalent participants
rated their romantic partner (M=7. 19) lower than their cthensexjmstjriend (M=7.28) [t
(26) = .39, p=ns]. Furthermore, their rating for their romantic partner (M=7.32) was not
significantly higher compared to their samzscxmstfimd (M=7.10) [t (61) = 1.18,
p=ns].

D4. It was predicted that across all three styles of attachment, individuals would
have higher ratings on Bassicn for their romantic partners compared to their
best fiiends.

This hypothesis was supported for both comparisons. Individuals rated their
romantic partner (M=7.28) significantly higher on Passion compared to their amen-sex
W (M=4.6l) [t (239) = 20.31, p<.001]. They also rated their romantic partner
significantly higher on Passion (M=7.52) compared to their same;5_ex_best_fi:iend
(M=4.34) [t (550) = 40.90, p<.001].

D5. It was predicted that across the three attachment styles, individuals would have
higher ratings on Ccnflietflmbiyalence for their romantic partners compared

to their best friends.

99

This hypothesis was supported for both comparisons. Individuals rated their
mmantiepattner (M=2.97) significantly higher on Conflict-Ambivalence compared to
their WM (M=2.31) [t (239) = 5.71, p<.001]. They also rated their
mmantiepartner significantly higher on Conflict-Ambivalence (M=2.84) compared to
their WW (M=2.04) [t (550) = 10.97, p<.001].

The differential hypotheses D6-D9 were conducted via planned comparisons
within the framework of the 2(Sex) X 3(Attachment Style) X 3(Relationship Type)
design when analyzing the Between-Subjects Sample. For the Within-Subjects Sample,
paired t-tests were conducted. Means, SDs, and NS for these hypotheses can be found in
Table 9.

D6. It was hypothesized that anxious-ambivalent individuals would have higher
ratings on Viability for romantic partner and other-sex best friend compared to
same-sex best friend.

This hypothesis was marginally supported for the Between-Subjects Sample.
Anxious-ambivalent individuals rated their romantic partner (M=7.36) and their other-sex
best friend (M=7.28) significantly higher on Viability than their same-sex best fiiend
(M=7.01) [t (717) = 1.36, p<.10]. However, inspection of the means for the Within-
Subjects Sample showed that they were not in the predicted directions. (See Table 9)
D7. It was hypothesized that avoidant individuals would have higher ratings on

Viability for same-sex best friend, followed by other-sex best fiiend, and

finally by romantic partner.

100

This hypothesis was partially supported for the Between-Subjects Sample.
Avoidant individuals rated their same-sex best friend (M=7.33) marginally higher than
their other-sex best friend (M=7.05) on Viability [t (717) = 1.38, p<. 10]. However,
avoidant individuals did not rate their other-sex best friend (M=7.05) significantly higher
than their romantic partner (M=6.99) [t (717) = 0.32, p=ns]. For the Within-Subjects
Sample, the hypothesis was partially supported. Avoidant individuals did not
differentiate between their same-sex best fiiend (M=7.34) and their other-sex best fiiend
(M=7.32). However, they had higher ratings for other-sex best friend (M=7.32) than their
romantic partner (M=6.84) [t (55) = 2.73, p<.01].

D8. It was hypothesized that secure and avoidant females would have higher
ratings on Intimacy for same-sex best friend, then romantic partner, then other-
sex best fiiend.

For secure females, this hypothesis was not supported in the Between-Subj ects
Sample. They did not rate their same-sex best fiiend (M=7.62) significantly higher on
Intimacy compared to their romantic partner (M=7.56) [t (717) = 0.29, p=ns]. They also
did not rate their romantic partner (M=7.56) significantly higher than their other-sex best
friend (M=7.19) [t (717) = 1.27, ns]. In the Within-Subjects Sample, the hypothesis was
partially supported. Secure females had higher ratings for their same-sex best friend
(M=7.73) than their romantic partner (M=7.16)[t (105) = 3.91, p<.001]. However,
inspections of the means for romantic partner and other-sex best fiiend showed that they

were not in the predicted direction. (See Table 9)

101

For avoidant females, this hypothesis was partially supported in the Between-
Subjects Sample. They did not rate their same-sex best fiiend (M=7.30) significantly
higher onjntimacy compared to their romantic partner (M=7.21) [t (717) = 0.39, p=ns].
However, they did rate their romantic partner (M=7.21) marginally significantly higher
than their other-sex best friend (M=6.8l) [t (717) = 1.60, p<.056]. In the Within-Subjects
Sample, the hypothesis was partially supported. Avoidant females had higher ratings for
their same-sex best friend (M=7.53) than their romantic partner (M=6.l3) [t (40) =
5.51,p<.001]. However, inspections of the means for romantic partner and other-sex best
friend showed that they were not in the predicted direction. (See Table 9)

D9. It was hypothesized that anxious-ambivalent individuals would have higher
ratings on Intimacy for romantic partner and other-sex best friend compared to
same-sex best friend.

This hypothesis was not supported for the Between-Subjects Sample. Anxious-
ambivalent individuals did not rate their romantic partner (M=7.04) and their other-sex
best fiiend (M=6.96) significantly higher onlntimacy than their same-sex best fi'iend
(M=6.90) [t (717) = 0.34, p=ns]. The hypothesis was also not supported for the Within-
Subjects Sample. Inspections of the means revealed that they were not in the predicted
direction. (See Table 9)

The overall result for the differential hypotheses is that depending on the
attachment style, ratings of relationship components among romantic partner and best

friends tended to vary. About half of the hypotheses were not supported. Rather than

102

rate their romantic partner higher than their best friends on Intimacy, secure and avoidant
males did not differentiate on these relationships. In fact, the means were in the opposite
direction for same-sex best fiiend and romantic partner. Avoidant and anxious-
ambivalent individuals did not differentiate between their ratings on Global Satisfaction
for other-sex best friend, same-sex best friend, and romantic partner. Secure individuals
did rate their romantic partner significantly higher on Care than their best friends.
However, anxious-ambivalent individuals did not differentiate in their ratings for
romantic partner, other-sex best friend, and same-sex best fiiend. Secure and avoidant
males perceived intimacy to be as important for their best friends and their romantic
partner. However, avoidant and anxious-ambivalent participants perceived Global
Satisfaction to be as important for their best fiiends and their romantic partner. As
predicted, all participants rated their romantic partner higher on both Passion and
Conflict-Ambivalence compared to their best fiiends. All participants perceived Passion
and Conflict-Ambivalence to be more characteristics of their romantic relationships
compared to their fiiendships. In the BSS, anxious-ambivalent individuals had higher
ratings for Viability for their romantic partner and other-sex best friend compared to their
same-sex best friend. Avoidant individuals had higher ratings on Viability for their
same-sex best friend compared to their other-sex best friend. They had similar ratings on
Viability for their other-sex best friend and their romantic partner. In the WSS, avoidant
individuals had similar ratings on Viability for their same-sex and other-sex best fiiend.
Best fiiends’ ratings were higher than ratings for romantic partner. In the BSS, secure

females and anxious-ambivalent individuals also did not differentiate in their ratings for

103

their romantic partner, other-sex best fiiend, and same-sex best friend on Intimacy.
Avoidant females had similar ratings for their same-sex best friend and their romantic
partner on Intimacy. However, they had higher ratings for their romantic partner
compared to their other-sex best fiiend on Intimacy. In the WSS, secure and avoidant
females had higher ratings for their same-sex best fiiend compared to their romantic
partner on Intimacy. These individuals did not differentiate in their ratings between their
romantic partner and their other-sex best friend on Intimacy. Anxious-ambivalent
individuals did not differentiate in their ratings for their romantic partner, other-sex best

friend, and same-sex best friend on Intimacy.

104

 

 

 

 

 

 

Table}
11 E E. 15 l I-I E l l l . S 1
Group Variable RP vs.OSBF Rva. SSBF
Intimacy
(Secure Males) M 7.00 6.93 7.01 7.09
SD 1.26 1.22 1.35 1.37
N 43 91
(Avoidant Males) M 6.30 5.87 6.30 6.51
SD 1.35 1.35 1.26 1.25
N 16 32
Global Satisfaction
(Avoidant Males) M 6.82 6.68 7.21 7.24
SD 1.51 1.65 1.45 1.40
N 60 120
(Anxious-Ambivalence M 7.08 7.37 7.10 7.36
Males) SD 1.58 1.22 1.67 1.21
N 26 61
Care
(Secure) M 7.54* 7.28 7.76" 7.47
SD 1.28 1.12 1.23 1.06
N 152 368
(Anxious- M 7.19 7.28 7.32 7.10
Ambivalent) SD 1.27 1.06 1.21 1.29
N. 27 62
Passion
(All) M 7.28” 4.61 7.52" 4.34
SD 1.23 1.72 1.16 1.59
N 240 55
Conflict-Ambivalence
(All) M 2.97" 2.31 2.84" 2.04
SD 1.55 1.20 1.59 1.07
N 240 551

 

RP = Romantic Partner; OSBF = Other-Sex Best Friend; SSBF = Same-Sex Best Friend.
** p<.001, *p<.05

105

Iahl:_2 .

U‘e' . v. ‘9 {HO' ' ‘. 0 I‘ A .l I", 0‘. {"0’

 

Group Variable RP vs. OSBF RP vs. SSBF OSBF vs. SSBF RP/OSBF vs. SSBF

 

 

 

Viability
Anxious-Ambivalent
M 7.07 7.35
SD 0.94 1.09
N 26
Viability
Avoidant
M 6.86 7.32* 7.32 7.34
SD 1.11 1.06 1.06 1.14
N 55 55
Intimacy

SecureFemales
M 7.16 7.16 7.16 7.73"
SD 1.33 1.15 1.33 0.92
N 105 105

 

Intimacy
Avoidant Females
M 6.13 6.87 6.13 7.53"
SD 1.46 1.31 1.46 1.07

 

N 40 40
Intimacy
Anxious-Ambivalent
M 6.86 7.27
SD 1.08 1.12
N 26

 

RP= Romantic Partner, OSBF= Other-Sex Best Friend; SSBF= Same-Sex Best Friend,
RP/OSBF= Romantic Partner/Other-Sex Best Friend.
"p<.001, *p<.01

106
l l I. . l H l

The following section contains additional hypotheses. Hypotheses Al-A3
compared males’ and females’ ratings of their same-sex best friend on Intimacy , Global
Satisfaction, and Care. Hypotheses A4-A5 examined the effects of attachment style on
having a romantic partner and best friends. These analyses were conducted on the
Analysis Sample. T-Tests were conducted on hypotheses A1-A3. Chi-squares were
conducted on hypotheses A4-A5.

Al. It was hypothesized that across all attachment styles, women would have
higher ratings on Intimacy for their same-sex best fiiend compared to men.

A t-test was conducted witlrlntimacy ratings of same-sex best fiiend as the
dependent variable and sex as the independent variable. In support of this hypothesis,
across all three attachment styles women had significantly higher ratings on Intimacy for
their same-sex best friend (M=7.56) compared to men (M=6.88) [t (736) = 7.38, p<.001].

A2. It was hypothesized that overall, across the three attachment styles, women
would have higher ratings on GlebaLSatisfacticn for their same-sex best friend
compared to men.

A t-test was conducted with _Glcbal_Satisfaeticn ratings on same-sex best fi'iend as
the dependent variable, and sex as the independent variable. In support of this
hypothesis, across all three attachment styles women had significantly higher ratings on
Wafacticn for their same-sex best fiiend (M=7.57) compared to men (M=6.99) [t

(736) = 5.98, p<.001].

107

A3. It was hypothesized that overall, across the three attachment styles, women
would have higher ratings on Care for their same-sex best friend compared to
men.

A t-test was conducted with Care ratings on same-sex best fiiend as the dependent
variable and sex as the independent variable. In support of this hypothesis, across all
three attachment styles women had significantly higher ratings on Care for their same-sex
best friend (M=7.46) compared to men (M=7.l6) [1(736) = 3.15, p<.005].

A4. It was hypothesized that secure individuals would be more likely to have a
romantic partner compared to anxious-ambivalent and avoidant individuals.

A chi-square was conducted with W (yes, no) as the
dependent variable and attachment style (secure, anxious-ambivalent, avoidant) as the
independent variable. In support of this hypothesis, secure individuals (Yes: 80%, No:
20%) were more likely to have a romantic partner than either anxious-ambivalent (Yes:
70%, No: 30%) and avoidant individuals (Yes: 67%, No: 33%) [x2 (2, N = 796) = 15.33,
p<.001].

A5. It was hypothesized that secure individuals would be more likely to have a best
friend (other-sex, same-sex) compared to both avoidant and anxious-
arnbivalent individuals.

A chi-square was conducted with hayeamstfricnd (yes, no) as the dependent
variable and attachment style (secure, anxious-ambivalent, avoidant) as the independent

variable. This hypothesis was not fully supported. Secure individuals (Yes: 84%, No:

108

16%) were more likely to have an other-sex best fiiend compared to either anxious-
ambivalent (Yes: 76%, No: 24%) and avoidant individuals (Yes: 77%, No: 23%) [x2 (2,
N = 796) = 5.84, p<.055]. However, secure individuals (Yes: 94%, No: 6%) were not
more likely to have a same-sex best friend compared to both anxious-ambivalent (Yes:
97%, No: 3%) and avoidant individuals (Yes: 93%, No: 7%) [x2 (2, N = 796) = 1.48,
p=ns].

Almost all of these hypotheses were supported. As predicted, women had higher
ratings on Intimacy, Global Satisfaction, and Care for their same-sex best fiiend than did
men. Secure participants were more likely to have a romantic partner and an other-sex
best friend compared to anxious-ambivalent and avoidant participants. However, all
attachment styles were found to be similar on their probability of having a same-sex best
fiiend.
Wises

These analyses were conducted within the 2(Sex) X 3(Attachment Style) X
3(Relationship Type) framework. This 2 X 3 X 3 design was instrumental in answering
an important question within the study of attachment style and relationship. This
question was “were there differences between the romantic partner and the other-sex best
friend and the same-sex best friend?” This design was utilized to extend the findings of

previous research, to look at whether different relationship types affected ratings for

109

different relationship variables from the three attachment styles. Both the Between-
Subjects Sample and the Within-Subjects Sample were employed for these analyses.

Because of the expansion of this design, no specific predictions were made.
Instead, general predictions were made for sex and attachment style. It was predicted that
females were expected to have higher ratings on these relationship variables compared to
men. Secure individuals were expected to have higher ratings on these relationship
variables compared to individuals in the two insecure groups.

A total of twelve 2 X 3 X 3 AN OVAS were conducted. Six were on the Between-
Subjects Sample (BSS) and six were on the Within-Subjects Sample (WSS). Thus, two 2
X 3 X 3 AN OVAS were conducted on each relationship variable, one for each sample.
Tukey-HSD post hoc tests were conducted on all significant attachment style and
relationship type main effects for both the BSS and the WSS. Means for the 2 X 3 X 3
AN OVAs for the Between-Subjects Sample and the Within-Subjects Sample can be
found in Tables 10 and 11 respectively. The 2 X 3 X 3 AN OVAs can be found in Table
12 for the Between-Subjects Sample and in Table 13 for the Within-Subjects Sample.
Attachment style means and main effects for the Between-Subj ects Sample and the
Within-Subjects Sample can be found in Table 14. Relationship type means and main
effects for the Between-Subjects Sample and the Within-Subjects Sample can be found in
Table 15. Simple effects for sex by relationship type interactions for the Within-Subj ects
Sample can be found in Table 16. Simple effects for attachment style by relationship

type interaction for the Between-Subjects Sample can be found in Table 17.

110
Viability
For Viability, the BSS showed only two significant main effects and no
interactions. The WSS showed two main effects and one marginally significant
interaction. There was a main effect for sex only for the BSS [E(1,717) = 4.43, p<.05].
Females had higher ratings on Viability than did males. There was a main effect for
anachmcntsfll: for both the BSS [E(2,717) = 9.48, p<.001] and the WSS [E(2,221) =
3.93, p<.05]. For attachment style, secure participants had a significantly higher rating
on Viability compared to anxious-ambivalent and avoidant participants. There was a
main effect for relaticnsliiiLtype only for the WSS [E(2,442) = 6.56, p<.005]. For
relationship type, participants had significantly higher ratings on Viability for their same-
sex and other-sex best friend compared to their romantic partner. Ratings for the same-
sex and other-sex best fiiend did not differ significantly from each other.
There was a marginally significant scxbyielaticnsliintym interaction for the

WSS [E(2,442) = 2.96, p<.055]. Simple effects tests conducted on the interaction showed
that females had a significantly higher rating for same-sex best friend on Viability
compared to their romantic partner. Ratings for other-sex best fiiend did not differ from
the other two groups. Males rated their other-sex best friend significantly higher than
their romantic partner. Ratings for the same-sex best fiiend did not differ from the other
two groups.
Intimacy

For Intimacy, the BSS showed three significant main effects but no interactions.

The WSS showed three main effects and one significant interaction. There was a main

111

effect for sex for both the BSS [E(1,717) = 23.34, p<.001] and the WSS [E(1,221) = 8.81,
p<.005]. Females had higher ratings on Intimacy than did males. There was a main
effect for W1: for both the BSS [E(2,717) = 10.58, p<.001] and the WSS
[E(2,221) = 7.86, p<.005]. For attachment style, secure participants had a significantly
higher rating on Intimacy compared to anxious-ambivalent and avoidant participants for
the BSS. Anxious-ambivalent participants had a significantly higher rating compared to
avoidant participants. For the WSS, secure and anxious-ambivalent participants had
significantly higher ratings compared to avoidant participants. Ratings for secure and
anxious-ambivalent participants did not differ from each other. There was a main effect
for relationshimize for both the BSS [E(2,717) = 3.42, p<.05] and the WSS [E(2,442) =
8.84, p<.001]. For relationship type, for the BSS participants had significant higher
ratings for their romantic partner and their same-sex best fiiend compared to their other-
sex best fiiend. Ratings for their same-sex best friend and romantic partner did not differ
significantly from each other. For the WSS, participants had a significantly higher rating
for their same-sex best friend compared to both their other-sex best friend and their
romantic partner. Ratings for their other-sex best friend were significantly higher
compared to their romantic partner.

There was a sexbynelaticnshim interaction for the WSS [E(2,442) = 4.21 ,
p<.05]. Simple effects tests conducted on the interaction Showed that females had
significantly higher ratings for their same-sex best friend onlntimacy compared to both
their other-sex best fiiend and romantic partner. Ratings for their other-sex best fiiend

and their romantic partner did not differ significantly from each other. Males had a

112

significantly higher rating for their same-sex best friend compared to their other-sex best
fiiend and their romantic partner. Ratings for their romantic partner and their other-sex
best friend did not differ from each other. The difference between the three types are
greater for females.
Global Satisfaction

For Global Satisfaction, the BSS Showed three significant main effects but no
interactions. The WSS showed two main effects and one significant interaction. There
was a main effect for sex for both the BSS [E(1,717) = 12.48, p<.001] and the WSS
[E(1,221) = 4.83, p<.05]. Females had higher ratings on Glebalfiatisfacticn than did
males. There was a main effect for attachmanstyle for both the BSS [E(2,717) = 10.77,
p<.001] and the WSS [E(2,221) = 4.28, p<.05]. For attachment style, secure participants
had a significantly higher rating on GlebaLSatisfacticn compared to both anxious-
ambivalent and avoidant participants for the BSS. Anxious-ambivalent and avoidant
participants did not differ from each other. In the WSS, secure and anxious-ambivalent
participants had significant higher ratings on GlcbaLSatisfacticii compared to avoidant
participants. Secure and anxious-ambivalent participants did not differ from each other.
There was a main effect for relationshipjjme only for the BSS [E(2,717) = 11.26,
p<.001]. For relationship type, participants had a significantly higher rating on Global
Satisfaction for their romantic partner compared to both their same-sex and other-sex best
fiiend. Ratings for their same-sex best friend were also significantly higher than their

other-sex best fiiend.

113

There was a mhxrelaucnshintxn: interaction for the WSS [E(2,442) = 3.35,

p<.05]. Simple effects tests conducted on the interaction showed females had
significantly higher ratings for their same-sex best fiiend on GlebaLSatisfacticn
compared to both their other-sex best friend and their romantic partner. Ratings for their
other-sex best friend and their romantic partner did not differ significantly from each
other. Males had significantly higher ratings for their romantic partner and their other-
sex best friend compared to their same-sex best friend. Ratings for romantic partner and
other-sex best friend did not differ Significantly from each other.
Care

For Care, the BSS showed three significant main effects and a marginally
significant interaction. The WSS showed only two main effects and no interactions.
There was a main effect for sex only for the BSS [E(1,717) = 12.55, p<.001]. Females
had higher ratings on Care than did males. There was a main effect for attacbmmLstyle
for both the BSS [E(2,717) = 13.93, p<.001] and the WSS [E(2,221) = 3.39, p<.05]. For
attachment style, for the BSS secure participants had a significantly higher rating on Care
compared to both anxious-ambivalent and avoidant participants. Anxious-ambivalent
and avoidant participants did not differ significantly from each other. For the WSS,
secure participants had a significantly higher rating on Care compared to avoidant
participants. Anxious-ambivalent participants did not differ from the other two groups.
There was a main effect for relationshimype for both the BSS [E(2,717) = 8.38, p<.001]
and the WSS [E(2,221) = 2.87, p<.06]. For relationship type, participants had a

significantly higher rating on Care for their romantic partner compared to both their

114

same-sex and other-sex best friend in the BSS. Ratings for their same-sex best friend was
also significantly higher than their other-sex best friend. In the WSS, participants had
significantly higher ratings for their same-sex best fiiend compared to both their romantic
partner and their other-sex best friend. Ratings for their romantic partner were also
significantly higher compared to their other-sex best friend.

There was an attachmentstylebyrclaticnsbipfipe interaction for the BSS
[E(4,717) = 2.34, p<.055]. Simple effects tests conducted on the interaction found that
avoidant participants had significantly higher ratings on (lane for their romantic partner
compared to both their other-sex and same-sex best friend. They also rated their same-
sex best fiiend significant higher than their other-sex best friend. Anxious-ambivalent
participants had significantly higher ratings on Care for their romantic partner compared
to both their other-sex and same-sex best friend. They also rated their other-sex best
friend significantly higher than their same-sex best fiiend. Secure participants had
significantly higher ratings on Care for their romantic partner compared to both their
other-sex and same-sex best friend. They also rated their same-sex best friend
significantly higher than their other-sex best friend.

Passion

For Passion, the BSS showed three Significant main effects but no interactions.
The WSS showed three main effects and one interaction. There was a main effect for sex
for both the BSS [E(1,717) = 12.91, p<.001] and the WSS [E(1,221) = 3.94, p<.05).
Females had higher ratings on Passion than did males. There was a main effect for

attachmentstyle for both the BSS [E(2,717) = 10.34, p<.001] and the WSS [E(2,221) =

115

7.91, p<.001]. For attachment style, anxious-ambivalent participants had a significantly
higher rating on Passicn compared to secure and avoidant participants. Secure
participants had a significantly higher rating compared to avoidant participants. There
was a main effect for rclaticnship_type for the BSS [E(2,717) = 219.25, p<.001] and the
WSS [E(2,221) = 192.63, p<.001]. For relationship type, participants had a significantly
higher rating on Bassicn for their romantic partner compared to both their same-sex and
other-sex best friend. Ratings for the other-sex best fiiend were significantly higher
compared to the same-sex best friend.

There was a sexbyrelaticnshipjim interaction for the WSS [E(2,442) = 3.95,
p<.05]. Simple effects conducted on the interaction found that females had a
significantly higher rating on Bassicn for their romantic partner compared to both their
other-sex and same-sex best friend. Ratings for the same-sex and other-sex best friend
were not significantly different from each other. Males had a significantly higher rating
on Bassicn for their romantic partner compared to their other-sex best friend and their
same-sex best friend. Ratings for the other-sex best friend were significantly higher
compared to their same-sex best friend.

Conflict-Ambivalence

For Conflict-Ambivalence, both the BSS and the WSS showed two significant
main effects and no interactions. There was a main effect for attachmeantyle for the
BSS [E(2,717) = 5.27, p<.01] and the WSS [E(2,221) = 5.22, p<.01]. For attachment
style, for the BSS anxious-ambivalent participants had a significantly higher rating on

Conflict-Ambivalence compared to avoidant and secure participants. Avoidant

116

participants also had a significantly higher rating compared to secure participants. For
the WSS, anxious-ambivalent participants had a significantly higher rating on Conflict—
Ambivalence compared to secure and avoidant participants. Secure and avoidant
participants did not differ significantly from each other. There was a main effect for
relaticnshiiytype for the BSS [E(2,717) = 14.42, p<.001] and the WSS [E(2,221) = 23.31,
p<.001]. For relationship type, participants had significantly higher ratings for their
romantic partner compared to both their same-sex best friend and their other-sex best
friend. Ratings for the other-sex best fiiend were also significantly higher compared to
the same-sex best friend.

WW

Overall, the results revealed many interesting findings. There were many
variables examined here within this 2(Sex) X 3(Attachment Style) X 3(Relationship
Type) framework that have not been examined in previous research. For the six
relationship variables (i.e., viability, intimacy, global satisfaction, care, passion, and
conflict-ambivalence), there were main effects for sex, attachment style, and relationship
type for both samples. There were also Significant interactions.

For the BSS, there were five main effects for sex. Conflict-ambivalence was the
only variable that did not show sex differences. All main effects were Significant for
attachment style. There were five main effects for relationship type. Viability was the
only variable that did not have relationship type differences. Overall, this set of analyses
showed only one marginally significant interaction; for attachment style by relationship

type on Care.

117

For the WSS, there were three main effects for sex. These variables were
intimacy, global satisfaction, and passion. All main effects were significant for
attachment style. There were five main effects for relationship type. Global satisfaction
was the only variable that did not show relationship type differences. Overall, this set of
analyses showed four significant interactions; all of sex by relationship type. These
interactions were on Viability, Intimacy, Global Satisfaction, and Passion.

For these analyses, two samples were employed, the BSS and the WSS. The WSS
has more power and is more sensitive to treatment/nonsubj ect effects compared to the
BSS. Individuals within each sample were different in that for the WSS, individuals had
to have all three types of relationships. These individuals have more intimate
relationships compared to those in the BSS. For the BSS, individuals could have one,
two, or three types of relationships but were only selected for one type of relationship for
analyses. There were benefits to conducting parallel analyses for these two samples.
Findings could be considered as more robust and consistent across different types of
designs.

In summary, both samples yielded similar results for main effects. They revealed
exactly the same main effects results for attachment style. Secure individuals had the
highest ratings for most of the positive relationship variables compared to anxious-
ambivalent and avoidant individuals. Effects for attachment style were similar to
previous research findings (e.g., Collins & Read, 1990; Hazan & Shaver, 1987). These
analyses revealed a recurring pattern for attachment style as predicted in earlier

hypotheses (i.e., planned comparisons) in that secure individuals received higher ratings

118

for positive relationship variables compared to their insecure counterparts. For Passion
and Conflict-Ambivalence, anxious-ambivalent had the highest ratings which is similar to
Kirkpatrick and Davis (1994).

These analyses also revealed a recurring pattern Of sex differences. There was an
overall sex effect in that across all three attachment styles women tended to have higher
ratings for these relationship characteristics compared to men. There were no sex
differences for Conflict-Ambivalence. The two samples shared three similar main effects
findings for sex. Effects for sex were also similar to previous research findings. Davis,
Todd, and Denneny (1987) found sex differences for Intimacy and Care/ Support where
females had higher ratings on these variables compared to males.

These analyses also revealed a recurring pattern for main effects on relationship
type differences for the two samples. For the BSS, romantic partner had significantly
higher ratings for Global Satisfaction, Care, Passion, and Conflict-Ambivalence
compared to same-sex best friend and other-sex best friend. Romantic partner and same-
sex best friend had significantly higher ratings for Intimacy compared to other-sex best
friend. For the WSS, same-sex best fiiend had significantly higher ratings for intimacy
and care compared to romantic partner and other-sex best friend. Same-sex best friend
and other-sex best friend had significantly higher ratings on Viability compared to
romantic partner. For both samples, on Passion and Conflict-Ambivalence, romantic
partner had significantly higher ratings compared to other-sex and same-sex best fiiend.
Other-sex best friend also had significantly higher ratings on these variables compared to

same-sex best fiiend.

119

Because the WSS was more powerful and sensitive to treatment effects, it
revealed four significant interactions compared to one marginally significant interaction
for the BSS. The four WSS interactions concerned Sex by Relationship Type.
Attachment style was not an issue. Therefore, any which way you examine it, these
analyses show that attachment style and relationship type patterns are robust.

Overall, the findings here are important in that attachment style does make a
difference. The difference is apparent in all three types of relationships examined here.
This generalizability of attachment style should shed new light on this area of research as
well as support the theory of attachment in that attachment styles are consistent across
relationships. I have replicated previous research on attachment style and relationship
variables for romantic partner as well as showing that these patterns are also true for other

close relationships.

120

Wateran

 

Relationship Variables

 

Viability Intimacy Global Care Passion Conflict-

 

 

 

 

Satisfaction Ambivalence
M M M M M M
Sex
Females 7.50 7.33 7.63 7.53 5.78 2.31
Males 7.23 6.77 7.14 7.11 5.16 2.50
Attachment Style
Avoidant 7.20 6.90 7.19 7.14 5.32 2.52
Anxious- 7.21 7.03 7.26 7.08 6.06 2.69
Ambivalence
Secure 7.58 7.34 7.69 7.61 5.65 2.22
Relationship Type

Romantic Partner 7.43 7.28 7.88 7.73 7.72 2.74

Other-Sex Best 7.38 6.92 7.13 7.13 4.70 2.29
Friend

Same-Sex Best 7.47 7.34 7.48 7.39 4.42 2.06
Friend

 

Netezn=735

121

11 fl 1 22132135110!!! E l Elli-51' S l

 

Relationship Variables

 

Viability Intimacy Global Care Passion Conflict-

 

 

 

 

Satisfaction Ambivalence
M M M M M M
Sex
Females 7.40 7.21 7.45 7.41 5.54 2.33
Males 7.22 6.75 7.00 7.15 5.12 2.55
Attachment Style
Avoidant 7.17 6.67 6.95 7.03 5.04 2.47
Anxious- 7.16 7.00 7.30 7.26 5.96 2.90
Ambivalence
Secure . 7.45 7.24 7.46 7.46 5.47 2.28
Relationship Type
Romantic Partner 7.11 6.82 7.31 7.35 7.27 2.98
Other-Sex Best 7.42 6.99 7.18 7.17 4.62 2.23
Friend

Same-Sex Best 7.52 7.41 7.46 7.48 4.37 1.98
Friend

 

Nam: df = 227

122

 

 

 

 

 

 

 

Iah1:_12
. 6U III-I I‘ . Ii‘.IIIII II.‘ it. an“
II. .' =. II -_' ':, II =II III -;III.‘I ' I I‘I‘h‘vi‘I-
Sampl:
Source Dependent Variables E df p<
Viability
Sex 4.43 1 .05
Attachment Style 9.48 2 .001
Relationship Type 0.04 2 ns
Sex X Attachment Style 0.87 2 ns
Sex X Relationship Type 0.04 2 ns
Attachment Style X Relationship Type 1.28 4 ns
Sex X Attachment Style X Relationship Type 0.43 4 ns
Intimacy
Sex 23.34 1 .001
Attachment Style 10.58 2 .001
Relationship Type 3.42 2 .05
Sex X Attachment Style 1.25 2 nS
Sex X Relationship Type 0.01 2 ns
Attachment Style X Relationship Type 1.08 4 ns
Sex X Attachment Style X Relationship Type 0.47 4 us
Global Satisfaction
Sex 12.48 1 .001
Attachment Style 10.77 2 .001
Relationship Type 11.26 2 .001
Sex X Attachment Style 0.49 2 ns
Sex X Relationship Type 0.00 2 ns
Attachment Style X Relationship Type 1.89 4 ns
Sex X Attachment Style X Relationship Type 1.12 4 ns

 

NQI::df=717

123

 

 

 

 

 

 

 

Iah1:12_(:omldl.
. at... III‘I s' . Ci‘-..II II AI‘ at. 6“
Io. .' -. II «T '._, II -II I!" é-qu't' I .I‘:‘h"1'
Sample
Source Dependent Variables E df p<
Care
Sex 12.55 1 .001
Attachment Style 13.93 2 .001
Relationship Type 8.38 2 .001
Sex X Attachment Style 0.25 2 ns
Sex X Relationship Type 0.07 2 ns
Attachment Style X Relationship Type 2.34 4 .055
Sex X Attachment Style X Relationship Type 0.89 4 ns
Passion
Sex 12.91 1 .001
Attachment Style 10.34 2 .001
Relationship Type 219.25 2 .001
Sex X Attachment Style 1.43 2 ns
Sex X Relationship Type 1.65 2 ns
Attachment Style X Relationship Type 1.36 4 ns
Sex X Attachment Style X Relationship Type 1.03 4 ns
Conflict-Ambivalence
Sex 1.09 1 ns
Attachment Style 5.27 2 .01
Relationship Type 14.42 2 .001
Sex X Attachment Style 0.09 2 ns
Sex X Relationship Type 0.18 2 ns
Attachment Style X Relationship Type 1.53 4 ns
Sex X Attachment Style X Relationship Type 0.68 4 ns

 

Nutmdf= 717

124

 

 

 

 

 

 

 

13121113
‘ . it... III‘I I‘ . ‘1‘..IIII AI.‘ in. 5AI.I:'|[ ‘
U's. ‘ II' Cue. II-., .' ._ II -'. II :II III
11°] fl llll'l'-Sl’ S 1
Source Dependent Variables E df p<
Viability
Between-Subjects Effects
Sex 0.84 1 ns
Attachment Style 3.93 2 .05
Sex X Attachment Style 1.41 2 ns
Within-Subj ects Effects
Relationship Type 6.56 2 .005
Sex X Relationship Type 2.96 2 .055
Attachment Style X Relationship Type 0.26 4 ns
Sex X Attachment Style X Relationship Type 1.08 4 ns
Intimacy
Between-Subj ects Effects
Sex 8.81 1 .005
Attachment Style 7.86 2 .005
Sex X Attachment Style 0.28 2 ns
Within-Subj ects Effects
Relationship Type 8.84 2 .001
Sex X Relationship Type 4.21 2 .05
Attachment Style X Relationship Type 1.69 4 ns
Sex X Attachment Style X Relationship Type 1.50 4 ns
Global Satisfaction
Between-Subj ects Effects
Sex 4.83 1 .05
Attachment Style 4.28 2 .05
Sex X Attachment Style 0.27 2 ns
Within-Subjects Effects
Relationship Type 1.22 2 ns
Sex X Relationship Type 3.35 2 .05
Attachment Style X Relationship Type 1.59 4 ns
Sex X Attachment Style X Relationship Type 0.99 4 ns

 

Nam: Between-Subjects Effects (df = 221), Within-Subjects Effects (df = 442)

125

 

 

 

 

 

 

 

Iahl:_13_(:cnt1dl.

‘. . 5L! III‘I 1‘ . Ci‘eIIII II; “I ; lili‘t‘e‘
U‘q. ,‘ II' III“ .09 ,. . II ‘ '. II :II III
ll'l E l llC'L-Sl' S 1

Source Dependent Variables E df p<
Care
Between-Subjects Effects
Sex 0.64 1 ns
Attachment Style 3.39 2 .05
Sex X Attachment Style 1.12 2 ns
Within-Subjects Effects
Relationship Type 2.87 2 .06
Sex X Relationship Type 0.15 2 ns
Attachment Style X Relationship Type 0.61 4 ns
Sex X Attachment Style X Relationship Type 2.15 4 ns
Passion
Between-Subjects Effects
Sex 3.94 1 .05
Attachment Style 7.91 2 .001
Sex X Attachment Style 0.51 2 ns
Within-Subjects Effects
Relationship Type 192.63 2 .001
Sex X Relationship Type 3.95 2 .05
Attachment Style X Relationship Type 0.63 4 ns
Sex X Attachment Style X Relationship Type 2.05 4 ns
Conflict-Ambivalence
Between-Subjects Effects
Sex 1.76 1 ns
Attachment Style 5.22 2 .01
Sex X Attachment Style 0.09 2 ns
Within-Subjects Effects
Relationship Type 23.31 2 .001
Sex X Relationship Type 1.14 2 ns
Attachment Style X Relationship Type 0.21 4 ns
Sex X Attachment Style X Relationship Type 1.56 4 ns

 

Nate: Between-Subjects Effects (df = 221), Within-Subjects Effects (df = 442)

126

 

 

 

Iablelfi
a1... III‘I \ 'uni in UL! i‘. o 1' 3'I.“i- o act 1 '1-
Subjectsfiamples
Attachment Style
Avoidant Anxious- Secure
Ambivalence

Variables Samples M M _M E (dt)
Viability

Between-SS 720° 721“ 7.58” 9.48 (2,717)

Within-SS 7.17a 7.16“ 7.45b 3.93 (2,221)
Intimacy

Between-SS 6.90“ 7.03” 734° 10.58 (2,717)

Within-SS 6.67“ 7.00” 7.24" 7.86 (2,221)
Global Satisfaction

Between-SS 7.19“ 7.26“ 7.69” 10.77 (2,717)

Within-SS 6.95“ 7.30b 7.46b 4.28 (2,221)
Care

Between-SS 7.14a 7.08” 7.61b 13.93 (2,717)

Within-SS 7.03a 7.26“” 7.46b 3.39 (2,221)
Passion

Between-SS 532° 606° 5.65” 10.34 (2,717)

Within-SS 5.04° 5.96° 5.47b 7.91 (2,221)
Conflict-Ambivalence

Between-SS 2.52b 269° 2.22' 5.27 (2,717)

Within-SS 2.47' 2.90” 2.28“ 5.22 (2,221)

 

N919: All values identified with different superscripts are significantly different from

each other at the_p<.05 level.

127

 

 

 

131219.15
{‘HII II II; URI-1.0 Uol I I I :‘l I- I I 1 .II'
SubjectsSamples
Relationship Type
Romantic Other-Sex Same-Sex
Partner Best Friend Best Friend

Variables Samples M M _M E (df)
Viability

Between-SS 7.43 7.38 7.47 0.04 (2,717)

Within-SS 7.11” 7.42a 7.52a 6.56 (2,442)
Intimacy

Between-SS 728° 6.92b 7.34a 3.42 (2,717)

Within-SS 682° 6.99” 7.41a 8.84 (2,442)
Global Satisfaction

Between-SS 7.88” 7.13° 7.48” 11.26 (2,717)

Within-SS 7.31 7.18 7.46 1.22 (2,442)
Care

Between-SS 7.73a 7.13° 7.39b 8.38 (2,717)

Within-SS 7.35b 7.17° 7.48“ 2.37 (2,442)
Passion

Between-SS 7.72a 4.70” 4.42° 219.25 (2,717)

Within-SS 7.27‘ 4.62b 4.37° 192.63 (2,442)
Conflict-Ambivalence

Between-SS 2.74“ 2.29b 206° 14.42 (2,717)

Within-SS 2.98“ 2.23b 1.98° 23.31(2,442)

 

N919: All values identified with different superscripts are significantly different fi'om
each other at the .p<.OS level.

128

 

 

 

121121116
III"i‘ I : .‘III II [If I‘t‘, II I I‘ .II.- I}. .III‘
Relationship Type
Romantic Other-Sex Same-Sex
Partner Best Friend Best Friend
Variables Sex M M M E (dt)
Viability
Females 7.10b 7.45ab 7.66“ 19.52 (2,320)
Males 7.13 7.35 7.18 0.90 (2,130)
Intimacy
Females 6.86b 7.10” 766‘ 26.69 (2,320)
Males 6.73 6.72 6.80 0.11 (2,130)
Global Satisfaction
Females 7.40b 7.24” 7.70“ 8.34 (2,320)
Males 7.09 7.04 6.88 0.63 (2,130)
Passion
Females 7.33“ 4.69b 4.61b 236.81 (2,320)
Males 7.13a 4.46” 377° 11 1.48 (2,130)

 

Ngm: All values, in the same rows, identified with different superscripts are significantly
different from each other at the_p<.05 level.

129

 

 

 

Iablszll
III"i‘. I 6|.IIII'I 5‘: :.€..'. II; I‘I: II I I‘:‘\l“I-
Subjectsfiample
Relationship Type
Romantic Other-Sex Same-Sex
Partner Best Friend Best Friend
Variable M M M E ((10
Attachment Style
Care
Avoidant 7.38“ 6.81b 7.23““ 3.64 (2,201)
Anxious-Ambivalent 7.48al 7.11“” 6.66b 3.20 (2,87)
Secure 7.94“ 729° 7.60b 14.66 (2,438)

 

Note: All values, in the same rows, identified with different superscripts are significantly
different from each other at the _p<.05 level.

130

SupplementalAnalxses

Supplemental analyses were performed on the Analysis Sample for questions
pertaining to relationships that were on the Demographic and Social History
Questionnaire. This questionnaire can be found in Appendix B-2. Responses were
analyzed in the fiamework of a 2(Sex) X 3(Attachment Style) AN OVA. The questions
pertained to all three relationship types (i.e., romantic partner, other-sex best friend,
same-sex best friend). Questions examined duration of relationships (Length of Time),
closeness (Closeness), expectation of still being with this person in the future
(Expectation of Future Togethemess), and number of previous relationship of this types
(Past Numbers).

For Lengflmflimmmmmanfigpamfi, there were two main effects but no
interaction. There was an effect for sex [E(1,628) = 4.00, p<.05]. Females had higher
ratings for Length of Time than did males. There was an effect for attachment style
[E(2,628) = 3.54, p<.05]. Secure individuals had higher ratings for Length of Time than
did avoidant or anxious-ambivalent individuals. For W, there was a
marginally significant main effect for attachment style [E(2,641) = 2.41, p<.09] but no
interaction. Secure individuals tended to have higher ratings for Length of Time than did
avoidant individuals. For W, there was a main effect for attachment
style [E(2,746) = 3.02, p<.05] but no interaction. Secure individuals had higher ratings
for Length of Time than did avoidant individuals. These means can be found in Table 18.

For MW, there was a main effect for attachment style

[E(2,624) = 5.39, p<.01] but no interaction. Secure individuals had higher ratings on

131

Closeness than did avoidant individuals. For WW, there were two main
effects but no interaction. There was an effect for sex [E(1,651) = 8.02, p<.01]. Females
had higher ratings on Closeness than did males. There was an effect for attachment style
[E(2,651) = 8.35, p<.001]. Secure and anxious-ambivalent individuals had higher ratings
on Closeness than did avoidant individuals. For WW, there were two
main effects and one interaction. There was an effect for sex [E(1,754) = 22.81, p<.001].
Females had higher ratings on Closeness than did males. There was an effect for
attachment style [E(2,754) = 5.46, p<.01]. Secure and anxious-ambivalent individuals
had higher ratings on Closeness than did avoidant individuals. There was also an
interaction for sex by attachment style [E(2,754) = 3.38, p<.05]. With each attachment
style, males had lower ratings than did females. These sex difierences were significant
for secure and avoidant groups but not for the anxious-ambivalent group. These means
can be found in Table 19.

For WWW, there was a main
effect and a marginally significant interaction. There was an effect for attachment style
[E(2,514) = 4.43, p<.05]. Secure individuals had higher ratings on Expectations of Future
Togethemess than did avoidant individuals. There was also a marginally significant
interaction for sex by attachment style [E(2,514) = 2.80, p<.07]. Secure females had
significantly higher ratings compared to anxious-ambivalent males and avoidant females.
For WW, there were two main effects but no interaction. There was an
effect for sex [E(1,611) = 4.58, p<.05]. Females had higher ratings on Expectations of

Future Togethemess than did males. There was an effect for attachment style [E(2,611) =

132

5.31, p<.005]. Secure individuals had higher ratings on Expectations of Future
Togethemess than did avoidant individuals. For WW, there were no main
effects or interaction for Expectations of Future Togethemess. These means can be found

in Table 20.

For WWW, there was a marginally significant main

effect for sex [E(2,715) = 3.36, p<.07] but no interaction. Males had a higher number of
past romantic partners than did females. For Past Numbers of W and
W, there were no main effects or interactions. These means can be

found in Table 21.

133

IahlLLS

U‘-.II¢.II° II ‘IuII II‘aIiIII-I 9'..- .JI" ’.:"f IO’I' -_II

W822;

 

 

 

Relationship Type
Romantic Other-Sex Same-Sex
Partner Best Friend Best Friend
Variable
Length of Time
Secure
Females 3.28 4.01 4.60
Males 2.89 4.06 4.64
Anxious-Ambivalent
Females 2.93 3.84 4.54
Males 2.27 4.04 4.56
Avoidant
Females 2.80 3.77 4.36
Males 2.89 3.72 4.45

 

Note: A rating of: 1 = 1-3 months; 2 = 4-6 months; 3 = 7-11 months; 4 = 1-2 years;
5 = 3+ years.

134

 

 

 

Iahlralfi
U'qIfiI‘ II I‘I' llIiIIIiI 'IIII' .I'-'.:"f I'II III
BestifiendthttaehmentfiflleandSex
Relationship Type
Romantic Other-Sex Same-Sex
Partner Best Friend Best Friend
Variable
Closeness
Secure
Females 7.71 7.46 7.89
Males 7.38 7.21 7.33
Anxious-Ambivalent
Females 7.52 7.80 7.80
Males 6.84 7.20 7.48
Avoidant
Females 6.69 7.02 7.76
Males 7.03 6.28 6.53

 

Note: Ratings can range from 1 = Not at all to 9 = Extremely.

135

U'eIti. I9 II .II‘ .c. III ,1.’ I°‘.I'II‘ .IIiIII-II 'IIII‘ I

 

 

 

 

Relationship Type
Romantic Other-Sex Same-Sex
Partner Best Friend Best Friend
Variable
Expectation of Future Togethemess
Secure
Females 2.88 3.28 3.52
Males 2.79 3.21 3.42
Anxious-Ambivalent
Females 2.69 3.39 3.54
Males 1.75 2.78 3.07
Avoidant
Females 2.36 2.90 3.39
Males 2.66 2.70 3.44

 

Note: A rating of: 1 = 1-2 years; 2 = 3-4 years; 3 = 5-6 years; 4 = 7+ years.

136

 

 

 

IahlszZl
U'el -_I° II'-.. \_III1‘ I {IIIeI .9!" .I - 3‘ I'II III III"
SexBestEriendhLAttachmenLSmlLandjex
Relationship Type
Romantic Other-Sex Same-Sex
Partner Best Friend Best Friend
Variable
Past Numbers
Secure
Females 3.55 2.71 3.78
Males 3.81 3.27 3.86
Anxious-Ambivalent
Females 3.75 2.68 4.35
Males 4.58 3.16 3.94
Avoidant
Females 3.77 3.10 3.87
Males 3.96 2.95 3.96

 

Note: Ratings can range from 1 to 9 persons.

CHAPTER IV

DISCUSSION

The goal of the present study was to further understand the relationship between
attachment style and close interpersonal relationships. Specifically examined was how
individuals within each attachment style perceive different relationship components among
different types of interpersonal relationships, such as romantic relationships and best
friendships. The three attachment styles were secure, anxious-ambivalent, and avoidant
and the three relationship types were romantic relationship, other-sex best fiiend, and same-
sex best fiiend.

Most researchers in this area have looked at attachment style and romantic
relationships. In the present study, I examined attachment style and three types of close
relationships (i.e., romantic relationship, best fi'iendships of the other-sex and same-sex).
This study, then, embeds adult attachment research into a broader interpersonal relationship
matrix by expanding the generalizability of attachment style to different types of
relationships. Thus, I have not only replicated findings for romantic partner (e. g., Collins &
Read, 1990; Feeney & Noller, 1990; Hazan & Shaver, 1987; Kirkpatrick & Davis, 1994;
Simpson, 1990) but have found that this pattern, between attachment style and romantic

relationships is also true for other types of relationships (i.e., other-sex best friend, same-sex

137

138

best friend). Researchers such as Bartholomew and Horowitz (1991) have suggested that
mental models are constant across close relationships. My findings indicate that mental
models are constant across romantic partner, other-sex best fi'iend, and same-sex best fiiend.

' Two different designs were examined (i.e., between-subjects, within-subjects) and
regardless of the design type, the results were similar for the same attachment style. Over
all relationship types, secure individuals had higher ratings for Viability, Intimacy, Global
Satisfaction, and Care compared to either anxious-ambivalent or avoidant individuals.
Anxious-ambivalent individuals had higher ratings for Passion and Conflict-Ambivalence
compared to secure individuals.

For the relationship variables of Viability and Intimacy all styles had higher ratings
for their same-sex best fi'iend compared to their romantic partner. For variables more
related to romantic relationships such as Passion and Conflict-Ambivalence, or associated
with romantic relationships from previous research (e. g., Kirkpatrick & Davis, 1994) such
as Global Satisfaction and Care, individuals in all three attachment styles had higher ratings
for their romantic partner compared to either their other-sex best friend or their same-sex
best fi'iend.

E l' .

The main purpose of these hypotheses was to replicate previous findings (e. g.,
Kirkpatrick & Davis, 1994; Levy & Davis, 1988) on attachment style and romantic
partners. For most of the relationship components examined in this study-Viability,
Intimacy, Global Satisfaction, and Careusecure individuals were predicted, and found, to

have higher ratings for their romantic partner compared to both anxious-ambivalent and

139

avoidant individuals. Those hypotheses that were not supported included the following.
For the Care dimension, attachment style and sex were predicted to interact but did not.
Based on research by Kirkpatrick and Davis (1994), it was predicted that secure and
anxious-ambivalent males would have higher ratings on Care for their romantic partner
compared to avoidant men. It was also predicted, based on Kirkpatrick and Davis (1994),
that secure and avoidant women would have higher ratings on Care for their romantic
partner compared to anxious-ambivalent women. Neither Care hypothesis was supported.
It was found that secure males had significantly higher ratings than either anxious-
ambivalent and avoidant males. Secure females had the highest ratings followed by
anxious-ambivalent females, then by avoidant females. Overall, secure males and females
had higher ratings on Care than anxious-ambivalent and avoidant individuals.

Two other relationship components were examined. For Passion, anxious-
arnbivalent individuals were predicted, and found, to have higher ratings compared to both
secure and avoidant individuals. This finding supports those of Kirkpatrick and Davis
(1994), who found anxious-ambivalent individuals to have the highest ratings on passion
compared to secure and avoidant individuals. The desire and obsession that anxious-
ambivalent individuals have about their romantic partner may bring about the high level of
intensity and passion within their relationships. For Conflict-Ambivalence, avoidant
individuals were predicted to have higher ratings compared to both secure and anxious-
ambivalent individuals; however, this hypothesis was not supported. Instead, anxious-
ambivalent individuals had higher ratings compared to secure individuals. The reason for

these findings may be that anxious-ambivalent individuals, because of their perceptions of

140

others and relationships as being inconsistent and unreliable, could become very intense in
their relationships. This intensity could cause the level of both passion and conflict-
ambivalence to be very high. The anxious-ambivalent people’s perceptions that others and
relationships have a limited positive time fiame may cause them to try to get as much out of
their partner as possible. Draining positive affects from their partner in a short span of time
would lead the relationship to enter the conflict state sooner than typical. Because of their
perceptions of others and relationships, such people experience more highs and lows in a
shorter span of time than secure and avoidant individuals. However, because they expect
this to happen, they keep perpetuating this relationship style since it is what they know and
expect from others and relationships. As such, there could be a self-fulfilling prophecy
operating for anxious-ambivalent individuals in relationships with others.

Brief profiles of individuals within each attachment style are discussed next. Secure
individuals are discussed first, followed by anxious-ambivalent individuals, then by
avoidant individuals. Findings from the present study are discussed with regard to both
how they either support or do not support previous research findings and how they extend
previous research findings.

Previous research has shown that secure people were more positive in their
perceptions of relationships, had more trust, had a more positive attitude, were more
satisfied with their relationships, had more positive experiences, had more positive
emotions with relationships, emphasized more openness, understanding and caring
compared to their insecure counterparts (e.g., Hazan & Shaver, 1987; Mikulincer &

Nachshon, 1991; Simpson, 1990). The present results also show that secure individuals had

141

these positive perceptions for their romantic partner, thus replicating previous research. In
addition, responses to relationship questions revealed that secure individuals were more
likely to have a romantic partner compared to anxious-ambivalent and avoidant individuals.
Secure individuals were also more likely to have been involved with their partner for a
longer span of time, were closer to this person, and expected to have this person as a
romantic partner for a longer time period in the future. Secure individuals were more likely
to have a romantic partner who was also their best fi'iend compared to anxious-ambivalent
and avoidant individuals. This result supports the idea that secure individuals have a more
positive perception of relationships and thus should be more likely to nominate their
romantic partner as also being their best friend. There was a higher percentage of secure
individuals in all three types of relationships. These findings are consistent with those of
previous research (e.g., Feeney & Noller, 1990; Hazan & Shaver, 1987) about secure
individuals and their romantic relationships.

Previous research findings showed that anxious-ambivalent people had higher
ratings for their romantic partner on Passion and Conflict-Ambivalence compared to secure
individuals (Kirkpatrick & Davis, 1994). Anxious-ambivalent people tend to be more
extreme in their relationship emotions; their love relationships have been characterized by
obsession, idealization of partner, desire for reciprocation and union, extreme sexual
attraction and jealousy, and fear of being abandoned (e. g., Feeney & Noller, 1990; Hazan &
Shaver, 1987). The obsession and idealization of the partner could create a high level of

passion and conflict, which are intense emotions. In the present study, it was found that

142

anxious-ambivalent individuals had higher ratings for passion and conflict-ambivalence
compared to secure and avoidant individuals, which replicates previous findings.

A related explanation for the findings could be that passion is desired by these
anxious-ambivalent individuals and thus the need to maintain passion, which because of its
intensity can only last for a short span of time, would require renewing the source of it.
Renewing the source would require replacing one partner with another in a relatively short
span of time. The high level of conflict-ambivalence would also support this suggested
need to replace the romantic partner after a short span of time.

Also, previous research (Brennan & Shaver, 1990; Hazan & Shaver, 1987) has
found a higher rate of relationship dissolution for anxious-ambivalent individuals. It was
found in the present study that anxious-ambivalent individuals had their romantic partner
for a shorter span of time, had not expected to have this person as their romantic partner for
a long span of time, and had had the highest number of previous romantic partners.

Previous findings showed that avoidant individuals tend to have low levels of trust,
satisfaction, and intimacy in their relationships (e.g., Feeney & Noller, 1990; Levy & Davis,
1988). Avoidant individuals’ love relationships have been characterized by fear of intimacy
and a low level of satisfaction (e.g., Hazan & Shaver, 1987). These people have been found
to prefer to maintain distance in their relationships (e.g., F eeney & Noller, 1990; Simpson,
1990), to be self-reliant, and perceive their significant other as being unreliable and overly
eager to commit to relationships (e. g., Feeney & Noller, 1990). These findings were
replicated in the present research. Avoidant individuals were found to have lower ratings

on Viability, Intimacy, Global Satisfaction, and Care for their romantic partner, compared

143

to secure individuals. They had lower ratings on passion compared to both secure and
anxious-ambivalent individuals. However, they had higher ratings for conflict-ambivalence
compared to secure individuals. These results support the idea that avoidant individuals are
not as relationship-oriented as those with one of the other two attachment styles. Responses
to relationship questions also revealed that avoidant participants were lower in their
closeness ratings for their romantic partner. This result supports research findings that
avoidant individuals have a desire for maintaining distance in their relationships by not
getting too close and committed to anyone.

This replication section was important in supporting previous findings on
attachment style and romantic relationships. The expected differences among the three
attachment styles in how they perceived their romantic partner on various relationship
variables were found. Secure people had a more positive perception of their partner, being
more trusting, caring, and satisfied with their partner and the relationship compared to
anxious-ambivalent and avoidant people.

The means and differences for the three attachment styles are very similar to those
of Kirkpatrick and Davis (1994). The mean ratings on the 1-9 scale ranged from 6.56 to
7.87, with 6 meaning “very much," 7 meaning “a great deal," and 8 meaning “strongly,
almost always.” Ratings for Conflict-Ambivalence ranged from 3.20 to 2.66 with 3
meaning “slightly or rarely” and 2 meaning “very little.” Thus, all ratings were relatively
positive. The difference fi'om the highest to lowest mean scores for any relationship
component was less than one point. The differences are small but they are statistically

meaningful. The findings are robust because they have been found throughout previous

144

research. My findings replicated those of previous research in both the pattern of results
and the mean differences. My data show that, for all three attachment styles, the means
were on the high end of the rating scale ranging fiom ‘very much’ to ‘strongly, almost
always.’ I cannot conclude that anxious-ambivalent and avoidant individuals are unhealthy
in their perception of others and relationships but rather that they are not as healthy as
secure individuals. In psychological terms, the meaningful differences are not on a
continuum of unhealthy to healthy but instead from healthy to more healthy.

Extension

The main purpose of these hypotheses was to extend the study of adult attachment
to different types of close relationships (i.e., other-sex best friend and same-sex best friend).
Hypotheses for other-sex and same-sex best friends were similar. It was predicted that
secure individuals would have higher ratings for Viability, Intimacy, and Care compared to
anxious-ambivalent and avoidant individuals. For Global Satisfaction, it was predicted that
both secure and avoidant individuals would have higher ratings compared to anxious-
ambivalent individuals.

This section was important in that it extended the findings from attachment style
and romantic relationships to attachment style and best fiiendships. Similar to the
replication findings, the three attachment styles difiered in their ratings of relationship
variables for both their other-sex best fi'iend and their same-sex best friend. Responses to
relationship questions revealed that secrue individuals were more likely to have an other-sex
best friend compared to anxious-ambivalent and avoidant individuals. Even within the

domain of friendship, secure people had a more positive perception of their best friends

145

compared to anxious-ambivalent and avoidant people. The secure people tended to be more
trusting, accepting, understanding, and caring for their best friends compared to the insecure
people.

Two designs were employed, a between-subj ects and a within-subj ects. A major
advantage of the within-subj ects design is that there is control of subject heterogeneity (i.e.,
individual differences; Keppel, 1982). Individual differences on ratings of each relationship
type is thus reduced. Therefore, the error variance is reduced to the extent that responses
across relationship types are correlated and the design is more powerful and sensitive to
relationship/ attachment effects compared to the between-subj ects design. However,
regardless of design type, the findings for other-sex best friend and same-sex best friends
were similar to previous findings for attaclunent style and romantic partner. Secure
individuals had higher ratings on important relationship components for both other-sex and
same-sex best friend compared to avoidant and anxious-ambivalent individuals.

The mean ratings for the positive relationship variables (i.e., all but conflict-
ambivalence) on the 1-9 scale for the other-sex best fi‘iend ranged from 6.56-7.50 and the
mean ratings for the same-sex best friend ranged from 6.66-7.62. Recall that 6 meant “very
much," 7 meant “a great deal," and 8 meant “strongly, almost always.” These findings were
very similar to those of romantic partner in that all ratings were relatively positive and the
difference from the highest to lowest mean scores for any relationship component was less
than one point. Although small, the differences were statistically meaningful. My data
show that for all three attachment styles, the means were on the high end of the rating scale

ranging fi'om ‘very much’ to ‘strongly, almost always.’

146

The findings for other-sex and same-sex best friends are important in that they
extend our understanding of attachment style and relationships with best fiiends where
previously we only knew about romantic partners. These findings also facilitate our
understanding of the consistency of mental models of others across various close
relationships. It was found that individuals within each attachment style had similar
perceptions for their same-sex and other-sex best fiiend as for their romantic partner. It is
important to note that the means and differences for these relationship variables for best
friends were very similar to those of romantic partners. Again, I conclude that anxious-
ambivalent and avoidant individuals are not unhealthy, but only somewhat less healthy in
their perceptions of others and relationships compared to secure individuals.

- ~. ‘ - - - - .‘ o o ‘ . o .- o o
'I I .I .1... IIII I . -.II\ -_.IIII AI. III-II“ I k -. IIII

Components

The purpose of these analyses was to extend the research area by comparing three
relationship types (i.e., romantic, best fiiend other-sex and same-sex) and sex of participants
within the area of adult attachment on various relationship components.

There were differences among the ratings of the three relationship types with the
different relationship components. However, because best fiiendships are very similar to
love relationships, the characteristics of self-disclosure, intimacy, trust, and satisfaction
should exist in both types of relationships. The BSS and the WSS were employed for these
analyses. They show similar findings for some components while other components had
dissimilar findings between the two samples. These findings may be in part biased by the

population sample here which was undergraduate college students, most of whom are

147

between the ages of 18 and 21. The developmental stage is critical because most
individuals are more likely to have a longer and more trusting relationship with a same-sex
best friend compared to a romantic partner. During the college years, individuals often date
a number of people in search of the ‘right’ person. Thus, many may not be involved in a
committed, long-term romantic relationship. These findings revealed that individuals with
both a same-sex best fiiend and a romantic partner tend to trust and confide in their best
friend more than their romantic partner. The findings could be different if the study had
been conducted using 30-45 year old individuals, many of whom are likely to be involved
in a long-term, committed romantic relationship.

For Viability, individuals in the BSS did not differentiate among the three
relationships for ratings while individuals in the WSS had a main effect for relationship
type on Viability. It was found that individuals rated their other-sex and same-sex best
fiiend higher on Viability compared to their romantic partner. Individuals who have all
three relationships tend to trust, respect, and accept/tolerate their best fiiends more than
their romantic partner. This supports previous research (e.g., Davis & Todd, 1982; Steck et
al., 1979) that fiiendship is based more on reciprocal trust and respect. Overall, as reported
earlier, participants in the study had their best friends longer than their romantic partner
which is another explanation for higher ratings of traits within the Viability factor.

For Intimacy, individuals rated their romantic partner and their same-sex best fiiend
higher than their other-sex best fiiend for the BSS but participants rated their same-sex best
fiiend higher than their other-sex best fiiend and their romantic partner for the WSS. Their

other-sex best fiiend was also rated higher than their romantic partner. Individuals with all

148

three relationships perceived that they could confide and be understood more by their same-
sex best fiiend compared to their other-sex best friend and their romantic partner. These
individuals could compare the three relationships on confiding and trust and thus found that
same-sex best fiiend was more pertinent for these dimensions in relationships (e. g., Steck et
al., 1979). In contrast, most individuals with either one or two relationships perceived that
they could confide and be understood more by either their romantic partner or same-sex
best fiiend than by their other-sex best fiiend.

For Global Satisfaction for the BSS, individuals rated their romantic partner higher
compared to their same-sex best fiiend and other-sex best fiiend. Participants also rated
their same-sex best friend higher compared to their other-sex best fiiend. These individuals
were perceived to have more relationship satisfaction and traits relating to this such as
reciprocity and enjoyment with their romantic partner compared to their fiiends.

Individuals in the WSS did not differentiate among their romantic partner, same-sex best
fiiend, and other-sex best fiiend for ratings on Global Satisfaction.

For Care for the BSS, individuals rated their romantic partner higher compared to
their other-sex and same-sex best fiiend. Participants also rated their same-sex best friend
higher compared to their other-sex best friend. For the WSS, participants rated their same-
sex best fiiend higher than their romantic partner and their other-sex best fiiend. They rated
their romantic partner higher than their other-sex best friend. The findings for the BSS on
Care also support previous research (Davis & Todd, 1982) that the variables for Care were
more romantic-oriented than fiiend-oriented. An individual would more likely ‘give the

utmost’ to a lover compared to a best fiiend, although some individuals would do the latter.

149

This was found in the BSS. Individuals in the WSS, who have all three relationships,
differentiate among them. They reported that they would do anything or give the utmost to
their same-sex best friend followed by their romantic partner then by their other-sex best
fiiend.

The above relationship components are important to both romantic relationships and
best friendships. However, when comparing WSS and BSS participants, the WSS
participants must have all three types of relationships compared to the BSS who can have
one, two or all three types of relationships. To equally compare all three relationships, the
more valid findings may be from the WSS participants because they can compare all three
relationship types on all these relationship components. Unlike the WSS participants, the
BSS participants are less likely to have all three relationship types to compare at the same
time and thus may be biased in their ratings because of the fewer number of relationship
types.

For Passion, both the BSS and the WSS generated similar findings. Participants
rated their romantic partner highest, followed by their other-sex best fiiend, then by their
same-sex best fiiend. This pattern supports previous findings by Davis & Todd (1982) in
that participants had higher ratings on Passion for their romantic partner compared to their
best friends. Besides the sexual aspects of passion, best friendships may not have the
emotional intensities and emotional expressions within the realm of desire and sexual
attraction, that romantic relationships have.

For Conflict-Ambivalence, both the BSS and the WSS yielded similar findings.

Participants rated their romantic partner highest, followed by their other-sex best fiiend,

150

then by their same-sex best fiiend. Again, regardless of the number of relationships an
individual has, perceptions of conflict and ambivalence are more prominent in the romantic
relationship. Conflict may be more acknowledged or expected as a normal or typical aspect
within romantic relationships than within friendships although conflicts do exist among
fi'iendships.

The overall analyses help put the replication hypotheses in a broader interpersonal
relationship matrix. Most researchers have looked at attachment style and romantic parhrer
and I have extended the research into the area of best fiiendships. This extension expands
the generalizability of attachment style to different types of relationships. I have replicated
previous findings for romantic partner and my findings reveal that this pattern is also true
for other relationships such as other-sex best fiiend and same-sex best friend.

There was a marginally significant interaction for attachment style and relationship
type for the BSS. Secure, avoidant, and anxious-ambivalent individuals, all rated their
romantic partner higher on Care compared to their same-sex and other-sex best fiiend.
Secure and avoidant individuals also rated their same-sex best fi'iend higher than their other-
sex best friend. In contrast, anxious-ambivalent individuals rated their other-sex best fiiend
higher than their same-sex best fiiend. The major difference was the ratings between the
same-sex and other-sex best fiiend. It is suggested that anxious-ambivalent people rated
their other-sex best fiiend higher than their same-sex best friend because they may perceive
the other-sex best fiiend as a potential romantic partner. As stated earlier, anxious-

ambivalent individuals perceive their romantic partner as being unreliable and unwilling to

151

commit to a relationship, therefore they may always be on the lookout for a future or
potential romantic partner.

There were sex by relationship type interactions for Viability, Intimacy, Global
Satisfaction, and Passion only for the WSS. Across three of the four relationship
components, females had the highest ratings for their same-sex best friend. For males, their
ratings varied depending on the relationship component. Both females and males rated their
romantic partner higher on Passion compared to their same-sex and other-sex best friend.

Females had higher ratings for their same-sex best fiiend on Viability compared to
their romantic partner. Females rated their same-sex best fiiend higher on Intimacy and
Global Satisfaction than their romantic partner and their other-sex best friend. Females in
the present study perceived their same-sex best friendship to be more trusting, respecting,
accepting, confiding, understanding, enjoyable, successful, reciprocal than their other
relationships.

It has been well documented that females consider their same-sex best friend as
closer, as more supportive, as more intimate, confide in them more, trust them more, are
more satisfied, and perceive that they understand them better than their romantic partner and
other-sex best fiiend (e.g., Aries & Johnson, 1983; Caldwell & Peplau, 1982; Dickens &
Perlrnan, 1981; Helgeson et al., 1987; Reisman, 1981). An explanation for this closeness
between women may be that there is more societal acceptance and approval for the
closeness and intimacy in women’s same-sex friendships (Rose, 1985). This societal
acceptance can also facilitate the closeness and intimacy in women’s same-sex friendships,

especially for these women in a new social milieu.

152

Males tended to have different ratings for the three relationship types depending on
the relationship component. They rated their other-sex best fiiend higher on Viability
compared to their same-sex best fiiend and their romantic partner. In contrast to previous
findings, they rated their same-sex best friend higher on Intimacy than their romantic
partner and their other-sex best friend. They rated their romantic partner and other-sex best
fiiend higher on Global Satisfaction compared to their same-sex best friend. Males had
higher ratings on trust, respect, and acceptance for the other-sex best fiiend which may
support the notion that these other-sex best friends may be perceived as potential romantic
partners. The high ratings for same-sex best fiiend on intimacy is contrary to previous
findings that men were more intimate and confide more in their other-sex best friend
(Komarovsky, 1974). It has been found that men tend to be closer and more intimate with
their other-sex friend (e.g., Rubin, 1985). This different finding in the present study may be
due to Davis and Todd’s (1982, 1985) definition of Intimacy. Intimacy is defined as
confiding and understanding. Other researchers (e.g., Wheeler, et al., 1983) have defined
intimacy as closeness. Also, this finding may be a result of a cohort effect. In the present
time, it is more acceptable for young men to be close to and confide in their same-sex best
fiiend compared to a generation ago where this was considered to be not as common or not
as accepted.

The findings above for males and females may be explained by their age group and
social milieu. At least half of the sample population consisted of first year college students
who may be living away from home for the first time. These individuals may be closer to

their same-sex best fiiend because they know them better, or at least feel more comfortable

153

confiding in them for the present time, compared to their romantic partner. The romantic
relationship may be new. Therefore an individual has too much to lose by confiding or
getting too intimate with the other if there is no reciprocation. These young adults may also
have more experience with fiiendships than with romantic relationships. Findings that men
are more close and intimate with their romantic partner may come about at a later point in
development where this relationship type is more secure and they feel more comfortable in
them. Most first year college students live in the dormitories where they share living
quarters with others of the same sex. This kind of environment promotes more interactions
with others, especially individuals of the same sex. This may facilitate more understanding,
comfortableness, and sharing of intimate details with roommates and floorrnates.

Both males and females had higher ratings on Passion for their romantic partner
compared to their other-sex and same-sex best fiiend. This pattern suggests that the
population in the present study was similar to other populations in previous studies because
Passion is a romantic relationship variable and was rated higher for the romantic partner
compared to the best friends. Interestingly, males rated their other-sex best fiiend higher
than their same-sex best fiiend. In contrast, females did not difi’erentiate their ratings
between their same-sex best fiiend and other-sex best fiiend. Again, it is suggested that
males may perceive their other-sex best friend as a ‘potential romantic partner’ whereas
females did not make this perception. This supports males’ higher rating on Passion for

their other-sex best friend.

1 54

SexDiffercnces

Sex differences were also examined because previous research has not consistently
examined this issue. Females were expected to have higher ratings on these relationship
components compared to males. Sex differences were found for many of the variables.
Overall, across attachment styles, females had higher ratings on Viability, Intimacy, Global
Satisfaction, Care, and Passion compared to males. Although males had higher ratings for
Conflict-Ambivalence across all three relationship types compared to females, the ratings
were not significantly different. These findings support previous relationship literature
findings in that females are more likely to have a more positive perception about
relationships and tend to rate positive relationship variables higher. Relationships tend to
be more central to women’s self-definitions (Gilligan, 1982). Females have also been found
to spend more time talking, sharing their feelings, helping, and supporting the other in their
relationships (e.g., Aries & Johnson, 1983; Caldwell & Peplau, 1982; Weiss & Lowenthal,
1975). They also have a higher level of intimacy compared to males (e.g., Wright, 1982).
Women have also had closer and more intimate fiiendships with each other compared to
men (Bell, 1981). These findings support previous research on same-sex fiiendships in that
females have been found to be closer, more understanding, more intimate, and more
satisfied with their friends compared to males (e.g., Burda, Vaux, & Schill, 1984; Rands &
Levinger, 1979; Wheeler et al., 1983).

Overall, women were more likely than men to have a romantic partner, have this
person for a longer span of time, be closer to the partner, expect this person to be their

romantic partner in the future for a longer span of time, and had a lower number of previous

155

romantic partners. This is consistent with the literature for women and relationships.

Among the three attachment styles, sex differences also appeared for relationship
questions pertaining to relationship types. Responses to these questions revealed that more
females had a romantic partner who was also their other-sex best friend compared to males.
This finding is in contrast to the literature on romantic partner and best fiiend which showed
that females were more likely to have a best fiiend who was not also their romantic partner
while in a serious relationship or marriage, while the reverse was true for males. Males
have been found to derive most of their intimacy needs fi'om their romantic partner (e.g.,
Rubin, 1985). Again, the developmental stage of the participants may have created this
finding. The romantic relationship may not be serious or intense enough for males to
disclose and be intimate with the partner and to also associate this partner as the best friend.
The numbers for the best fiiend also being the romantic partner for males should increase
with time.

Comparing responses to relationship questions, sex differences occurred more with
romantic relationships compared to best fiiendships. For length of Time, females had a
longer time span with their partner compared to males. There were no differences for other-
sex and same-sex best fiiends between males and females. For Closeness, females had a
higher rating compared to males for both other-sex and same-sex best fiiends. Females and
males did not differentiate in their ratings for their romantic partner. For Expectation of
Future Togethemess, females had higher ratings than males only for other-sex best fiiends.
There were no differences in the ratings for same-sex best friend and romantic partner for

females and males. Males had a higher number of past romantic partner compared to

156

females. Males and females were similar in their past numbers of other-sex and same-sex
best fiiends.
mm:

It is suggested here that for individuals during late adolescence/young adulthood,
their best fiiend, mostly same-sex best fiiend, may serve as a major attachment figure for
this transition period between adolescence and adulthood. Weiss (1991) has suggested that
individuals may go through stages where their attachment figures may be different people
depending on the age and developmental stage of the individual. In infancy and childhood,
the attachment figure is most likely to be the parent or caregiver (e.g., Ainsworth et al.,
1978; Bowlby, 1979). In adulthood, the attachment figure is most likely the romantic
partner (e.g., Hazan & Shaver, 1987). Around college age, most individuals have
relinquished their parents as attachment figures (Weiss, 1991). As individuals increase in
age, there is a decrease in intimacy with parents (e.g., Hortacsu, 1989; Wright & Keple,
1981). Furthermore, adolescents/young adults tend to be closer to their fiiends and confide
in them more and perceive that their fiiends understand them more than their parents (e. g.,
Wright & Keple, 1981). Thus, best friends may play a prominent role at this stage of
development because they are more likely to provide or be perceived to provide the
essential functions to become an attachment figure (e.g., a secure base and a safe haven)
(Hazan & Shaver, 1994). Coleman (1974) found that security is one of the most important
properties of adolescents’ relationships with their friends. This security can be linked to a
secure base or safe haven which is provided by the attachment figure. Because these new

attachment figures would be in the same environment and be within physical and

157

psychological proximity, individuals can further explore the environment knowing that
safety and security are nearby.

Although some individuals have romantic partners, because of their short duration
in this type of relationship, the relationship is not perceived to be as secure or as stable as
the fiiendships. It is suggested that when individuals are older and/or more established in
their romantic relationship, their romantic partner may become more important as an
attachment figure. However, some individuals may always continue to have their best
fiiend as one primary attachment figure; just as some individuals may always perceive their
parent/caregiver to be an important attachment figure regardless of their age or
developmental stage.

It was found that almost all individuals in the present study reported having a same-
sex best fiiend while a majority reported having an other-sex best fiiend. Although a
majority reported having a romantic partner, from this group over a third stated that their
romantic partner was also their best fiiend. Individuals reported having their same-sex and
other-sex best friend for 1-2 years. Those who had a romantic partner reported having this
relationship for only 7-11 months. Same-sex best fiiend seems to be the predominant
relationships for these undergraduates. These individuals have been found to be most likely
to have a same-sex best fiiend and to have had them for a longer period of time. At this
point in development, individuals may perceive parents to be important but their fiiends to
be more important in their lives (e.g., Wright & Keple, 1981). Hortacsu, Oral, and Yasak-
Gilltekin (1991) found that university students reported being closest to their same-sex best

fiiend.

158

There were several limitations to this study. First, almost all of the participants
were between the ages of 18 and 22. Developmental issues faced by these young adults
concerning relationships may have resulted in some findings not found in other age
groups. One such finding is that these individuals did not differentiate much between
romantic partner and best fiiends. Perhaps differences would occur in an older age group.
Also, at least half of the sample consisted of freshmen and being a first year college
student has its difficulties. The environment is new and the recent move away from home
may bring about changes that affects relationships and/or perceptions of relationships.
Some individuals may have separated fiom significant others while others developed new
significant relationships. Thus, certain findings in this study may be specific to this
developmental transition.

A second limitation involves the sample in this study which was comprised
largely of mostly White individuals. However, this is a problem for many researchers
conducting research with the college environment. Thus, the generalizability of these
results is limited to Caucasians/Whites.

A third limitation is a problem that has been faced by many adult attachment
researchers. This involves the adult attachment style measurements. A widely used scale
is Hazan and Shaver’s which is a single-item scale. Hazan and Shaver (1987) have
expressed concern about this problem and have encouraged researchers to develop a new
adult attachment scale with multiple items. In the present study, this problem was

remedied using two methods. One, by combining two adult attachment scales, Hazan and

159

Shaver’s and Bartholomew’s scales, individuals’ attachment style in the present study
should be more valid and reliable. Several studies (e.g., Brennan et al., 1991) have
employed this combination method to increase the accuracy of rating individuals’
attachment style. Two, correlations were conducted on each attachment style and the
equivalent 1-7 Likert ratings to demonstrate that the assumed differentiation did in fact
exist.
W

A future direction for research would be to require best friends to participate in
the study. The best fiiends would complete all measures just like the participants.
Researchers can then examine the attachment styles and perceptions about relationships
from the perspective of both the subjects and the significant others. It would be
important to find out if best fiiends share similar expectations about relationships and if
their attachment styles confirm the other’s in terms of expectation of another in a
relationship. Also, an individual’s different levels of satisfaction and trust level in a
relationship may depend on the other person’s behavior and perception about the
relationship. For example, Kirkpatrick and Davis (1994) found that the low level of
satisfaction that the subjects had was caused by the romantic partner’s behavior within
the relationship.

Along this line, researchers can also examine whether the best fiiends have
similar or complementary attachment styles compared to the participants. It would be
important to know whether the perception of others and relationships extend out to best

friends similarly to romantic partner. Past researchers (e.g., Kobak & Hazan, 1991;

160

Senchak & Leonard, 1992) have found that secure individuals tend to be in relationships
with other secure individuals and that anxious-ambivalent and avoidant individuals tend
to be attracted to each other. These individuals confirm the other’s expectations of what
significant others are like in relationships. It would be important to recognize and
identify whether these processes are the same for best fi'iendships and if they have similar
results to those of romantic partners.

A second direction for future research should be to examine the relationship
between attachment style, sex, and close interpersonal relationships with different age
groups and developmental stages (e.g., 35-50 years old, middle-aged). This would extend
our understanding about mental models to determine if perceptions of others and
relationships remain constant across the different developmental stages and the lifespan.

A third future direction would be to study developmental issues using longitudinal
studies. We would learn much from observing individuals throughout their lives and
investigate what types and kinds of relationship individuals in different attachment
establishes and how all this may be related or affected by attachment styles. Along the
longitudinal line of future research, we could follow individuals from early adolescence
to middle adulthood to examine how certain attachment styles develop and become
involved in different types of relationships such as romantic relationships and fiiendships.
This may help us answer or be more aware of the issue of changes in attachment styles
through the life span and fi'om one relationship to another. For example, researchers such
as Kobak and Hazan (1991) and Senchak and Leonard (1992) found in their studies that

there was a higher proportion of secure individuals in marital relationships compared to

161

the general population. Kobak and Hazan (1991) suggested that people who are insecure
may become secure once they are in a relationship. Or that marital relationships attract
secure people. Which ever the case may be, this presents a different issue to attachment
style. To extend this line of reasoning, a question exists which states that “Could a
person develop both a secure and an insecure attachment style depending on the situation
and individual(s) involved?” In other words, could an avoidant person be a secure person
in their marriage but when interacting with others outside of the marriage fi'amework be
an avoidant person? Could attachment style be situation-specific or be a variation of this
concept?
Conclusion

The findings in this study are important because they show that attachment style
impacts an individual’s perception of relationship components. This impact is apparent
and consistent across the three different types of relationships examined here (i.e.,
romantic relationship, best fiiendships). Secure individuals had higher ratings on positive
relationship components and lower ratings on negative relationship components
compared to anxious-ambivalent and avoidant individuals. This pattern of findings
extends our understanding beyond the impact of attachment style on romantic
relationships to the impact of attachment style on other close relationships. This places
adult attachment research into a broader interpersonal relationship matrix by expanding
the generalizability of attachment style to different types of relationships. This
generalizability of attachment style should shed new light in this area of research as well

as support the theory of attachment. Bowlby did not set parameters on attachment as it

162

pertains to adult relationships and thus the present findings strengthen the theory by
showing a generalization of results beyond the parameters of romantic relationships. I
have replicated previous research on the impact of attachment style on relationship
variables for romantic partner as well as showing that these patterns are also true for other
close relationships. The findings here are potential starting points for many fruitful future

inquiries.

LIST OF REFERENCES

LIST OF REFERENCES

Ainsworth, M. D. S. (1982). Attachment. Retrospect and prospect. In C. M.

Parkers & J Stevenson-Hinde (EdS) Ihoolaoeofattaohmentinhumanhehayior (pp 3-
30). New York. Basic Books.

Ainsworth, M. D. S. (1989). Attachment beyond infancy. American
13W 709-716.

Ainsworth, M. D. S. Blehar, M. C., Waters, 13., &Walls, S. (1978). Banemsof
anachmontAosychologioalsmdLofflmangosimmion. Hillsdale, NJ: Erlbaum

Allan, G. & Adams, R. G. (1989). Aging and the structure of fiiendship. In R.
G. Adams & R. Blieszner (Eds), QldeLAdulLEriendship, Newbury Park, CA: Sage.

Argyle, M. & Henderson, M. (1984). The rules of fiiendship. W31
andPorsonalRelatronshlsz, 211-237.

Aries, E. J. & Johnson, F. L. (1983). Close fiiendship in adulthood:
Conversational content between same-sex fiiends. W, 1183-1196.

Bakan, D. (1966). Ihodualiflofhumanoxistonoe. Chicago: Rand McNally.

Bartholomew, K. (1990). Avoidance of intimacy: An attachment perspective.
loumalofSooiaLandPersonalRelationshipsl, 147-1 7 8

Bartholomew, K. & Horowitz, L. M. (1991). Attachment styles among young

adults. A test of a four-category model. W
6.1, 226-244.

Bell, R. R. (1981). MW Beverly Hills, CA: Sage.

Belsky, J. & Cassidy, J. (1994). Attachment and Close Relationships: All
Individual-Difference Perspective. PsychologicaLInqnm, 27- 30.

Berg, J. H. (1984). Development of fiiendship between roommates. lonmalof

RersonahtxandSoclalPsxoholongo 346-356
163

164

Berrnan, W. H., Marcus, L., & Berrnan, E. R. (1994). Attachment 1n Marital
Relations. In M. B. Sperling & W. H. Berrnan (Eds), AttachmemJnAdnlts, New
York/London: The Guildford Press.

Berscheid, E. & Walster, E. (1969). Won, Reading, MA:
Addison-Wesley.

Blieszner, R. & Adams, R. G. (1992). AMEfiendship, Newbury Park, CA:
Sage.

Blau, Z. S. (1973). Oldagejnamangingsociety. New York: New Viewpoints.

Booth, A. & Hess, E. (1974). Cross-sex fiiendship. .[omnalofMamaggandjhe
W. 38-47.

Bowlby. J. (1969). AnaohmemandlossfloLLAttachmem. New York: Basic
Books.

Bowlby, J. (1973).
New York: Basic Books.

Bowlby, J. (1979). Thomakingandhmakingofaffeotionalmm London:

Tavistock.

 

Bowlby, J. (1980). W New York: Random

House.

Bowlby,J. (1982). AttachmmtandJosssloLLAttachmmtflndsdl. New
York: Basic Books.

Bowlby,J. (1988). . - ._ '
Moment, New York: Basic Books.

Brennan, K. A. & Shaver, P. R. (1995). Dimensions of attachment, affect

regulation, and romantic relationship firnction. Bmonahfimdfioorallsxchology
BMW, 267-283.

Brennan, K., Shaver, P. R., & Tobey, A. E. (1991). Attachment styles, gender,
and parental problem drinking. loumalofSooialondPersonaLRelationshiosfi, 451-466.

Buhrke, R. A. & F uqua, D. R. (1987). Sex differences in same- and cross-sex
supportive relationships. W, 339-352.

 

165

Buhrrnester, D. & F urrnan, W. (1986). The changing functions of fiiends in
childhood: A neo-Sullivian perspective. In V. J. Derlega & B. A. Winstead (Eds),

EriendshioandsooiaLintcraotion (pp. 41-62). New York: Springer-Verlag.
Bulmer, M. (1987). WWW. London: Allen and

Unwin.

Burda, P. C., Vaux, A., & Schill, T. (1984). Social support resources. Variation
across sex and sex roles Bermalflandfimialfisycholmflulm 119126.

Burhrrnester, D. & Furhman, W. (1986). The changing filnctions of fiiends in
childhood. a neo-Sullivan perspective. In V. J. Derlega, B. A. Winstead (Eds),

Enendshioanismmmmion (pp. 41-62). New York: Springer-Verlag.

Byme, D., Clore, G. L., & Worchel, P. (1966). Effect of economic similarity-

dissimilarity on interpersonal attraction lounnloffiersonalinLandSociaLthologLA.
220- 224.

Caldwell, M. A. & Peplau, L. A. (1982). Sex differences in same-sex fiiendship.
SortRolosj, 721-732-

Campos, J. J., Barrett, K. C., Lamb, M. E., Goldsmith, H. H. & Stemberg, C.
(1983). Socioemotional development. In P. H. Mussen (Ed.,) Handlmokofflhfld

BsxcholongoLJLJnfanoLandDexelopmontalPsxchohiologL M. M. Haith & J. J.
Campos (Eds), New York: Wiley.

Candy, S. G., Troll, L. E., & Levy, S. G. (1981). A developmental exploration of
friendship functions in women. WW1“; 456-472.

Chodorow, N. (1978). ThikeprodnflionofMchcrhood. Berkeley, CA:

University of California Press.

Clark, M. S. & Mills, J. (1979). Interpersonal attraction in exchange and
communal relationships. lorrmalofflersomlrtrLandSoelaLBsysholongZ, 12-24.

Cobb, S. (1976). Social support as a moderator of life stress. Psychosomatic
Medioinsfl 300-314.

Cohen, S. & McKay, G. (1984). Social support, stress, and the buffer hypothesis.

In A. Baum, S. Taylor, & J. E. Singer (Eds), Handhogkoflpsychologxandheahh (V 01.
IV). Hillsdale, NJ: Erlbaum.

Coleman, J. (1974). RelationshipsinAdolescence. London: Routledge & Kegan
Paul. '

166

Collins, N. L. & Read, S. J. (1990). Adult attachment, working models, and

relationship quality in dating couples. WW3,
644-663.

Davis, K. E. & Latty-Mann, H. (1987). Love styles and relationship quality: A

contribution to validation loumalofSocidandPersonalRelationshiosA. 409-428
Davis, K. E. & Todd, M. J. (1982). Friendship and love relationships. In K. E.

Davis & T Mitchell (Eds ). AdxanoesindoscriotixsosxchologyWol. 2, pp. 79-122).
Greenwich, CT. JAI.

Davis, K. E. & Todd, M. J. (1985). Assessing fiiendship: Prototypes, paradigm
cases and relationship description. In S. W. Duck and D. Perlman (Eds) Understanding

PersonaLRelationshiosLAnlnterdisoiolinanLAoproaoh. London: Sage, pp 17-38

Dickens, J. W. & Perlman, D. (1981) Friendship over the life cycle. In S. Duck
& R. Gilmour (Eds.,) Dexelomngmrsonaljolanonshlos London: Academic Press.

Duck, S. (1983). WE. Great Britain: The Harvester Press.

Duck. 8- (1984). PersonalRelanonsthoamngRersonalRelanonshrps
Academic Press: London & New York.

Duck, S. (1991). Undorstandingfielationships New York, NY: Guildford Press.

Egeland, B. & F arber, E. A. (1984). Infant-mother attachment: Factors related to
its development and changes over time. ChildflexolopmenLfi, 753-771.

Elicker, J ., Englund, M., & Sroufe, L. A. (1992). Predicting peer competence and

peer relationships in childhood from early parent-child relationships. In R. Parke & G.
Ladd (Eds), Growing points of attachment theory and research. Monographs“:

WWW 147-166.

Feeney, J. A. & Noller, P. (1990). Attachment styles as a predictor of adult
romantic relationships. lotrmalofRt:rsonalit:La11d_8.05:iaLBsxohologxtiBs 281-291

F eeney, J. A., Noller, P., & Hanrahan, M. (1994). Assessing adult attachment:
Developments in the conceptualization of security and insecurity. In M. B. Sperling &

W. H. Bermans (Eds), WW
W5, New York. Guilford.

Feeney, J. A., Noller, P., & Patty, J. (1993). Adolescents' interactions with the

opposite sex: Influence of attachment style and gender. lonmalofAdolosoomio, 169-
1 86.

167

Fischer, J. L. & Narus, L. R., Jr. (1981). Sex roles and intimacy in same sex and

other sex relationships. Bsyohologlofiflomonflumorm; 444-455.

Fox, M., Gibbs, M., & Auerback, D. (1985). Age and gender dimensions of
friend-ship- Bsxchologxofflomenfluarterlsfl. 489-502.

George, C.. Kaplan, N. & Main. M. (1984)- Attachmentjntendmfonadults
Unpublished manuscript, University of California, Berkeley.

Gilligan, C. (1982). Inadrfferentxolcolisxchologloauheontandflomeru
doyolopmem. Cambridge, MA: Harvard University Press.

Ginsberg, D. Gottman, J. M., & Parker, J. G. (1986). The importance of

friendship. In J. M. Gottman and J. G. Parker (Eds.,) ConwaationsofEfionds.
Cambridge: Cambridge University Press.

Gottlieb, B. H. (1994). Social Support. In A. L. Weber & J. H. Harvey (Eds)
BorspcotixesonflosoRelationships (pp. 307-324). London: Allyn and Bacon.

Grossman, K. E. & Grossman, K. (1991). Attachment quality as organizer of
emotional and behavioral responses in a longitudinal perspective. In C. M. Parkes, J.

Stevenson-Hinde, & P. Marris (Eds) WWW
London: Tavistock/Routledge.

Hatfield, E. (1984). The dangers of mtimacy. In V. Derlaga (Ed)
Commumoanommumacxaandolosorelanonshrpstom New York. Academic

Press.

Hays, R. B. (1988). Friendship. In S. W. Duck (Ed.), Handbookofpersonal
relationships (pp. 391-408). Chichester, England: Wiley.

Hays, R. B. (1989). 'The day-to-day functioning of close versus casual
friendships'. loumalofSooraLandPorsonaLRelatronshrsz. 21 -37.

Hazan, C. & Shaver, P. R. (1987). Romantic love conceptualized as an
attachment process. WWW, 511-524.

Hazan, C. & Shaver, P. R. (1994). Attachment as an organization framework for
research on close relationships. Bsychologioaljnqm, 1-22.

Helgeson, V. S. Shaver, P, & Dyer, M. (1987). Prototypes of mtimacy and
distance rn same-sex and other-sex relationships. lmrmalofSooialondeonal
Relationshiosfi 195-233

168

Henderson, S., Byme, D. G., & Duncan-Jones, P. (1981). Neurosisandjhe
SooiaLEnyiLonment. Sydney: Academic Press.

Holmes, J. G. & Rempel, J. K. (1989). Trustinclose relationships. InC.

Hendrick (Ed.,) WWW
Newbury Park, CA. Sage.

Hortacsu, N. (1989). Targets of communication during adolescence. loumaLof
AdolescenosJZ. 253-263-

Hortacsu, N., Oral, A., & Yasak- Gfiltekin, Y. (1991). Factors affecting
relationships of Turkish adolescents with parents and same-sex fiiends.1he_loumaLof

SootaLEsxeholongllG), 413-426.

Howes, C. (1981, April). Banomuffiiondshio. Paper presented at the Biennial
Meeting of the Society for Research in Child Development, Boston, MA.

Huston, T. L., Surra, C. A., Fitzgerald, N. M., & Cate, R. M. (1981). From
courtship to marriage: Mate selection as an interpersonal process. In S. Duck & R.

Gilmour (Eds), Bomonalfielafionshios (pp. 53-88). San Diego, CA: Academic Press.

Jones, D. C. (1991). Friendship satisfaction and gender: An examination of sex
differences in contributors to fi'iendship satisfaction. lonmaloffiooialandPorsonal

Relationshisz 167-185

Jones, W. H. & Burdette, M. P. (1994). Betrayal in relationships. In Weber, A.

L. &Harvey,J H (Eds.,) WWW Boston. MA:
Allyn and Bacon.

Jourard, S. M. (1974). IheHealIhLPensoualim. New York: Macmillan
Publishing.

Keelan, J. P., Dion, K. & Dion, K. (1994). Attachment style and heterosexual
relationships among young adults: A short-term panel study. .[oumalofSooialjnd

Personalfielatioushiosll 201-214

Keppel, G. (1982). QesigrrandanalxsisLAreseareheL’shandhook (2nd ed).
Englewood Cliffs, NJ: Prentice-Hall.

Kirkpatrick, L. A. & Davis, K. E. (1994). Attachment style, gender, and

relationship stability: A longitudinal analysis. W
Bmhologsfifi, 502-512.

169

Kobak, R. R. & Hazan, C. (1991). Attachment in marriage: Effects of security

and accuracy of working models. loumaloffiersonalmLandSoelaLEsyeholongQ, 861-
869.

Kobak, R. R. & Sceery, A. (1988). Attachment in late adolescence: working
models, affect regulation, and representations of self and others. Childfiexelopmow,
135-146.

Komarovsky, M. (1974). Patterns of self-disclosure of male undergraduates.
lorrmalofMarnageandflleEamlw 677-686

La Gaipa, J. J. (1977). Testing a multidimensional approach to friendship. In S.
W. Duck (Ed) Women. London: Academic Press

Levitt, M. J. (1991). Attachment and close relationships: A life-span perspective.
W L Gerwitz& W M Kurtines (Eds ).IntersecrionudthAttarhmem. Hillsdale, NJ.
Lawrence Erlbaum Associates.

Levy, M. B. & Davis, K. E. (1988). Lovestyles and attachment styles compared.
Their relations to each other and to various relationship characteristics. .[onmdoffiooial

andPersonalRelatlonshlsz 439-471

Limb, S. (1989). Female Friendship. In R. Porter and S. Tomaselli (Eds. ),Ihe
DialefliosofEdondshio, New York, NY. Routledge.

Main, M. & Cassidy, J. (1988). Categories of response with the parent at age six:
predicted from infant attachment classifications and stable over a one month period.

DexelopmentaLBsxcholongA, 415-426.

Main, M., Kaplan, N., & Cassidy, J. (1985). Security in infancy, childhood, and
adulthood. A move to the level of representation. In I. Bretherton and E. Waters (Eds)

.II'IlII. I-.l.-. III III II III “w. ItlIIIunII I I’ I ‘\ I

Reseamhjuflhildflmlopmflmw, 66-104. Chicago: University of Chicago

Press.

Mikulincer, M. & Nachshon, O. (1991). Attachment styles and patterns of self-
disclosure. loumalofifiersonalloaaniSoclalPsxehologxtol, 321 -331.

Monsour, M. (1992). Meanings of mtimacy in cross- and same-sex fiiendships.

JoumaLofSoeialandRersonaLRelationshipsfl, 277-295.

Morgan, B. S. (1976). Intimacy of disclosure topics and sex differences in self-

disclosure. SoxRolosZ, 161-166.

170

Newcomb, M. D. (1981). Heterosexual cohabitation relationships. In S. Duck &
R. Gilmour(Eds.,) RusonalRelationshioleSmdyinghrsonalRelationshins. New

York: Academic.

Noller & Feeney (1994). Whither attachment theory. Attachment to our
caregivers or to our models? Bsyohologmaljmi, 51 5-6.

O'Connor, P. (1992). Edendshioshoflaeonflomen, London: The Guilford Press.

Parker, S. & De Vries, B. (1993). Patterns of fiiendship for women and men in

same and cross-sex relationships .IoumalofSoeiaLandRersonalRelationshiosJfl, 617-
626.

Pietromonaco, P. & Camelley, K. B. (1994). Gender and working models of
attachment: Consequences for perceptions of self and romantic relationships. Personal

RelationshiosJ. 63-82-

Pistole, C. (1989). Attachment 1n adult romantic relationships. Style of conflict

resolution and relationship satisfaction. WWW,
505-510.

Pogrebin, L. C. (1987). Amongfxieuds. New York, NY: McGraw-Hill Book
Company.

Rands, M. & Levinger, G. (1979). Implicit theories of relationship. An
intergenerational study. loumalofBersonalitlLandSmiaLBsxehologLfl. 645-661

Rawlins, W. K. (1992). EdondshioMans-xs, New York: Walter de Gruyter, Inc.

Reis, H. T. (1984). Social interaction and well-being. In S. Duck (Ed.),
WWW. London: Academic Press.

Reis, H. T. & Shaver, P. (1988). Intimacy as an interpersonal process. In S.

Duck (Ed), Handhookofresearehinpersonalrelationshios (pp. 367- 389). London:
Wiley.

Reis, H. T., Senchak, M. ,& Solomon, B. (1985). Sex differences in the intimacy
of social interaction: Further examination of potential explanations. loumaLof

WW, 1204-1217.

Reisman, J. M. (1981). Adult Friendships. In S. Duck & R. Gilmour (Eds),
DexelopmgPersonalBelationships. London: Academic Press

Reisman, J. M. (1990). Intimacy in same-sex fiiendships. W, 65-82.

171

Reohr,J.R (1991). EriendshinAnexplorationofstmemreandpmeess. New
York: Garland Publishing, Inc.

Rose, S. M. (1985). Same- and cross-sex fiiendships and the psychology of
homosociality. SoxRolosJZ, 63-74.

Rubin, L. B. (1985). Woods, New York, NY: Harper & Row.

Rubin, Z. (1970). Measurement of romantic love. MW
Mayhem, 265-273.

Rubin. Z- (1973). LikingandLordnsWhoMoSom’alPsyehology. New
York: Holt, Rinehart, and Winston, Inc.

Senchak, M. & Leonard, K. E. (1992). Attachment styles and marital adjustment
among newlywed couples loumalofSoeralandPersonaLRelatlonshloLQ, 51-64

Shaver, P. R. & Brennan, K. (1992). Attachment styles and the "big five"
personality traits: Their connections with each other and with romantic relationship

outcomes. BersonalinhandSoeialPsxchologLBullethfi 536-545.

Shaver, P. R., Hazan, C., & Bradshaw, D. (1988). Love as attachment: The
integration of three behavioral systems. In R. J. Stemberg & M. L. Barnes (Eds), The

psychologLofloye (pp. 68-99). New Haven, CT: Yale University Press.

Shea, J. A. & Adams, G. R. (1984). Correlates of romantic attachment: A path
analysis study. loumaloflouthandAdoleseehele. 27-44-

Shulrnan, S., Elicker, J. & Sroufe, L. A. (1994). Stages of fiiendship growth in
preadolescence as related to attachment history. WW
Relationshiosll 341 -361

Simpson, J. A. (1990). Influence of attachment styles on romantic relationships.
WWW 971-980

Sroufe, L. A. & Fleeson, J. (1986). Attachment and the construction of

relationships. In W. Hartup and Z. Rubin (Eds.,) Relationshinsandoeyelopment (pp. 51-
71). Hillsdale, NJ: Erlbaum.

Steck, L., Levitan, D., McLane, D., & Kelley, H. H. (1979). Care, need, and
conceptions of love. loumalofliersonahmandSoolalPsxehologLAl, 481-491.

172

Stemberg, R J. & Barnes, M. (1985). Real and ideal others 1n romantic

relationships: Is four a crowd? W, 1586-
l 608.

Stemberg, R. J. & Grajek, S. (1984). The nature of love. Joumaloffiersonalim
andSm'ralPsyehologLAl, 312- 329.

Sullivan, H- H- (1953). IheJutetpetsonatheonLostxehiatm. New York!

Norton.

Swain, S. (1989). Covert intimacy. Closeness in men's friendships. In B. J.

Risman & P. Schwartz (Eds.,) Genderjnintimaterelationshios (pp. 71 -.86) Belmont,
CA: Wadsworth.

Thomas, L. (1987). Friendship. W,217-236.

Wanner U G (1936) Attachmentmmfanexandatagesmaandehrldrensself
coneeotLAfollonLuooiafiermanlonginldinaLsmdx. Doctoral Dissertation, University

of Virginia.

Weiss, R. S. (1982). Attachment rn adult life. In C. M. Parkes & J. Stevenson-

Hinde (Eds), WW (pp. 171-184). New York.
Basic Books, Inc.

Weiss, R. S. (1991). The attachment bond 1n childhood and adulthood. In C. M.

Parkes, J. Stevenson-Hinde, and P. Marris (Eds.,) AmohmonLAorosthoLlfeficho
(pp. 66-76). London: Routledge.

Weiss, L. & Lowenthal, M. F. (1975). Life-course perspectives on fiiendships.
In M. F. Lowenthal, M. Thumher, & D. Chiriboga (Eds), EouLSIagoLofLJfo (pp. 48-
61). San Francisco: Jossey-Bass.

Wheeler, L. Reis, H., & Nezlek, J. (1983). Loneliness, social interaction, and
sex roles. IoumalofPersonalioLandSoeiaLPsxchologyAi, 943- 953

Williams, D. G. (1985). Gender, masculinity-femininity, and emotional intimacy
in same-sex friendship. SoxRolosJZ, 587-600.

Wong, M. M. & Csikszentrnihalyi, M. (1991). Affiliation motivation and daily

experience: Some 1ssues on gender differences. loumloffiersonaliflmdfigmal

Wright, P. H. (1982). Men's fiiendships, women's fiiendships, and the alleged
inferiority of the latter. SoxRolosJ, 1-20.

173

Wright, P. H. (1984). Self-referent motivation and the intrinsic quality of
friendship. loumalofSoeialandPersonalRelationshiosJ, 115- 130.

Wright, P. H. (1988). Interpreting research on gender differences 1n friendship: A

case for moderation and a plea for caution. loumalofSooialonifimonalRolationships,
5, 367- 373.

Wright, P. H. & Keple, T. W. (1981). Friends and parents of a sample of high
school juniors: An exploratory study of relationship intensity and interpersonal rewards.

lorrmalofMarriageandtheEamihaAl. 559-570.

APPENDICES

1 74
APPENDIX A-l

CLOSE RELATIONSHIPS QUESTIONNAIRE Part 1

1) Following are descriptions of three general relationship styles that people often report.
Read each of the three self-descriptions below (A, B, and C) and then place a check mark
next to the letter corresponding to the style that best describes you or is closest to the way
you are. (Note: The term "close" and "intimate" refer to psychological or emotional
closeness, not necessarily to sexual intimacy.)

A. I am somewhat uncomfortable being close to others; I find it difficult to trust
them completely, difficult to allow myself to depend on them. I am nervous
when anyone gets too close, and often, love partners want me to be more
intimate than I feel comfortable being.

B. I find that others are reluctant to get as close as I would like. I often worry that
my partner doesn't really love me or won't want to stay with me. I want to
merge completely with another person, and this sometimes scares people away.

C. I find it relatively easy to get close to others and am comfortable depending on
them and having them depend on me. I don't often worry about being
abandoned or about someone getting too close to me.

Now please rate each of the relationship styles above according to the extent to which
you think each description corresponds to your general relationship style.

Not at all Somewhat Very much
like me like me like me
2. Style A. l 2 3 4 5 6 7
3. Style B. 1 2 3 4 5 6 7

4. StyleC. 1 2 3 4 5 6 7

1 75
APPENDIX A-2

CLOSE RELATIONSHIPS QUESTIONNAIRE Part 2

5) Following are descriptions of four general relationship styles that people often report.
Read each of the four self-descriptions below (A, B, C, and D) and then place a check
mark next to the letter corresponding to the style that best describes you or is closest to
the way you are. (Note: The term "close" and "intimate" refer to psychological or
emotional closeness, not necessarily to sexual intimacy.)

A. It is easy for me to become emotionally close to others. I am comfortable
depending on others and having others depend on me. I don't worry about
being alone or having others not accept me.

B. I am uncomfortable getting close to others. I want emotionally close
relationships, but I find it difficult to trust others completely, or to depend on
them. I worry that I will be hurt if I allow myself to become close to others.

C. I want to be completely emotionally intimate with others, but I often find that
others are reluctant to get as close as I would like. I am uncomfortable being
without close relationships, but I sometimes worry that others don't value me
as much as I value them.

D. I am comfortable without close emotional relationships. It is very important
to me to feel independent and self-sufficient, and I prefer not to depend on
others or have them depend on me.

Now please rate each of the relationship styles above according to the extent to which
you think each description corresponds to your general relationship style.

Not at all Somewhat Very much
like me like me like me
6. Style A. 1 2 3 4 5 6 7
7. Style B. 1 2 3 4 5 6 7
8. Style C. 1 2 3 4 5 6 7

9. Style D. 1 2 3 4 5 6 7

176

APPENDIX B-l

The Short Form of the DSHQ
Please circle the answer that applies to you for each item below.

10. Are you a first term Freshmen? (1). Yes (2). No
11. What is your sex? (1). Female (2). Male
12. Do you currently have a romantic partner? (1). Yes (2). No
(this is a person that you are dating or going out with)
13. If No, did you have a romantic partner a year ago or recently? (1). Yes (2). No
14. Do you currently have an other-sex best friend? (1). Yes (2). No
(this is a person whom you consider to be the primary and closest other-sex fiiend to
yon)
15. Is he or she your romantic partner? (1). Yes (2). No
16. If you donot currently have an other-sex best friend, did you have one a year ago or
recently? (1). Yes (2). No
17. Do you have a same-sex best friend? (1). Yes (2). No
(this is a person whom you consider to be the primary and closest same-sex friend to
yon)
18. Is he or she your romantic partner? (1). Yes (2). No
19. If you donot currently have a same-sex best friend, did you have one a year ago or
recently? (1). Yes (2). No
20. Do you have an other-sex close friend? (1). Yes (2). No
(this is a person whom you consider to be a close fiiend of the other-sex but not a best
fiiend)
21. Is he or she your romantic partner? (1). Yes (2). No
22. If you do not currently have an other-sex close friend, did you have one a year ago or
recently? (1). Yes (2). No
23. Do you have a same-sex close friend? (1). Yes (2). No
(this is a person whom you consider to be a close fiiend of the same-sex but not a best
fiiend)
24. Is he or she your romantic partner? (1). Yes (2). No
25. If you donot currently have a same-sex close friend, did you have one a year ago or
recently? (1). Yes (2). No

177

APPENDIX B-2

Demographic and Social History Questionnaire
Please circle the answer that applies to you for each item below.
1. How old are you? (1). 18 yrs old (2). 19 yrs old (3). 20 yrs old (4). 21 yrs old (5). 22+
2. What is your race? (1). White/Caucasian (2). Black/African-American
(3). Asian/Pacific Islander (4). Hispanic/Mexican American (5). Native American
3. What is your class level? ( l). Freshmen (2). Sophomore (3). Junior (4). Senior (5). Other
4. What is your sex? (1 ). Female (2). Male
5. Do you have a romantic partner? (1). Yes (2). No
(this is a person that you are dating or going out with)
6. What is the sex of your romantic partner? (1). Female (2). Male
7. Have you had a romantic partner? ( 1). Always (2). Frequently (3). Sometimes
(4). Never
If you check Never - skip questions 8-13
8. How long has this particular person been your romantic partner?
(1). 1-3 months (2). 4-6 months (3). 7-11 months (4). 1-2 years (5). 3+ years
9. How close are you to your romantic partner? 1 2 3 4 5 6 7 8 9
Not at all Extremely
10. Do you expect to still have this person as your romantic partner in the future?
(1). 1-2 years (2). 3-4 years (3). 5-6 years (4). 7 years +
l 1. If you donot have a romantic partner, did you have one a year ago or recently?
( 1). Yes (2). No
12. Iers, how close were you to this person? 1 2 3 4 5 6 7 8 9
Not at all Extremely
13. How many romantic partner(s) have you had? 1 2 3 4 5 6 7 8 9
14. Do you have a same-sex best friend? (1). Yes (2). No
(this is a person whom you consider to be the primary and closest same-sex friend to you)
15. Is he or she your romantic partner? (1). Yes (2). No
16. What is your relation to this person? (1). Friend (2). Roommate (3). Mother/Father
(4). Sister\Brother (5). Another Relative

178

APPENDIX B-2 (cont’d)

17. Have you had a same-sex best friend? (1).Always (2).Frequently (3).Sometimes
(4).Never
If you check Never - skip questions 18-23
18. How long has this particular person been your same-sex best friend?
(1). 1-3 months (2). 4-6 months (3). 7-11 months (4). 1-2 years (5). 3+ years
19. How close are you to your same-sex best friend? 1 2 3 4 5 6 7 8 9
Not at all Extremely
20. Do you expect to still have this person as your same-sex best friend in the future?
(1). 1-2 years (2). 3-4 years (3). 5-6 years (4). 7 years +
21. If you donot have a same-sex best friend, did you have one a year ago or recently?
(1). Yes (2). No
22. If Yes, how close were you to this person? 1 2 3 4 5 6 7 8 9
Not at all Extremely
23. How many same-sex beat friend(s) have you had? 1 2 3 4 5 6 7 8 9
24. Do you have an other-sex best friend? (1). Yes (2). No
(this is a person whom you consider to be the primary and closest other-sex friend to you)
25. Is he or she your romantic partner? (1). Yes (2). No
26. What is your relation to this person? (1). Friend (2). Roommate (3). Mother/Father
(4). Sister\Brother (5). Another Relative
27. Have you had an other-sex best friend? (1). Always (2). Frequently (3). Sometimes
(4).Never
If you check Never - skip questions 28-33
28. How long has this particular person been your other-sex best friend?
(1). 1-3 months (2). 4-6 months (3). 7-11 months (4). 1-2 years (5). 3+ years
29. How close are you to your other-sex best friend? 1 2 3 4 5 6 7 8 9
Not at all Extremely
30. Do you expect to still have this person as your other-sex best friend in the future?
(1). 1-2 years (2). 3-4 years (3). 5-6 years (4). 7 years +
31. If you donot have an other-sex best friend, did you have one a year ago or recently?
(1). Yes (2). No
32. If Yes, how close were you to this person? 1 2 3 4 5 6 7 8 9
Not at all Extremely

179

APPENDIX B-2 (cont’d)

33. How many other-sex belt friend(s) have you had? 1 2 3 4 5 6 7 8 9
34. Among your romantic partner, same-sex best friend, and other-sex best friend, who do you
consider yourself to be the closest to?
(1). Romantic Partner (2). Same-sex Best Friend (3). Other-sex Best Friend
35. Do you have a same-sex close friend? (1). Yes (2). No
(this is a person whom you consider to be a close friend of the same-sex but not a best friend)
36. Is he or she your romantic partner? (1). Yes (2). No
37. How close do you feel to this person? 1 2 3 4 5 6 7 8 9
Not at all Extremely
38. If you donot have a same-sex close friend, did you have one a year ago or recently?
(1). Yes (2). No
39. If Yes, how close were you to this person? 1 2 3 4 5 6 7 8 9
Not at all Extremely
40. Do you have an other-sex close friend? (1). Yes (2). No
(this is a person whom you consider to be a close friend of the other-sex but not a best friend)
41. Is he or she your romantic partner? (1). Yes (2). No
42. How close do you feel to this person? 1 2 3 4 5 6 7 8 9
Not at all Extremely

43. If you donot have an other-sex close friend, did you have one a year ago or recently?
(1). Yes (2). No
44. If Yes, how close were you to this person? 1 2 3 4 5 6 7 8 9
Not at all Extremely

45. How many female friends do you have? 1 2 3 4 5 6 7 8 9+
46. How many male friends do you have? 1 2 3 4 5 6 7 8 9+

Signed

180

APPENDIX C

MICHIGAN STATE UNIVERSITY
Department of Psychology
DEPARTMENTAL RESEARCH CONSENT FORM #1

. I have freely consented to take part in a scientific study being conducted by

Carole N. Hodge under the supervision of Dr. Joel Aronoff and Dr. Ellen Strommen,
Professors.

. The study has been explained to me and I understand that I am being requested to
complete three short questionnaires which will take about 5 minutes.

. I understand that I am free to discontinue my participation in the study at any time
without penalty.

. I understand that the results of the study will be treated in strict confidence and that I
will remain anonymous. Within these restrictions, results of the study will be
available to me at my request.

. I understand that my participation in the study does not guarantee any beneficial
results to me.

. I understand that at my request, I can receive additional explanation of the study after
my participation is completed.

. I understand that my participation today will give me a chance to participate in a future
study which will take about 1 hour and 30 minutes later on in the semester. I can
receive 3 extra credit points toward my psychology course for my participation in this
future study.

. I understand that in the event that I will not be able to participate in this experiment to
earn extra credit points, there are alternative ways for me to earn extra credit points for
my psychology class which will vary according to the instructor that I have. I can
receive information about these alternative ways from my psychology instructor.

Title of study: The Social Relationship Study

Date:
Print Name here:

 

 

181

APPENDIX D

Future Participation Sheet

Please complete this sheet if you would like to participate in the Social Relationship
Study for 3 extra credit points in your psychology class. This study will take place within
the next few weeks. This study take one hour and 30 minutes to complete.

 

 

 

Student ID #
(please print)
Name:
Local phone number:
Best time to reach me by phone: (check all that apply)
in the early morning (8am - 10am) in the late morning (10am - 11:30am)
in the afiemoon (12pm - 5pm) in the evening (6pm - 8pm)
late at night (9pm - 11:30)
(check one)
I am in Prof. Henderson's Class I am in Prof. D. Gerard's Class
I am in Prof. Jackson's Class Other Professors:

 

I am in Prof. L. Gerard's Class
I am in Prof. Hyman's Class

1 82
APPENDIX E-l

Davis-Todd's Relationship Rating Form

Please complete this questionnaire with your Same-Sex Best Friend in mind.

Rate each item below on a scale of 1-9.

1 = Not at all 4 = Somewhat (not often) 7 = A great deal
2 = Very little 5 = A fair amount 8 = Strongly (almost always)
3 = Slightly (or rarely) 6 = Very much 9 = Completely or extremely

‘ ‘N

N
9‘

1°?°>'.°‘.V‘:“1”

fl
0

u—II—II—Iu-II—Iu—I
MANN—'O

NI—II—II—II—I
cease

NNNN
Lh-bWNI-I

N
\I

Do you accept this person as she/he is?

Are you willing to ignore this person's small sins because of the way you feel about
her/him?

Is it easy for you to forgive this person?

Does this person disappoint you?

Do you respect this person?

Does this person make bad judgements on important matters?

Does this person bring out the best in you?

Is this person a good sounding board for your ideas and plans?

Do you trust this person?

. Can you count on this person in times of need?

. Does this person ever forget your welfare?

. Does this person use things against you that she or he shouldn't?
. Do you and this person openly discuss person matters?

. Do you confide in this person?

Do you feel that there are things about you that this person just would not
understand?

Do you feel some things about yourself are none of this person's business?
Do you know what kind of person she/he is?

Is this person's behavior surprising or puzzling to you?

Do you know this person's faults and shortcomings?

. Do you know about this person's past?

. Does this person dominate your thoughts?

. Does it give you pleasure just to watch or look at this person?

. Do you think about this person even when you are not with him/her?

. Are there things that you do only with this person?

. Do you have feelings about this person that you couldn't have about others?

Would you feel betrayed or hurt if this person had the same relationship with
someone else that she/he has with you?

. Do you and this person have your own way of doing things?

183
APPENDIX E—l (cont’d)

Rate each item below on a scale of 1-9.

1 = Not at all 4 = Somewhat (not often) 7 = A great deal
2 = Very little 5 = A fair amount 8 = Strongly (almost always)
3 = Slightly (or rarely) 6 = Very much 9 = Completely or extremely
31. Can you count on this person to lend you a substantial sum of money?
32. Can you count on this person to risk personal safety to help you if you were in
danger?
33. Can you count on this person to give the utmost on your behalf?
34. Are you prepared to make a significant sacrifice on this person's behalf?
35. Can you count on this person to let you know how others feel about you?
36. Can you count on this person to support you in an argument or dispute with others?
37. Can you count on this person to champion your interests where there is a conflict
between your interests and those of others?
38. Can you count on this person to come to your aid when you need help?
39. Can this person count on you for help when she/he is in need?
40. Can you count on this person to tell you what she/he really thinks about issues
regardless of whether he or she agrees with you?
41. Do you tell this person exactly what you think about important issues regardless of
whether he or she agrees with you?
42. Are you happy in your relationship with this person?
43. Has your relationship with this person satisfied your needs?
44. Has your relationship with this person been a success?
45. Do you enjoy doing things with this person more than with others?
46. Do you enjoy doing things with this person that you otherwise would not enjoy?
47. Do you enjoy this person's company?
48. Does your partner share the same feeling for you that you have for him/her?
49. Does this person really care about you as a person?
50. Do you feel that your partner cares for you as much as you care for him or her?
51. Does your partner make you feel worthwhile and special?
52. Does your partner make you feel proud of yourself?
53. Do you fight and argue with this person?
54. Does this person treat you in unfair ways?
55. Is there tension in your relationship with this person?
56. Are you confused or unsure of your feelings towards this person?
57. Do you feel that this person demands too much of your time?
58. Do you feel trapped in this relationship?
59. Do you talk with this person about your relationship?
60. Do you and this person try to work out difficulties that occur between you?
61. Are you trying to change things that you do to make the relationship better between
the two of you.
62. Are you committed to staying in your relationship?
63. Does this person measure up to your ideals for a life partner?
64. How likely is it that your relationship will be permanent?
65. How committed is your partner in this relationship?
66. Has your partner ever forced you to do something that you did not want to do?
67. Have you ever forced your partner to do something that she/he did not want to do?
68. Is your relationship one of equals?

l 84
APPENDIX E-2

Davis-Todd's Relationship Rating Form

Please complete this questionnaire with your Other-Sex Best Friend in mind.

Rate each item below on a scale of 1-9.

1 = Not at all 4 = Somewhat (not often) 7 = A great deal
2 = Very little 5 = A fair amount 8 = Strongly (almost always)
3 = Slightly (or rarely) 6 = Very much 9 = Completely or extremely

ewsewew

fl
0

10.
. Does this person ever forget your welfare?

. Does this person use things against you that she or he shouldn't?

. Do you and this person openly discuss person matters?

. Do you confide in this person?

. Do you feel that there are things about you that this person just would not

~H—i—l—l
UIAUJNI—

h—‘I—Ilflfl
00¢qu

NNNN

GUI-h

N
\I

Do you accept this person as she/he is?

Are you willing to ignore this person's small sins because of the way you feel about
her/him?

Is it easy for you to forgive this person?

Does this person disappoint you?

Do you respect this person?

Does this person make bad judgements on important matters?
Does this person bring out the best in you?

Is this person a good sounding board for your ideas and plans?
Do you trust this person?

Can you count on this person in times of need?

understand?
Do you feel some things about yourself are none of this person's business?

. Do you know what kind of person she/he is?
. Is this person's behavior surprising or puzzling to you?
. Do you know this person's faults and shortcomings?

Do you know about this person's past?

. Does this person dominate your thoughts?

. Does it give you pleasure just to watch or look at this person?

. Do you think about this person even when you are not with him/her?

. Are there things that you do only with this person?

. Do you have feelings about this person that you couldn't have about others?

. Would you feel betrayed or hurt if this person had the same relationship with

someone else that she/he has with you?

. Do you and this person have your own way of doing things?

185

APPENDIX E-2 (cont’d)
Rate each item below on a scale of 1-9.
1 = Not at all 4 = Somewhat (not often) 7 = A great deal
2 = Very little 5 = A fair amount 8 = Strongly (almost always)
3 = Slightly (or rarely) 6 = Very much 9 = Completely or extremely
31. Can you count on this person to lend you a substantial sum of money?
32. Can you count on this person to risk personal safety to help you if you were in
danger?
33. Can you count on this person to give the utmost on your behalf?
34. Are you prepared to make a significant sacrifice on this person's behalf?
35. Can you count on this person to let you know how others feel about you?
36. Can you count on this person to support you in an argument or dispute with others?
37 . Can you count on this person to champion your interests where there is a conflict
between your interests and those of others?
38. Can you count on this person to come to your aid when you need help?
39. Can this person count on you for help when she/he is in need?
40. Can you count on this person to tell you what she/he really thinks about issues
regardless of whether he or she agrees with you?
41. Do you tell this person exactly what you think about important issues regardless of
whether he or she agrees with you?
42. Are you happy in your relationship with this person?
43. Has your relationship with this person satisfied your needs?
44. Has your relationship with this person been a success?
45. Do you enjoy doing things with this person more than with others?
46. Do you enjoy doing things with this person that you otherwise would not enjoy?
47. Do you enjoy this person's company?
48. Does your partner share the same feeling for you that you have for him/her?
49. Does this person really care about you as a person?
50. Do you feel that your partner cares for you as much as you care for him or her?
51. Does your partner make you feel worthwhile and special?
52. Does your partner make you feel proud of yourself?
53. Do you fight and argue with this person?
54. Does this person treat you in unfair ways?
55. Is there tension in your relationship with this person?
56. Are you confused or unsure of your feelings towards this person?
57. Do you feel that this person demands too much of your time?
5 8. Do you feel trapped in this relationship?
59. Do you talk with this person about your relationship?
60. Do you and this person try to work out difficulties that occur between you?
61. Are you trying to change things that you do to make the relationship better between
the two of you.
62. Are you committed to staying in your relationship?
63. Does this person measure up to your ideals for a life partner?
64. How likely is it that your relationship will be permanent?
65. How committed is your partner in this relationship?
66. Has your partner ever forced you to do something that you did not want to do?
67. Have you ever forced your partner to do something that she/he did not want to do?
68. Is your relationship one of equals?

1 86
APPENDIX E-3

Davis-Todd's Relationship Rating Form

Please complete this questionnaire with your Romantic Partner in mind.
Rate each item below on a scale of 1-9.

1 = Not at all 4 = Somewhat (not often) 7 = A great deal
2 = Very little 5 = A fair amount 8 = Strongly (almost always)
3 = Slightly (or rarely) 6 = Very much 9 = Completely or extremely

Plfilfl‘Ql9‘P’P

I—II—Iu—II—II—II—I
Lh-hbJN—‘C

O \O M \l 0‘

NNN
aware

N
.°‘

MN
00%

m r—I

mix)
90

Do you accept this person as she/he is?

Are you willing to ignore this person's small sins because of the way you feel about
her/him?

Is it easy for you to forgive this person?

Does this person disappoint you?

Do you respect this person?

Does this person make bad judgements on important matters?

Does this person bring out the best in you?

Is this person a good sounding board for your ideas and plans?

Do you trust this person?

. Can you count on this person in times of need?

. Does this person ever forget your welfare?

. Does this person use things against you that she or he shouldn't?

. Do you and this person openly discuss person matters?

. Do you confide in this person?

. Do you feel that there are things about you that this person just would not

understand?

. Do you feel some things about yourself are none of this person's business?
. Do you know what kind of person she/he is?

. Is this person's behavior surprising or puzzling to you?

. Do you know this person's faults and shortcomings?

. Do you know about this person's past?

. Does this person dominate your thoughts?

Does it give you pleasure just to watch or look at this person?

. Do you think about this person even when you are not with him/her?

Are there things that you do only with this person?

. Do you have feelings about this person that you couldn't have about others?

Would you feel betrayed or hurt if this person had the same relationship with
someone else that she/he has with you?

. Do you and this person have your own way of doing things?
. Are you sexually intimate with this person?
. Do you find this person sexually attractive?

Do you enjoy being touched by this person and touching him/her?

187

APPENDIX E-3 (cont’d)
Rate each item below on a scale of 1-9.
1 = Not at all 4 = Somewhat (not often) 7 = A great deal
2 = Very little 5 = A fair amount 8 = Strongly (almost always)
3 = Slightly (or rarely) 6 = Very much 9 = Completely or extremely
31. Can you count on this person to lend you a substantial sum of money?
32. Can you count on this person to risk personal safety to help you if you were in
danger?
33. Can you count on this person to give the utmost on your behalf?
34. Are you prepared to make a significant sacrifice on this person's behalf?
35. Can you count on this person to let you know how others feel about you?
36. Can you count on this person to support you in an argument or dispute with others?
37. Can you count on this person to champion your interests where there is a conflict
between your interests and those of others?
38. Can you count on this person to come to your aid when you need help?
39. Can this person count on you for help when she/he is in need?
40. Can you count on this person to tell you what she/he really thinks about issues
regardless of whether he or she agrees with you?
41. Do you tell this person exactly what you think about important issues regardless of
whether he or she agrees with you?
42. Are you happy in your relationship with this person?
43. Has your relationship with this person satisfied your needs?
44. Has your relationship with this person been a success?
45. Do you enjoy doing things with this person more than with others?
46. Do you enjoy doing things with this person that you otherwise would not enjoy?
47. Do you enjoy this person's company?
48. Does your partner share the same feeling for you that you have for him/her?
49. Does this person really care about you as a person?
50. Do you feel that your partner cares for you as much as you care for him or her?
51. Does your partner make you feel worthwhile and special?
52. Does your partner make you feel proud of yourself?
53. Do you fight and argue with this person?
54. Does this person treat you in unfair ways?
55. Is there tension in your relationship with this person?
56. Are you confused or unsure of your feelings towards this person?
57. Do you feel that this person demands too much of your time?
58. Do you feel trapped in this relationship?
59. Do you talk with this person about your relationship?
60. Do you and this person try to work out difficulties that occur between you?
61. Are you trying to change things that you do to make the relationship better between
the two of you.
62. Are you committed to staying in your relationship?
63. Does this person measure up to your ideals for a life partner?
64. How likely is it that your relationship will be permanent?
65. How committed is your partner in this relationship?
66. Has your partner ever forced you to do something that you did not want to do?
67. Have you ever forced your partner to do something that she/he did not want to do?
68. Is your relationship one of equals?

1 88
APPENDIX F

MICHIGAN STATE UNIVERSITY
Department of Psychology
DEPARTMENTAL RESEARCH CONSENT FORM #2

I have freely consented to take part in a scientific study being conducted by
Carole N. Hodge under the supervision of Dr. Joel Aronoff and Dr. Ellen
Strommen, Professors.

 

2. The study has been explained to me and I understand that I am being requested to
complete a questionnaire packet on fiiendships and romantic relationships.

3. I understand that I am free to discontinue my participation in the study at any time
without penalty.

4. I understand that the results of the study will be treated in strict confidence and that
I will remain anonymous. Within these restrictions, results of the study will be
available to me at my request.

5. I understand that my participation in the study does not guarantee any beneficial
results to me.

6. I understand that at my request, I can receive additional explanation of the study
after my participation is completed.

7. I understand that my participation will take about 1 hour and 30 minutes. At the
conclusion of this study, I can receive 3 extra credit points toward my psychology
course for my participation.

Title of study: The Social Relationship Study

Signed

Date:

 

Print Name here:

 

I 89
APPENDIX G

PARTICIPANT INFORMATION SHEET

This study is concerned with fi'iendship and romantic relationships in the context of
attachment style. The study is interested in examining how individuals in different
attachment styles will rate their best friend and their romantic partner on certain
relationship characteristics (i.e., viability, intimacy, and satisfaction).

Attachment refers to affectional bonds that individuals display toward others. An
individual develops an attachment style from their early relationship with their primary
caregiver. There are three attachment styles - secure, anxious-ambivalent, and avoidant.
Secure people have a positive view of relationships and others. They perceive others as
good and trustworthy. Anxious-ambivalent people have a negative view of relationships.
They want to be in a relationship and worry about being abandoned. They perceive
others as being unreliable and unwilling to commit to a relationship. Avoidant
individuals have a negative view of relationship and of others. They perceive others as
being unreliable and overly eager to commit to a relationship. They try to become
emotionally self-sufficient.

The relationship characteristics that I am interested in are viability, intimacy, and
satisfaction. Viability measures the relationship characteristics of trust, respect, and
acceptance/tolerance. Intimacy measures the relationship characteristics of confiding and
understanding. Satisfaction measures the relationship characteristics of reciprocity,
success, esteem, and commitment.

Individuals within different attachment styles are predicted to have different ratings
for each of the relationship characteristics for their same-sex best friend, other-sex best
fiiend, and their romantic partner.

For more information about this study, you can contact Carole Hodge at 353-5324
or at 410 Baker Hall. Thank you again for participating in today's study.

190

APPENDIX H-l

Research Script

Mass Testing during the first few days of classes

Hi, my name is Carole Hodge. I am a graduate student in psychology and I am
working on my dissertation. I am here today to conduct part of my dissertation study.
This is an interesting study. I am studying how people meet, get to know each other, and
form relationships. I am looking to study people in all types of relationships. This study
should be of interest to most people because humans are social creatures and we form
some type of relationship with others. I need to have a large array of people in different
relationships. This prescreening is a way of identifying a broad population for the study.
I need your help and by completing these few short questions today, you can help me
gather my prescreening data and have a chance to participate in the Social Relationship
Study later on in the semester.

Now, I will be distributing three short questionnaires, a future participation sheet,
and a consent form. As soon as you receive your questionnaire packet, please read and
Sign the consent form first, then complete the questionnaires. By completing these today,
it will take you about 5 minutes, you will be eligible to participate in the Social
Relationship Study later on in the semester for 3 extra credit points toward your
psychology class. If you would like to participate in the Social Relationship Study for the
3 extra credit points, please complete the Future Participation Sheet which asks for your
name, your phone number, the best time to reach you by phone, and the name of your
Introductory Psychology instructor.

Please note that if you do not wish to participate in this study later on in the
semester, you will be able to earn extra credit points for your psychology course through
alternative ways. You can receive information about these alternative ways from your
psychology instructor. Thank you again for your participation.

191

APPENDIX H-2

Research Script

Social Relationship Study

Hi, my name is Carole Hodge and I will be your experimenter. This is my dissertation and
I would really like to thank you for coming to today's study. There are a few good things about
participating in a study that is a dissertation: one, as a participant, you can feel that you are a part
of something, a project; two, you helped a person, me, complete their Ph.D.; three, when I write
this up, you can find yourself listed as a participant in the dissertation, there will be a copy at the
library.

This study is called the Social Relationship Study. I am interested in your responses to
questions about friendships and romantic relationships. The two friendships that I am
particularly interested in are the other-sex best friend and the same-sex best friend. I am also
interested in your responses to a few personality variables, as well as variables that are perceived
as important to most relationships. More explanation about the study will be given at the end of
the study and there is also a Participant Information Sheet for you to take home with you.
Today's study will require you to complete a questionnaire packet which will take about 1 hour
and 30 minutes. At the end of the study, please return your completed questionnaire packet to
me. At that point, for your time and participation, you will receive three extra credit points
toward your psychology class. Bring your psychology credit card up with you and I will stamp
the card.

Okay, let's begin. I am now administering the consent form. Please read it and if you
wish to participate in today's study, sign the consent form. After I collect the consent forms, I
will administer the questionnaire packet. Each packet contains several questionnaires. All
participants will receive the same number of questionnaires, however the order of the
questionnaires is different for each of you. When everyone has a packet, I will go over some
instructions with you. Please pay attention because you must follow these instructions carefully
to complete the packet correctly. Remember to read the directions for each questionnaire before
answering it and read each question carefully and respond to it appropriately. You have up to 1
hour and 30 minutes to complete all the questionnaires.

This part will not be read out to participants

After all the participants are finished, the experimenter will read out loud the

Participant Information Sheet and answer any questions that any participants may have

about the study.

[** If any participants say that they have two or more best fiiends of the same or
opposite sex, they will be told to pick the person that they consider to be their primary
best friend and answer the questionnaires thinking of this person] [** If any participants
say that they have two romantic partners, they will be told to pick the person that they
consider to be their primary romantic partner and answer the questionnaires thinking of
this person.]

192

FOOTNOTE

1 Similar results were found for parallel analyses conducted on the Between-
Subjects Sample and the Within-Subjects Sample.

"Illlllllllllllllllli