Say»:
#.;;...a: ..

«J
1..x.t...t'
T. q .23!!!

. £35525
(.1: .
..

Mama: . ..

 

.. AR.

.2» 1:-
l...

m T}... “u .1;
«UN 5-941}; 2
z... ,
.v. 5!?
33.00...» i 5.3.1...
I a . 4. (All...
.p 00?“
w {a}... 6h...
. “mark...“ E
1....

. ._ .74.. v.
1:. c. ling... .
.9 1...]: t :1.

 

 

THES’S

llfllllllll

 

lllllllllllll\llllllllllllllll

3 1293 01413

This is to certify that the

dissertation entitled

Attachment Styles and Gender Roles:
Contributions to Problem Solving Communication
in Dating Couples

presented by

Jennie Ann Leskela

has been accepted towards fulfillment
of the requirements for

Doctoral degree in Counseling Psychology

 
 

Major professor V d/

Date 4-18—96

 

MSU is an Affirmative Action/Equal Opportunity Institution 0-12771

 

LIBRARY

Michigan State
University

 

 

 

cu.

PLACE IN RETURN BOX to roman this chockout from your rocord.
TO AVOID F INES return on or bdoro duo duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

lw

 

1m] —— ——j
'L_______J
r: ——:l’-:I,

MSU It An Afflnnotlvo Action/Equal Opportmlty lnotltwon
Wyn-9.1

 

 

 

 

 

 

ATTACHMENT STYLES AND GENDER ROLES:
CONTRIBUTIONS TO PROBLEM SOLVING COMMUNICATION IN
DATING COUPLES
By

Jennie Ann Leskela

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Counseling, Educational Psychology
and Special Education
1996

ABSTRACT
ATTACHMENT STYLES AND GENDER ROLES:
CONTRIBUTIONS TO PROBLEM SOLVING COMMUNICATION IN
DATING COUPLES
By

Jennie Ann Leskela

In an effort to expand the research on couple functioning in close relationships, this
study adopted Bowlby's attachment theory as a framework for identifying predisposing
factors in couple communication patterns. Initial studies have suggested that attachment
style, developed in infancy, continues into adulthood. Furthermore, prior research has
found attachment style differences in dating and manied couples play a role in couple
interaction and relationship functioning. The purpose of this study was to examine the
contribution of adult attachment styles to specific problem solving communication exhibited
by dating couples. This study (a) examined the relationship of participants' attachment
styles to their problem solving communication; (b) explored the extent to which sex and
gender role affects the relationship of attachment style and problem solving communication;
and (c) identified couple's conjoint attachment style-related contributions to their problem
solving communication. Forty couples completed both attachment style and gender role
measures, and then participated in an audiotaped problem solving task designed to measure
problem solving communication styles. Data analyses examined individual participant
differences in attachment style, gender role, and problem solving communication styles.
Finally, the contributions of couple attachment style pairings to problem solving
communication outcomes was examined. Limitations of the study and implications for

further research and counseling were discussed.

Copyright by
J ENNIE ANN LESKELA
1996

ACKNOWLEDGMENTS

The accomplishment of this dissertation was made more manageable and enjoyable
thanks to the contributions and support of several individuals. My advisor and dissertation
committee chairperson, Dr. Fred Lopez, provided the necessary balance between support
and challenge, while being consistently available during the stress of data collection and
post hoc analyses and editing Of the dissertation. I especially extend my thanks and
appreciation to Dr. Lopez for helping me develop my professional career and acquire "self-
efficacy" for research and writing. I thank the members from my dissertation committee
for their feedback on this project: Drs. Ann Austin, Robert Boger, and Linda Forrest.
Furthermore, I could not have conducted my statistical analyses as proficiently without the
humor and expertise from Dr. Ira Washington, 111 and Sean Nugent. In addition to
thanking the couples, raters and transcriptionists who participated in this project, I thank
my friends who provided help and support in this process: Connie Devantier, John
O'Brien, David Sagula, Lisa Schirmer, and Louise Sommer. I thank the Attachment Style
Research Team for their feedback during the project's design and implementation.

I thank my family for their long standing support which helped me achieve my goals
(Edna and Elroy Leskela, Ben and Jenni Leskela, Stanley and Judy Syria, Penny and Pat
Hill, and Mary and Carroll Syria). I extend my thanks to my wonderful friends who
supported me during all the ups and downs of graduate school, life and the dissertation
project: Lynn Chambers, Marianne Dunn, Mary Frey, Cindy KOk, Sue Leskela, Ellen
Narusis, Sukhvender Nijjer, Cynthia Riggs, Lisa Schirmer, and Louise Sommer. And, a
final thank you to the mentors throughout my weer, who guided my professional and

personal development; Kenneth Kelley, Ellie Vanderlinde and Leonard Vanderlinde.

TABLE OF CONTENTS

LIST OF TABLES
LIST OF FIGURES

Introduction

Problem Statement

Review of the Literature

Attachment Theory . .
Theoretical Foundations
Key Assumptions. .

Internal Working Model

Ainsworth's Empirical Support
Patterns of Attachment .

Continuity of Attachment Styles . .

Conceptualization and Measurement of Adult Attachment .
Three Group Categorization of Adult Attachment Styles .
Four Group Categorization of Adult Attachment Styles
Summary. . .

Individual Differences in Adult Attachment Styles

Adult Attachment and Relationship Outcome .

Summ . .

Adult Attachment and Relationship PrOcess
Summ . .

Attachment Style, Sex, and Gender ROle .

Summ .

Attachment Style and Problem Solving
Summ .

Attachment Style and ConjOint Interaction
Summ .

Sex/Gender Role and Couple Communication and FunctiOning

Summary and Limitations

Definitions .

Adult Attachment Styles

Dating Relationship . .

Problem Solving Communication

Sex and Gender Role .
Hypotheses

Relationship Among Attachment Styles, Sex, Gender Role and Problem .

Solving Communication .
Relationship Between the Couple's Conjoint Attachment Style and
Problem Solving Behavior . . . . . .

Methodology .

é:

M

\O\O®\I\IO\O\ A.—

Participants
Procedures
Instruments .
Demographic and Background InfOnnation FOrrn
Attachment Style . .
Bartholomew' 3 Relationship Questionnaire
Simpson's Adult Attachment Style Inventory
Gender Role . .
Bem Sex Role Inventory .
Problem Solving Task (CIT).
Interrater Reliability
Research Hypotheses .
Data Analysis .

Results .
Descriptive Statistics
Correlational Findings
CIT Scores . .
Overall Problem SOlving Scores f or Raters and .COupIes .
Masculinity and Femininity . . . .
Demographic/Background Variables .
Research Hypotheses . .
Relationships of Attachment Styles, Sex, and Gender Role to Problem
Solving Communication. . . . .
Relationship Between Couple' 3 Conjoint Attachment Style and
Problem Solving Behavior . . . .
Data Analyses for the Main Hypotheses
Hypotheses 1 -3 .
Hypothesis 4 .
Hypothesis 5 .
Post Hoc Analyses .
Correlational Findings .
Stress and Success Scores of the CIT
Femininity .
Relationship SatisfactiOn .
Relationship Seriousness .
Attachment Style . . .
Demographic/Background Variables .
Hypothesesl 1-3 Re- Analyzed. . .
Sex Differences and Problem Solving Effectiveness .
Couple Analyses . .
Gender Role of the Couple . .
Connectedness/Individuality (Separateness)
Individual Gender Roles of the Sample
Discussion .
Individual AttaChment Style and Problem Solving Commimication

Sex, Gender Role, Attachment Style, and Problem Solving CommunicatiOn
Couple' 8 Conjoint Attachment Style and Problem Solving Communication .

Limitations of the Study

CIT Limitations . .

Sampling Limitations .

Other Limitations .
Implications for Practice . .
Recommendations for Future Research

vi

118

122

Appendices

Appendix A

Request for ParticipatiOn and Participation Consent FOrm.

Appendix B . .
Request of Participation, Phase 2.
Appendix C

Request forParticipatiOn. and Participation Consent FOrm.

Appendix D .

Demographic and Background Inf orrnatron
Appendix E .

Bartholomew Relationship QueinOnnaire
Appendix F .

Simpson‘ 8 Adult Attachment Style InventOry.
Appendix G . .

Bern Sex Role Inventory .
Appendix H .

Activity Planning Form
Appendix I .

Transcript Format.
Appendix J . .

Transcript Guidelines .
Appendix K . .

Audiotape Coding Manual
Appendix L . .

Sample Coding Sheet.
Appendix M . .

Sample Coding Sheet 2
Appendix N . . .

Problem Solving COmrnunication Totals .
Appendix 0 . .

Couple Problem Solving Rating Seore
Appendix P . .

Couple Problem SOlving Rating Score (Rater)

References

vii

Table 1.
Table 2.
Table 3.
Table 4.

Table 5.

Table 6.
Table 7.
Table 8.
Table 9.

Table 10.

Table 11.

Table 12.

Table 13.

Table 14.

Table 15.

Table 16.

Table 17.

LIST OF TABLES

Hazan and Shaver's Attachment Style Inventory

Sample Demographic Information - Phase 1

Attachment Style Rate of Change

Intercorrelations of Simpson's Adult Attachment Style Inventory at
Phase 1 and Phase 2 . . . .

Means on Simpson's Adult Attachment Style Inventory Subscales by
Attachment Style Self —Class1frcatron .

Sample Demographic Information - Phase 2
Intercorrelations of Demographic and Research Variables .
Summary of the Analysis of Covariance Testing Hypotheses 1-3 .

CIT Subscale Means by Attachment Style Group (Secure/Insecure)
and Sex

Summary of the Hierarchical Multiple Regression Analyses of CIT
Subscale Scores .

Summary of the Analyses of Covariance of Couple's Combined
Mutuality and Problem-Solving Effectiveness Scores and Rater's
Problem-Solving Effectiveness Scores (Hypothesis 5).

Couple Attachment Style Pair Means on Couple's Combined
Mutuality and Problem Solving Effectiveness Scores and on Rater's
Problem Solving Effectiveness Scores . . . . . .

Summary of the Analysis of Covariance Testing Hypotheses] 1-3
(Revised). . .

CIT Subscale Means by (Revised) Attachment Style Group and Sex .

Analysis of Covariance of Overall Problem Solving Effectiveness by

Attachment Style Group and Sex with Length of Dating Relationship as

the Covariate .

Means of Overall Problem Solving Effectiveness by Attachment Style

Group and Sex with Length of Dating Relationship as the Covariate .

Summary of the Univariate Analyses of CIT Subscale Scores by
Attachment Style and Gender Role Group of the Couple .

viii

12
45
52

55

69

78

89

91

94

Table 18.

Table 19.
Table 20.

Table 21.

Table 22.

Table 23.

Means of CIT Subscale Scores by Attachment Style and Gender Role
Group of the Couple . . . . 96

Means of CIT Subscale Scores by Individual Statement and Sex . . 97

Summary of the Univariate Analyses for Connectedness and Individuality
(Separateness) Scores . . . . . . . . . 102

Means of Connectedness and Individuality (Separateness) Scores by
Attachment Style of the Couple, Gender Role of the Couple and Sex. 103

Summary of the Analysis of Variance of the Relative Variation of
Gender Role of the Group (Sample) by Gender Role of the Individual
and Sex . . 106

Means of Relative Variation of Gender Role of the Group (Sample)
by Gender Role of the Individual and Sex . . . 107

ix

LIST OF FIGURES

Figure 1. Hierarchical Regression Scatterplot

Figure 2. Mutuality: Sex by Individual Statements

Figure 3. Permeability: Sex by Individual Statements

Figure 4. Individual Gender Role Of Sample: Sex by Individual Categories

82

100
108

CHAPTER I

Introduction

Research on close relationships is a relevant societal issue (Holmes & Boon, 1990).
Increasing incidents of domestic violence and divorce rates highlight the need for more
information on couple functioning and interaction (Miller, 1991; Weitzman, 1985). Theory
and research in this area has received increasing attention in the past decade with a recent
influx of studies. Research has consistently indicated that effective communication patterns
relate to couple satisfaction (Billings, 1979; Koren & Carlton, 1980; Lopez, 1993;
Margolin & Wampold, 1981; Rusbult, Johnson, & Morrow, 1986; Rusbult, Verette,
Whitney, Slovik, & Lipkus, 1991; White, 1989). Despite the best intentions, all partners
in close relationships eventually communicate inef f ectively and experience relationship
conflict. Research further indicates that distressed couples exhibit more negative problem
solving communication patterns (Billings, 1979). Based on these results, additional
information is needed regarding the specific couple interaction patterns that promote healthy
functioning. One strategy for accomplishing this involves identifying factors in the early
stages of relationship development that place couples at risk. If areas of incompatibility can
be identified eariy, couples can address communication and interaction difficulties before
they become locked into more severe, recurrent patterns of conflict (Holmes et al., 1990).

Currently, there has been a growing number of studies examining close
relationships within the framework of Bowlby's attachment theory. This research
demonstrates that one's adult attachment style affects couple interaction and functioning
(Shaver & Hazan, 1993). This area of research provides a new approach in which love is
conceptualized as an attachment process. Specifically, how an infant attaches to a primary

caregiver may be similar to how an adult attaches to their dating and marital partner.

1

2
Information on factors that predispose adults to engage in specific positive or negative

behaviors may provide useful information on distressed and non-distressed relationships.
Future research in this area may help mental health professionals increase their
understanding of close relationships and identify areas for prevention and therapeutic
interventions for couple counseling.

Attachment theory developed from the work of John Bowlby (1969/ 1982) and
Mary Ainsworth (1982). Hazan and Shaver (1987) used attachment theory as a framework
for studying romantic relationships. They provided theoretical and empirical support for
attachment style differences in adult intimate relationships. Hazan and Shaver proposed
that affectional bonds developed in infancy between the infant and primary caregiver predict
adult attachment styles. The specific attachment style from infancy is believed to be
relatively enduring and manifests itself in the way adults think, feel, and behave in adult
romantic relationships. Individuals endorsing different attachment styles reported
differences in their descriptions of child/parent relationship histories, themselves, and
others. Furthermore, individuals exhibiting attachment style differences reported different
types of romantic love experiences. Hazan and Shaver demonstrated that the experience of
adult romantic relationships is compatible with Bowlby's theory of attachment.

Hazan and Shaver's (1987) article generated numerous studies that replicated and
extended their initial work on individual attachment style differences in romantic
relationships. This literature indicates that attachment style differences in couples are
related to type of commitment, conflict style resolution, degree of intimacy, level of self -
disclosure, problem solving behavior, relationship satisfaction, self-esteem management,
support-seeking and support-giving behaviors, trust, use of power tactics, and views on
relationship quality (Bartholomew, 1990; Carnelley, Pietromonaco, & Jaffe, 1994; Collins
& Read, 1990; Feeney & Noller, 1990; Feeney & Noller, 1991; Kobak & Hazan, 1991;
Mikulincer & Erev, 1991; Mikulincer & Nachshon, 1991; Pistole, 1989; Pistole, 1993;
Rothbard, Roberts, Leonard, & Eiden, 1993; Senchak & Leonard, 1992; Simpson, 1990;

3
Simpson, Rholes, & Nelligan, 1992). These preliminary studies suggest that attachment

style differences in dating and married couples play an important role in couple interaction
and functioning.

Although these findings have promising implications for couple interaction and
functioning, several conceptual weaknesses exist in the current literature. First, the
majority of dating and marital relationship studies used the three category typology of
secure, anxious, and avoidant based on Ainsworth's work (Ainsworth et al., 1978) and
Hazan and Shaver's (1987) extension of the typology to adult functioning. Currently,
Shaver and Hazan (1993) support Bartholomew's ( 1990) newer model of attachment styles
which identifies two types of avoidant styles: dismissive and fearful. The introduction of a
fourth attachment category provides for finer distinctions among the attachment styles.

Second, the literature reports inconsistent results regarding sex and gender role
differences. Several studies suggest that sex and/or gender role may moderate the
relationship of attachment style and relationship functioning. Collins and Read (1990)
reported that for women, their partner's comfort with closeness predicted relationship
satisfaction. For men, relationship satisfaction related to their partner's level of anxiety
about abandonment. Both partners were less satisfied with the relationship when the male
had an avoidant attachment style and the female had an anxious attachment style. Simpson
(1990) reported that males with an avoidant attachment style experienced less emotional
distress after relationship termination. Males reported a decreased level of satisfaction
when the females exhibited an anxious attachment style. Simpson, Rholes, and Nelligan
(1992) examined how adult attachment style moderated behavior in couples when the
female member was confronted with an anxiety provoking situation. Feeney, Noller, and
Patty (1993) reported that adolescent females with an avoidant attachment style and males
with an anxious/ambivalent (preoccupied) attachment style were less likely to report
involvement in sexual intercourse. Several studies reported that men identified as avoidant

and women identified as preoccupied (anxious) experienced lower levels of relationship

4
satisfaction (Fiala & Pietromonaco, 1991; Kirkpatrick & Davis, 1994; Pietromonaco &

Carnelley, 1994) and higher levels of relationship conflict and ambivalence (Fiala &
Pietromonaco, 1991; Kirkpatrick & Davis, 1994). Further studies need to clarify a) if the
reported male/female differences in relationship functioning are related to sex and/or gender
role; and b) the inter-relationships of sex, gender role, attachment styles, and relationship
functioning.

Third, the conjoint contribution of each partner's attachment style to relationship
functioning has been largely overlooked. Most studies have examined the effect of each
partner's attachment style separately. Further research in this area may support
Kirkpatrick and Davis' (1994) argument that specific attachment styles are not ”universally
good or bad.” Instead, different attachment pairings may result in successful relationship
experiences.

Fourth, efforts to examine the relationship of adult attachment styles to dyadic
functioning has generally relied on self -reports. Memory recall and self-report are both
vulnerable to distortions. With regard to problem solving differences in particular, several
studies asked the couples to recall or imagine a conflict or problem (e. g. Fiala &
Pietromonaco, 1991; Pistole, 1989). Other studies asked the couple to identify their own
problem (e. g. Kobak & Hazan, 1991; Rothbard et al., 1993). Recalling a problem is
different from experiencing it directly. How one perceives what they would do may be
different from what they would actually do. Furthermore, when couples are given a choice
of problem identification, they may identify a problem that most likely has been rehearsed
by numerous discussions about the concern. Research employing in-vivo methods for
observing spontaneous couple interaction is needed. Direct Observation may provide new
information on couple's problem solving communication.

Problem St_atement
There is a need for more controlled research studies on the impact of attachment

style on couple interaction that extends beyond the realm Of self -report. To enhance the

5
current research base, future studies need to use the four category attachment style typology

while concurrently examining the influence of sex and gender role as moderating variables.
Examining the role of attachment styles and their interaction with sex and gender role to
specific problem solving communications may elaborate our understanding of problematic
couple functioning.

The purpose of this study is to examine the contribution of adult attachment styles
to specific problem solving communication exhibited by dating couples. This study will (a)
examine the relationship of participants' attachment styles to their in vivo problem solving
communication; (b) explore the extent to which sex and gender role affects the relationship
of attachment style and problem solving communication; and (0) identify couple's conjoint
attachment style-related contributions to their problem solving communication.

Should support be found for the hypotheses, greater attention can be given to
considering the issues of sex, gender role, and individual and conjoint couple attachment
style differences when working with couples in counseling settings. Information on
predisposing factors and early relationship functioning will be helpful in identifying
treatment interventions and preventive methods. A lack of support for the hypotheses
would suggest that specific attachment style combinations and/or gender role do not predict
actual problem-solving communications. In this case, other variables related to differences
in couple interaction and functioning may need to be explored. Furthermore, if attachment
style differences only reflect gender role differences, we only need to understand gender

role and not pursue attachment style differences further.

CHAPTER II

Review of the Literature

This chapter will focus on ten areas of related research including: (a) Bowlby's
attachment theory; (b) Ainsworth's empirical work; (c) conceptualization and measurement
of adult attachment; ((1) individual differences in adult attachment style; (e) adult attachment
and relationship outcome; (1) adult attachment and relationship process; (g) attachment
style, sex, and gender role; (h) attachment style and problem solving; (i) attachment style
and conjoint interaction; and (j) sex/ gender role and couple communication and functioning.

Attachment Theory

Attachment theory developed from the joint work of John Bowlby and Mary
Ainsworth (Ainsworth & Bowlby, 1991; Bretherton, 1992). The theory proposes an
ethological approach to personality development which is based on ideas from cybemetics,
information processing, developmental psychology, and psychoanalysis (Bretherton,
1992). Attachment theory explains the role that infant-caregiver emotional bond plays
throughout the course of one's life. Bowlby's classic trilogy on attachment, separation,
and loss (1969/ 1982, 1973, 1980) provided the theoretical explanations Of how parental
availability and responsiveness provides a sense of security which guides the child's
development. When the infant is threatened, he/she innately seeks proximity with his/her
primary attachment figure. How this figure responds to this proximity-seeking shapes the
nature of the inf ant-caregiver attachment bond. From these early attachment experiences,
the child develops internal representations of self and others that guide his or her future
expectations and behaviors in relationships. Bowlby refers to this internal representation as
”internal working models." Ainsworth provided the empirical support for Bowlby's ideas

via her controlled observational studies of mother—infant interaction (Ainsworth, 1967;

6

7
Ainsworth, Blehar, Waters, & Wall, 1978; Ainsworth, 1982). She proposed that the

mother's accessibility and responsiveness to an infant's behavioral cues related to the
formation of different attachment styles (Ainsworth et al., 1991). Currently, attachment
theory and research is receiving renewed interest and is being applied to adult development.
Theoretical Foundations

Bowlby's attachment theory developed from his early research on the impact of
maternal separation and inadequate maternal care on orphaned infants and juvenile
delinquents. He developed the connection between early parental deprivation and loss with
later childhood emotional disturbances. Initially, Bowlby noted that infant primates and
infants and young children demonstrated similar distressed reactions when separated from
their primary caregiver. Due to the similar behavioral responses, Bowlby concluded that
the nature of early attachment bonds was innate and internally motivated and that these
bonds had evolutionary significance (Bowlby, 1988). Bowlby hypothesized that
attachment results from a biologically-driven system designed to ensure the protection and
care for the infant. He proposed that attachment is a drive related behavioral system similar
to the drive behavior of feeding, mating, and exploration. The attachment system tends to
be activated during periods of stress and enhances the likelihood of survival and
reproduction of the species (Bowlby, 1979).
Key Assumptions

The attachment process includes three defining features: secure base, safe haven,
and proximity seeking and maintenance behaviors. The theory assumes that when an
attachment figure is available and responsive (secure base), the child is more likely to
engage in exploratory behavior with others in his or her environment. The neamess of the
caregiver provides a safe haven in which to return if the infant feels threatened during
exploration activities. Infants seek proximity and protest separation from the caregiver
when they need the assurance of security and safety (proximity seeking and maintenance)

(Ainsworth et al., 1978; Bowlby, 1979; Bowlby, 1988; Shaver & Hazan, 1993). Bowlby

8
stated that the concept of secure base continues into adulthood. An adult shows the same

patterns from infancy Of moving away and maintaining contact with those he/she loves
(Bowlby, 1979).

In addition, the child exhibits emotional reactions when separated from his or her
primary caregiver. These reactions are designed to reestablish proximity. The first reaction
is protest which involves crying, active searching, and resistance to other's soothing. If
such reactions are unsuccessful in restoring contact with the caregiver, a subsequent
reaction is despair in which the infant exhibits passivity and sadness. If these affective
reactions are unsuccessful on restoring proximity with the caregiver, the final reaction is
emotional detachment which is an active, defensive disregard for the primary caregiver
upon return (Bowlby, 1979). The child demonstrates protest to bring the primary caregiver
near. The despair reaction helps the child avoid physical exhaustion. Detachment allows
for the return to normal activities. Bowlby hypothesized that these innate attachment
behaviors and feelings are genetically determined and promote physical maintenance with
the primary caregiver. Once proximity is achieved, the child can return to exploratory or
feeding behaviors.

The attachment process develops over a two to three year period. At the
completion, the child and primary caregiver form a goal-corrected partnership. At this
point, the child is able to maintain mental representations of the primary caregiver and is
more capable of understanding the caregiver's point of view. The relationship takes on
new goals with the addition of the child's verbal communication and cognitive advances
(Ainsworth et al., 1978; Bowlby, 1979, 1982; Hazan & Shaver, 1994).

Intem_al Working Model. Bowlby's theory not only explains the development of
infants and children but also explains how early attachment experiences affect the behavior
of adolescents and adults (Bowlby, 1988). Bowlby's construct of ”internal working
models” explains how children develop views of themselves and other important

individuals in their lives that operate outside of conscious awareness. The internal working

9
model refers to the beliefs and expectations about whether the primary caretaker is caring,

responsive, and available and whether one perceives him or herself as competent and
lovable. Furthermore, internal working models help the child recognize the absence and
return of the primary attachment figure. These working models, initially developed in the
context of the relationship with the primary caregiver, are subsequently applied to other
relationships in the child's life. Bowlby believed that internal working models of self and
others play an important part in determining and regulating a person's behavior,
expectations, feelings, and perceptions about relationships across the life span.
Ainsworth's Empirical Supmrt

The pioneering research conducted by Ainsworth further supports Bowlby's work
on infant attachment behaviors (Ainsworth, 1967; Ainsworth et al., 1978; Ainsworth,
1982). She and her colleagues observed mother and child interactions within the first year
of the infant's life using a standardized observational research paradigm called the "Strange
Situation." This methodology involved eliciting and monitoring the child's attachment
behaviors and the mother's responsiveness to the child through repeated separations from
and reunions with the mother and through observed interactions with a stranger.
Furthermore, she hoped to activate the child's exploration system by using toys in the
environment. Results of these investigations identified individual differences within these
attachment relationships. Ainsworth identified three distinct styles of mother/child
attachment which included secure, anxious/ambivalent and avoidant.

Patternst Att_achment. The "Strange Situation” studies revealed three distinct
patterns of mother-inf ant attachment behavior. The secure child explored the environment,
experienced distress when separated from mother, maintained contact with her upon
reunion, and appeared readily comforted by her. His or her caregiver was Observed to be
warm, sensitive, consistently available, and responsive. The anxious/ambivalent child
appeared difficult to comfort after separation, distressed prior to separation, and

preoccupied with the mother's availability. His or her caregiver was frequently

10
unpredictable, inconsistent, sometimes unresponsive and intrusive. The avoidant child

focused his or her attention on other objects in their environment, exhibited little distress
upon separation, and presented as unresponsive to mother. His or her caregiver was often
observed to be emotionally unavailable, rejecting, hostile, and rigid. Ainsworth concluded
that the mother's responsiveness to her child's signals, behaviors, and needs during the
first year of life related to differing attachment styles (Ainsworth et al., 1978; Hazan &
Shaver, 1994; Karen, 1994; Shaver & Hazan, 1993).
Continuity of Attachment Styles

Although Bowlby's and Ainsworth's work focused on child development, they
believed that attachment was a life long process. Bowlby (1988) stated that in families
where care giving arrangements were stable, attachment styles developed in childhood, and
once established, tended to persist throughout the lifespan. Furthermore, earlier attachment
patterns are then imposed onto new relationships.

Parkes (1972) and Weiss (1982) were the first to provide support for the continuity
Of attachment into adulthood. Parkes' empirical research demonstrated similarities between
the child's attachment process and the bereavement process in grieving adults. Weiss
argued that adults established relationships in a manner similar to how an infant attaches to
a primary caregiver. He noted, however, differences between infant and adult attachment
dynamics. First, adult attachments develop in reciprocal relationships while child
attachments involve a caretaker relationship. Second, adult attachment is not as easily
overwhelmed as it is in childhood. Adults are able to attend to other relationships and
concerns despite threats to the relationship with the primary attachment figure. Third, the
adult's attachment relationship usually involves a sexual relationship. In conclusion,
Bowlby's and Ainsworth's theoretical assumptions and the empirical work of Parkes and
Weiss provided the conceptual foundation for extending attachment theory to adult
development. Current research and methodological advancements have enhanced

attachment and adult development further.

1 l
Conceptualizationgpd Measurement of Adult Attachment

Three Group Categorization of Adult Attachment Styles

Hazan and Shaver (1987) were the first to examine adult romantic relationships
based on attachment theory. The major goal of their initial study was to apply Ainsworth's
et al. (1978) attachment style categories to adult romantic relationships. A second goal was
to explore Bowlby's idea that one's adult relationship styles consists of mental models of
self and relationship developed from childhood. Finally, they explored the possibility that
differences in early parent-child relationships, as identified by Ainsworth et al's. (1978)
three attachment styles, predicted differences in adult romantic attachment styles. (See also
Shaver & Hazan, 1987; Shaver & Hazan, 1988; Shaver & Hazan, 1993; Shaver, Hazan, &
Bradshaw, 1988). Through this research, Hazan and Shaver developed the first self -report
measure of adult attachment styles in romantic relationships (See Table 1). The measure
translated Ainsworth's et al. (1978) infant attachment style categories (secure,
anxious/ambivalent, and avoidant) into three paragraphs describing adult romantic
relationships and asked respondents to indicate which paragraph was most descriptive of
their adult intimate relationships. Results supported their initial hypotheses. Proportions
of the three attachment categories were stable across the three studies. Participants' self -
classifications were as follows: 56% secure, 20% anxious/ambivalent, and 23-25%
avoidant. The distribution was similar for Ainsworth's attachment style category
proportions (62% secure, 23% avoidant, and 15% anxious/ambivalent). Individuals with
different attachment styles reported different beliefs about romantic love, love experiences,
and descriptions of childhood relationships with parents, self, and others.

Since its development, Hazan and Shaver's self categorization has received much
criticism. The first criticism related to it being a discrete, categorical measure. The second
criticism related to respondents being forced to accept an entire description that may not
accurately reflect themselves and their relationships (Collins et al., 1990). Other

researchers changed the measure into rating scales designed for correlational research.

D.
12 ‘,
Table 1. Hazan and Shaver's Attachment Styl’é Inventory

 

Which of the following best describes how you feel about close relationships?

Secure
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.

Avoidant
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.

Anxious/Ambivalent
I find that others are reluctant to get as close as I would like. I of ten 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 desire sometimes scares people away.

 

13
First, Levy and Davis (1988) added a Likert-type scale to each of Hazan and Shaver's three

attachment style paragraph descriptions asking participants how similar they were or
weren't to each description. Collins and Read (1990), Simpson (1990), and Mikulincer,
Florian and Tolmacz (1990) formed multi-item scales by decomposing Hazan and Shaver's
original paragraph descriptions into a larger set of item statements, factor analyzing the
results, and identifying dimensions underlying the attachment items. Collins and Read
(1990) added additional attachment dimensions to the original descriptions and factor
analyzed them, producing three dimensions, respectively labeled "close”, ”depend" and
"anxiety". A discriminant function analysis suggested the possibility of two types of
anxious participants. Simpson's (1990) results from using factor analysis indicated that
two or three dimensions underlie attachment style differences. Mikulincer et al. (1990)
added a Likert-type scale to 15 items of Hazan and Shaver's (1987) measure for concurrent
validity. From a sample of 80 subjects, only five mismatches resulted from the two
instrument comparisons. This approach attempted to overcome the problem of assuming
that the three adult attachment styles were mutually exclusive and moved the measure
beyond the forced-choice classification.

Shaver and Hazan (1993) concluded from this research that two dimensions
underlie their measure: one dimension reflecting the level of expressed comfort with
interpersonal closeness and dependency; the other tapping the degree of tension or worry

about distance and separateness in these relationships.

Four Group Categorization of Adult Attachment Styles

 

In other research, Bartholomew and Horowitz proposed a f our-category typology
of adult attachment styles (Bartholomew 1990; Bartholomew & Horowitz, 1991). This
model examined attachment style on two dimensions: view of self and view of others.
This is based on Bowlby's internal working models of self and other (perception of

attachment f i gures' availability and their own view of worthiness in receiving support and

14/

love). These dimensions result in four attac ment styles: secure, preoccupied, dismissive,
and fcarful.

Bartholomew (1990) argued that the avoidant group consists of fearful and
dismissing adults, and that adult avoidance developed from parental rejection. She stated
that

Adverse experiences result in negative models of others that are hypothesized to

mediate adult avoidance of relationships. Unlike children, adults differ in their

conscious awareness of unfulfilled attachment needs” (p. 173).

The distinction between dismissive and fearful styles relates to two differing views of self.
The dismissive style includes a positive view of self while minimizing the need for close

relationships. Dismissive individuals view independence as more important than close

relationships. The fearful st le involves a view of self that is undeservin of love
support while avoiding close relationships for fear of re'ection. Fearful individuals turn to

others for validation of their self worth. The dismissive and fearful styles both have a
negative view of others. Bartholomew and Horowitz (1991) provided empirical support
for the four group rationale and developed a single item categorical self -report measure
based on four attachment styles similar to Hazan and Shaver's measure. Griffin and
Bartholomew (1994) examined the validity of the two dimensions (view of self and view of
others) hypothesized to underlie the proposed four attachment styles. Based on five
different measures of assessment, their results indicated that the self and other dimensions
of attachment had construct, discriminant, and predictive validity.

Brennan, Shaver, and Tobey (1991) compared Hazan and Shaver's (1987) three
category model with Bartholomew's (1990) four category model. Results from this study
indicated that the same two dimensions underlie both the three and four category models.
In comparing the three and four category scales, most secure participants classified
themselves as secure on both scales. However, some of Hazan and Shaver's secure

participants classified themselves as dismissive on Bartholomew's scale. Participants

'W
. 15
classifying themselves as anxious-ambitialemoy/H/azan and Shaver's scale distributed
themselves into Bartholomew's fearful and preoccupied categories. Hazan and Shaver's
avoidant participants distributed themselves into Bartholomew's fearful and dismissive
categories. The results indicated that Hazan and Shaver's measure may have forced
participants to misclassif y themselves. The use of Bartholomew's four attachment style
typology provided finer attachment style distinctions based on the two dimensions: view of
self and view of other. These results support the use of the four category typology.
Summm

In their review Of the extant literature, Shaver and Hazan ( 1993) supported the use
of Bartholomew's four attachment styles. Self -report measures for adult attachment styles
remain in the preliminary phases of development. Future self -report measures based on the
four-group taxonomy need to be assessed for reliability and validity.

Individu_al_Differences in Adult Attachment Styles

Shaver and Hazan (1993) support Bartholomew's (1990) newer model Of four

adult attachment styles. Research describes the four attachment style categories as follows:

r Secure Adults

Secure adults are comfortable with intimacy yd autonomy. They adopt a positive
view of self and others (Bartholomew et al., 1991). Secure attachment includes comfort
with closeness and minimal levels of jealousy (Shaver et al., 1993; Mikulincer & Erev,
1991). Secure adults report love experiences that are happy, friendly, trusting, and
supportive. They are able to accept and support their partner despite faults. Their
relationships are believed to be more enduring. This style includes a positive view of self
and other. Secure adults report positive childhood family memories with loving, available,
and responsive parenting (Kobak & Sceery, 1988).

Preoccupied Adths.
Preoccupied adults are overly concerned with their relationships. They have a

negative view of self and a positive view of others (Bartholomew & Horowitz, 1991). The

16
adult with this style exhibits mood lability, partner obsession, and extreme jealousy

(Shaver et al., 1993; Mikulincer et al., 1991). They believe it is easy to fall in love quite
often, however, they rarely find true love. They report that few other people are as willing
to commit themselves to long term relations. This style includes a negative view Of self and
a positive view of others. Preoccupied adults report negative childhood family memories
which include inconsistent parenting, lack of parental supportiveness, and role reversal
(Feeney et al., 1990; Kobak et al., 1988).
If“ Avoidant Adults
Avoidant adults (anxious avoidant, dismissive, and fearful) are fearful of intimacy.
They exhibit difficulty with closeness and reliance on others. The adult with this style
reports the lowest incidence of positive relationships (Shaver et al., 1993). Avoidant adults
often report the experience of important childhood separations from their mothers (Feeney
et al., 1990), poor childhood memory recall, and lack of parental love (Kobak et al.,
1988).
r4“. Dismissive avoidaprt Adults with a dismissive attachment style dismiss the
importance of intimacy. They are frequently counterdependent in their relationships
(Bartholomew et al., 1991; Mikulincer et al., 1991). This style includes a positive view of
self and a negative view of others. Dismissive adults expect a partner to be unavailable and
non-responsive (Bartholomew et al., 1991).
~ Fearful avoidant. Adults with a fearful attachment style are fearful of intimacy and
are more likely to be socially avoidant in their relationships. This style includes a negative
view of self and others. The fearful adult fears attachment and expects his or her partner to
be rejecting (Bartholomew et al., 1991).
Adult Attachment a_r_Ig Relationship Outcome
Hazan and Shaver's (1987) results generated numerous replications and extensions
of attachment theory as a framework for researching adult romantic relationships. Initial

studies incorporated Hazan and Shaver's ( 1987) measure to examine romantic love.

4 {1
Several studies used the original categOrical rn ure to examine group differences in

couples. J

Two studies examined the utility of attachment style as a predictor of adult romantic
relationships with an intemational population. Feeney and Noller (1990) replicated Hazan
and Shaver's (1987) measure with undergraduate students from the University of
Queensland, Australia Results indicated that secure participants exhibited higher levels of
self -esteem than insecure participants, avoidant participants idealized their partners less than
anxious/ambivalent participants, and anxious/ambivalent participants exhibited extreme
forms of love. Mikulincer and Nachshon (1991) administered a Hebrew translation of
Hazan and Shaver's (1987) measure to undergraduate students at Bar-Ilan University in
Israel. Their results suggested that the different attachment groups may have different
working models that guide social behaviors. Secure and anxious/ambivalent participants
exhibited higher levels of self -disclosure. For both studies, the proportions of attachment
style category distributions were comparable to those observed in Hazan and Shaver's
(1987) American samples (5463% secure, 22-31% avoidant, and 15%
anxious/ambivalent).

Several studies used the original Hazan and Shaver (1987) categorical measure to
examine relationship functioning and satisfaction. These self-report studies provided
preliminary support for the applicability of examining adult attachment styles within a
romantic relationship model. Hazan and Shaver (1990) examined the possibility that love

\nd work in adulthood are similar to attachment and exploration of childhood. Securely
K§ attached participants reported a positive and confident work approach. Anxious/ambivalent

@ \ partici ts indicated a fear ' tion for poor work performance and wonies am

rpmamimlationships interfering with work Avoidant participants used work as a way to
v w

 

 

avoid social interactions outside of work Feeney and Noller ( 1991) examined attachment
style differences in verbal descriptions of dating partners. Secure participants emphasized

positive relationship characteristics. Anxious participants exhibited a demanding,

18
overinvolved relationship style in which they idealized their partner. Avoidant participants

exhibited low levels of emotional intensity and greater relationship distance. Pistole (1989;
1991) conducted two studies on adult attachment style and relationship functioning. The
first study revealed that securely attached participants reported higher relationship
satisfaction and exhibited mutually focused conflict strategies. Anxious/ambivalent
participants complied with their partner's wishes more than did avoidant participants. The
second study revealed that secure and anxious respondents reported higher levels of
comfort with self -disclosure than did avoidant respondents.

Several early studies modified or changed the original Hazan and Shaver measure to
assess attachment style differences. Simpson's (1990) correlational results indicated that
secure participants reported greater relationship satisfaction, relationship interdependence,
commitment, trust, and positive emotions. In a six month follow-up, avoidant participants
reported the least emotional distress following relationship break-up. Mikulincer and Erev
(1991) reported that secure participants achieved relationship intimacy and had partners
who shared similar relationship views. Anxious participants valued passionate love,
however, they failed to experience secure love. Avoidant participants desired intimacy less
and perceived less intimacy and commitment in their partners. Feeney and Noller (1992)
examined the relationship between attachment style and the process of relationship
dissolution. Results indicated that avoidant adults reported the highest levels Of
relationship dissolution and least distress over these losses. Anxious-ambivalent adults
reported the highest levels of surprise and upset over the relationship dissolution and
exhibited the highest levels of involvement in a new relationship.

Summary

Early studies of attachment style differences were replicated with international
populations (Feeney & Noller, 1990; Mikulincer & Nachshon, 1991). Several studies
used the categorical Hazan and Shaver (1987) original attachment style measure to assess

group differences. Group differences were observed with regard to attitudes towards

19
friends and family, conflict resolution patterns, emotional regulation, self-disclosure

patterns, trust level, and work styles (Hazan & Shaver, 1990; Feeney & Noller, 1991;
Pistole, 1989; Pistole, 1991). Several subsequent studies addressed the initial criticisms of
the Hazan and Shaver's (1987) categorical measure by developing and employing
continuous measures of adult attachment styles. These investigations also observed
attachment style differences in beliefs about human nature, commitment, emotional
experiences, expressiveness, instrumentality, perception of others, relationship dissolution,
relationship satisfaction, self -esteem, styles of loving, and trust (Collins & Read, 1990;
Feeney & Noller, 1992; Mikulincer & Erev, 1991; Simpson, 1990). The early studies
suggested that attachment style plays a role in couple interaction and functioning. Much of
this work used the three- group instead of the f our- group categorization of adult attachment
styles. As both theory and subsequent research indicate, the four group categorization
provides a finer delineation of adult attachment styles. In addition, the early studies used
primarily correlational designs which precluded cause-effect interpretations.
Adult Attachment and Relationship Process

More recently, several studies have examined links between adult attachment styles
and the relationship processes with more sophisticated instruments and research designs.
Some of these studies have also employed participant pools other than undergraduate dating
couples. Kobak and Hazan (1991) used the original Hazan and Shaver (1987) measure
and videotaped married couple interactions. The study examined if partner's attachment
security was related to emotion regulation during problem solving and confiding behaviors.
Results indicated that insecure partners exhibited negative affect and secure partners
exhibited greater constructive management of emotions and marital adjustment.

Simpson, Rholes, and Nelligan (1992) employed an experimental design and a
continuous measure of adult attachment orientation to investigate whether attachment style
moderated the relationship of anxiety to spontaneous behavior (support-seeking and

support giving) in dating couples. The researchers unobtrusively videotaped undergraduate

20
dating couples while the female member waited her turn to participate in an anxiety

provoking activity. Secure women sought support from their partner as their anxiety
increased while avoidant women distanced from their partners. At lower levels of anxiety,
more avoidant women sought support from their partners than secure women. Secure men
offered more support than avoidant men. These results are, however, only applicable to
female participants with a secure or avoidant attachment styles due to sampling limitations.

Rothbard, Roberts, Leonard, and Eiden ( 1993) used Bartholomew's four
attachment style categories. They used the Relationship Scales Questionnaire (RSQ) which
is a 30-item measure developed from Hazan and Shaver's (1987) categorical descriptions,
Bartholomew and Horowitz's (1991) categorical descriptions, and Collins and Read's
(1990) dimensions of close, depend, and anxiety. Married couples completed the RSQ,
and then participated in a 15 minute videotaped naturalistic discussion of a couple-selected
current, unresolved marital problem. Results demonstrated that husband and wife
attachment style differences were related to actual couple behaviors during a problem
solving discussion. The attachment style measure used in this study (RSQ) included two
attachment dimensions which accounted for significant amounts of variance in the couple's
problem solving behaviors. This Observed relationship persisted even after marital
satisfaction and marital aggression were statistically controlled. Results suggested that
avoidant husbands exhibited hostility and detachment during the conflict discussion and
their wives exhibited defensiveness. Secure husbands exhibited more reflective listening
and their wives exhibited less defensiveness. Results suggested that husbands' attachment
style and problem solving behavior style may have influenced their wives' responsivity.
Specifically, fearful husbands' wives exhibited low levels of disagreement and validation
of their partner's position. Preoccupied husbands' wives exhibited the highest levels of
disagreement. Dismissive husbands' wives exhibited high levels of validation.

Pistole, Clark, and Tubbs (1995) examined the relationship of adult attachment

styles and relationship investment. Based on Hazan and Shaver's (1987) three group

21
attachment style model, attachment style differences were expected regarding relationship

satisfaction, rewards, costs, investment, alternatives, and commitment. As hypothesized,
securely attached individuals reported greater relationship satisfaction, fewer costs, and
greater commitment than the insecure groups. The avoidantly attached individual reported
the lowest level of relationship investment while the anxious-ambivalently attached
individual reported the highest relationship cost (negative relationship behavior). Based on
the results, problematic relationships may be more evident for the insecure individual due to
decreased relationship commitment, investment and negative relationship perceptions.
Summag

Later studies began addressing the criticisms related to the original Hazan and
Shaver (1987) measure by using continuous measures to assess attachment styles. Several
used experimental designs and naturalistic observation. However, only one study used the
Bartholomew and Horowitz (1991) four category typology. Results indicated that
attachment style played a role in dating and married couple's functioning and interactions.
Secure participants report more enduring, satisfying, and less problematic relationships,
greater levels of commitment, and more effective affect regulation; avoidant participants
report less intense, more distant love relationships with lower levels of relationship
investment, and exhibited more hostility and detachment during conflict discussions; and
anxious participants report more frequent and less satisfying love relationships.

Attachment Style, Sex, and Gender Role

The literature on adult attachment style reveals inconsistent results regarding sex

 

and gender role differences. Several recent studies suggest that sex may moderate the
influence of attachment style on relationship functioning. Collins and Read (1990)
indicated that both partners were less satisfied with the relationship when the male was
avoidant and when the female was anxious/ambivalent. Simpson (1990) reported sex
differences in a longitudinal investigation of dating couples. Avoidant males reported less

emotional distress following a relationship termination. Males reported a decreased level of

22
satisfaction when the female exhibited an anxious attachment style. Senchak and Leonard

(1992) reported sex and attachment style differences in newlywed problem solving
behavior. Insecure husbands in couple pairings with a secure wife exhibited a low
frequency of problem solving behaviors. Anxious husbands reported shorter premarital
relationships.

Cohn, Silver, Cowan, Cowan, and Pearson (1992) examined attachment
differences in marital satisfaction and observed couple interaction. Post hoc findings that
examined couples' joint attachment classifications, including insecure-insecure, insecure-
secure, and secure-secure dyads, suggested that insecure women married to secure men
reported more positive relationships than insecure women married to insecure men. These
authors suggested that the husband's early attachment experiences had a greater influence
on couple functioning than did the wife's early attachment experiences. They raised the
possibility that the husband's secure attachment style may help transform an insecure
wife's attachment style to a secure model. These results should be cautiously interpreted
due to the low number of insecure men and an overall low sample size.

Scharf e and Bartholomew (1995) examined the relationship between individual
attachment style differences based on Bartholomew's four category model and the use of
accommodation strategies among young couples. Accommodation strategies were defined
as constructive responses to potentially destructive behavior exhibited by a romantic
partner. Results indicated that a secure attachment was positively associated with the use of
constructive strategies which included active discussion of the problem for men and passive
waiting for an improvement in the partner's behavior for women. A secure attachment was
negatively associated with destructive strategies which, for both men and women, included
leaving or threatening to leave the relationship, for men only, also included ignoring the
problem or partner. These results suggested that a secure attachment style may lead
individuals to respond to threatening partner behavior with constructive, accommodating

behaviors. Conversely, f earf ulness was positively associated with destructive responses of

23
ignoring the problem and leaving or threatening to leave the relationship for men. Fearful

men also exhibited a negative association with discussing the problem. Associations for
preoccupied and dismissive were less consistent. Finally, there was minimal support that
the partner's attachment style predicted one's use of accommodation independent of
individual attachment style. However, the partners of fearful and preoccupied males
discussed problems less of ten than other attachment styles. Partners of anxious females
threatened to leave or terminated the relationship more of ten than did partners of individuals
with other attachment styles. Finally, male and female anxiety related to an increase in
destructive relationship behaviors for their partners.

A few studies assessed attachment styles and sex differences with sex as one of the
independent variables. In these studies, sex and not gender role was considered. Feeney,
Noller, and Patty (1993) examined attachment style and sex differences among
undergraduate dating couples. In comparing male and female participants from all three
attachment style groups, female avoidants and male anxious/ambivalents (preoccupied)
were the least likely to report involvement in sexual intercourse. This finding supports the
joint influence of attachment style and sex on relationship functioning.

Fiala and Pietromonaco (1991) examined how internal working models of others
influenced how individuals perceive and respond in romantic relationships. They asked
secure, preoccupied, and avoidant men and women to read about and imagine a relationship
with a partner displaying one of three attachment style behaviors. This was the first study
that directly examined the role of sex as a moderating variable. Individuals conforming to
stereotypic gender roles (i.e. avoidant men and preoccupied women) experienced the most
negative feelings about themselves. And, stereotypic men (avoidant) saw themselves in a
future relationship with a stereotypic female (preoccupied). Avoidant females did not see
themselves in a relationship with a preoccupied male. A preoccupied female fits the
prescribed gender role for women, however, the same behavior in a male does not f it the

prescribed gender role and hence may have been unacceptable to these females. Avoidant

24
women were less positive about a future relationship with an avoidant partner than were

secure or preoccupied woman. Results suggested that the relationship between internal
working models of others and emotional reactions and perceptions about imagined
relationships are moderated by sex and partner's attachment behavior.

Kirkpatrick and Davis (1994) examined male/female attachment style pairings, the
role of each partner's attachment style on relationship satisfaction, commitment, and
conflict and the role that attachment style plays in predicting of relationship break-up.
Males and females rated the relationship negatively when the female was anxious. When
the male was avoidant, the male partner rated the relationship negatively. Avoidant men
reported a stable relationship at the time of the first follow-up despite the avoidant males'
negative rating Of the relationship. Anxious women reported stable relationships at the time
of the second follow-up despite their negative relationships.

Summgy

Based on the Hazan and Shaver (1987) measure, the distribution of attachment
styles appears to be independent of sex (Brennan, Shaver & Tobey, 1991; Hazan &
Shaver, 1987; Levy & Davis, 1988). Only one study using Bartholomew's four category
typology of adult attachment style demonstrated sex differences (Brennan, Shaver &
Tobey, 1991). In this study, there were proportionally more males in the dismissive
category and more females in the fearful category. Kunce and Shaver (1994) subsequently
compared Hazan and Shaver's three category model and Bartholomew and Horowitz's
(1991) four category model of adult attachment styles. These investigators reported no sex
differences in the self reports of attachment style using either measure.

Still, evidence exists that respondents' sex may moderate how attachment style
affect relationship perceptions and behaviors. Based on the reviewed literature of sex
differences and attachment styles, both partners were less satisfied when the male had an
avoidant attachment style and the female had an anxious attachment style (Collins & Read,

1990; Fiala & Pietromonaco, 1991; Kirkpatrick & Davis, 1994; Simpson, 1990).

25
Avoidant males reported less emotional distress following relationship termination

(Simpson, 1990) and a greater likelihood that they would marry an anxious female (Fiala &
Pietromonaco, 1991). Avoidant women retracted from emotional support when anxious
(Simpson, Rholes, & Nelligan, 1992) and reported that they were less likely to participate
in sexual intercourse (Feeney, Noller, & Patty, 1993). Anxious men reported the lowest
amount of relationship stability (Kirkpatrick & Davis, 1994); shorter premarital relationship
lengths (Senchak & Leonard, 1992); and that they were less likely to participate in sexual
intercourse (Feeney, Noller, & Patty, 1993). Anxious women reported the highest levels
of relationship stability. Insecure men paired with secure wives exhibited the lowest
frequencies of problem solving behaviors (Senchak & Leonard, 1992). Insecure women
married to secure men reported more positive relationships than insecure women married to
insecure men (Cohn et al., 1992). Secure women used their partner for support when
anxious and secure men offered more support to their partners than avoidant men (Simpson
et al., 1992). Stereotypic gender role behavior exhibited in avoidant males and anxious
females resulted in negative relationship experiences for both partners. Secure attachment
styles in males and females generated the highest levels of relationship satisfaction.

In conclusion, preliminary results suggest that sex and/or gender role may moderate
the influence of adult attachment styles on relationship functioning. Stereotypical gender
role behavior appears to present itself in certain attachment styles (Kirkpatrick &

Davis, 1994; Feeney, 1994). This is evidenced in stereotypical avoidant men and
stereotypical anxious women. For example, avoidant men are prone to distancing behavior
in relationships which is considered to be a stereotypically masculine behavior. Anxious
females are prone to being preoccupied with their relationships which is considered to be a
stereotypically feminine behavior (Fiala & Pietromonaco, 1991; Pistole, 1994).

Several criticisms are noted with the current studies that examined sex and gender
role. Few of the studies used Bartholomew's four attachment style typology. Several

studies were not able to compare the three groups adequately due to small sample sizes for

26
one of the three attachment style groups. Finally, most studies examined sex and not

gender role. What remains unclear is whether it is sex or gender role that moderates the

attachment style influence on relationship functioning. Sex may not moderate attachment

style effects on relationship functioning when gender role differences are controlled.
_A_t;_achment Style aid Problem Solving

Several studies examined participants' problem solving as a dependent variable
when examining adult attachment style differences in romantic couples. Levy and Davis
(1988) reported that the secure style was positively correlated with constructive approaches
such as compromising and integrating conflict styles. The anxious/ambivalent style was
positively correlated with a dominating style. The avoidant style was negatively correlated
with compromising and integrating conflict styles. Pistole (1989) reported that secure
participants used a strategy that was mutually focused while anxious/ambivalent
participants reported a strategy that focused on the other's perspective. Secure participants
applied constructive conflict resolution styles and anxious ambivalent and avoidant adults
applied negative approaches to problem solving.

Two studies asked couples to identify their own problems to discuss with their
partner. Kobak and Hazan (1991) videotaped marital couples discussing a problem that the
couple selected. Husbands with positive expectations of their wives' availability exhibited
less rejection and more support. More rejection was exhibited when wives described their
husband as less psychologically available and when they relied less on their husbands.
Rothbard, Roberts, Leonard, and Eiden (1991) videotaped couples engaging in discussion
of a current marital conflict. These authors compared couples who differed with regard to
husbands' aggressive status (aggressive and nonaggressive). Attachment style dimensions
accounted for significant variance in problem solving behavior after marital satisfaction and
marital aggression were controlled. Avoidant husbands and wives exhibited an increase of
attack and a decrease of reflective listening and attentiveness during these discussions.

Secure husbands scored the highest on reflective listening whereas their wives scored the

27
lowest on defensive behavior. Fearful avoidant husbands' wives scored the lowest on

disagree/disapprove and validation of partner's position. Preoccupied husbands' wives
scored the highest on disagree/disapprove. Dismissive husbands' wives scored highest on
validation of partner's position. Husbands' and wives' attachment styles were
independently related to actual behaviors during a problem solving interaction.

Summgy

The literature indicates that significant, meaningful differences in problem solving
behavior are associated with adult attachment style differences in dating and married
couples. Three of the four studies were based on Hazan and Shaver's three category
attachment style typology. Results from those studies indicated that secure attachment
styles were related to more constructive and effective problem solving behaviors and that
insecure styles exhibited less effective problem solving behavior and greater conflict.
When using an alternate attachment style measure based on the four category attachment
style typology, similar results were obtained. One observed difference involved the
interaction between the dismissive husbands' wives' behavior. Within these pairings, the
wife was more validating of her partner's position. This may have been a positive problem
solving behavior or it may have been a way to increase the husband's involvement.

The studies relied on self-report of conflict styles, recalling or imagining a problem,
or identifying a problem to discuss. These behaviors may provide preliminary information
on problem solving behavior; however, they may not reflect actual problem-solving
behavior. Recalling or imagining about a problem is different from experiencing it.
Furthermore, identifying a problem may result in couples selecting a problem that is ”safe"
and possibly rehearsed. Current studies have neglected direct observation of spontaneous
couple problem-solving communication.

Mment Sgle and Conjoint Interaction
The conjoint contribution of each partner's attachment style to relationship

functioning has been largely overlooked. More specifically, studies have generally not

28
examined how the internal working models of others influence the emotional reactions,

relationship perceptions, and subsequent behavior of the relationship partner. Several
studies briefly examined attachment style dyadic matching. For example, Collins and Read
(1990) reported stable dating relationships for secure-secure partner matching. Simpson
(1990) reported that anxious men were involved with insecure females and avoidant men
were paired with anxious women.

Fiala and Pietromonaco's (1991) research demonstrated dyadic attachment style
differences in mental representations when participants imagined a relationship with a
secure, anxious, or avoidant partner. When participants, regardless of attachment style,
imagined a relationship with a secure partner, they reported more positive feelings than
when they imagined a relationship with a preoccupied or avoidant partner. Participants
reported the most negative feelings after imagining a relationship with an avoidant partner.
In general, secure participants were more positive about their future relationships. All
participants reported greater perceived conflict when imagining a relationship with a
preoccupied or avoidant partner. Insecure participants responded more favorably when
imagining they were paired with a dissimilar insecure partner (i.e. avoidant participant was
more favorable about a preoccupied partner than an avoidant partner). The results provided
evidence that participants of different attachments styles reported different emotional
reactions when imagining a romantic relationship.

Several studies examined differences in couple interactions in married and dating
couples with similar and dissimilar attachment styles. Cohn, Silver, Cowan, Cowan, and
Pearson (1992) examined attachment differences in marital satisfaction and observed couple
interaction. Insecure females married to secure men reported more positive and less
conflictual relationships. Insecure-insecure dyads exhibited more anger and conflict than
the other dyad combinations and were perceived as less well functioning than were either
the secure-secure or the secure-insecure couple. The results suggested that when at least

one of the partner's has a secure attachment style, less negative couple interaction is

29
reported. A secure partner may buffer negative effects of a mate's insecure attachment

style.

Senchak and Leonard (1992) examined attachment style differences and
interpersonal functioning in newlywed couples. Results indicated that partners tended to
marry based on attachment similarity. However, both secure and insecure partners were
more likely to marry a secure partner. This may have related to the greater sample number
of secure participants. Findings suggested that secure-secure couples reported better
overall marital adj ustrnent than insecure couples or mixed couples. Insecure attachment
style similarity did not lead to better marital adjustment. Insecure couples reported more
negative partner conflict resolution behavior than did secure couples. Couples in which the
husband reported an ambivalent attachment style had shorter premarital relationships than
did couples in which the husband had a secure or avoidant attachment style. Reported sex
differences indicated that wives reported more intimacy and evaluated their husbands more
favorably regardless of couple type. Results suggested that it is the nature of the pairing
instcad of attachment style pairing similarity that related to couple functioning. Couples in
which both partners were secure reported better overall marital adjustment than insecure
couples or mixed couples.

Kirkpatrick and Davis's (1994) large scale survey of undergraduate dating couples
indicated an over-representation of secure couples and an absence of avoidant-avoidant and
anxious-anxious couples. Their results need to be interpreted with caution since there was
an underrepresentation of anxious and avoidant adults. There was no evidence of any
interactions between the individual partner's attachment style in predicting the relationship
ratings. In couples with an anxious female, both partners rated the relationship negatively
on satisfaction, viability, and conflict ambivalence. In couples with an avoidant male, only
men rated the relationship negatively in all relationship dimensions. The female partners of
the avoidant males, however, reported greater passion and less conflict than those females

paired with an anxious male. These ratings were given regardless of the female partner's

3O
attachment style. In examining the dating relationship at Time 2, the relationships that

received the highest negative ratings at Time 1 (avoidant men and anxious women)
described their relationships as stable as the Time 1 secure subjects. Anxious men and
avoidant women reported the highest break-up rates over time. The results indicated that
relationship rating differences related to attachment style pairings, stage of the relationship,
and sex.

Feeney (1994) examined the relationship between attachment style and relationship
satisfaction across the marital life cycle. Her study examined the influence of each
individual partner's attachment style and the conjoint interaction of participants' attachment
styles. In addition, she assessed whether communication variables (mutuality, coercion,
destructive process, and post conflict distress) mediated the relationship between
attachment and relationship satisfaction. Secure persons (high in comfort with closeness
and low in anxiety) were more often paired with secure partners. The secure-secure couple
pairing reported higher levels of satisfaction for both husbands and wives. Satisfaction
was negatively correlated with anxiety for both spouses. This effect was primarily noted
when the husband was uncomfortable with closeness and the wife reported increased
anxiety. Examination of attachment style pairings indicated that a problematic combination
in marriages of short duration were couples wherein the husband reported low scores with
comfort and the wife reported high anxiety. Among wives, the relationship between
attachment and relationship satisfaction was largely mediated by communication patterns,
yet among husbands only partially so. The communication pattern of mutuality (open and
constructive communication and lack of avoidance and withdrawal) was the strongest
correlate of relationship satisfaction.

Feeney, Noller, and Callan (1994) used a longitudinal design to examine concurrent
relations between participants' and their partners' attachment style and communication
patterns and marital satisfaction. Couples completed interaction diaries for a two year

period. Attachment style was based on self report measures of comfort with closeness and

3 1
anxiety over abandonment. Husbands scoring higher on comfort with closeness

demonstrated more constructive patterns of communication. These couples reported higher
levels of recognition, involvement, disclosure, and satisfaction with their communication.
Husbands scoring higher on anxiety reported lower levels of self -disclosure in their
marriages. Wives scoring higher on the anxiety scale exhibited decreased couple
involvement and satisfaction with communication and increased marital conflict and
domination of the partner. For both husbands and wives, anxiety related to conflict
patterns of coercion, withdrawal, feelings of guilt and hurt, and lack of mutual expression
and understanding. Anxiety about attachment related to negative and destructive patterns of
communication. Relationship satisfaction and anxiety were negatively correlated.
Husbands and wives both reported that husbands' discomfort with closeness and wives'
anxiety over abandonment were associated with negative relationship outcomes. The
results demonstrated differences in communication patterns and relationship satisfaction
based on dyadic attachment style differences among couples in established relationships.
Summa_ry

In examining types of dyadic attachment style pairings, the literature has indicated a
predominance of secure-secure pairings and a relative absence of anxious-anxious and
avoidant—avoidant couples. Reasons for the absence of similar insecure pairings may relate
to a violation of expectations. The avoidant individual expects his or her partner to be
clingy and demanding while the anxious individual expects their partner to be distant and
withdrawn. Therefore, to meet these expectations, they choose partners with dissimilar
insecure attachment styles. The interlocking expectations of avoidant and anxious
individuals are in line with consistent findings that avoidant individuals tend to couple with
anxious partners instead of with same attachment style partners (Collins & Read, 1990;
Kirkpatrick & Davis, 1994; Simpson, 1990). The avoidant-anxious attachment pairing
may also be perceived as more acceptable since it is intertwined with the culturally

stereotypical behavioral patterns for men and women.

32
Overall, in examining both partners‘ attachment styles with the dyadic relationship,

insecure-insecure dyads reported the highest levels of anger and conflict as compared to the
secure-secure dyads. More specifically, avoidant men and preoccupied women reported
the lowest relationship satisfaction and highest level of conflict and ambivalence (Fiala &
Pietromonaco, 1991). Secure-secure partners reported the highest levels of stability and
relationship adjustment as compared to insecure couples or mixed (i.e., secure-insecure)
couples. Most studies, however, overlooked the conjoint contribution that individual
attachment style may play in relationship functioning. In the studies that did examine the
conjoint attachment, most used Hazan and Shaver's (1987) three category typology or
Collins and Read's (1990) three attachment style dimensions to measure attachment style.
Both Feeney (1994) and Feeney et al. (1994) used two factor-analytically—derived scales to
measure attachment, comfort with closeness and anxiety over relationships. Preliminary
results provided initial support for the role that conjoint attachment style pairings play in
couple functioning. The use of the three category attachment style model instead of the
four category model, small sample sizes for the anxious and avoidant categories, and
minimal emphasis on the interaction of the couple's attachment styles limit the conclusions
that can be drawn from these studies.
Sex/Gender Rolgaad Couple Communicatigapd Functioning

Attachment style appears to affect the type of problem solving style an individual
exhibits. Previous studies have found that sex and/or gender role appears to moderate the
relationship of attachment style and dyadic behavior. This study addresses a gap in the
literature which will hopefully clarify relationships among all four variables: attachment
style, sex, gender role, and problem solving communication. One final area of research
will be presented that provides support for this line of inquiry-- the relationship between
sex/ gender role and problem solving communication in couples.

There is confusion and ambiguity in the literature about the labeling and meaning of

the terms sex and gender role. The research in the area of male and female differences used

33
biological sex and societal gender roles interchangeably. Sex is the biological term of being

born male or female. Gender role refers to the psychological phenomenon of behaviors,
expectations and roles of the societal definitions of masculinity and femininity (Mintz &
O'Neil, 1990; Winstead & Derlega, 1993). During the past two decades, research on sex
and gender role differences has increased. In the studies that used sex of the participant as
the independent variable, fewer differences were found than expected. The amount of
variance accounted for by sex was often small. When differences related to sex were
found, the characteristic of the specific task was found to be a critical factor in eliciting or
suppressing sex differences (Deaux, 1984; Maccoby, 1990). Maccoby (1990) reported
that the research on sex differences has become more methodologically sophisticated over
the last 15 years. Despite the changes in research design and analyses, similar conclusions
regarding sex differences remain constant. There were still moderate sex differences in
performance on tests of mathematical and spatial abilities, however, previously reported
sex differences in verbal abilities had decreased when the sex differences were analyzed
with more sophisticated methodologies and research designs . Overall, the results indicated
that sex differences in personality traits did not differ.

With the limitations of sex as an independent variable, subsequent research began to
examine gender role differences (Bern, 1974). Gender role was assessed as a continuous
variable ranging from high masculinity to high femininity. Research supported that
masculinity and femininity predicted differences in instrumental and expressive domains.
And, research supported that gender role stereotypes exist and relate to judgment and
evaluations of males and females.

Whereas research in the area of male and female sex differences is beyond the scope
of this paper, sex differences on intellectual performance and personality traits are rarely
found (Maccoby, 1990). When male and female differences are reported, these differences
are influenced by the cultural stereotypes of gender role and their resulting discriminatory

behavior. Despite the findings of gender role differences, the research remains fragmented.

34
Further research needs to clarify the role of gender role differences by examining the

influence of individual choices that males and females make, and how these choices are
influenced by the stereotypes and discriminatory behavior that may affect the decision-
making process (encouraging some choices and discouraging others) (Deaux, 1984).
Based on the current research, the present study will test the hypotheses that male and
female differences in the problem solving communication patterns of dating couples are
related to gender role and not sex.

The psychological literature has noted and described the different socialization
processes impacting men and women (Chodorow, 1978; Gilligan, 1982). Men are
socialized to be emotionally inhibited, assertive, powerful, independent and to compare
sexuality with intimacy, manliness, and self esteem. Women are socialized to be
emotional, nurturing and to direct their sense of worth and achievement through affiliation
with others (Kaplan, 1979; Gilbert, 1987). These different processes in socialization may
create male and female differences in couple interaction and functioning. Recent theories Of
women's development suggest that these differences relate to the themes of separateness
(focus on independence in relationships) and connectedness (focus on emotional intimacy
in relationships) (Chodorow, 1978; Gilligan, 1982). Women's developmental theory
proposes that men define themselves through themes of separateness whereas women
define themselves through themes of connectedness in relationships (Gilligan, 1982).
These different self definitions result in different relationship expectations and behaviors
for men and women. Lang-Takac and Osterweil (1992) found empirical support for this
theory indicating that men are more emotionally differentiated from others and independent
whereas women are more empathic and desired higher intimacy. From these results, one
can speculate that males may be more attuned to individual goals whereas women may be
more attuned to relationship maintenance. This gender role difference may explain

differences noted in relationship process variables for men and women.

35

Several studies have examined the role of sex, gender role and problem solving
communication patterns, conflict resolution, and decision making styles. Sayers and
Baucom (1991) examined the role of gender-stereotypic behavior in distressed couple's
communication. Higher levels of femininity among women in distressed marriages
correlated with greater negativity in the marital interactions. Femininity levels in males
were less related to communication patterns; however, men's femininity was related to a
tendency to terminate fewer negative interactions in comparison with the females.
Masculinity was not predictive of communication pattern differences between men and
women. The results were interpreted as suggesting that spouses in distressed marriages
with higher levels of femininity may take more responsibility for ensuring that conflictual
issues are addressed in the relationship. This supports socialization theory assumptions
that women may focus more on maintaining connectedness in the relationship as
demonstrated by engaging in the discussion of negative, emotionally laden topics during
problem solving discussions.

White (1989) examined conflict-resolution patterns in couples. In marriages where
either the male or female partner expressed marital dissatisfaction, men were more likely to
exhibit a coercive style of conflict resolution. Women exhibited an affiliative style of
interaction when either they or their spouses were unhappy. Males and females responded
differently to marital dissatisfaction in that males sought distance and females sought
connection. Heavey, Layne, and Christensen (1993) reported that women were more
demanding and men were more withdrawing during conflictual discussion regarding
wive's and not husband's issues. Furthermore, when couples exhibited the gender-
stereotyped roles of females demanding and males withdrawing, the females reported
higher relationship dissatisfaction. Rusbult, Johnson, and Morrow (1986) reported that
women demonstrated engaging and loyalty (waiting for conditions to improve) behaviors

whereas men tended to not engage in interpersonal issues during problem solving.

36
Lamke, Sollie, Durbin, and Fitzpatrick (1994) examined the relationships among

masculinity, femininity, and relationship satisfaction in dating couples. Individuals who
identified with feminine traits described themselves as being comfortable in engaging in
emotionally supportive behaviors in intimate relationships (expressive competence).
Individuals who identified with masculine traits described themselves as being comfortable
in initiating behaviors that involved discussing negative thoughts and feelings (instrumental
competence). For both males and females, relationship satisfaction related to identifying
oneself as capable of expressive competence and perceiving one's partner as feminine.
This study provides support for the view that femininity indirectly influences relationship
satisfaction and expressive competence (stereotypically feminine traits) and that perceptions
of one's partner as feminine directly influences dating relationship satisfaction.

A final article indirectly relevant to the discussion of sex/ gender role and problem
solving, and directly relevant to this study because of its use of the same problem-solving
communication task is Grotevant and Cooper's (1985) study of the concepts of
individuality and connectedness within family relationships. Individuality was defined as
verbal behavior that reflects separateness and self -assertion, and operationalized as in
verbal expressions that distinguished self from others and that clarified one's own
viewpoint. Connectedness was defined as verbal behavior that reflected mutuality and
permeability, and operationalized as verbal expressions of sensitivity and respect for the
other's viewpoint and of openness and responsiveness to these viewpoints. Their study
explored how these concepts were demonstrated through the f amily's communication
process. Their findings suggested that both individuality and connectedness are important
predictors of individual problem-solving competence. Despite the sex differences noted in
these variables, additional research suggested that healthy relationships need a balance
between individuality and connectedness (Feeney & Noller, 1991; Grotevant & Cooper,
1985; Markrnan, Silvem, Clements, & Kraft-Hanak, 1993). Current research indicated

that individuals with different attachment styles tended to exhibit tendencies toward extreme

37
individuality or extreme connectedness (Feeney & Noller, 1991; Hazan & Shaver, 1987;

Pistole, 1994). Since a balance of the two concepts is desired, an absence of this balance
in relationships may explain problems in relationship functioning that the research has
uncovered.

In conclusion, the research reviewed in this section provides support for the view
that sex and/or gender role influences communication patterns, conflict resolution, and
problem solving behavior. Lamke et al. (1994) demonstrated that gender role influenced
relationship satisfaction in dating couples. In addition, the research on sex, gender role and
problem solving patterns and style supports the view that men seek distance during conflict
resolution and problem solving whereas women attempt to maintain communication and
attend to the relationship (Heavey, Layne, & Christenson, 1993; Sayers & Baucom, 1991;
White, 1989). Although the research reviewed for this section indicated that male and
female differences in self definitions, communication patterns, conflict resolution style,
problem solving patterns, and decision making styles affect relationship behavior, the
relative contributions of sex versus gender role differences were not clarified. Other
research suggests that despite sex differences in certain task specific areas, sex differences
in relationship functioning are actually related to stereotypical gender role behaviors (i.e.,
masculinity and femininity) of couple participants. This study attempted to clarify the
relationship between sex and gender role and relationship functioning, and specifically
problem solving behavior.

X A final point is that the research on couple functioning highlighted that W
focus on inWrs (separateness) within the relationship whereas women tend

X.
to focus on relationship maintenance (connectednessfiflhis different focus within the

 

relationship creates problems with closeness-distance regulation in couple communication
and interactions. Pistole (1994) proposed that attachment theory may be one way to gain
further understanding of this behavior. Individuals differ in their desires for closeness and

distance in relationships. This difference is exaggerated further during times of increased

38
anxiety. Pistole suggested that these differences are demonstrated in the attachment style

categories. The individual with an avoidant attachment style (fearful and dismissive) seeks
distance during times Of stress whereas an individual with a preoccupied attachment style
seeks increased closeness. This suggests that sex and gender role differences may only be
one part of the explanation for differences in couple communication, functioning, and
relationship satisfaction.

Summm and Limitations

Attachment theory provides a new framework within which to examine dating and
marital couple functioning and interaction. Hazan and Shaver (1987) provided theoretical
and empirical support for attachment style differences in adult intimate relationships.
Bartholomew and Horowitz (1991) provided theoretical and empirical support for the
existence of four adult attachment styles instead of the three styles that were initially
proposed. Their four group model includes secure, preoccupied, dismissive, and fearful
styles.

Numerous studies have replicated and extended the initial work on individual
attachment style differences in romantic relationships. Preliminary studies suggested that
attachment style plays a role in several areas Of dating and marital couples interaction and
functioning. later studies used improved adult attachment measures and more
sophisticated research designs. The studies continued to be problematic since they did not
use Bartholomew's four category typology. Only recently have researchers implemented in
vivo methods to assess attachment style and couple functioning.

The literature has yielded inconsistent results regarding the relationship of sex and
gender role and attachment style to couple interactions. Few studies have specifically
examined and controlled for sex or gender role influences. Stereotypical gender role
behaviors are present in the male dismissive adult and the female preoccupied adult. Future
research needs to control for sex and gender role while examining the affects of attachment

style differences on relationship processes and outcomes.

39

Few studies have examined the conjoint contribution of each partner's attachment
style to romantic couple functioning. Future research needs to clarify the conjoint effects of
attachment styles in couple pairings on relationship behaviors.

A study that concurrently examines participants' sex, gender role, and attachment
style will expand our overall understanding of the individual and conjoint roles that each
variable plays in the role of couple functioning. Examining couples' conjoint attachment
style contributions and clarifying the role of sex and gender role differences as moderators
of couple communication patterns will extend and clarify and enhance the current research
on adult attachment style differences in romantic couples.

Definitions
Adult A4ttachment Styles

In this study, Shaver and Hazan's (1993) and Bartholomew's (1990) adult
attachment styles categories are used and defined as follows:

Secure adults are comfortable with intimacy and autonomy. They adopt a positive
view of self and others. Secure attachment includes comfort with both closeness and
separateness. Secure adults report positive childhood family memories with loving and
available parents.

Preoccupied adults (anxious/ambivalent, anxious resistant) are preoccupied with
relationships. They have a negative view of self and a positive view of others.
Preoccupied attachment includes Obsession, excessive dependency, and extreme jealousy.
Preoccupied adults report negative childhood family memories which include inconsistent
parenting, lack of parental support, and role reversal.

Avoidant adults (anxious avoidant, dismissive, and fearful) are fearful of intimacy.
They exhibit difficulty with closeness and reliance on others. Avoidant adults report
childhood separation from their mother, poor childhood memory recall, and lack of parental
love. This study divides the avoidant into two distinct categories of dismissive and fearful
attachment styles. Dismissive avoidant adults dismiss intimacy and they are
counterdependent in their relationships. This style includes a positive view of self and a
negative view of others. They deny the need for attachment. Dismissive adults expect the
partner to be unavailable and non-responsive. Fearful avoidant adults are fearful of

intimacy and socially avoidant. This style includes a negative view of self and Others. The
fearful adult fears attachment and expects their partner to be rejecting

Dating Relationship
In this study, dating relationship is defined as a heterosexual couple that has been

dating for at least one month.

40
Problem Solving Communication
In this study, Grotevant and Cooper's (1985) communication behavior will be used

to identify varying styles of problem solving communication. The problem solving styles
are defined as follows:

Self Assertion involves direct suggestions. The participants displays awareness of
their own point of view and accept responsibility for communicating it clearly.

Permeability involves acknowledgment, requests for information, agreement,

making relevant comment(s), and compliance with requests for action. Participants

express responsiveness to the views of others.

Mutuality involves indirect suggestion, initiation Of compromise, statements

of other's feelings, and answer to requests for information. Participants show

sensitivity and respects for others views.

Separateness involves requests for action, direct disagreement, indirect

disagreement, and irrelevant comment. Participants express distinctiveness of self

from others.
Sex and Gender Role

In this study, gag refers to the biological and physiological differences between
males and females. It does not refer to psychological differences between males and
females. Gender role refers to an individual's endorsement of psychologically feminine
and masculine characteristics created by societal and cultural influences.

Hymtheses

There is continuing need to understand the factors involved in couple interaction
and functioning. Attachment theory provides a new framework for exploring these issues
further. Specifically, attachment theory provides a framework for understanding the effect
that gender role behavior and adult attachment styles may have on problem solving
communication in dating couples. As a result of these investigations, a number of

unanswered questions and literature gaps will be explored further. The following

hypotheses will be investigated in this study:

41
Relationships among attachment styles, sex, gender role and problem solving

communication

The first four hypotheses address how the individual participants' attachment styles
affect their problem solving communication and explore how sex and gender role influence
this relationship.

Hymthesis 1: Controlling for gender role variation, there will be a significant
relationship between individual attachment style and problem solving communication.
Hypothesis la: Secure and dismissive individuals will exhibit higher scores on self-
assertion than preoccupied or fearful individuals. Self -assertion involves making direct
suggestions, displaying an awareness of one's point of view and accepting responsibility
for communicating it clearly. Individuals who exhibit higher scores on self-assertion need
to feel confident in themselves. The secure and dismissive participant are more likely to
have this level of confidence due to their positive view of self. Hymthesis lb: Secure and
preoccupied individuals will exhibit higher scores on permeability than dismissive or
fearful individuals. Secure and preoccupied individuals have a positive model of others.
Permeability refers to communications and behaviors that support a relationship partner.
Hypgthesis 1c: Secure individuals will exhibit the highest score on mutuality relative to
individuals in the other attachment style categories. Mutuality involves showing sensitivity
and respect of other's viewpoints. Due to a positive view of self and other, secure
individuals will be more empathic and better able to attend to other's thoughts and feelings
than will insecure individuals. Hypothesis ld' Dismissive individuals will exhibit the
highest scores on separateness relative to individuals in the other attachment style
categories. The dismissive individual has a positive view of self with a low regard for
others. They value self sufficiency and are more likely to make separate statements which
create a distinction between themselves and others than secure, preoccupied, and fearful

individuals.

42
Hymthesis 2: When gender role differences are controlled, sex will not be

significantly related to problem solving communication.

Hypgthesis 3: There will be a significant relationship between gender role and
problem solving communication. Hypothesis 3a: Individuals scoring higher on the
masculinity scale will exhibit higher scores on self assertion and separateness. The
reviewed literature supports the view that masculinity is related to greater emphasis on
individuation in relationships. Hypothesis 3b: Individuals scoring higher on the femininity
scales will score higher on permeability and mutuality. The reviewed literature supports the
view that femininity is related to greater emphasis on maintaining connection in
relationships.

Hymthesis 4: Attachment style and gender role will interact to affect problem
solving behavior. Hypothesis 4a; Secure and dismissive individuals with high masculinity
scores will score higher on self -assertion and separateness than secure and dismissive
individuals with a high femininity score. Hypothesis 41x Secure and preoccupied
individuals with high femininity scores will score higher on permeability than dismissive or
fearful individuals with a high masculinity score. Secure and preoccupied attachment styles
exhibit a positive view of others and femininity relates to a greater emphasis on connection
in relationships. Permeability involves a responsiveness to the views of others in a
relationship. Relative to their peers in other groups, dismissive and fearful individuals with
higher masculinity scores will score the lowest on permeability. Dismissive and fearful
attachment styles have a negative view of others and masculinity scores relate to a greater
emphasis on individuality in relationships. H4pothesis 4c: Secure individuals with high
femininity scores will score higher on mutuality than will secure individuals with a high
masculinity score. Hypothesis 4d Dismissive individuals with high masculinity scores
will score higher on separateness than will dismissive individuals with high femininity

SCOTCS.

43
Relationship between couple's conjoint attachment style and problem solving behavior

The final hypothesis addresses how the couple's conjoint attachment style
influences problem solving communication.

Hypothesis 5: There will be a significant relationship between conjoint attachment
style and how participants' rate their overall problem solving effectiveness. Hypothesis 5a:
The secure/secure couples will receive the highest scores on couple's overall problem
solving effectiveness. Hypothesis Sb: The secure/secure couples will receive the highest

scores on rater's overall problem solving effectiveness.

CHAPTER III
Methodology

Particiwts

Students enrolled in College of Education classes at Michigan State University were
asked to voluntarily participate in this study. Presentations for recruitment were
administered to 12 undergraduate courses and one graduate course. For phase one,
students were asked to attend a data collection session where they completed three
questionnaires. As an incentive for participation, each student received extra course credit
points. During phase one, 188 students volunteered to participate and 157 students
attended the data collection sessions.

Table 2 contains descriptive demographic information of the sample from phase
one. The sample included 156 (30 men and 126 women) participants. One student did not
correctly complete the inventories and was thus dropped from the study. The sample
consisted primarily of seniors (54%) and juniors (21%) with a mean age of 21 (SD=1.88).
Participants were predominantly Caucasian (74%) and never married (99%). Attachment
style self classifications were as follows: secure (N=82, 53%), preoccupied (fl=l7, 11%),
fearful (_I\_l=42, 27%), and dismissive (I_\l_=15, 10%). These frequencies are comparable to
other studies using the Bartholomew Relationship Questionnaire with college samples
(Bartholomew & Horowitz, 1991; Duggan & Brennan, 1994; Kunce & Shaver, 1994;
Lopez, Gover, Leskela, Sauer, Schirmer, & Wyssmann, 1995; Pistole, 1994).

Sixty-seven percent of the overall sample indicated they were currently in a dating
relationship. Of this group, participant's attachment style self classifications were as
follows: secure (60%), preoccupied (13%), fearful (24%), and dismissive (3%). Dating
participants indicated they had known their partner for slightly less than three years

44

45
Table 2. Sample Demographic Information - Phase 1

 

 

Variable
Overall Response Rate #
Signed up for Phase 1 188
Attended Phase 1 157 (one participant did not identify attachment
style; N-156)
Sex and Attachment Style Total % Male % Patrols %
Secure 82 53% 11 37% 71 56%
Preoccupied 17 11% 4 13% 13 10%
Fearful 42 27% 11 37% 31 25%
Dismissive 15 10% 4 13% ll 9%
Total 156 30 19% 126 81%
Age Mean Standard Deviation
21 1.88
Year in College Total %
Freshman 10 6%
Sophomore 26 17%
Junior 32 21%
Senior 85 54%
Other 3 2%
Ethnicity Total %
African American 29 19%
Asian American 6 4%
Caucasian/White 116 74%
Hispanic/Latino 3 2%
Native knerican 1 1%
Other 1 1%
Marital Status Total %
Married 0 0%
Divorced 1 1%

Currently in 4: Dating Relationship Total % Male % Panels %
Secure 58 60% 8 47% 50 57%
Preoccupied 13 13% 3 18% 10 11%
Fearful 23 24% 4 24% 19 22%
Diamissive 10 3% 2 12% 8 9%
Total 104 67% 17 58% 87 69%

Volunteered for Phase 2 Total % Male % Panels %

Secure 42 71% 5 50% 37 76%
Preoccupied 7 12% 3 30% 4 8%
Fearful 6 10% 1 10% 5 10%
Dismissive 4 7% 1 10% 3 6%
Total 59 10 49
Length of Time Knew Dating Partner (months)
Mean Sample Deviation

Secure 36.1 26.3

Preoccupied 43 31.3

Fearful 20.1 17.7

Dismissive 31.8 28.7

Total 33.1 26.2

Table 2. (continued)

 

 

Sample Deviation
17.7
23.5
14.9
20.1
18.6

Seriousness of Dating Relationship (knot at all, very casual; 5-very serious)

Sample Deviation
1 . 0

Satisfied with this Relationship (l-not at all; 5=very satisfied)

Sample Deviation
. 75

l . 1

1 . 0

1 .2

Variable
Length of Time of Dating Relationship (months)
Mean
Secure 21.9
Preoccupied 31.2
tearful 14
Dismissive 20.9
Total 21.3
Mean
Secure 4.5
Preoccupied 4.4
Fearful 3.9
Dismissive 3.6
Total 4.3
Mean
Secure 4.4
Preocmmied 3.9
Fearful 4
Dismissive 3.7
Total 4.2

Exclusivity of Dating Relationship

.93

Exclusive Not Exclusive
# % # %
Secure 56 97% 2 3%
preoccupied 12 92% 1 8%
Fearful 19 83% 4 17%
Dismissive 7 70% 3 30%
Total 94 90% 10 10%
Parents' Marital Status
Married Separated Divorced Mother Father
Deceased Deceased
# % # % # % # % # %
Secure 58 70% 3 4% 17 21% 1 1% 3 4%
Preoccupied 12 71% O 0% 5 29% 0 0% 0 0%
Fearful 30 71% 2 5% 7 17% 2 5% 1 2%
Dismissive 9 71% 1 7% 4 27% 0 0% 1 7%
Total 109 68% 6 4% 33 21% 3 2% 5 3%
Quality of Parents' Marriage
Very Poor Poor Average Good Very Good
# % # % # % # % # %
Secure 6 8% 9 12% 12 15% 14 18% 37 47%
Preoccupied 3 18% 2 12% 3 18% 6 35% 3 18%
Fearful 1 3% 3 8% 15 38% 9 23% 12 30%
Dismissive 2 13% 1 7% 2 13% 2 13% 8 53%
Total 12 8% 15 1% 32 21% 31 21% 60 40%

47
(mean=33.1 months; SD=26.2) with preoccupied participants indicating the longest time

period (mean=43; SD=31.3) and fearful participant indicating the shortest time period
(mean=20. 1; SD=17.7). Dating participants reported they had dated their current partner
for slightly less than two years (mean=21.3 months; SD=18.6) with preoccupied partners
reporting the longest time period (mean=31.2; SD=23.5) and fearful participants indicating
the shortest time period (mean: 14; SD: 14.9). Kirkpatrick and Davis (1994) reported a
similar finding regarding preoccupied partners reporting the longest term relationships.
One could, however, hypothesize that secure subjects would have demonstrated the
greatest length of time in a dating relationship. The unidentified partner's attachment style
may be a factor influencing this result. Participants in dating relationships indicated that
their relationships were serious (mean=4.3; SD=1.0) in which they were satisfied
(mm=4.2; SD=.93). Secure participants reported the highest degree of seriousness
(mean=4.5; SD=1.0) and satisfaction (mean=4.4, SD=.75). Dismissive participants
reported the lowest degree of seriousness (mean=3.6; SD=1.3) and satisfaction with the
relationship (mean=3.7; SD=1.2). In examining the overall parental status of the
participants, 68% reported that their parents were still married and 25% reported their
parents were divorced or separated. Sixty-one percent described their parents marriage as
good to very good.

The packet for phase one included a form recruiting participation for phase two of
the study. Students in a heterosexual dating relationship of at least one month in duration
were eligible. From the original dating sample, 57% volunteered to participated in phase
two which resulted in a total of 60 participants. Several participants indicated they could
not participate in phase two since their partner resided outside of the immediate area. On
the basis of responses to Bartholomew's Relationship Questionnaire, ten individuals from
each attachment group were to be chosen to participate in the second part of the study. The
60 participants, who volunteered to participate in the second phase, divided as follows:
four dismissive, six fearful, eight preoccupied, and 42 secure participants. From this

48
sample, 40 participants in heterosexual dating relationships were selected to complete the

second part of the study. Sampling decisions included recruiting all students in the
dismissive, fearful, and preoccupied categories to allow for a diverse sample. Three of the
students from this group indicated that their partners lived outside of the immediate area and
were added to an alternate list In identifying the remainder of the participants for phase
two, males with a secure attachment styles and minority students were identified first
followed by a random selection from the females with a secure attachment style. The
sample included 3 dismissive, 4 fearful, 8 preoccupied, and 30 secure participants. Five
additional secure participants were scheduled for phase two to allow for no-shows, audio
equipment difficulties, and unusable data packets. Participants were contacted by telephone
to schedule a data collection session for phase two.

Forty-two students agreed to schedule a data collection session for phase two. Two
males (one identifying as preoccupied and the other identifying as dismissive) indicated that
their girlfriends ended their relationship and they were unable to participate in phase two.
Six secure female participants stated they were unable to participate since their boyfriends
were not able to attend or did not want to attend phase two. Three secure participants (two
females and one male) did not return phone messages. One secure female provided an
incorrect phone number. The final scheduled sample for phase two included: four fearful
(one male and three females), seven preoccupied (two males and five females), two
dismissive females, and 29 secure (three males and 26 females). Six students did not show
for their scheduled data collection time (one male with a fearful attachment style and five
females with a secure attachment style). Two females with a secure attachment styles
canceled their scheduled time due to scheduling conflicts and they were unable to
reschedule. To address the subject attrition, participants were asked to volunteer friends
that may be interested in participating in the project. In addition, participants were recruited
from another education undergraduate course. Seven new couples were added to the

study. Forty-two couples participated in the project and forty couples were used for data

49
analyses. Final sampling decisions included discarding one packet for incomplete

information and randomly discarding one of the secure-secure couple packets since that
category had the highest number of couples.

During the second part of the study (phase two), 80 participants (40 couples) were
asked to complete three questionnaires. The partner that was not originally recruited from a
class also completed a demographic/background form. Both participants from the seven
couples that did not participate in phase one completed the demographic/background form
and the three questionnaires. Immediately following the completion of the inventories,
each couple participated in a twenty-minute audiotaped problem solving task. Following
the task, each participant completed two questions on joint problem solving. Couples
participating in the second part of the study were entered in a drawing for four $25.00
awards. Informed consent forms (See Appendices A and C) were completed by all
students participating in this study.

Procedures

An overview and stated purpose of the study were introduced at the start of the
semester to each class from which participants were initially recruited. Students were
informed that the study would examine relationship beliefs and problem solving
communication in dating couples. Students selected an administration time to attend for the
first phase of the study. During phase one, participants were provided with an introduction
letter explaining the purpose of the study, requesting participation, explaining informed
consent, and assuring confidentiality (See Appendix A). In addition, individuals received a
request for participation in phase two (See Appendix B). After both forms were
completed, participants were given a demographic/background form, Bartholomew's
Relationship Questionnaire, and Simpson's Adult Attachment Style Inventory. Participants
completed the packet in 10-15 minutes.

Following completion of phase one, participants' attachment style was identified via

their responses to Bartholomew's Relationship Questionnaire. Participants were contacted

50
by phone to schedule phase two of the study. Reminder letters were mailed one week in

advance and phone calls were made the day before the couple participated in phase two.

During phase two, the couple was initially placed in separate rooms while they
completed the following: informed consent for audiotaping (See Appendix C),
Bartholomew's Relationship Questionnaire, Simpson's Adult Attachment Style Inventory,
and Bern's Sex Role Inventory. The partner that did not attend phase one also completed
the demographic/background form. The couple jointly completed the 20 minute Couple
Interaction Task. After the couple was escorted to a private room, the proctor read the task
and alerted the couple to the 20 minute time limit A written form of the task was left with
the couple. The proctor started the tape recorder. They were told that the proctor would
knock on the door to provide a five minute warning. Following completion of the task, the
couple completed two questions about the task. A short oral debriefing followed the
completion of phase two.

Instruments

This study included a demographic and background information form; two self -
report measures of adult attachment (Bartholomew's Relationship Questionnaire; Simpson
Adult Attachment Style Inventory); one self-report measure of gender role (Bern Sex Role
Inventory); and an in vivo problem solving communication task (Couple Interaction Task).
Demographic and BaLkgLound Infomtion Form

Participants were asked to provide the following demographic information: sex,
age, year in college, ethnicity, and marital or dating status. In addition to demographic
information, this form inquired about the following: length of time partners knew each
other; length of time dating current partner; seriousness of relationship; relationship
satisfaction; exclusivity of relationship; parents' marital status; and quality of parents'
marriage (See Appendix D).
A_ttaLchment Style

Two instruments were used to assess adult attachment styles in this study.

51
Bartholomew's Relationship Questionnaire (BRQ) is a self -report, categorical

measure of the four adult attachment styles (identified by Bartholomew, 1990 and
Bartholomew & Horowitz, 1991). The measure is a adapted from Hazan and Shaver's
(1987) original measure. Participants are asked to select one of four descriptive paragraphs
that best describe their feelings about closeness or intimacy in romantic relationships. The
four paragraphs represented adult versions of Ainsworth's three attachment styles: secure,
anxious, and avoidant with the avoidant category divided into two styles, dismissive and
fearful. Paragraph one represented the secure attachment style; paragraph two represented
the dismissive attachment style; paragraph three represented the preoccupied attachment
style; and paragraph four represented the fearful attachment style (See Appendix E).

Scharf e and Bartholomew (1994a, 1994b) reported that the BRQ self -classif ication
demonstrated moderate stability over an eight month period and a two year period with test—
retest correlations by attachment style group ranging from .49 to .71, and from .30 to .67,
respectively. In this present study, stability was examined over a one month time period.
The overall rate of attachment style rating change was 22%. Due to the low sample size for
dismissive, fearful, and preoccupied participants, the attachment style categories were
collapsed into an insecure category. Secure subjects changed self ratings 4.2%. Insecure
participants attachment style rate of change was 58%, with 71% of the insecure participants
changing to one of the other three insecure categories (See Table 3). The latter group's
overall rate of change was comparable to that observed by previous studies (Baldwin &
Fehr, 1992; Pistole, 1989).

The BRQ has some problematic properties. The single item forced choice measure
results in participants responding to mutually exclusive categories. Differences in
individual variability is overlooked. Hence, a second continuously scaled attachment style
inventory (Simpson's Adult Attachment Style lnventory) was used in this study as a way to

examine the concurrent validity of self -reported attachment style classification. Thirty six

52
Table 3. Attachment Style Rate of Change

 

 

number Changed Rate of Change Total Subjects
Secure* 1 4.2% 24
Insecure** 7 58% 12
liar 2 33% 6
Warren 6 20% 30
Total 8 22% 36

 

*Changed to an insecure category

**TWo insecure categories changed to a secure category and five insecure categories
changed to one of the other three insecure categories .

53
subjects completed the BRQ during phase one and again one month later during phase two.

Baldwin and Fehr (1992) recommended that due to the instability of attachment style
classification findings, subjects need to classify themselves at the time of the experimental
session. Data analysis used the BRQ at phase two during the problem solving session of
data collection.

Simpson's Adult Attachment Style Inventory (S-AAS; Simpson, 1990; Simpson et
al., 1992) is a 13-item continuous measure of adult attachment styles (See Appendix F).
The S-AAS decomposed Hazan and Shaver's (1987) attachment style descriptions into 13
individual sentences. The participant answers each sentence on a 7—point Likert—type scale
(1: Strongly disagree and 7 = Strongly agree). Three sentences were changed to a
negative direction to avoid acquiescence response biases. The S-AAS contains two factor-
analytically—derived subscales: avoidance/security and anxiety. Higher scores represent
greater attachment related avoidance and anxiety, respectively. Simpson et al. (1992)
reported Cronbach alpha reliability coefficients of .81 for the avoidance/secure scale and
Cronbach alphas ranging from .58 to .61 for the anxiety subscale. Lopez, Gover, Leskela,
Sauer, Schirmer, and Wyssman (1995) reported Cronbach alpha reliability coefficients of
.83 for the avoidance/secure scale and .70 for the anxiety scale. In the present study,
obtained Cronbach alphas for phase one avoidance/secure and anxiety scores were .73 and
.83 respectively, and alphas for phase two avoidance/secure and anxiety scores were .86
and .82 respectively.

Test-retest scores of S-AAS subscale scores between phase one and phase two (one
month apart) were computed to assess their temporal stability. In examining the
relationship between the subscales, there was a high positive correlation between the
avoidant/secure phase one and phase two scales (p.79, p<.01) and between the anxiety
phase one and phase two scales (r=.86, p<01). There was a moderate positive correlation
between anxiety phase one and avoidant/secure phase two (r=.43, p<.01), and a negligible

correlation between the two subscales (anxiety and avoidant/secure) when compared in the

54
same time frame. This indicated that participants who were identified as anxious during

phase one were less secure (more avoidant) during phase two. Over a one month period,
anxiety about one's relationship may erode one's sense of security in the relationship (See
Table 4).

A one way ANOVA was computed to examine the relationship between the BRQ
and the S-AAS. The S-AAS is a self-reported continuously scaled attachment style
inventory that was used in this study to provide concurrent validation of the BRQ which is
a self-reported, categorical attachment style inventory. In addition, the S-AAS was used to
measure adult attachment styles for some of the regression analyses in this study. Highly
significant differences were found for the avoidant/secure subscale, E13, 76) = 28.34,
p<.0001. Scheffe post hoc tests revealed four significant pair-wise group differences for
the BRQ. Based on mean scores (See Table 5), significant group differences were found.
Results indicated that fearful and dismissive respondents scored higher on the
avoidant/secure subscale (higher avoidance) than did secure and preoccupied respondents.
Significant differences were found on the anxiety subscale, E( 3, 76) = 8.04, p<.0001.
Scheffe post hoc tests revealed four significant pair-wise group differences for the BRQ.
Based on the mean scores (See Table 5), significant group differences were found. Results
indicated that fearful and preoccupied respondents scored higher on the anxiety subscale
than did secure and dismissive respondents. Taken together, these results indicated that the
attachment style self classifications derived from the BRQ categorical measure were
consistent with the dimensional scores supplied by the continuous indices on adult
attachment orientation.

Gender Role

Bem Sex Role Inventory (BSRI; Bern, 1974) was used to assess gender role

differences in this study. The BSRI measures an individual's endorsement of gender role

attributes. Initially, femininity and masculinity were conceptualized as opposite ends of a

55
Table 4. Intercorrelations of Simpson's Adult Attachment Style Inventory at Phase 1 and

 

Phase 2

Advsec Anxiety Advsec ( 2) Anxiety( 2 )
Advsec . 19 .79** .20
Anxiety .44Mr .86**
Advsec(2) .21

 

Note: Avdsec-Avoidance/Security; Advsec (2 )=Avoidance/Security, time 2; Anxiety
(2)-Anxiety, time 2.

* - Signif. LE .05 ** - Signif. LE .01 (2-tailed)

56

Table 5. Means on Simpson's Adult Attachment Style Inventory Subscales by Attachment
Style Self Classification

 

 

8m Avoidant/Secure Anxiety

than SD 11 than SD n
Secure 20.88 5.3 49 12.86 3.7 49
Dismissive 34 .60 8 .9 10 12 .70 3 .9 10
Preoccupied 22.88 3.1 8 18.75 4.8 8
Fearful 33.77 5.5 13 18.23 6.9 13

 

57
single continuum. Bem (1974, 1977) argued that masculinity and femininity are two

separate dimensions. Furthermore, an individual could identify as having both masculine
and feminine characteristics. Hiller and Philliber (1985) argued that the BSRI also
measures two clusters of personality traits, assertiveness and sensitivity. Spence and
Helmreich (1981) argued that the BSRI measured self images of instrumental and
expressive personality traits associated with masculine and feminine traits, respectively.
Despite the debate regarding the BSRI, the inventory has been used extensively in the
literature to measure gender role orientation. Bieger (1985) stated that the instrument is a
valuable research measure.

Bern (1974) reported internal consistencies, based on coefficient alpha, ranging
from .70 to .86. Masculinity and femininity scores are free to vary independently. Two
samples indicated that they are empirically independent Bern reported that test-retest
reliability ranged from .89 to .93 over a one month period. Hiller and Philliber (1985)
indicated that the BSRI correlated with education, age, occupational sex type, housework
participation, and attitudes about who should do housework and earn income. Ballad-
Reisch and Elton (1992) reassessed the reliability and validity of the BSRI. Cronbach
alpha for the masculinity subscale was .89 and for the feminine subscale was .86. These
researchers concluded that the factor structure of the BSRI was reliable. In addition, the
researchers concluded that the BSRI measures differing personality characteristics but these
characteristics may be better described as instrumental and expressive.

The BSRI is a self -report inventory consisting of 60 attributes. The respondents
indicate how well each attribute describes them based on a 7-point scale. The scale ranges
from 1 (never or almost never true) to 7 (always or almost always true). Twenty attributes
reflect the cultural view of masculinity, 20 reflect the cultural view of femininity, and 20 are
desirable for both (neutral fillers) (Bern, 1984). The attributes were judged to be most
desirable for one sex or the other in the American culture (Bern, 1977). Each respondent

receives two scores, one on the masculinity scale and one on the femininity scale.

58
Androgynous scores are derived from the two subscales, masculinity and femininity. Bem

developed the BSRI with data from a college student population. In the present study,
obtained Cronbach alphas for the masculinity and femininity scores were .90 and .78,
respectively.

Problem Solving Task (CIT)

Watzlawick (1966) stated that direct questioning is often an ineffective means of
obtaining information on behavioral functioning. Furthermore, ”significance lies not only
in the content of communication, but mainly in the specific process of communication” (p.
256). Based on these ideas, this study obtained information on problem solving
communication in couples with a spontaneous problem solving task.

Grotevant and Cooper (1985) adapted the Family Intervention Task from
Watzlawick's (1966) Plan Something Together Task. This task was originally part of a
Structured Family Interview designed to observe the process of family communication
patterns and to shorten the time required to identify family interaction patterns. Grotevant
and Cooper (1985) used this task to elicit the expression and coordination of vieWpoints
from all family members on a particular topic. They believed that the task would elicit
statements that suggested individuality (suggestions or disagreements about activities) and
connectedness (agreements, questions, or initiation of compromises about activities) from
the individual family members. The task avoided focusing on family differences and,
therefore, encouraged power sharing and mutual decision making. Generalization of
findings were interpreted with caution since communication was observed during only one
task.

For this study, the Plan Something Together Task! Family Intervention Task was
modified for use with dating couples and referred to as the Couple Interaction Task (CIT).
Condon, Cooper, and Grotevant (1984) described the directions as follows: The proctor
told the couple that they have two weeks and unlimited funds to spend on a vacation

together. Their task was to plan (within a 20 minute time limit) a day-by-day itinerary for

59
the vacation. They were asked to write down the location and activities planned for each

day (See Appendix H), and told that their interaction would be tape recorded. The proctor
provided the following instructions:

I have a problem that I would like you to solve together. I would like you to

imagine that you have two weeks and unlimited funds available for a [couple]

vacation. Your job will be to plan the vacation day-by-day. Here is a sheet of
paper with 14 spaces, one for each day of the two weeks. In the left-hand column
of each space, I would like you to record the geographic location of each day's
activity. In the right-hand column of each space, list the specific activity or
activities planned for each day. You will have 20 minutes to make your

decisions. When you are ready, I will turn on the tape recorder and go in the other

room. [I will knock at the door to let you know that you have 5 minutes remaining

to complete the task]. (p.8).

The proctor answered any questions and left a copy of the task with the couple. The tape
recorder was turned on. After 15 minutes, the couple was alerted to the fact that only five
minutes remained to complete the task.

The coding of the audiotapes was based on a system developed by Grotevant and
Cooper (1985). This system was based on developmental and clinical research and on
principles of speech and conversation analysis. Condon, Cooper and Grotevant (1984)
described this system in detail. In an earlier family project of 444 participants, family
interaction was coded with 14 communication categories. Based on factor analyses, the
results indicated that the 14 communication categories loaded on four constructs: self -
assertion, separateness, mutuality, and permeability.

Grotevant and Cooper's (1985) results demonstrated sex differences for adolescent
males and females. Males made more requests for information and females provided more
answers to requests for information. Correlations were computed between problem

solving communication behaviors and identity exploration ratings for adolescents. For

60
male adolescents, direct suggestions and expressed separateness with their fathers were

positively correlated with identity exploration. Female adolescents who rated higher on
identity exploration exhibited more indirect suggestions with their fathers. Female
adolescents' identity exploration was negatively related with expressed mutuality toward
their mothers. The adolescent f emales' higher identity exploration was positively
associated with indirect suggestions and negatively associated with direct suggestions.
Grotevant and Cooper concluded that adolescents who scored the highest in role taking
skill interacted with at least one parent in a relationship characterized by a co-occurrence of
permeability and separateness. Adolescents with low scores on identity exploration and
role taking skill exhibited low levels of disagreement and high degrees of permeability.

Three transcriptionists transcribed the audiotapes (See Appendix I for Transcript
Format and Appendix J for Transcript Guidelines). Four raters blind to the hypotheses and
individual attachment styles coded the audiotapes (See Appendix K for Audiotape Coding
Manual). Two teams of two raters each rated 20 of the 40 audiotapes. Before the raters
coded the transcript, the speech was divided into codable chunks. Each chunk was
numbered serially. Condon, Cooper, and Grotevant (1984) defined a chunk as ”all
independent clauses together with any dependent clauses that are connected to it" (p. 13).
The first 300 utterances were coded. Cooper, Grotevant, and Condon (1982) reported that
the correlation between the first 300 utterances and the entire session ranged from .54 and
.88. This suggested that the first 300 utterances were sufficiently representative of the
whole task. If the couple finished early and there were fewer than 300 utterances, the
entire tape was coded.

Raters coded the chunks into one of 14 categories. The 14 categories fell under the
headings of the four main constructs: self -assertion (direct suggestion); permeability
(acknowledgment, request for information, agreement, relevant comment, compliance with
request for action); mutuality (indirect suggestion, initiation of compromise, statement of

others feelings, answer to request for information); and separateness (request for action,

61
direct disagreement, indirect disagreement, and irrelevant comment). Scores on the four

categories were obtained by summing the recorded frequencies in each specific category
(See Appendices L, M, and N for coding sheets).

In_terr_ater Reliability. The raters attended a two hour training session prior to rating
the transcripts. Each rater coded a sample transcript Correspondence scores were
completed for each team of raters. In the case of any coding discrepancies for any of the
300 utterances, a third rater made a final decision. Grotevant and Cooper (1985) reported
interrater reliabilities exceeding .75 for all the but three categories (initiates compromise,
acknowledgment, and agreement). The reliability results for those three categories were as
follows: initiates compromise, .52; acknowledgment, .64; and agreement, .72.

For this study, the raters' sample transcript was compared to the project
coordinator's coded transcript. After their two hour training session, the reliability results
were as follows: Rater A, .75; Rater B, .62; Rater C, .69; and Rater D, .64. Each rater
exhibited consistency in the errors they exhibited. Extensive feedback on the problematic
areas was provided in verbal and written form for each rater. Reliability results were
examined a second time. The raters completed a second transcript which was compared to
the project coordinator's coded transcript The following reliability results were obtained:
Rater A, .77; Rater B, .62; Rater C, .76, and Rater D, .67. Extensive feedback on the
areas that differed from the project coordinator's coded transcript were provided in written
form for each rater. Reliability results were obtained for each team of raters at the end of
the coding process. Each team coded 2O transcripts based on coding 300 utterances per
transcript with 14 CIT categories. Eleven transcripts had less that 300 responses ranging
from 112 to 299 responses. Both teams obtained interrater reliability results of .58.

Generalizability of findings need to be interpreted with caution since the couple was
observed in one communication task. In addition, the task was designed to focus on power

sharing patterns during a problem solving discussion.

62
One additional component was included in the overall CIT procedure. Following

the problem-solving discussion, each partner individually answered two additional
questions: ”How successful were you and your partner in completing this task?" and
"How stressful did you find this task?" Comparisons were made regarding each partner's
individual and the couple's conjoint responses to their effectiveness in completing the
problem solving task and regarding their views on the stressfulness of the task (See
Appendix 0). In addition, the raters independently answered the question on problem
solving effectiveness and indicated how effective they viewed the couple's ability to
complete the task (See Appendix P).
Research Hypotheses
The results of the data analyses were expected to confirm a relationship between
attachment style and problem solving communication in couples and to show that this
relationship was moderated by gender role. Specifically, a participant's individual
attachment style was expected to affect the style of problem solving communication pattern
he/she exhibited in an in vivo couple interaction task. Gender role, and not sex, was
expected to influence the problem solving communication patterns. Couples' conjoint
attachment style-related contributions were expected to influence problem solving
communication patterns with certain attachment style combinations scoring higher and other
attachment style combinations scoring lower on a composite problem solving score.
Data Analysis
1. For each of the appropriate demographic and background variables, the following
descriptive statistics were computed: mean, standard deviation, and range.
2. To measure internal consistency, Cronbach alpha was computed for the continuously-
scaled instruments used in this study.
3. Other reliability analyses included the following: interrater reliabilities for the CIT were
compared with those reported by Grotevant and Cooper (1985); test-retest scores on the

Simpson for 36 participants between phase one and phase two were computed; and

63
attachment style rate of change scores for the BRQ for 36 participants between phase one

and phase two.

4. Correlation matrices were computed to examine interrelationships among the following
variables: sex, attachment style, Bern masculinity and femininity scores, length of time
since met dating partner, relationship length, relationship satisfaction, relationship
seriousness, parents marital status, quality of parents marriage, CIT scores, overall
problem solving score for couples, and overall problem solving score for raters.

5. Relationship history variables that significantly correlated with CIT scores were used as
covariates in subsequent analysis.

6. An ANOVA was computed to examine the relationship between the Bartholomew
Relationship Questionnaire and the Simpson Adult Attachment Inventory. The Simpson
Adult Attachment Inventory is a continuously scaled attachment style inventory that was
used in this study as a way to check the validity of Bartholomew's Relationship
Questionnaire which is a sel f -reported, categorical attachment style inventory.

7. The first step of data analyses examined individual participant differences. Four
ANCOVA's included sex and attachment style as the independent variables, Bern
femininity and masculinity scores as covariates, and individual CIT scores as the dependent
variables. Four relationships were analyzed regarding individual participant relationship
behavior“. a) covariates and CIT scores; b) sex and attachment style and CIT scores while
controlling for Bem femininity and masculinity scores; and c) gender role and attachment
style interaction. In addition, a regression analyses was used to assess if the Bern
masculinity and femininity scores moderated the attachment style-CIT relationships.

8. The second stage of data analyses examined the couple scores. An ANCOVA included
couple attachment style pair group as the independent variable; male Bem femininity and
masculinity scores and female Bem femininity and masculinity scores as the covariate; and
couple average rating regarding problem solving effectiveness, rater average rating

regarding problem solving effectiveness, and the couple's combined scores on the

64
mutuality scale (problem solving communication category) as the dependent variables.

Two relationships were analyzed: a) covariate (Bem femininity and masculinity scores) and
the dependent variables; b) attachment style pair groupings and dependent variables while
controlling for Bern femininity and masculinity scores.

9. Post hoc analyses explored whether male and female problem solving effectiveness

scores differentially correlated with individual's attachment styles.

CHAPTER 4

Results

Descriptive St_atistics

For the following descriptive statistics, there were two missing values and no
invalid responses for the respondent sample. For the two missing values, the average scale
score was substituted. Descriptive statistics for the demographic and background variables
in the study are reported in Table 6.

The overall sample from phase two included the following _I\_I_s_ and frequencies of
attachment style self classifications: secure (flz49, 61%), preoccupied (13:8, 10%),
fearful (91:13, 16%), and dismissive (flle, 13%). These frequencies are comparable to
other samples with a college sample using the Bartholomew Relationship Questionnaire
(Bartholomew & Horowitz, 1991; Duggan & Brennan, 1994; Kunce & Shaver, 1994;
Lopez, Gover, Leskela, Sauer, Schirmer, & Wyssmann, 1995; Pistole, 1994). Forty
dating couples participated in phase two of this study (40 men, 40 women; mean
age=21.9). The sample consisted primarily of seniors (53.8%); participants were
predominantly Caucasian (87.5%) and never married (97.5%). Almost all participants in
the sample (97.5%) reported that their current dating relationship was exclusive.

Participants indicated they had known their dating partner for 33.3 months
(SD=25.6) with preoccupied participants indicating the longest time period (mean=40.3,
SD=33.1) and fearful participants indicating the shortest time period (mean=27.4,
SD=23.2). Dating participants reported they had dated their current partner for 25.6
months (SD=21.8) with preoccupied partners reporting the longest time period
(mean=36.3, SD=27.6) and fearful participants indicating the shortest time period
(mean=23.2, SD=23.1). One-half of the sample reported they had been dating for more

than two years. Dating partners indicated that this was a serious relationship (mean=4.5,
65

66
Table 6. Sample Demographic Information - Phase 2

 

 

Variable
Sex and Attachment Style Total % Male % Panels %
Secure 49 61% 22 53% 27 68%
Preoccupied 8 10% 4 5% 4 10%
Fearful 13 16% 7 18% 6 15%
Dismissive 10 13% 7 18% 3 8%
Total 80 40 40
Age than so Range
21.9 3.1 21
Year in College Total %
Freshnnn 5 6.3%
Sophanore 9 11.3%
Junior 8 10%
Senior 43 53.8%
Other 15 18.8%
Ethnicity Total %
African knerican 1 1.3%
Asian American 5 6.3%
Caucasian/White 70 87 . 5%
Hispanic/ Latino 1 1 . 3%
Other 3 3.8%
Marital Status Total %
Married 0 0%
Divorced 2 2.5%
Single/Never Married 78 97.5%
Length of Time Knew Dating Partner (months)
than SD Range
Secure 33.6 24.4 94
Preoccupied 40 . 3 33 . 1 92
Fearful 27.4 23.2 75
Dismissive 33.5 30.2 90
Total 33.3 25.6 94
length of Time of Dating Relationship (months)
than SD Range
Secure 23.7 19.4 71
Preoccupied 36 . 3 27 . 6 68
Fearful 23.2 23.1 83
Dismissive 29.3 26.5 77
Total 25.6 21.8 83

Seriousness of Dating Relationship (1=not at all, very casual; 5=very serious,
cannitted)

than SD Range
Secure 4 . 5 . 71 2
Preoccupied 5 0 0
Fearful 4 .5 . 78 2
Dismissive 4 . 4 . 70 2
Male 4 .5 . 72 2
ranale 4 . 6 . 68 2

Total 4.5 .69 2

Table 6. (continued)

67

 

 

ararro artisans}:
0

Net Exclusive

Mother

Deceased

%

1%
0%
0%
0%

variable
Satisfied with this Relationship (1-not at all; 5-very satisfied)
Mean SD
Secure 4.6 .65
Preoccupied. 4.8 .46
Fearful 4.2 .93
Dismissive 4.4 .88
Male 4.4 .74
Panele 4 . 6 . 71
Total 4.5 .73
Exclusivity of Dating Relationship
Exclusive
# %
Secure 48 98.8%
Preoccupied 8 100%
Fearful 12 98.8%
Dismissive 10 100%
Total 78 97.5%
Parents' Marital Status
Married Separated
# % # %
Secure 34 43% 0 0%
Preoccupied 6 8% 0 0%
tearful 7 9% 0 0%
Dismissive 7 9% 0 0%
Total 54 68% 0 0%

*One participant's parents divorced and the father was deceased.

Age at Parent's Divorce

Mean SD

Secure 10 5.2
Preoccupied 17 0
Fearful 5.5 3.8
Dismissive 13 5.7
Total 9.96 5.5
Quality of Parents' Marriage

very Poor Poor

# % # %
Secure 1 1% 9 11%
Preoccupied 1 1% 1 1%
tearful 4 5% 3 4%
Dismissive 0 0% 1 1%
Total 6 8% 14 18%

# %

1 1.25%
0 0%

1 1.25%
0 0%

2 2.5%
Divorced
# %

14 18%

2 3%

6 8%

2 3%

24 30%

Total
14
2
6
2
24
Average
# %
14 18%
1 1%
1 1%
3 4%
19 24%

HOOOH‘

‘OfiéN‘O

1%

11%
3%
5%
5%

24%

Father

Deceased

# %

1* 1.3%
0 0%

0 0%

1 1.3%

2 2.6%

very Good

# %

16 20%
3 4%
1 1%
2 3%

22 28%

68
SD=.69) in which they were satisfied (mean=4.5, SD=.73). Preoccupied participants

reported the highest degree of seriousness (mean=5, SD=0) and satisfaction (mean=4.8,
SD=.46). Dismissive participants reported the lowest degree of seriousness (mean=4.4,
SD=.70) and satisfaction with the relationship (mean=4. 1, SD=.88). In examining the
overall parental status of the participants, 68% reported that their parents were still married.
Secure participants reported the highest degree of parental marital stability (43%) as
compared to the insecure attachment groups. Thirty percent of the overall sample reported
that their parents were divorced. Fifty-one percent of the entire sample described their

parents marriage as good to very good and 26% described their parent's marriage as poor

to very poor.

Correlational Findings
CIT Scores

The Pearson intercorrelations among all the key demographic and research variables
are reported in Table 7. Three variables correlated significantly with CIT permeability
scores. There was a modest positive correlation between 'permeability' and 'separateness'
(r_=.24, p<.05). This indicated that participants that exhibited more permeability statements
exhibited more statements of separateness. There was a similar relationship between
'permeability' and 'femininity' (r=.24, p<.05). This was an expected relationship based
on Hypotheses 3b. Individuals scoring higher on the femininity scale were expected to
obtain a higher permeability score. Finally, there was a moderate positive correlation
between 'permeability' and 'Simpson's anxiety subscale' (_r;=.28, p<.05). This
relationship was expected based on Hypothesis 1b. Preoccupied individuals (those scoring
higher on Simpson's anxiety subscale) were expected to obtain a higher permeability score.

There was a modest negative correlation between 'separateness' and 'Iength of
dating relationship' (Lt-.24, p<.05). This indicated that participants that reported longer
term relationships obtained lower scores on separateness.

A modest positive correlation was found between 'self-assertion' and 'Simpson's

69

Hue—o 5. Sansone—sec: om UnBochEo can W888: Sense—om

 

 

5.5—bows H n u o u a 5 a o no 5 5 5
H gal—Mew
n wagon—Haw. .5
u ”Dacha mnoo mo”: .3. .5
a nocvwo wnoo monk .3. .Nm. .3
m esophageal r.o~ .2; Jun .3
a négdh .35 be» .Ho 23. 33
5 rounds cannon .3 32 .34 .3 the. .8
a EQHEQ .5 Jon Jam .5 .5 the. Joe
m mange. guano Ibo I.ou .Hu .3 Joe .on .5. .05
Ho 95H meadow EH .ou .ow I.oo I.Hm .oo .ou t.o~ I.ou the!»
up wowndwosnowv

munhnnsonwonp I.om .3 .Ho .09; .05 .NN .Nes I.om .oe I.o~
5 mownrvnnofinwon .3 .om .5 .0» 38 Jo.» 33 .5 low Lu .15
5 rough—.0356

menu—beacons .35 .3 .Nu .ums .oa .5 .3: r.ou .mo 1.3 .3: r.-
we 349.535

moocHHQ .3 .33 r.oo Jun .3 fun: .2 Lo .3 r.o~ 33: .3: .1:
Ha fish .5 .Nms .ON .NN .No .305 I.om .He .ou IL.» 1.5 .ow I.ou
pm 350 .393

Mg .ou .OH .Nw .Nq ILb .om .mm»: I.~H .om I.ou_. .NH I.oo .uqss
5 00:95.6 menses

”Inga .om I.~o IL—w I.Nm I.om I.o~ r.om .He lbw; .Nu r151. .05 fume
Ho Hun whoops mo“: .0» .om .NN . I.o~ .Nw» .qu .om .HN I.ow .Nut I.o~ .Hm
Ho Hon mdflonn ”toga .Hm I.oo lbw; I.~u I.ow I.NH 1.0» .Ho thus .om fumes .Hu lbwsa
no Ea whoops-s

mom—3&5 .HQ .3. I.w~ . Lb .ums .No .NH .NN ILm .uus I.ou .ums
3 «go 3.0on:—

m0H<HDo .No .Nu .muw . I.ou .wu» .3 .5 .Ho ILu .334 .om .uw»
- Ea menace

”Devon .uw .35 3mm: t.~u I.ou 1.0m I.ou .om the» .N» that .0» the:
nu page meson

”Down—n I.Hw I.Hu .uN I.~u 1.0m .ou I.om .5 IL: .5 I32; .ow I.~H
no egg

gapew . .5 .3 .8 .o~ r.om r.o~ km 3.5 .No .3 .5 1.2..
we won the...» .5 J: . .8 .3: r.o~ rho» Joe .8 .5 r.ou .om

s m M .03 .1 w M .o: .. . .. we oak—ween Mn 9 000930an caused on 086559

7O

Hugo 5. 803253

 

 

cshwnowe as pm pm we we we ~o up - mu we nu
Hm ssxhnnh .NH
Hm Ewan users
manseH .ow r.om
H5 noeowo.n menses
”Deena .Ha .Hu I.Ho
we Hunh<wncnw wnoowna
mapcpsn r.~us .oo .oo .
we anh<wn=sw menace
weapon .No .wws r.om . r.~ut
no zaps MHOUHQB
mOH<Hun I.Nw .NN .Hm I.w~s . I.~5
NH mmBDHo mucobms
mOF<Fum I.~m .Hm .uu» I.~H . I.Hm .mm»s
nu zaps menace
weapon .H5 .H5 I.oe . I.~5 . I.~5 I.Hm
~u anaswo manna
”Davao .om .ow I.He . I.~5 . t.~5 I.Hw .ums
no noaownon
xfincbwwdw .om I.ou .om .No .oa .ue» .0» Ho .wos 1.05
nu max I.om I.oe .ou . I.ou I.om . . . . .

s m M .ou» 4* w M .ou

o
h

. . . Hm Unwbdea Mn 9 oommmwowosd asbbow om ongocece

71

Table 7. (continued)

1, 2, 5, 12.Couple Interaction Task (CIT) was coded with four scores: (Mutuality, Permeability,

Separateness, and Self-Assertion).

3. Overall problem solving score for raters was coded from the rater's mean scores (Rater's Problem
Solving).

4. Overall problem solving scores for couples was coded from the couple's individual mam scores
(Couple Problem Solving).

6,8. Bern was coded as masculinity and femininity scores (Masculinity, Femininity).

7. Relationship length ranged from 1 month to 84 months (Length Dating).

9. Parent's marital status was coded as (Parent's Marriage): l=married; 2=separated; 3=divorced;
4=mother deceased; 5=father deceased

10. Quality of parent's marriage was coded as (Quality Parent's Marriage): lzvery poor; 2=poor;
3=average; 4=good; 5=very good

11. Relationship satisfaction was coded on a 5 point Likert-type scale (i.e., l=not at all satisfied to
5=very satisfied) (Relationship Satisfaction).

13. Relationship seriousness was coded on a 5 point Likert-type scale (i.e., l=not at all; very casual to
5=very serious; we're committed) ( Relationship Seriousness).

14,15 Attachment style was coded with the Simpson Attachment Style Inventory Subscales,
Avoidant! Secure and Anxiety, Phase two (Avoidant/ Security, Anxiety).

16. Time since met dating partner ranged from 2 months to 96 months (Time Known Partner).

25. Sex was coded with male as 'l' and female as '2' (Sex)

Post Hoc Variables

17. Overall stress rating for the CTT coded from the couple's mean scores was coded from the couple's
individual mean scores (Couple's Stress Rating).

18. Individual problem solving score was coded on a 7 point Likert-type scale (i.e., 1=not at all
successful to 7=very successful) (Individual Problem Solving).

l9. Individual's stress rating for the CIT was coded on a 7 point Likert-type scale (i.e., l=not at all
stressful to 7=very stressful) (Individual Stress Rating).

20. Male's individual problem solving score was coded on a 7 point Likert-type scale (i.e., 1=not at all
successful to 7=very successful) (Male Problem Solving).

21. Female's individual problem solving score was coded on a 7 point Likert-type scale (i.e., l=not at
all successful to 7=very successful) (Female Problem Solving).

22. Male's individual stress rating for the CIT was coded on a 7 point Likert-type scale (i.e., 1=not at
all stressful to 7=very stressful) (Male Stress Rating).

23. Female's individual stress rating for the CTT was coded on a 7 point Likert-type scale (i.e., 1=not
at all stressful to 7=very stressful) (Female Stress Rating).

24. Couple's combined mutuality score was coded from the couple's individual mutuality mean scores

(Combined Mutuality).

Table 7. (continued)

 

 

Variable N than SD Range
1. Mutuality 80 9.63 2.73 14.75
2. Permeability 80 15.39 4.52 20.40
3. Rater's Problan

Solving 40 4 .59 . 99 5 . 00
4 . Couple Problan

Solving 40 6 . 36 . 88 . 88
5. Separateness 80 2.43 1.48 6.75
6. ranininity 80 4.91 .59 3.05
7. Length Dating 80 25.55 21.79 83.00
8. Masculinity 80 5.21 .77 4.35
9. Parent Marriage 80 1.69 1.03
10. Quality Parent's

Marriage 80 3.46 1.27
11 . Relationship

Satisfaction 80 4 . 46 . 73 2 . 00
12. Self-Assertion 80 6.88 4.88 21.00
13. Relationship

Seriousness 80 4 . 54 . 69 2 . 00
14. Avoidant/

Security 80 24 . 89 8 . 18 39 . 00
15. Anxiety 80 14.30 5.06 28.00
16. Time Known

Partner 80 33.26 25.61 94.00
17. Couple's Stress

Rating 40 1.35 .50 .50
18. Individual Problem

Solving 80 6.36 .96 5.00
19 . Individual Stress

Rating 80 1.35 .60 3.00
20. Male Problen

Solving 40 6.38 .84 3.00
21 . Fanale Problan

Solving 40 6 . 35 1 . 08 5 . 00
22. Male Stress

Rating 40 1.40 .67 3.00
23. Fanale Stress

Rating 40 1.30 .52 2.00
24. Carbined

Mutuality 40 9 . 63 1 . 89 10 . 25

 

73
avoidant/security subscale' (p.22, p<.05). This finding suggested that the more avoidant

a participant is, the higher his or her score is on self -assertion.

The CIT variable, 'mutuality ' did not correlate with any expected variables (See
Table 7). In particular, there was no correlation between 'Simpson's avoidant/security'
and 'mutuality', or between 'femininity' and 'mutuality' scores. 'Mutuality' did correlate
in a moderate direction with 'sex' (r=-.30, p<.05), indicating that men reported higher
mutuality scores.
Oveyrjall Problem Solving Scores for Raters and Couples

'Rater's overall problem solving score' correlated in a moderate positive direction
with 'Iength of time in a dating relationship‘ (1:32, p<.05) suggesting that the increased
length of time that a couple dated, the higher the rater's indicated the couple's effectiveness
in completing the CIT. There was a moderate positive correlation between 'overall couple
problem solving score' and 'femininity' (p.36, p<.05). This relationship indicated that
the higher participants scored on the femininity scales, the higher the couple rated their
successfulness in completing the CIT. 'Overall couples' problem solving' moderately
correlated with both 'seriousness' (_r=.38, p<.05) and 'satisfaction’ (r=.42, p<.01). These
relationships indicated that couples who rated their overall problem solving on the CIT as
more effective also exhibited higher scores on seriousness and satisfaction with their
relationship.
Masculinity and Femipinity

The predictor variables from the Bem inventory, 'masculinity' and 'femininity'
were negatively correlated as expected (5:24, p<.05). A substantial negative correlation
was found between 'femininity' and 'Simpson's avoidant/secure scale'(r_=-.52, p<.01).
This indicated that as one's score increased on the avoidant/secure scale (more avoidant),
their score decreased on the femininity scale. 'Sex' was moderately correlated with both

'masculinity' (F-.28, p<.05) and 'femininity' (1:41, p<.01). These relationships

74
indicated that the males in the sample obtained generally higher masculinity scores whereas

females obtained higher femininity scores. Both were expected relationships.
Demographic/Background Vamfliles

Several demographic/ background variables were intercorrelated. There was a
moderate negative correlation between 'relationship satisfaction' and 'Simpson's
avoidant/secure scale' (r=-.41, p<.01). This indicated that the higher a participant scored
on the avoidant/secure scale (more avoidant), the less satisfied they reported to be in their
dating relationship. There was a modest positive correlation between 'relationship
satisfaction' and 'Iength of dating relationship' (LL-.24, p<.05). This indicated that the
longer a couple dated, the more satisfied they were with their relationship. Finally, there
was a moderate positive correlation between 'relationship satisfaction' and 'relationship
seriousness' (_r=.61, p<.01). This indicated that the more an individual was satisfied in
their dating relationship, the more an individual was serious about the relationship.

Relationship seriousness correlated with two additional variables. There was a
moderate positive correlation between 'relationship seriousness' and 'Iength of dating
relationship' (r=.41, p<.01). This indicated that the longer a couple dated, the more
serious they were about their relationship.

A high positive correlation was obtained between 'Iength of dating relationship' and
'Iength of time a couple knew each other' (588, p<.01). This finding indicated that the
longer a couple knew each other, the longer they reported to be dating. Finally, as might
be expected, there was a significant negative correlation existed 'quality of parent's
marriage' and 'parent's marital status' (r_=-.59, p<.01), indicating that participants from
intact family households (i.e., both parents married and living together) rated their parents'

marriage more positively than did students from divorced family backgrounds.

75
Research H theses

Relationship Among Attachment Styles, Sex, Gender Role and Problem Solving

Communication.

The first four hypotheses addressed how the individual participants' attachment
styles affected their problem solving communication and explored how sex and gender role
influenced this relationship. To restate hypotheses 1, 2, 3, and 4:

Hypothesis 1. Controlling for gender role variation, there will be a significant
relationship between individual attachment style and problem solving communication.
Hypothesis La - Secure and dismissive individuals will exhibit higher scores on self-
assertion than preoccupied or fearful individuals. Hypothesis 1b - Secure and preoccupied
individuals will exhibit higher scores on permeability than dismissive or fearful individuals.
Hypothesis 1c - Secure individuals will exhibit the highest score on mutuality. Hymthesis
1_c1 - Dismissive individuals will exhibit the highest scores on separateness.

Hymthesis 2. When gender role differences are controlled, sex will not be
significantly related to problem solving communication.

Hypothesis 3. There will be a significant relationship between gender role and
problem solving communication. Hypothesis 3a - Individuals scoring higher on the
masculinity scale will exhibit higher scores on self -assertion and separateness.

Hymthesis 3b - Individuals scoring higher on the femininity scales will score higher on
permeability and mutuality.

Hypothesis 4. Attachment style and gender role will interact to affect problem
solving behavior. Hypothesis 4a - Secure and dismissive individuals with high masculinity
scores will score higher on self -assertion and separateness than secure and dismissive
individuals with a high femininity score. Iiypothesis 4b - Secure and preoccupied
individuals with high femininity scores will score higher on permeability than dismissive or
fearful individuals with a high masculinity score. Dismissive and fearful individuals with a

higher masculinity score will score the lowest on permeability. Hypothesis 4c - Secure

76
individuals with high femininity scores will score higher on mutuality than secure

individuals with a high masculinity score . Hypothesis 4d — Disrrrissive individuals with
high masculinity scores will score higher on separateness than dismissive individuals with
high femininity scores.
Relationship Between Couple's Conioint Attachment Style and Problem Solving Behavior

Hypothesis five addressed how the couple's conjoint attachment style affected their
problem solving effectiveness score. To restate: Hypothesis 5. There will be a significant
relationship between conjoint attachment style and how participants' rate their problem
solving effectiveness. Hypothesis 5a - The secure/secure couples will receive the highest
score on couple's problem solving effectiveness. Hypothesis 5b - The secure/secure
couples will receive the highest score on rater's problem solving effectiveness.

Data Analyses for the Main Hymtheses

Hypotheses 1-3

To test the first three hypotheses which examined individual participant differences,
four ANCOVAs were conducted. Due to low male preoccupied cell sizes and low female
dismissive and preoccupied cell sizes, the four groups were collapsed to two groups,
secure and insecure. This resulted in an inability to analyze the following sub—hypotheses:
1a, 1b, and 1d. Prior to conducting the four ANCOVAs, a correlation matrix was
conducted. Relationship history variables that significantly correlated with CIT variables
were analyzed. The length of the couple's dating relationship was the only variable that
correlated with any of the four CIT variables (separateness). For the ANCOVA with
separateness as the dependent variable, length of dating relationship was added as a second
covariate.

As reported in Tables 8 and 9, hypotheses 1, 2, and 3 were not supported. There
were no significant CIT differences due to attachment style, even when gender role was

controlled. Sex was, however, significant in predicting mutuality EU, 74) = 4.53, p<.05

77
Table 8. Summary of the Analysis of Covariance Testing Hypotheses 1-3

 

Sig
Source of Variation SS DE MS E of E

 

Mutuality by Attachment Style and Sex with Ben Gender Role as the Covariate

FENININITY 7.30 1 7.30 1.04 .31

MASCULINITY 14.55 1 14.55 2.07 .15

ATTACHMENT STYLE 4.31 1 4.31 .61 .44

SEX 31.81 1 31.81 4.53 .04*
ATTACHMENT STYLE BY SEX 1.67 1 1.67 .24 .63

Residual 519.74 74 7.02

 

Permeability by Attachment Style and Sex with Ben Gender Role as the Covariate

FENININITY 100.29 1 100.29 5.12 .03*
NASCULINITY 4.89 1 4.89 .25 .62

ATTACHMENT STYLE 49.19 1 49.19 2.51 .12

SEX 10.46 1 10.46 .53 .47

ATTACHMENT STYLE BY SEX 3.35 1 3.35 .17 .68

Residual 1449.36 74 19.59

 

Self-Assertion by Attachment Style and Sex with Ben Gender Role as the Covariate

FENININITY .36 1 .36 .01 .91
MASCULINITY 19.10 1 19.10 .77 .38
ATTACHMENT STYLE 5.71 1 5.71 .23 .63
SEX .41 1 .41 .02 .90
ATTACHMENT STYLE BY SEX 13.79 1 13.79 .55 .46
Residual 1840.97 74 24.88

 

Separateness by Attachment Style and Sex with Ben Gender Role and length of Dating
Relationship as the Covariates .

 

EEMIRINITY . 12 1 . 12 . 06 . 81

MASCULINIT‘! 2.19 1 2.19 1.08 .30

m or DATING RELATIONSHIP 9.23 1 9.23 4.53 .04*
mm STYLE 6.78 1 6.78 3.33 .07

SEX 1.90 1 1.903 .93 .34

ATTACHMENT STYLE BY SEX 1.94 1 1.94 .95 .33

Residual 148 . 65 73 2 . 04

Note. n-eo

*p_<.05.

78
Table 9. CIT Subscale Means by Attachment Style Group (Secure/Insecure) and Sex

 

 

 

 

 

 

 

 

Attachment Style than SD n
Mutuality
Secure 9 . 69 2 . 7 49
Insecure 9 .55 2 .9 31
Permeability
Secure 15 . 12 4 .5 49
Insecure 15 . 81 4 . 6 31
Self—Assertion
Secure 6 . 61 5 . 0 49
Insecure 7 . 29 4 . 7 31
Separateness
Secure 2 . 21 1 .3 49
Insecure 2 . 77 1 . 7 31
Sex than SD n
Mutuality
Males 10 .46 2 .8 40
Eanales 8 . 81 2 .4 40
Permeability
Males 14 . 64 4 . 3 40
lees 16 . 14 4 . 6 40
Self-Assertion
Males 7 . 10 . 79 40
Fanales 6 . 65 . 69 40
Separateness
Males 2 . 33 1 . 4 40
Eanales 2 . 52 1 . 6 40
Attachment Style Secure Insecure
Sit m .52 .9. than; 3.? a
Mutuality
Males 10.53 3.2 22 10.36 2.4 18
Fanales 9.00 1.9 27 8.42 3.2 13
Pennsability
Males 14.39 4.6 22 14.94 4.1 18
Renales 15 . 72 4 .4 27 17 . 02 5 . 1 13
Self-Assertion
Males 7.23 5.4 22 6.94 4.6 18
Fanales 6.11 4.8 27 7.77 4.9 13
Separateness
Males 1.95 1.2 22 2.79 1.5 18
Rules 2.42 1.4 27 2.73 1.9 13

 

79
with males exhibiting higher scores than females. In addition, femininity scores

significantly predicted permeability scores E1, 74 ) = 5.121, p<.05. Including length of
date as a second covariate to the ANCOVA for separateness (CIT dependent variable) did
not result in any significant effects. Specifically, there were no significant differences on
the separateness scale due to attachment style or sex when gender role and length of dating
relationship were controlled. The effect for attachment security, however, approached
significance 131, 73) = 3.33, p<.07 with insecure participants reporting higher CIT
separateness scores (mean=2.77) than did secure subjects (mean=2.21). Despite the lack
of significant results, several means were in the hypothesized direction. As expected,
secure participants scored higher on the mutuality scale than did insecure participants.
Women scored higher than men on the permeability scale. Men scored higher on the self -
assertion scale than did women. Additional hypothesized relationships could not be
examined due to collapsed attachment style categories.
Hymthesis 4

To analyze the relationship regarding gender role and attachment style interaction, a
hierarchical multiple regression analysis was used to assess whether Bern masculinity and
femininity scores moderated attachment style-CIT relationships (See Table 10). Due to low
individual attachment cell size, attachment style categories were again collapsed, and only
secure and insecure groups were compared. This resulted in an inability to analyze the
following sub-hypotheses: 4a, 4b, and 4d. (Since the BRQ was used to measure
attachment styles, appropriate dummy variables were created for this regression.)

Hypothesis 4 was not supported by the hierarchical multiple regression analysis.
Specifically, gender role did not moderate any of the attachment style/CIT relationships for
any of the four problem solving communication patterns of the CIT. Additionally, there
was no significant interaction of gender role (femininity and masculinity) and attachment
style (secure and insecure). There was one significant relationship from the hierarchical

regression which was the relationship between femininity and the CIT dependent variable,

80
Table 10. Summary of the Hierarchical Multiple Regression Analyses of CIT Subscale

 

 

 

 

 

Scores
Variable Beta R qu Change in r Sig
qu of E
Mutuality
ATTACH STYLE 1.75 .03 .00 .00 .05 .82
FEMININITY -.00 .18 .03 .03 2.50 .12
HASCULINITY .18 .24 .06 .03 2.07 .15
w x m -015 024 006 .00 001 091
E x MAS “1.56 029 008 002 2011 015
Permeability
ATTACH STYLE -.88 .08 .01 .01 .44 .51
FEMININITY .37 .30 .09 .09 7.25 .01*
MASCULINITY .15 .31 .09 .00 .21 .65
m x m -069 032 010 001 e72 040
AS X HAS -.41 .32 .10 .00 .15 .70
Separateness
ATTACH STYLE -.86 .19 .03 .03 2.75 .10
FEMININITY .09 .19 .03 .00 .00 1.00
MASCULINITY .20 .22 .05 .02 1.27 .26
B x a“ -049 023 005 000 028 060
m x ms -058 024 .06 00° .28 060
Self-Assertion

ATTACH STYLE .79 .07 .05 .05 .36 .55
FEMININITY .06 .07 .05 .00 .36 .90
MASCULINITY .15 .12 .02 .01 .76 .39
m x m -034 013 002 000 012 074
AS X MAS -.54 .14 .02 .00 .23 .63

 

Note:

*The variables were entered in the following order: ATTACHMENT STYLE, FEMININITY,
MASCULINITY, INTERACTIONI (ATTACHMENT X FEMININITY) , AND INTERACTIONZ (ATTACHMENT X
nascunrnrn) .

*p<.01

N-80

81
permeability (See Figure 1). This result indicated that individuals with higher femininity

scores exhibited higher permeability scores which provided support for hypothesis 3b.
Hymthesis 5

To test Hypothesis 5, a one-way ANCOVA was conducted. Three attachment style
pair groupings were analyzed: secure/secure (pair 1), secure/insecure (pair 2), and
insecure/insecure (pair 3). As reported in Tables 11 and 12, hypothesis 5 was not
supported. There were no significant differences for any of the dependent variables due to
conjoint attachment style even when gender role was controlled. Gender role (femininity)
was significant in predicting couple's problem solving effectiveness score E(1, 35) = 6.03,
p<.05. The gender role covariate was not significant in predicting rater's problem solving
effectiveness scores or couple's combined mutuality score. Despite the lack of significant
findings, the means were in the hypothesized direction. Secure pairs scored higher than
secure/insecure and insecure pairs on all three dependent variables.

Post Hoc An_alyses

To further explore the possible bases for these findings, several post hoc analyses
were conducted. First, additional variables on how couple's viewed the CIT were added to
the matrix. Bowlby indicated that one's attachment style is activated under stress. The
couple had been asked how stressful they viewed the task and how successful they
believed they completed the task. It was hoped that this could provide further insights on
the couple's task performance. Second, hypotheses 1—3 were reanalyzed to assess if the
insecure categories had been correctly collapsed due to the initial cell size. Insecure
categories were recollapsed and subsequently reanalyzed. Third, as initially planned, post
hoc testing was conducted to determine if female and male problem solving differences
were related to different attachment styles. Fourth, since the data were collected with the
couple as pairs, it was decided that further analyses were needed to explore possible couple
differences in problem solving communication. Couple attachment style differences and

couple gender role identity were examined. Finally, the individual masculinity and

82
Figure 1: Hierarchical Regression Scatterplot

 

3O

'7

 

 

 

 

83

Table 11. Summary of the Analyses of Covariance of Couple's Combined Mutuality and
Problem-Solving Effectiveness Scores and Rater's Problem-Solving
Effectiveness Scores (Hypothesis 5)

 

 

 

 

 

Source of Variation 55 DE MS 1' gig!
Couple's Canbined Mutuality Score

FEMININITY .66 1 .66 .19 .66
MASCULINITY 10.74 1 10.74 3.13 .09
PAIR 7.57 2 3.79 1.10 .34
Residual 120.15 35 3.43

Couple's Problem Solving Effectiveness Score
FEMININITY 4.14 1 4.14 6.04 .02*
MASCULINITY 1.01 1 1.01 1.47 .23
PAIR 1.00 2 .50 .73 .49
Residual 24.01 35 .69

Rater's Problen Solving Effectiveness Score
FEMININITY .03 1 .03 .02 .88
MASCULINITY .12 1 .12 .11 .74
PAIR .86 2 .43 .40 .67
Residual 37.44 35 1.07
N-40

*p< .05

84

Table 12. Couple Attachment Style Pair Means on Couple's Combined Mutuality and
Problem Solving Effectiveness Scores and on Rater's Problem Solving
Effectiveness Scores

 

Couple ' s Attachment Style Pair Group than SD n

 

Couple ' s Canbined Mutuality Score

 

 

Secure/Secure 9.99 1.7 17

Secure/ Insecure 9.37 2.1 15

Insecure/ Insecure 9 . 38 1 . 9 8
Couple's Problen Solving Effectiveness Score

Secure/Secure 6.59 .69 17

Secure/ Insecure 6.10 1.04 15

Insecure/Insecure 6.38 .88 8
Problan Solving Effectiveness Score

Secure/Secure 4.74 1.36 17

Secure/ Insecure 4.57 .68 15

Insecure] Insecure 4 . 31 . 46 8

 

85
femininity scores of the entire sample were examined to provide further information on the

population used in this study.
Correlational Findings

To provide further clarification on how the couple viewed the CIT and how
couples' mutually approached the task, three additional variables were added to the
correlation matrix. These variables included individual problem solving score (how each
partner viewed their successful completion of the task); stressfulness of the task (how each
partner and how the couple together viewed the stressfulness of the task); and couple's
combined mutuality score (how the couple jointly demonstrated sensitivity and respect for
their partner during the problem solving discussion). Post hoc intercorrelations are
reported in Table 7. In examining the relationships among successful task completion
(problem solving score), stressfulness of the task, and couple's combined mutuality score,
there were several interesting significant relationships.

Stress and Success Scores of the CIT. Male and female problem solving scores
were substantially intercorrelated (_r,=.68, p<.01). Male and female ratings of CIT-related
stress were moderately intercorrelated (1:.38, p<.01). Individual ratings on 'successful
task completion of the CIT' and 'stress level of the CIT' were modestly correlated (F--23.
p<.05). This relationship indicated that the higher individuals rated their successfulness in
completing the CIT, the lower they rated the stressfulness of this task.

Femininity. There was a modest positive correlation between 'individual problem
solving score' and the 'femininity scale' (r_:.23, p<.05). This was similar for men (r_=.35,
p<.05) and women (p.33, p<.05). This relationship indicated that the higher men and
women rated their successfulness in completing the CIT, the higher they scored on the
femininity scale.

Relationship Satisfaction. There was a moderate positive correlation between
'successfulness of completing the CIT' and 'satisf action in the relationship' (p.25, p<.01)

for both men (p.35, p<.05) and women (5.41, p<.01). This indicated that for men and

86
women reporting higher levels of relationship satisfaction also reported greater success in

completion of the CIT. There was a moderate negative correlation between 'stress level of
the CIT' and 'satisf action in the relationship'. Individual's stress level reports on the CIT
(p: -.29, p<.01) for both men (_r=-.36, p<.05) and women (_r=-.44, p<.010) and the
'couple stress level score' (1:247, p<.01) correlated with 'relationship satisfaction in a
negative direction. This indicated that for both men and women (individually and jointly)
the higher their relationship satisfaction, the lower they rated the stress level of the CIT.

Relationship Seriousness. 'Relationship seriousness' correlated with both
'successfulness in completing the CIT' and 'stressfulness of the CIT'. Both men (_r=.36,
p<.05) and women (LL-.33, p<.05) reported that the more serious and committed they were
in their relationships, the more successful they rated their CIT performance. 'Stress level
on the CIT' for males (F'-37. p<.05) and couples (r=-.36, p<.05) correlated negatively
with 'relationship seriousness'. This indicated that for males and couples that reported
higher levels of seriousness and relationship commitment reported that they viewed the CIT
as less stressful.

Alachment Style. 'Simpson's avoidant/secure subscale' correlated negatively with
the 'individual success rating of the CIT' (F43. p<.05). This indicated that individuals
who scored in the avoidant direction reported lower levels of success in completing the
CIT. 'Simpson's anxiety subscale' correlated positively with the 'stress rating of the CIT'
(_r=.23, p<.05). This indicated that individuals who scored higher on the anxiety scale
rated the CIT as more stressful.

DemoggpLhic/Background Vafles. 'Stress and success level ratings of the CIT
correlated with several demographic/background variables. 'Success level' correlated
positively with 'Iength of the dating relationship' (5:.27, p<.05). This relationship
indicated that the longer a couple dated, the more successful they rated their performance on
the CIT. 'Female's success ratings of the CIT were positively correlated with the 'Iength

of time a couple knew each other' (1:32, p<.05). This relationship indicated that females

87
that knew their dating partner longer rated the couple's performance on the CIT as more

successful. Finally, there was a positive correlation between the 'couple's combined
mutuality score' and 'stress level of the task' (r=.34, p<.05). This relationship indicated
that the lower the couple rated the CIT as stressful, the lower the combined mutuality scale
score.
Hypotheses 1-3 Re-An_alyzed

Four post hoc ANCOVAs were conducted to analyze hypotheses 1-3 since the first
ANCOVA did not provide support Initially, dismissive, preoccupied, and fearful
categories were collapsed due to insufficient cell size. Simpson's anxiety subscale was,
however, noted to correlate with permeability (p.28, p<.01) providing support for a
relationship between attachment style and one of the CIT subscales. Theory supports the
view that both preoccupied and fearful attachment styles should score high on Simpson's
anxiety subscale. For this analysis, these two categories were collapsed to one category
and the dismissive category was eliminated due to insufficient cell size. The insecure group
(fearful and preoccupied; combined N=21) was compared to the secure group (N=49).
Sex and attachment style were the independent variables, Bem femininity and masculinity
scores were the covariates, and individual CIT scores were the dependent variables. Since
length of dating relationship had a significant correlation with separateness, it was added as
the second covariate for that dependent variable. Three relationships were analyzed
regarding individual participant relationship behavior: a) covariates and CIT scores; b) sex
and CIT scores while controlling for Bern femininity and masculinity scores and attachment
style; and c) attachment style while controlling for Bem femininity and masculinity scores
and sex.

Hypotheses 1, 2, and 3 were not supported with the post hoc ANCOVA which was
analyzed by eliminating the dismissive attachment category and collapsing the preoccupied

and fearful attachment style categories (See Tables 13 and 14). There were no significant

88
Table 13 Summary of the Analysis of Covariance Testing Hypotheses 1—3 (Revised)

 

Sig
Source of Variation 88 DE MS P of E

 

Mutuality by Attachment Style (Secure, Preoccupied] Fearful) and Sex with Ben Gender
Role as the Covariate

FEMININITY 12.60 1 12.56 1.84 .18
MASCULINITY 8.40 1 8.40 1.23 .27
ATTACHMENT STYLE 2.71 1 2.71 .40 .53
SEX 18.33 1 18.33 2.68 .11
ATTACHMENT STYLE BY SEX .18 1 .18 .03 .87
Residual 437.77 64 6.84

 

Permeability by Attachment Style (Secure, Preoccupied/ Fearful) and Sex with Ben Gender
Role as the Covariate

FEMININITY 70.53 1 70.53 3.32 .07
MASCULINITY 2.08 1 2.08 1.00 .76
ATTACHMENT STYLE 54.04 1 54.04 2.54 .12
SEX 21.30 1 21.30 1.00 .32
ATTACHMENT STYLE BY SEX 18.93 1 18.93 .89 .35
Residual 1361.53 64 21.27

 

Self-Assertion by Attachment Style (Secure, Preoccupied/Fearful) and Sex with Ben
Gender Role as the Covariate.

FEMININITY 2.54 1 2.54 .11 .75
MASCULINITY 24.28 1 24.28 1.00 .32
ATTACHMENT STYLE .01 1 .01 .00 .99
SEX 1.32 1 1.32 .05 .82
ATTACHMENT STYLE BY SEX 47.59 1 47.59 1.96 .17
Residual 1555.12 64 24.30

 

Separateness by Attachment Style (Secure, Preoccupied/ Fearful) and Sex with Ben Gender
Role and Length of Dating Relationship as the Covariates .

 

PEMININITY .02 1 .02 .01 .92

MASCULINITY .50 1 .50 .28 .60

m or DATING 8.27 1 8.27 4.65 .04*
ATTACHMENT STYLE 3.89 1 3.89 2.19 .14

SE)! .13 1 .13 .07 .79

ATTACHMENT STYLE BY SEX 6.20 1 6.20 3.48 .07

Residual 112.06 63 1.78

Note. N-80

*p<.05.

89
Table 14. CIT Subscale Means by (Revised) Attachment Style Group and Sex

 

 

 

 

 

 

 

 

Attachment Style Mean SD n
tmtuality
Secure 9 . 69 2 . 7 49
Preoccupied/ Fearful 9 . 57 2 . 7 2 1
Permeability
Secure 15 . 12 4 .5 49
Preoccupied/Fearful 16 . 47 5 . 2 21
Self-Assertion
Secure 6 . 61 5 .0 49
Preoccupied/Fearful 6 . 67 4 . 6 21
Separateness
Secure 2 .21 1 . 3 49
Preoccupied/Fearful 2 . 67 1 .5 21
Sex Mean SD n
Mutuality
Males 10 . 46 2 . 8 33
Panales 8 . 93 2 . 3 37
Permeability
Males 14 . 53 4 . 6 33
Fanales 16 . 14 4 .7 37
Self-Assertion
thles 6 . 61 4 . 9 33
Famles 6 . 65 4 . 9 37
Separateness
Males 2 . 31 1 .4 33
Fanales 2 . 38 1 .4 37
Attachment Style Secure Preoccupied/Fearful
.39.?! m §Q _n_ £39.11 Q E
Mutuality
thles 10.53 3.2 22 10.32 2.0 11
Perrrales 9.00 2.0 27 8.75 3.2 10
Permeability
Males 14.39 4.6 22 14.82 5.0 11
Eanales 15.72 4.4 27 18.28 5.1 10
Self-Assertion
Males 7.23 5.4 22 5.36 3.6 11
Pareles 6.11 4.8 27 8.10 5.3 10
Separateness
Males 1.95 1.2 22 3.02 1.5 11
ferrules 2.42 1 4 27 2.28 1.5 10

 

9O

CIT differences due to attachment style or sex even when gender role was controlled.
Length of dating relationship was correlated with separateness and was a significant
covariate E (1, 63) = 4.7, p<.05. When length of dating relationship was controlled,
however, attachment style still was not significant in predicting CIT differences. Despite
the lack of significant differences, several mean scores were in the hypothesized direction.
Secure participants scored higher than insecure participants on the mutuality scale. In
addition, women scored higher than men on the permeability scale.
Sex Differences and Problem Solving Effectiveness Scores

Post hoc analyses examined whether male and female problem solving effectiveness
scores (successfulness in completing the CIT) were significantly related to participants'
attachment styles. To test this post hoc analysis, one ANCOVA was conducted (See
Tables 1516). This included sex and attachment style as the independent variables, length
of dating as the covariate and overall problem solving effectiveness as the dependent
variable. Three relationships were analyzed regarding individual participant behavior: a)
covariates and CIT problem solving effectiveness ; b) sex and problem solving
effectiveness while controlling for length of dating relationship and attachment style; and c)
attachment style while controlling for length of dating relationship and sex. Due to low
insecure cell sizes, the four groups were collapsed to two groups: secure and insecure.

This post hoc analysis was not supported. There were no problem solving
effectiveness differences due to attachment style or sex. Length of dating relationship
correlated with problem solving effectiveness and was a significant covariate E (1, 75) =
6.06, p<.02. When this demographic variable was controlled, attachment style and sex
were not significant predictors of problem-solving effectiveness. Despite the lack of
significant differences, mean scores were in the expected direction with secure participants

scoring higher on problem solving effectiveness than insecure participants.

91

Table 15. Analysis of Covariance of Overall Problem Solving Effectiveness by Attachment
Style Group and Sex with length of Dating Relationship as the Covariate

 

 

 

Source of Variation SS DE MS 1' <83ng
m or DATING 5.31 1 5.31 6.07 .02*
SEX .05 1 .05 .05 .82
ATTACHMENT STYLE 1.57 1 1.57 1.79 .19
sax BY ATTACHMENT STYLE .06 1 .06 .07 .79
Residual 65.55 75 .87

Note. N=80

*p<.05.

92

Table 16. Means of Overall Problem Solving Effectiveness by Attachment Style Group

and Sex with Length of Dating Relationship as the Covariate

 

 

 

AMt Style than S_D 3;

Secure 6.45 98 49

Insecure 6.23 .92 21

a _@ S_D 1.1.

Males 6.38 84 40

Rules 6.35 1.20 40

Attachment Style Secure Insecure

& fl 3.3 9. & S_D n
Males 6.50 .80 22 6.22 .88 18
Panales 6.41 1.12 27 6.23 1.01 13

 

93
Couple Analyses

Gender Role of the Couple. The previous data analyses examined individual
differences on the CIT. Since the data on the CIT were collected with the couple as pairs,
further analyses were required to examine couple differences on the CIT. In addition, the
previous data analyses examined gender role as an independent variable for each individual.
The conjoint contribution of gender role differences for the couple was overlooked.
Finally, the CIT was examined further. Previous results from this study indicated a lack of
individual differences on the CIT. The CIT is composed of four categories: mutuality,
permeability, self -assertion, and separation. Each category is composed of individual
responses that are related to each category (See Appendix I). The question that remained
unanswered was whether differences occuning on the individual responses were canceling
each other out when the average CIT category score was examined. To address this
concern, individual statements was examined as a within-group factor to explore this issue
further.

A series of four univariate tests on the CIT measures were used for this post hoc
test. For these analyses, the within- group f actors were sex of partner and individual
responses for each of the CIT categories. The between-group factors were pair (attachment
style of couple) and gender role of the pair. The dependent variables were the four CIT
scores (mutuality, permeability, self -assertion, and separateness).

As reported in Tables 17-19, several significant results were found. First, relative
to males in the sample, females exhibited a higher number of mutuality responses. Second,
significant sex differences were observed with regard to types of individual statements
made by participants during the CIT. Scheffe post hoc tests revealed five significant pair—
wise group differences. Based on mean scores (See Tables 18 and 19), significant group
differences supported that females exhibited a higher number of answer and indirect
statements than initiation and statement of other's feelings. In addition, the interaction

effect of sex by individual statements on mutuality scores approached significance (p < .06)

Table 17. Summary of the Univariate Analyses of CIT Subscale Scores by Attachment

Style and Gender Role Group of the Couple

 

 

 

819

Source of Variation 88 or t8 1' of 1"
Mutuality
Sex 179.17 1 179.17 5.74 .02*
Individual Staturents 10798.86 3 3599.62 1.15 .00**
Attachnent Style of Pair 51.67 2 25.84 .83 .43
Gender Role of the Pair 4.26 1 4.26 .14 .71
Sex By Individual Statanents 239.41 3 79.80 2.56 .06
Pair By Sex 50.21 2 25.10 .80 .45
Gender Variation By Sex 17.49 1 17.49 .56 .46
Pair By Individual Statements 178.63 6 29.77 .95 .46
Gender Variation By Ind State 14.57 3 4.86 .16 .93
Pair by Garder Variation 41.43 2 20.72 .66 .52
Pair By Sex By Ind Statements 73.79 6 12.30 .39 .88
Gender Var By Sex By Ind Statements 44.82 3 14.94 .48 .70
Pair By Gender Variation By Sex 56.13 2 28.07 .90 .41
Pair By Gender Var By Ind Statements 88.66 6 14.76 .47 .83
Pair By Gen Var By Sex By Ind State 176.75 6 29.46 .94 .47
Residual 8496.59 272 31.24
Permeability

Sex 17.52 1 17.52 .22 .64
Individual Statenents 15259.54 4 3814.89 4.70 .00**
Attachment Style of Pair 121.76 2 60.88 .75 .47
Gender Role of the Pair 38.52 1 38.52 .47 .49
Sex By Individual Statanents 946.06 4 236.51 2.91 .02*
Pair By Sex 117.88 2 58.94 .73 .49
Gender Variation By Sex 66.93 1 66.93 .82 .37
Pair By Individual Statanents 476.61 8 59.58 .73 .66
Gender Variation By Ind State 191.95 4 47.99 .59 .67
Pair by Gender Variation 127.47 2 63.74 .78 .46
Pair By Sex By Ind Statanents 484.47 8 60.56 .75 .65
Gender Var By Sex By Ind Statanents 288.31 4 72.08 .89 .47
Pair By Gender Variation By Sex 103.47 2 51.73 .64 .53
Pair By Gender Var By Ind State 277.15 8 34.64 .43 .91
Pair By Gen Var By Sex By Ind State 790.03 8 98.75 1.22 .29
Residual 27610.72 340 81.21

 

Table 17. (continued)

 

 

 

819

Source of variation SS DP MS P of P
Selqussertion
Sex .02 1 .02 .00 .99
Attachment Style of Pair 138.85 2 69.43 2.83 .07
Gender Role of the Pair 13.93 1 13.93 .57 .45
Pair By Sex 13.11 2 6.56 .27 .77
Gender variation By Sex 3.90 1 3.90 .16 .69
Pair by Gender Variation 78.19 2 39.09 1.59 .21
Pair By Gender variation By Sex 3.79 2 1.90 .08 .93
Residual 1669 . 32 68 25 . 55
***Self-Assertion had only one individual statanent.
Separateness

Sex 1.67 1 1.67 .19 .66
Individual StatsltBnts 27.03 3 9.01 1.02 .38
Attachment Style of Pair 7.43 2 3.72 .42 .66
Gender Role of the Pair 5.02 1 5.02 .57 .45
Sex By Individual Statanents 7.85 3 2.62 .30 .83
Pair By Sex 3.80 2 1.90 .22 .81
Garder Variation By Sex 9.76 1 9.76 1.11 .29
Pair By Individual Statanents 29.20 6 4.87 .55 .77
Gender variation By Ind State 36.90 3 12.30 1.40 .24
Pair by Gender variation 2.73 2 1.37 .16 .86
Pair By Sex By Ind Statements 12.05 6 2.01 .23 .97
Gender var By Sex By Ind Statements 8.39 3 2.80 .32 .81
Pair By Gender variation By Sex 4.90 2 2.45 .28 .76
Pair By Gender var By Ind Statements 30.42 6 5.07 .58 .75
Pair By Gen var By Sex By Ind State 23.52 6 3.92 .45 .85
Residual 2394.75 272 8.80

 

Note: N840 Pairs

N880 Individuals
*p < .05
**p < .01

96
Table 18. Means of CIT Subscale Scores by Attachment Style and Gender Role Group of

 

 

 

 

the Couple

Gmder Role Mean so count
Mutuality

Panininity 10.0 1 . 19 88

Masculinity 9 . 5 11 . 40 232
Permeability

Panininity 1.54 1.18 110

Masculinity 1 . 54 1 . 44 290
Self—Assertion

Panininity 7.23 5 .83 22

Masculinity 6 . 74 4 .52 58
Separateness

Femininity 2 .05 1 .88 88

Masculinity 2 . 57 3 . 19 232
Pair Mean SD count
Mutuality

Secure-Secure 9.99 1.18 136

Secure-Insecure 9 . 37 1 . 17 120

Insecure-Insecure 9 . 38 1 . 07 64
Permeability

Secure-Secure 1 . 54 1 . 21 170

Secure-Insecure 1 . 50 14 . 52 150

Insecure-Insecure 16 . 01 1 . 53 80
Self-Assertion

Secure-Secure 7.32 5 .53 34

Secure-Insecure 5 .60 3 .70 30

Insecure-Insecure 8 . 31 5 . 08 16
Separateness

Secure-Secure 2 . 20 2 . 35 136

Secure-Insecure 2 . 33 3 . 28 120

Insecure-Insecure 3 . 08 3 . 11 64

 

97
Table 19. Means of CIT Subscale Scores by Individual Statement and Sex

 

 

 

 

 

 

Individual Statanent Mean SD count
Mutuality
Initiation . 3 1 . 59 80
Statement .46 1.43 80
Answers 13.40 6.93 80
Indirect 24.36 8.80 80
Permeability
Agreement 11.08 5.47 80
Acknowledgment 11 . 86 5 . 67 80
Request 21.38 1.13 80
Relevant 31.69 14.48 80
Canpliance . 95 1 . 38 80
Self-Assertion*
Separateness
Direct 2.23 2.15 80
Indirect 2.46 2.31 80
Request 2.61 2.38 80
Irrelevant 2 . 40 4 .25 80
Sex Mean SD count
Mutuality
Hales 8.81 1.05 160
Fanales 10 . 46 1 . 25 160
Permeability
Males 16 . 14 1 . 39 200
renales 14 . 64 1 . 35 200
Self-Assertion
Males 6.65 4.83 40
Pennies 7 . 10 4 . 98 40
Separateness
Males 2.52 3.04 160
Panales 2 . 33 2 . 74 160
Sex by Individual Statements
Mean SD count
Mutuality
Fanale Initiation .30 .65 40
Female Statement .25 .59 40
Female Answers 15.60 7.68 40
Panels Indirect 25.68 9.80 40
Male Initiation .33 .53 40
Male Statanent .68 1.93 40
Male Answers 11.20 5.33 40
Male Indirect 23.05 7.58 40
Permeability
female Agreenent 10.35 4.73 40
Panels Acknowledgmentlo.08 4.75 40
Fanale Request 19.38 9.43 40
Fanale Relevant 32.45 1.44 40
runale Caupliance .95 1.52 40
Hale Agreanent 11.80 6.09 40
Male Acknowledgmnt 13.65 5.99 40
Male Request 23.38 1.26 40
Male Relevant 30.93 1.48 40
Male Ccnpliance .95 1.24 40

 

*No Individual Statuaents for this Category.

98
with mean scores indicating that females made more answer and indirect statements than

did males (See Figure 2). In examining the specific types of responses included in the
mutuality category (initiation of compromise, statement of other's feelings, indirect
suggestion, and answers request for information and validation), females were more likely
than males to make more statements which involved responding to their partner's requests
for information and validation (answer statements) and indirectly introducing solutions to
tasks (indirect statements).

For the dependent variable penneability, there were two significant results. First,
the effect for individual statements was significant which indicated differences among the
five individual statements. Scheffe post hoc tests revealed nine significant pair-wise group
differences. Individuals made more request and relevant responses than agreement,
acknowledgment and compliance statements. Compliance statements had the lowest overall
observed frequency score. Second, the interaction effect of sex by individual statements on
permeability scores was significant which indicated sex differences among the five
individual statements. In examining the specific types of responses included in the
permeability category (agreement, acknowledgment, requesting information, providing
relevant comments, and complying with requests for action) males made more request
statements (seeking information, clarification and confirmation of opinions or decisions
from the partner to then accomplish the task) than females and females made more relevant
comments (factual statements) than males (See Figure 3).

For the dependent variable self -assertion, there were no significant between or
within group effects observed. Attachment style of pair, however, approached
significance. Insecure-insecure couples scored higher on self -assertion than did insecure-
secure and secure-secure pairs. The largest difference was between the insecure-
insecureand the insecure-secure pair. For the dependent variable separateness, there were
no significant between or within group differences. Despite the lack of significant

attachment style differences for couples, means were in the hypothesized

99

Figure 2. Mutuality: Sex by Individual Statements

CalflthyShtanmb

 

4o
35 -.
30 -
25 —-

20q

 

I Female
B Male

T—I—

Initiation

in;

Statement Answers

madam-.111“:

 

 

 

Indirect

100

Figure 3. Permeability: Sex by Individual Statements

 

 

 

 

101
direction. Secure-secure couples scored the highest on mutuality and insecure-insecure

couples scored the highest on separateness and self -assertion.

Connectedness/Individuality (Semteness). As previously discussed, women's
development theory (Chodorow, 1978; Gilligan, 1982) suggests that in an intimate
relationship, men define themselves through experiences of separateness in relationship
with others whereas women define themselves through relational connectedness. In
Grotevant and Cooper's (1985) study, connectedness was defined as mutuality and
permeability and individuality was defined as self -assertion and separateness. In
examining the CIT further, the categories of mutuality and permeability are similar to the
concept of connectedness as defined by women's development theory. Mutuality includes
sensitivity and respect for another's views and permeability includes an expression of
responsiveness to the view of others. Both relate to an attention to the
partner and maintenance of the relationship. The categories of self -assertion and
separateness are similar to the concept of separateness (or as Grotevant and Cooper (1985)
defined it as individuality). Self —assertion reflects an awareness of one's own point of
view and separateness expresses a distinctiveness of self from others. Both relate to an
attention to self and not the partner or the relationship. Based on these similarities, it was
decided to change the CIT from four categories to two: connectedness and individuality
(separateness) while examining possible differences related to attachment style of the
couple, gender role of the couple and sex.

A series of two univariate tests on the CIT measures were used for this post hoc
test. The within-group factors was sex within couple. The between-group factors were
attachment style of the pair and gender role of the pair (combined individual gender role
scores for the couple indicating that the couple was either more masculine or more feminine
as a couple). The dependent variables were connectedness and individuality
(separateness). The results from the MANOVAs are reported in Table 20 and 21. One

significant result was found. Based on mean scores, females scored higher on the

102

Table 20. Summary of the Univariate Analyses for Connectedness and Individuality
(Separateness) Scores

 

 

 

319

Source of Variation SS DP MS F of P
Connectedness
Sex 1251.54 1 1251.54 58.10 .00*
Attachment Style of Pair 25.34 2 12.67 .59 .56
Gender Role of Pair 2.28 1 2.28 .11 .75
Pair By Sex 4.12 2 2.06 .10 .91
Gender Role of Pair By Sex 11.92 1 11.92 .55 .46
Pair By Gender Role of Pair 26.99 2 13.49 .63 .54
Pair By Gen Role of Pair By Sex 21.18 2 10.59 .49 .61
Residual 1464 .77 68 21 .54
Individuality (Separateness)

Sex .24 1 .24 .10 .75
Attachment Style of Pair 7.73 2 3.86 1.64 .20
Gender Role of Pair 2.70 1 2.70 1.14 .29
Pair By Sex .26 2 .13 .05 .95
Gender Role of Pair By Sex .73 1 .73 .31 .58
Pair By Gender Role of Pair 1.55 2 .77 .33 .72
Pair By Gen Role of Pair By Sex 1.62 2 .81 .34 .71
Residual 160.56 68 2.36

 

NI40 Couples
8-80 Individuals
*9 < .01

103

Table 21. Means of Connectedness and Individuality (Separateness) Scores by Attachment
Style of the Couple, Gender Role of the Couple and Sex

 

 

 

 

 

 

Gender Role than SD count
Connectedness

ranininity 1 . 93 7 . 52 22

Masculinity 1 . 89 7 . 64 58
Individuality

Femininity 3 . 08 1 . 66 22

Masculinity 3 . 40 1 . 46 58
Pair than SD count
Connectedness

Secure-Secure 1 . 92 7 . 55 34

Secure-Insecure 1 . 87 8 . 22 30

Insecure-Insecure 1 . 90 6 . 70 16
Individuality

Secure-Secure 3 . 22 1 . 44 34

Secure-Insecure 2 . 99 1 . 50 30

Insecure-Insecure 4 . 13 1 . 49 16
Sex than SD count
Connectedness

Males 1 . 29 2 . 94 40

Fanales 25 . 10 5 . 55 40
Individuality

Males 3 . 35 1 . 61 40

Fanales 3.29 1.42 40

 

104
connectedness measure. This finding is consistent with the women's development theory

which presumes that relative to males, females are more focused on connectedness in an
intimate relationship. Males and females exhibited near equal mean scores on individuality
(separation) (See Table 21). Neither between-group factor (attachment style of pair and
gender role of the pair) significantly predicted connectedness or individuality (separateness)
scores. Despite the lack of significant attachment style differences for couples, means were
in the expected direction. The secure-secure couple scored the highest on connectedness
and the insecure-insecure couple scored the highest on individuality (separateness).
Individual Gender Roles of the Sample

One possibility for the study's unsupported research hypotheses was that the
sample differed from the larger population in some significant way. It was hypothesized
that a college age sample of males in longer term relationships may exhibit higher
femininity scores than masculinity scores. If this was supported, it may provide insight
into the lack of CIT individual differences. More specifically, males would be attending
more to the relationship and their partner and not exhibiting stereotypical masculine
relationship differences.

An ANOVA was used for this post hoc test. The within-group factors were sex
and gender role categories of the individual participants category (masculinity and
femininity). The individual participant category examined the relative strength of
masculinity/femininity within the couple. The dependent variable was relative variation of
masculinity and femininity of the group. The results are presented in Tables 22 and 23.
There were three significant results. First, there was a significant effect due to individual
gender role category which indicated that the entire sample was more masculine than
feminine. Second, there was an interaction effect of sex and individual gender role
categories (See Figure 4). Post hoc comparisons indicated three significant pairwise group
differences: a) females had higher masculinity scores than males; b) males had higher

femininity scores than females; and c) females' masculinity scores were higher than their

105
femininity scores. Within males, there were no gender role differences which indicated that

males were not predominantly more masculine or feminine.

106

Table 22. Summary of the Analysis of Variance of the Relative Variation of Gender Role
of the Group (Sample) by Gender Role of the Individual and Sex

 

 

Source of variation SS DP MS I gig!
Sex .02 1 .02 .05 .82
Individual Categories 3.48 1 3.48 8.30 .01*
Sex By Ind Categories 8.19 1 8.19 1.95 .00*
Residual 65.41 156 .42

u-so

 

*2 < .01

107
Table 23. Means of Relative Variation of Gender Role of the Group (Sample) by Gender

 

 

 

 

 

 

Role of the Individual and Sex
Individual Categories Mean SD n
Masculine 5.21 .77 80
Peninine 4.91 .59 80
Sex than S) n
Male 5.07 .62 80
Panale 5.05 .77 80
Sex By Ind Categories than SD 11
Male, Masculine 5.00 .69 40
Male, Paninine 5.15 .53 40
Panale, Masculine 5.42 .79 40

Female, Faninine 4 . 68 . 55 40

 

108
Figure 4. Individual Gender Role of Sample: Sex by Individual Categories

 

8
J I Female
71 [2 Male

GalleIRoleSme
A

  

 

 

 

 

 

/

/A

 

 

Masculine Feminine

CHAPTER 5

Discussion

This present study attempted to expand research on couple functioning in close
relationships by examining predisposing factors in couple's problem-solving
communication. Within the framework of Bowlby's attachment theory, this study
examined the influence of adult attachment styles on problem solving communication in
dating couples. This study used the four category attachment style typology (Bartholomew
& Horowitz, 1991) while concurrently examining the influence of sex and gender role on
problem-solving communication. In particular, this study (a) examined the relationship of
participants' attachment styles to their in vivo problem solving communication; (b) explored
the extent to which sex and gender role affected the relationship of attachment style and
problem solving; and (c) identified couple's conjoint attachment style-related contributions
to their problem solving communication.

Overall, the main hypotheses of this study were not statistically supported.
However, the directionalities of correlational results and group means relevant to several
hypotheses were in the expected direction. Post hoc testing provided support for the
influence of sex and gender role on the various CIT scales. In addition, the results
provided reliability and validity support for the attachment style instruments used for this
study. The BRQ demonstrated moderate attachment stability over a one month period.
Results were comparable to other samples (Baldwin & Fehr, 1992; Pistole, 1989; Scharfe
& Bartholomew, 1994a; Scharfe & Bartholomew, 1994b). The S-AAS exhibited high test-
retest scores over a one month period. Furthermore, the results of this study provided
convergent validity for the BRQ, a self -report categorical attachment style inventory.

This chapter will discuss the observed interrelationships of individual attachment

styles, sex and gender role, and couple's conjoint attachment styles on their problem
109

1 10
solving communication and behavior. First, observed relationships between individual

participants' attachment style and their problem solving communication will be discussed,
then, the observed relationships among sex, gender role, and attachment style and their
problem solving communication relationship will be discussed, followed by observed
relationships between couple's conjoint attachment styles and their problem solving
behavior. A discussion of the limitations of the study, implications for practice and
recommendations for future research will follow.

Individual Attachment Style and Problem Solving Communication.

By introducing an in vivo method for observing spontaneous couple interaction,
this study attempted to examine whether individual attachment styles contributed to
different styles of problem solving communication in dating couples. Patterns in the
observed variable relationships from the correlational matrix suggested that participants
with an anxious attachment orientation exhibited higher permeability scores (hypothesis 1b)
indicating that they responded more to their partner during the problem solving exercise
than would participants with an avoidant attachment style orientation. Based on attachment
theory, individuals with a preoccupied or anxious style of attachment tend to focus on
relationship maintenance due to an anxious fear of loss of security. They are described as
overly concerned with their relationships and as tending to idealize their partner and oblige
their partner's wishes (Bartholomew & Horowitz, 1991; Feeney & Noller, 1991; Pistole,
1989). In addition, participants with an anxious attachment orientation perceived the CIT
as more stressful. This is not surprising in that they already approach the relationship with
higher levels of anxiety than do persons with other attachment styles.

The correlational results further suggested that participants scoring higher on the
avoidant attachment index exhibited higher scores on the self -assertion scale (hypothesis
1a) indicating that they focused on their own point of view and exhibited more direct
suggestions as compared to the participant with a secure or avoidant attachment orientation.

Attachment theory indicates that individuals with a dismissive or avoidant attachment style

l 11
expect their partner to be unavailable and non-responsive. They reportedly focus on

meeting their own needs instead of focusing on their partner or on the relationship. In
addition, participants with an avoidant attachment orientation reported lower levels of
success in completing the CIT. As expected, the patterns in the observed variable
relationships suggested that they also reported the lowest degree of relationship satisfaction
and seriousness.

An unexpected pattern was that individuals scoring higher on the permeability scale
also scored higher on the separateness scale. Participants expressing a responsiveness to
the views of others also expressed a distinctiveness of self from others. Therefore,
participants appeared to be able to focus on attachment and separation during the problem
solving task This same result was noted in the original completion of the Family
Intervention Task (Grotevant & Cooper, 1984). In contrast, couples that dated for a longer
period of time exhibited a pattem of lower CIT separateness scores and higher relationship
satisfaction and relationship seriousness scores. These individuals made fewer distinctions
between themselves and others in the problem solving communication patterns.

Despite the significant results from the correlational matrix, further analysis
(analysis of covariance) did not provide additional support for the remaining hypotheses on
the relationship between individual attachment style and problem solving communication
even when gender role was controlled. Adding length of dating relationship as an
additional covariate in the analysis of separateness did not clarify this relationship further.
The CIT separateness scores, however, did register attachment style effects that approached
significance, with insecure participants reporting higher CIT separateness scores than
secure subjects (hypothesis 1d). This finding, along with the finding that individuals with
an avoidant attachment orientation reported lower relationship seriousness and satisfaction,
are in line with the results of previous studies which have found that avoidant individuals
exhibited greater relationship distance and relationship detachment (Feeney & Noller, 1991;
Rothbard, Roberts, Leonard, & Eiden, 1993). The mean scores for mutuality were in the

1 12
hypothesized direction with secure participants scoring higher than insecure participants

(hypothesis 1c) supporting previous studies that secure individuals exhibited mutually
focused conflict strategies and more reflective listening (Pistole, 1989; Pistole, 1991;
Rothbard, Roberts, Leonard, & Eiden, 1993). Due to the nonsignificant statistical results,
a post hoc analysis of covariance was conducted that recollapsed the insecure category
(preoccupied and fearful were collapsed to one insecure category and the dismissive
category was eliminated due to insufficient cell size). However, these analyses again did
not reveal significant attachment style effects on problem solving communication.

In conclusion, patterns in the observed variable relationships suggested that
participants with an anxious attachment orientation focused on their partner while viewing
the CIT as stressful, whereas participants with an avoidant attachment orientation focused
on their own point of view while reporting lower levels of success in completing the CIT.
These findings suggest that participants' attachment styles may have been activated during
the CIT. The results also suggest that insecure participants viewed themselves more
individually during the problem solving exercise while secure participants focused more on
their partner and conveyed respect for their partner's point of view. In addition, these
preliminary results suggest that secure and insecure participants demonstrated expected
problem solving behavior in this in vivo communication task. However, these
interpretations are quite tentative as they are based on correlational findings and on tests of
group differences that approached, but did not reach statistical significance.

The failure of participants' attachment style classification to predict their problem-
solving communication may have been due to several reasons. First, there was a
disproportionate representation of secure females from phase one who agreed to participate
in phase two. Since the participants were from the College of Education, they were more
likely to have been teacher education majors. Stereotypically, teacher education majors
may be more "feminine" relative to the general population, possibly resulting in a sample

more prone to individual conflict avoidance and instead focusing on relationship

1 13
maintenance. Second, length of dating relationship was higher than expected (mean=25.6

months) for a population of college dating couples. This may have resulted in a sample that
was more representative of long-term, stable, committed dating couples than dating couples
in general. Third, males and females that reported higher levels of relationship satisfaction,
seriousness, and degree of commitment reported higher levels of success on the in vivo
problem solving task (CIT). Couples rated their relationships as very serious and
committed (mean=4.5; 5=very serious on the Likert scale for this item) and satisfied
(mean=4.5; 5=very satisfied on the Likert scale for this item). These same individuals
reported that the CIT was less stressful. These differences may have resulted in an
interpersonally sensitive sample in which the couples approached the task with more
refined problem solving skills. In addition, the CIT was noted to be relatively low in stress
(mean=l.35; l=not at all stressful). Participants' attachment styles may not have been
sufficiently activated by the task used in this study due to the low stress rating. This
hypothesis should be re-examined within a larger, more balanced sample that is sufficiently
varied in regard to attachment style and length of dating experience. The task should be
restructured to increase stress levels in order that each participant's attachment style is
activated. Proposed ways to restructure the task will be discussed later in this chapter.

Sex, Gender Role, Attachment Sgle, and Problem Solving Communication.

This study also attempted to clarify the relationships of sex, gender role and
attachment style to problem solving communication. Sex was not expected to relate to
problem solving communication when gender role was controlled (hypothesis 2). It was
expected that there would be a significant relationship between gender role and problem
solving communication (hypothesis 3), and, that gender role and attachment style would
interact to affect problem solving behavior (hypothesis 4).

A hierarchical multiple regression revealed that gender role did not moderate the
attachment style/CIT relationship for any of the four CIT problem solving communication

pattern (hypothesis 3). Results from the hierarchical regression, correlational analyses, and

1 14
ANCOVA of CIT scores indicated that females and individuals of both sexes with higher

femininity scores were more responsive to the views of their partner (permeability)
(hypothesis 4b). Lamke, Sollie, Durbin, and Fitzpatnck (1994) also found that individuals
who identified with feminine traits described themselves as being comfortable in engaging
in emotionally supportive behaviors in intimate relationships. The mean scores for males,
however, suggested that they scored higher than females on self -assertion and mutuality
suggesting that male participants were able to focus on themselves and their partner's view.
Previous research has indicated that, relative to females, male participants tend to be more
independent or to avoid engaging in interpersonal issues during problem solving
discussions (Lang-Takac & Osterweil, 1992; Rusbult, Johnson, & Morrow, 1986). This
unexpected pattern for male participants may possibly relate to both the male and female
participants' report of higher scores on the successful completion of the problem solving
task. The correlational matrix also indicated that participants with higher attachment related
avoidance scored lower on the femininity scale. This finding is consistent with theoretical
expectations and with prior research which has demonstrated that individuals with an adult
avoidant attachment style seek interpersonal distance during times of stress (Pistole, 1994;
Simpson, Rholes, & Nelligan, 1992) and exhibit less focus on relationship connectedness
(lower femininity scores). Finally, individuals with higher femininity scores reported
higher scores on their successful completion of the CIT. This observed pattern from the
variable relationships suggests that individuals with higher femininity scores, and females
in general, (a) exhibited a communication style that was more focused and invested in the
relationship task, and (b) viewed their performance as a couple as more successful.

Several post hoc analyses were conducted to further explore the relationships
among attachment style, gender role, sex and problem solving communication. Overall,
females made more statements in which they responded to their partner's requests for
information and provided validation for their partner's responses (mutuality) as well as

statements that involved sensitivity and respect for their partner's views (connectedness).

115
In regard to sex and individual gender role differences (and not couple gender role

differences), females and individuals with higher femininity scores were more focused on
the relationship by attending and validating their partner. These results suggested that
individuals scoring higher on femininity reported more success on the problem solving task
while exhibiting more statements that demonstrated relationship connectedness. This
included showing validation, sensitivity and respect, and responsiveness for their partner's
views. Moreover, females, regardless of gender role, were more responsive and validating
of their partner than males. Previous research provides support for these findings. White
(1989), for example, found that females exhibited an af f iliative style of interaction with
their partner. Lamke, Sollie, Durbin, and Fitzpatrick (1994) reported that individuals who
identified with feminine traits engaged in emotionally supportive behaviors in intimate
relationships. The results for males were less conclusive. In this study, males made more
requests for specific behaviors to occur and were more task oriented (permeability).
Rusbult, Johnson, and Morrow (1986) also found that men tended not to engage in
interpersonal issues during problem solving discussions.

The lack of significant support for several of the hypotheses regarding the
relationships of sex, gender role and attachment style to problem solving communication
may be partially explained by a final post hoc analysis. An ANOVA was used to examine
the individual gender role of the sample. The gender role description, based on within
couple analyses of the sample, was contrary to expectations: females had higher masculine
scores than males; males had higher feminine scores than females; and females were overall
more masculine than feminine. The opposite gender role differences may have resulted in
several altering possibilities. For example, males may have responded more to their partner
and to the relationship. This may have created decreased disagreements regarding the task.
Based on this unexpected finding, it is recommended that these hypotheses be reexamined

with a more diverse sample in regards to gender role differences.

1 16
Couple's Conjoint Attachment Style and Problem Solving Behavior.

This study also examined the conjoint contribution of each partner's attachment

 

style to relationship functioning, a largely neglected area in the extant literature. The
majority of previous studies examined the effect of each partner's attachment style
separately. This study examined three different attachment style couple pair groups which
included secure-secure, secure-insecure, and insecure-insecure couple pairings. Results
from a one way ANCOVA indicated that there were no significant conjoint couple
attachment style differences for the couples' average rating regarding problem solving
effectiveness, on rater's average rating of problem solving effectiveness, or on the couple's
combined scores on the mutuality scale, even when gender role was controlled.
Nonetheless, the means were in the hypothesized direction, with secure-secure couples
scoring higher than secure-insecure and insecure-insecure couples on all three dependent
variables.

It is important to note here some findings for the correlational analyses. In
particular, the longer a couple dated, the more successful they rated their CIT performance.
In addition, the lower the couple rated the CIT as stressful, the lower their combined
mutuality score. The first result suggested that couples that have been together longer view
themselves as more successful. They may have had more practice in approaching and
resolving problem-solving tasks. The second result suggested that when the
communication task was perceived as less stressful, couples demonstrated a decreased need
for sensitivity and respect for their partner's views. One possible interpretation of these
findings is that couples have less of a need to work together when the issue or task is
perceived as non-stressful.

A post hoc MANOVA examined whether the attachment style of the couple and
gender role of the couple related to differences in problem solving communication. One
finding that approached significance indicated that insecure-insecure couples scored higher

on self -assertiveness than did insecure-secure and secure-secure pairs. This finding

l 17
suggests that insecure-insecure couples focused more on expressing their individual point

of view versus responding to their partner. Indeed, connectedness and individuality means
were in hypothesized directions with secure-secure couples scoring the highest on
mutuality and connectedness and insecure-insecure couples scoring the highest on
individuality (separateness and self -assertion).

Previous studies have reported that secure-secure couple pairings reported better
overall marital adjustments (Senchak & Leonard, 1992) and relationship satisfaction
(Feeney, 1994). The current results suggest that secure dyads showed more respect to
each other during the problem solving task and perceived themselves as more successful
while engaging in this task. Inspection of group means also indicated that secure-secure
couple pairs attended more to relationship maintenance during problem solving behavior
while insecure-insecure pairs focused more on individuality. This may begin to explain the
different reports on relationship adjustment and satisfaction based on the different
attachment style pairings. For example, if one of the couples in the insecure-insecure
couple pairing has a preoccupied attachment style while the other partner has a dismissive
attachment style, focusing on individuality will increase the level of anxiety for the
preoccupied partner due to fears related to loss of attachment. The increased level of
anxiety may cause the preoccupied partner to become demanding and clingy which will
cause the dismissive partner to distance further, thus creating more anxiety for the
preoccupied partner (Pistole, 1994). This results in a ”vicious cycle” in which each
partner's attempts to manage his or her discomfort with contrasting methods of distance
regulation, thus resulting in less relationship satisfaction. The secure-secure couple pairs
are better able to manage distance regulation, as they focus directly on problem solving and
relationship maintenance, as well as the need for both individuality and connectedness to
occur within the relationship. These factors may relate to higher reports of relationship

satisfaction for the secure couple.

 

1 18
Limitations of the Study.

Several methodological limitations likely affected the findings of the present study
and thus constrain the interpretability and generalizability of obtained results. These
limitations will be explored in greater detail

CIT Limitations. The Couple Intervention Task was adapted form Grotevant and
Cooper's (1985) Family Intervention Task and Watzlawick's (1966) Plan Something
Together Task. This task was constructed to permit observation of communication patterns
and to identify interaction patterns. Although steps were taken to implement the task as
previously described, there were several limitations with the CIT that may have affected the
results of this study. First, there is a question as to whether the CIT was perceived as a
sufficiently stressful task by the participants due to the low rating on the stress level
question (mean=l.35 on a 7 point Likert scale where 1=not at all stressful to 7=very
stressful). Participants provided their responses to the CIT success and stress questions in
front of their partner which may have also resulted in less reliable answers. If this task was
perceived as not stressful or one of low stress, the participant's attachment style may not
have been activated fully. Based on the results of this study, a low stress task still
produced results that were generally in the hypothesized direction suggesting that a task
with a higher stress level may produce significant results.

A second concern was that many couples planned unrealistic vacations in which
they visited a different location or country daily without providing adequate time for
transportation. Several couples finished the task in less than the allotted 20 minutes. These
approaches may have helped the couples decrease the stress level of the task and create a
situation where compromising strategies were not necessary. Future research using the
CIT should require that the planned vacation be realistic and that the couple use the entire
20 minutes to complete the task.

Another issue that may have affected the study was the original Family Intervention

Task required that a family plan a vacation together. The CIT required that the couple plan

 

l 19
a vacation together. It is possible that unmarried couples may have experienced anxiety

regarding the overnight nature of the task. It is undetermined if this was more difficult for
new couples creating a difference when comparing couples with various lengths of dating.

The coding process for the CIT was unable to consider nonverbal communication
and voice inflection, which may have affected coding reliability. Humor, sarcasm, anger,
and other emotional tones may have been overlooked. Statements may have been
incorrectly understood and coded in incorrect categories. In light of the fact that interrater
reliabilites were only marginally acceptable, the raters may indeed have had difficulty
coding the transcripts without access to nonverbal communication and voice inflection.
The study attempted to compensate for this concern by having a third rater make final
decisions on any coding differences.

Finally, couples were aware of the audiotaped process and at times made comments
in this regard. One concern is they may have been self -conscious in their interactions
presenting less realistic views of their communication styles. Example comments for the
couples included "this is easy for us," ”don't appear assertive," and ”I am suppose to
forget the recorder.” Couples may have interacted differently without the audiotaping
resulting in a more realistic problem solving discussion. It would be desirable to use
videotaping in future research.

Sampling Limitations Given that the study used a non-randomly selected college
sample, the present findings may not be generalizable to other young adult populations. All
subjects were enrolled in College of Education classes. College major was not assessed,
however, it is possible that the sample may have been unique in that the majority were
teacher education majors which may possibly draw more "feminine" persons relative to the
general population. This resulted in an unbalanced representation of gender during phase
one.

Phase two involved a self -selected sample in which most subjects self -classif ied

their adult attachment style as secure. This resulted in an overrepresentation of secure

120
females volunteering for phase two. This may have limited the generalizability of the

sample since the college population is more balanced in terms of gender and attachment
styles. The advantage of the current sampling procedure as compared to other studies was
that males did participate since their female partner brought them to phase two of the study.

Performance on the CIT may have been unintentionally influenced by partner
differences with respect to whether one or the other was the original study "volunteer.”
More specifically, the partner-recruited member of the couple may have been more willing
to defer to his/her partner since it was the partner's suggestions and efforts to participate in
the research project. The overrepresentation of secure females volunteering for phase two
resulted in the female partner more often being the "lead partner" in the CIT. It is possible
that the male more often deferred more often to the female based on this dynamic. This
could explain the unexpected gender role differences in that the males were more feminine
and the females were more masculine in comparison to their partner. The couple may
communicate differently if the other partner volunteered to participate in a communication
task or if the couple decided to participate jointly.

Additional limitations of sampling involved the lack of adequate representation in
each of the three insecure attachment style groups resulted in the collapsing the three
categories into a combined insecure category. Feeney et al. (1994) stated that this remains
a problem in couple research since many insecure subjects are underrepresented in dating
relationships or are reluctant to expose their relationship for research purposes. Another
reason for the low number of insecure participants may have been that insecure individuals
may describe themselves as insecure until they became involved in a long term dating
relationship, and, once so involved, may feel more secure and classify themselves as
secure. Regardless as to the reasons for the limited sampling, the study was unable to
assess for differences among the four attachment style groups and the different conjoint

attachment style pairings.

121
More specifically, the college sample for this study may represent a particular type

of student and dating couple. The majority of students were seniors and juniors who had
been dating an average of 25.6 months. The average length of dating was longer than
expected. This may have affected the findings in that the sample may have been more
representative of long-term, stable, committed dating couples rather than dating couples in
general. It is possible that these dating couples may have resolved early relationship
communication difficulties due to the length of time they were together. Therefore, they
were more successful in negotiating this task and experienced the task as unstressful. A
replication of this study is recommended within a larger, more diverse sample that is
sufficiently varied with respect to length of dating relationship, attachment style pairings,
ethnicity, age, and education level. Sampling from a general population which has greater
ethnic and cultural diversity would be useful. Thus far, few studies on adult attachment
have focused on special populations.

Other Limitations. The quasi-experimental design of the study and its reliance on
correlational analyses does not permit cause-effect conclusions to be derived.
Implications for Practice

Bowlby's attachment theory has been a useful conceptual framework for the
therapist. Byng-Hall (1995) stated that the therapist provides a temporary secure base for
the client The client uses the secure base to explore and solve problems. Satterfield and
Lyddon (1995) highlighted that clients with different working models of attachment will
evaluate the counseling relationship and their personal relationships differently. The
therapist needs to be aware of the individual differences in attachment behavior and how
this may impact relationship difficulties. Specifically, attachment theory provides the
therapist with a way to conceptualize problematic functioning in couple relationships.
Pistole (1993) argued that attachment theory helps the therapist understand
distance/regulation issues leading to the withdrawal/anxiety behaviors. As clients better

understand their attachment styles as well as those of their partners, they can view their

122
current relationship concerns from a new perspective. The couple can begin to clarify what

is possibly triggering the relationship disturbances without placing blame on each other.
The therapist can begin to help the couple regulate distance and intimacy issues more
effectively and increase felt security in themselves and in their relationship. Both changes
will help the couple develop more satisfying relationships. For example, the therapist can
clarify how the preoccupied partner's intense focus on an avoidant partner can create even
more distance in the relationship. Partner's expectations and differing needs can be
identified. The therapist can help the couple meet each other's differing needs and
relationship expectations regarding closeness and distance.

This current study encourages continued support for applying attachment theory to
the therapeutic process. The therapist may find attachment theory useful in helping couples
address problematic communication patterns, understand relationship dynamics, and
address distance-regulations issues within the relationship. Attachment theory also may be
useful in the areas of premarital counseling and preventive work.

Recommendations for Future Research

The present study attempted to examine the contributions of individuals' and
couple's attachment styles to their specific problem solving communication. The study
explored the extent to which sex and gender role affects the relationship of attachment style
and problem solving communication. Although the main hypotheses were not supported,
the means were in the hypothesized direction suggesting that secure individuals and secure-
secure couple pairings exhibited more successful relationship problem solving
communication styles. This area of research is in the preliminary stages of development
and will require further refinement and study. This initial study provides support for
continued research in the area of attachment theory and couple functioning while continuing
to clarify the influence of respondents' sex and gender role as possible moderating

variables.

123
Despite the usefulness of applying Bowlby's attachment theory to furthering our

understanding of couple problem-solving communication patterns, this theory may not be
the most parsimonious framework to efficiently predict these patterns. Future researchers
should also consider other theories to explain differences in couple functioning. For
example, Bowen's (1978) intergenerational model of family dynamics emphasizes the need
for individuals to adequately differentiate themselves from their families of origin in order
to successfully develop their individual expression of self and their capacities for
autonomous intimacy with others. Adults who have not adequately differentiated
themselves from their own parents are likely to ”project" their lack of differentiation to their
own children, thus contributing to their children's later problem solving difficulties in close
adult relationships.

In addition to considering attachment theory along with other possible theories to
explain couple communication pattern differences, the following recommendations may
also be helpful for future research. Investigators should continue examining the role that
attachment style differences play in couple functioning since previous rescarch has
generally overlooked these relationships. Efforts to examine the relationship of adult
attachment styles to dyadic functioning has generally relied on self -report measures. This
study attempted to address both issues by using in vivo methods and audiotaping. Results
suggested that audiotapin g may overlook the contributions of nonverbal behavior.
Videotaping is recommended to capture the finer details of communication. In addition, by
employing both tape transcripts and the videotape to code the couple's communication,
interrater reliability may be greatly enhanced.

According to Bowlby (1979), the attachment system tends to be activated during
periods of stress. For this study, it is likely that the CIT was not perceived as a sufficiently
stressful task by the participants. Consequently, there remains doubt as to whether the
participant's attachment style was appropriately activated during the problem solving

discussion. Future studies should employ a more stressful task in order to activate each

124
participant's attachment system such as introducing a stimulus event that couples find

stressful. Being mindful of research ethics, this may be limited to the couple's discussing
this event ”as if it happened.” Scenarios could include couple discussions involving
questioning the relationship commitment; observing one's partner flirting with an attractive
stranger; or discovering that one's partner lied. Simpson, Rholes, and Nelligan's (1992)
creative use of an experimental situation for increased anxiety for one member of a dating
couple may provide a framework for future research. They unobtrusively videotaped a
female in a waiting room after she was informed that she would soon be engaged in an
activity that would provoke anxiety in most people. She was subsequently observed as she
interacted with her male partner during the waiting period. If the CIT is used again,
researchers should establish parameters for the task where the couple establishes realistic
travel plans within the required time frame. This will increase the likelihood that the
participant's will find the task more stressful.

The contribution of conjoint couple attachment pairings needs to be examined
further. Kirkpatrick and Davis (1994) argued that specific attachment styles are not
"universally good or bad." Instead, different attachment pairings may result in successful
relationship experiences. Previous research found a predominance of secure-secure
pairings and a relative absence of anxious-anxious and avoidant-avoidant couples and that
avoidant individuals tend to couple with anxious partners (Collins & Read, 1990; Simpson,
1990; Kirkpatrick & Davis, 1994). The interactive affect of these attachment pairings has
been largely overlooked. This study attempted to clarify the contribution of conjoint couple
attachment pairings by using Bartholomew and Horowitz's (1991) four category typology.
The original plan was to examine the differences among the different possible attachment
style couple pairings (e. g., secure-secure, secure-preoccupied, secure-fearful, secure-
dismissive, preoccupied-fearful, preoccupied-dismissive, preoccupied-preoccupied,
dismissive-fearful, dismissive-dismissive, and fearful-fearful attachment style pairings).

Due to the limited number of insecure attachment styles in the present sample, this could

125
not be accomplished and the sample was collapsed to form three different couple attachment

pairings (secure-secure, secure-insecure, and insecure-insecure). Further research on
couple functioning within secure-secure, insecure-secure, and insecure-insecure couples as
well as within all the different possible attachment style couple pairings is needed. In
particular, it would be desirable to explore whether, in insecure-secure couples (e. g.,
secure individual coupled with a preoccupied individual), the secure partner may provide
the relationship skills to help the insecure partner cope with anxiety regarding closeness and
distance regulation thus allowing for higher relationship satisfaction and success.

Future studies should also employ larger, more diverse samples or explore
alternative sampling methods to increase the number of insecure participants. Subsequent
studies should also gather information from a more diverse population. This may include
sampling from other than college populations such as recruiting participants through
newspaper solicitations (as Hazan and Shaver (1987) initially did) or sampling from a
mental health outpatient clinic or employee assistance program at a university or
corporation. These sampling refinements should provide a greater opportunity to examine
the four stage model of attachment for individual and couple differences in couple
interactions.

Researchers should explore more specific areas of relationship functioning such as
differences in stages of relationship development It is possible that attachment style
differences affect relationship functioning differently at various relationship stages. For
example, an individual with a preoccupied attachment style may experience more difficulty
during the initial stages of a relationship, however, as the relationship moves toward the
committed phase, this individual may become more secure and less clingy and demanding.
Attachment style differences may be especially influential on relationship functioning
during significant life transitions such as when a couple experiences pregnancy, loss of a
parent, physical illness, job change/relocation, etc. Future research could clarify these

possible relationship developmental changes and transitions further.

126
Finally, investigators should consider using qualitative methods to explore

communication patterns within couples that vary with respect to their attachment styles.
Couple functioning and communication is a complex phenomenon, involving both verbal
and nonverbal communication. Qualitative methods could generate additional information
about the couple that may be limited by quantitative methodology. This, along with the
other recommendations, will provide an opportunity to further explore the relationship of
the individuals' attachment style and the couple's conjoint attachment style-related
contributions to their problem solving communication style and their overall relationship
functioning. In conclusion, attachment research opens a new and exciting arena in which

to understand couple interaction and functioning and should be pursued further.

APPENDICES

127
APPENDIX A

Request for Participation and Participation Consent Form
Dear student:

Thank you for your expressed interest in my study of the relationship experiences of
college students. This research is being conducted by Jennie Leskela, a doctoral candidate
in the Department of Counseling, Educational Psychology, and Special Education here at
Michigan State University.

The purpose of this research is to learn more about relationship beliefs and problem solving
communication patterns in dating couples. If you choose to participate, you will be given a
packet containing three brief self -report questionnaires to complete. The first questionnaire
will solicit important background information; the second and third questionnaires will ask
questions regarding attitudes and feelings about close relationships. I expect that it will
take you between 5-10 minutes to complete the questionnaires contained in the survey
packet. I do not anticipate that your participation will result in any physical or emotional
risk to you. As a benefit for your participation you will receive some extra credit points
from your instructor.

Please know that your responses to this survey will be kept completely confidential. Do
not put your name on the questionnaires. Code numbers will be assigned to your names.
The primary researcher, Jennie Leskela, will be the only person that has access to the code
number and names. The code numbers and names will be destroyed after the study is
completed. Information will only be presented in aggregate form and not by individual
results.

Your participation in this survey is strictly on a volunteer basis. You are free to withdraw
your consent and to stop participation at any time without penalty to you. If you decide to
participate, read the brief statement below and m and s_ign your name, and enter today's
date on the appropriate lines. I appreciate your participation in this project. This form will
be kept separate from your survey responses. If you have any questions, please f eel free to
ask the survey administrator or to call Jennie Leskela at 353-4734.

 

I agree to participate in the survey directed by Jennie Leskela. I understand the nature of
the project, the nature of my participation, that my participation is voluntary, and that I can
terminate my participation at any time without penalty.

 

 

 

PRINT your name here SIGN your name here Today's date

128
APPENDIX B

Request for Participation
Phase 2

Dear student:

This study will be conducted in two phases. You are completing Phase 1. After Phase 1,
40 couples will be selected to participate in Phase 2. Phase 2 will involve the completion of
three brief questionnaires; two on couple attitudes and feelings and one on description of
self. Following the completion of the surveys, couples will be privately audiotaped for 20
minutes while they are engaged in a structured communication task

I expect that it will take 30-40 minutes to complete the questionnaires and the couple
communication task I do not anticipate that your participation will rcsult in any physical or
emotional risk to you. There is the possibility that disagreements between you and your
partner about how to complete the task may occur. As a benefit for your participation you
and your partner will be entered in a drawing. Four $25.00 prizes will be awarded.
Winners will be contacted following the completion of data collection.

Your participation in this survey is strictly on a volunteer basis. If you are in a
heterosexual dating relationship for at least one month and you are interested in
participating with your partner in Phase 2 of this study, please sign below and indicate a
phone number where you can be contacted to schedule your conjoint participation in Phase
2.

 

 

 

PRINT your name SIGN your name Phone number

129
APPENDIX C

Request for Participation and Participation Consent Form
Phase 2
Dear student

Thank you for your expressed interest in completing Phase 2 of my study of the
relationship experiences of college students. This research is being conducted by Jennie
Leskela, a doctoral candidate in the Department of Counseling, Educational Psychology,
and Special Education here at Michigan State University.

The purpose of this research is to learn more about relationship beliefs and problem solving

communication patterns in dating couples. If you choose to participate, you will be given a I
packet containing three self -report questionnaires to complete on couple attitudes and
feelings and description of self. This will take 540 minutes to complete. Following, you
and your partner will participate in a 20 minute private audiotaped interaction on your
responses to a communication task. I do not anticipate that your participation will result in
any physical or emotional risk to you. However, disagreements between you and your
partner about how to complete the task may occur. As a benefit for your participation you
and your partner will be entered in a drawing. Four $25.00 prizes will be awarded.
Winners will be contacted following the completion of data collection.

Please know that your responses to this survey will be kept completely confidential. Do
not put your name on the questionnaires. This way your name cannot be connected to any
of your answers and your confidentiality can be assured. Code numbers will be assigned
to your names. The primary researcher, Jennie Leskela, will be the only person that has
access to the code number and names. The code numbers and names will be destroyed
after the study is completed. Information will only be presented in aggregate form and not
by individual results.

Your participation in this survey is strictly on a volunteer basis. You are free to withdraw
your consent, to stop participation, and withdraw your audiotape at any time without
penalty to you. If you decide to participate, read the brief statement below and pg'n_t and
S_ng your name, and enter today's date on the appropriate lines. I appreciate your
participation in this project. This form will be kept separate from your survey responses.
If you have any questions, please feel free to ask the survey administrator or to call Jennie
Leskela at 353-4734.

 

I agree to participate in the survey directed by Jennie Leskela. I understand the nature of
the project, the nature of my participation, that my participation is voluntary, and that I can
terminate my participation at any time without penalty.

 

 

 

PRINT your name here SIGN your name here Today's date

130
APPENDIX D

ID. Number
Demographic and Background Information

Thank you for deciding to participate in my project. The following questions ask about
your background. Please circle the appropriate number under each of the items below or
enter the correct information on the blank spaces that are provided.

 

 

 

 

1. Your sex: 2. Your current age:
( 1) Male
(2) Female
3. Your year in college: 4. Your ethnic/racial background:
(1) Freshman (1) African-American
(2) Sophomore (2) Asian-American
(3) Junior (3) Caucasian/White
(4) Senior (4) Hispanic/Latino
(5) Other (5) Other
please describe please describe
5. Are you currently married? (If yes, skip to question 13)
(1) Yes (2) No
6. Have you ever been married?
(1) Yes (2) No
7. Are you in a dating relationship? (If no, skip to question 13)
(1) Yes (2) No
8. How long have you known your dating partner?
9. How long have you and your partner been dating?
10. How serious do you consider this relationship?
1 2 3 4 5
not at all; very serious;
very casual we're committed
11. How satisfied are you with this relationship?
1 2 3 4 5
not at all very satisfied
satisfied

12. Are you currently dating anyone else, that is anyone other than the person you referred
to in questions 7-11? (1) Yes (2) No

13. What is your (biological or adoptive) parents' current marital status?
(I) married and still living together
(2) separated
(3) divorced (indicateo yo_ur age at time of parents' divorce: ___)
(4) mother deceased (your age at time of mother's death: )
(5) father deceased (your age at time of fathers death: )

14. The quality of my parent's marriage is/was (if deceased or divorced:
(1) very poor (4) good

(2) P00r (5) very good
(3) average

131
APPENDIX E

BRQ

Directions Please read each of the descriptive paragraphs below and place a checkka
next to the one that best describes how you feel about close relationships.

 

Most descriptive
of me
(Check one)
1. 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 > 1.

 

2. I am comfortable without close relationships. It is very
important for me to feel independent and self -suf ficient, and
I prefer not to depend on others or have others depend on me ------ > 2.

 

3. 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 > 3.

 

 

4. 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 too close to others > 4.

 

 

132
APPENDIX F

S-ASS

Directions: Using the scale adjacent to each of the items below, indicate (by circling the
appropriate number) to what extent the item describes how you have typically felt toward

romantic partners in general.

Strongly Strongly
was» Agree
1. I find it relatively easy to get 1 2 3 4 5 6 7
close to others
2. I'm not very comfortable having
to depend on other people 1 2 3 4 5 6 7
3. I'm comfortable having others
depend on me 1 2 3 4 5 6 7
4. I rarely worry about being
abandoned by others 1 2 3 4 5 6 7
5. I don't like people getting
to close to me l 2 3 4 5 6 7
6. I'm somewhat uncomfortable
being too close to others 1 2 3 4 5 6 7
7. I find it difficult to trust
others completely 1 2 3 4 5 6 7
8. I'm nervous whenever anyone
gets too close to me 1 2 3 4 5 6 7
9. Others often want me to be more
intimate than I feel comfortable
being 1 2 3 4 5 6 7
10. Others are often reluctant to
get as close as I would like 1 2 3 4 5 6 7
11. I often worry that my partner (3)
don't really love me 1 2 3 4 5 6 7
12. I rarely worry about my
partner (3) leaving me 1 2 3 4 5 6 7

13. I often want to merge with others,
and this desire sometimes scares
them away 1 2 3 4 5 6 7

133

APPENDIX G

BSRI

Instructions In this inventory, you will be presented with sixty personality characteristics.
You are to use those characteristics in order to describe yourself. That is, you are to
indicate, on a scale from 1 to 7, how true of you these various characteristics are. Please

do not leave any characteristic unmarked.

 

1 2 3 4 5 6 7
Never Usually Sometimes Occasionally Often Usually Always
or almost not but true true true or almost
never true true infrequently true
true
1. Self-reliant 31. Makes decisions
easily
2. Yielding 32. Compassionate
3. Helpful 33. Sincere
4. Defends own beliefs 34. Self-sufficient
5. Cheerful 35. Eager to soothe
hurt feelings
6. Moody 36. Conceited
7. Independent 37. Dominant
8. Shy 38. Soft-spoken
9. Conscientious 39. Likable
10. Athletic 40. Masculine
11. Affectionate 41. Warm
12. Theatrical 42. Solemn
13. Assertive 43. Willingtotakea
stand
14. Flatterable 44. Tender
15. Happy 45. Friendly
16. Has strong personality 46. Aggressive
17. Loyal 47. Gullible
18. Unpredictable 48. Inefficient
19. Forceful 49. Acts as a leader
20. Feminine 50. Childlike
21. Reliable 51. Adaptable
22. Analytical 52. Individualistic
23. Sympathetic 53. Does not use harsh
language
24. Jealous 54. Unsystematic
25. Has leadership abilities 55. Competitive
26. Sensitive to the needs 56. Loves children
of others
27. Truthful 57. Tactful
28. Willing to take risks 58. Ambitious
29. Understanding 59. Gentle
30. Secretive 60. Conventional

10.

11.

12.

13.

14.

Geographic Location

134
APPENDIX H

Activity Planning Form
ID. Number

Activity

135

 

 

 

APPENDIX I
Transcript Format
CIT Transcription Page of
Transcribed by ID. Number
Male. Ecmls
Okay
Let's go to Europe
Yes, let's go

Where should we go first?
I would like to go to France

136
APPENDIX J

Transcriptionist Guidelines
(adapted from Condon, Cooper, & Grotevant, 1984)

1. Follow the format with single-space typing. Type exactly as dictated.

2. Complete following information: code number from tape, page numbers, and your
name.

3. Male and female's comments are placed in the labeled column.

4. When the speaker changes, place the second speaker's dialogue in the appropriate
column and begin on a new line.

5. If the speakers talk at the same time, speech overlap is placed on the same lines and
underlined. With simultaneous speech, the exact timing is not important. Do not worry
about exactly determining the points of overlap. The underlining indicates who started
talking first and approximately where the overlap ends.

6. All utterances in the interaction are transcribed.

7. Coughs and sighs need not be transcribed.

8. Transcribe laughter as follows: (laughter).

9. Conversations between the couple and interviewer need not be transcribed.

10. Unintelligible speech represented by a question mark and underline: ? when
the couple is speaking at the same time.

1 1. Unintelligible speech which is mt simultaneous is represented by a question mark and
dashes: ?- - -.

12. Type long, run-on sentences as is.

13. Three periods... are used when speech trails off or when there is a long pause within
sentences.

14. Use a question mark when a question is asked.

137
APPENDIX K

AUDIOTAPE CODING MANUAL
Abridged Family Discourse Training Manual
(adapted for couples from Condon, Cooper, and Grotevant, 1984)

1. PROBLEM SOLVING COMMUNICATION PATTERNS

*The first 300 chunks of verbal communication will be assigned into one of 14 categories.
The 14 categories are categorized into 4 problem solving communication patterns.

 

Semrateness Permeability Mutuality Self Assertion
direct disagreement agreement initiation of direct suggestion
indirect disagreement acknowledgment compromise
requests for action requests for info statement of
irrelevant comment relevant comment other's feelings
compliance with indirect suggestion
request for action answers request for
info/validation

*The 4 problem solving categories are defined as follows:

Separateness: Ability to express differences between the self and others.
Willingness to accept responsibility for one's own feelings and thoughts.
Ability to communicate one's ideas clearly and directly and to differentiate
them from others.

Permeability: Responsiveness or openness that an individual displays to the ideas of
others.
When an individual gives permission and encouragement to others to
develop a point of view.

Mutuality: Individual demonstrates respect for the beliefs, feelings, and ideas of
others.

Self Assertion: Being aware of one's own point of view.
Taking responsibility for communicating point of view clearly.

138
2. CODING THE INTERACTION

*Move/Response
The code is organized into three sections:

(a) Move function: Moves the conversation forward
Can be picked up and moved elsewhere without problem.
(b) Response: Responds to what was previously said

(c) Other

*Coding categories are noted in each of the three sections:

Move Response
irrelevant comment initiates compromise
suggests action or location directly agrees-accepts-incorporate
suggests action or location indirectly disagrees directly
requests inf ormation/validation disagrees indirectly
requests action answers requcsts for inf ormation/validation
no clear move function complies with requests for action
acknowledgment
no clear reactive function
Other
relevant comment

mindreads/states other's feelings

*All chunks are assigned one category in the MOVE section and one category in the
RESPONSE section (This may, however, mean that it is assigned to the NO CLEAR
MOVE FUNCTION or NO CLEAR RESPONSE FUNCTION).

*A chunk is assigned to a code category be entering the letter which represents the speaker
of the chunk (MzMale; F=Female) in the square on the code sheet representing the
appropriate chunk (column) and corresponding chosen code category (row).

*The third section is the OTHER category. There are two other categories: RELEVANT
COMMENT and MINDREADS/STATES OTHER'S FEELINGS. It is optional to code
MINDREAD/STATES OTHER'S FEELINGS regardless of how it is coded in the MOVE
and RESPONSE categories as long as it f its the requirements for Mindreading. The
RELEVANT COMMENT category is different A chunk can be coded RELEVANT
COMMENT only if it has been assigned to both NO CLEAR MOVE FUNCTION and NO
CLEAR RESPONSE FUNCTION.

*Code as No Clear Move Function and No Clear Response Function
-expressions like well, let's see. or uh which serve no purpose except floor holding
or attention focusing.
-not a complete idea
-cannot understand the statement

*Code as Irrelevant Comment and No Clear Response Function
-statement not on task, not relevant to the discussion.

139
*Code High

If a chunk has more than one move or response function, code high. The code categories
are arranged according to hierarchy. Categories arranged spatially higher on the page
represent stronger functions. Whenever we have to choose among functions, we choose
the highest one. Example: "Is there anymore coffee? This is both an Irrelevant
Comment and a Request for Information. It is scored as Irrelevant Comment since
Irrelevant Comment is spatially higher in the code category hierarchical list. Example: "92
you want to go to Asmn"? This could be coded as Request for Information or
Indirect Suggestion. If Aspen had not been previously mentioned, it is scored as Indirect
Suggestion. If the partner answers "Yes", the coder has to decide if this is an Agreement
or Answer to Request for Information. If you coded the question as Indirect Suggestion,
you would code the " Yes " as Agreement If you coded the question as Rguest for
Information, you would code the "Yes" as An_swer to Request for Information.

 

*Complex Responses

A single answer to a single question may be a long utterance and continue over several
chunks. The long utterance contains only one RESPONSE category. First code the chunk
as Answer to a Request. The remaining chunks of the lengthy answer are coded as
Relevant Comments (other category). (Reminder: When Relevant Comment is used, the
No Clear Move Function and No Clear Response Function were also used). Example:
"No/I don't want to go there./It's too cold./" The first chunks coded as
Disagrees/Challenges Other's Idea Directly (and No Clear Response in the MOVE
category) and the next two chunks are coded as Relevant Comments.

In the case of Direct Disagreement, always code the Direct Disagreement, if there is a
choice. Example: flNe shouldn't go there./I don't want to./It's too cold./"
The first and third chunks can be coded as Relevant Comment and the middle chunk can be
coded as Disagree/Challenges Other's Idea Directly. Disagreements are the only case in
which the first chunk of a response is not coded for the that response.

 

*Observer Coding

You cannot determine what is going on in the participants' minds and you are not supposed
to consider what a speaker might "really mean" or ”really intend" by some remark.
Instead, rely on your understanding of the verbal interaction.

When examining a chunk, describe to yourself what has occurred in the context Translate
your description into the categories. Consider whether the chunk has a MOVE or

RESPONSE function or both. Remember that a MOVE function can be removed and place
elsewhere where a RESPONSE function is context-bound.

*Checking Process
1. Make sure every chunk has an entry in both the MOVE and RESPONSE categories.

2. Make sure that relevant comments have only NO CLEAR MOVE FUNCTION AND
NO CLEAR RESPONSE FUNCTION in the MOVE and RESPONSE categories.

3. Make sure that each chunk has the same letter for all entries (MzMale and F=Female).

140
BIC—“QM

MOVE FUNCTIONS

1. IRRELEVANT COMMENT are clearly not related to the task or to the discussion
that is directed toward completion of the task If the remark is not understandable at all, it
should be coded as No Clear Move Function.

Examples: What will we do after we finish with this project?
Did you bring your car in for repair today?

2. SUGGESTS ACTION OR LOCATION introduces an option of a location or
activity for the vacation. They present possible solutions to the task It can include a
particular time, day, or sequence to the activities.

There are two kinds of Suggestions:

Suggests Action or Location Directly uses the personal pronouns I or m to express
the speaker's responsibility for the proposal. Also coded as Direct Suggestions are
commands with no pronoun. These include commands to write something down.

Examples: I want to go to California.
Go to a ballgame.
Put fly to New York.
To me, fishing is the best thing to do.

Suggests Action or Location Indirectly use first person plural pronouns such as w_e_
or g. Also included as Indirect Suggestions are those with the first person singularl but
negative or indecisive verb forms, suggestions which name a place or activity alone with no
pronouns or verbs and suggestions in the form of questions.

Examples: Let's go to California.
We could go to Yellowstone.
I wouldn't mind going to a ball game.
I guess New York would be interesting.
Hawaii.
How about flying to New York?
Do we want to ski?
Fishing would be fun.

3. REQUESTS INFORMATION/VALIDATION involves seeking information,
clarification, confirmation, opinions, or decisions from the partner. The Request for
Validation seeks confirmation of a statement while the Request for Information seeks input
that is pertinent to the accomplishment of the task. They are usually in the form of
questions.

Examples: All right?
Do you want to learn to ski? (provided that skiing has already been
suggested)
How far is Yellowstone from San Francisco?
I wonder how long it takes to get there.
I wish I knew if we could fly there.

141

4. REQUESTS ACTION specifies some action which the speaker wants the bearer to
perform or to refrain from performing, where the action is relevant to the ongoing
discussion (not the pretend vacation). Usually, they pertain to the flow of talk or to the
action of writing answers on the answer sheet. We require that a verb be present to specify
the action requested or that a Iocative (there, here) be present to indicate that the spcaker is
requesting some action about where something is to be written on the answer sheet.

Examples: Write that down.
Be sure to get that
Would you read back what we already have?
Listen.
Wait a minute.
There, in that space.
First day, here.
We should write this down.

5. NO CLEAR MOVE FUNCTION is a catch-all category for chunks that do not
exhibit any of the strong MOVE functions. Items with no understandable content or
incomplete content due to interruption or trail-of f are coded in the No Clear Move Function
and the No Clear Response Function categories. Items with Response functions only are
coded as No Clear Move Function. Many particles such as h_ey, we_ll, Q, and you know
are coded in this category if they are uttered separately from any more meaningful remarks
and they have no clear function of their own.

Examples: Well, uh....
Of course, I'm also....
Let's see now .....
yeah. (where y;ca_h has a purely response function as Agreement or
Acknowledgment).
This is fun (where the chunk is being coded as Relevant Comment)

142
RESPONSE FUNCTIONS

 

1. INITIATES COMPROMISE functions to resolve disagreement about solutions to
the task or about managing the task. Compromises should integrate ideas which were not
originally associated. The ideas should be explicitly mentioned or referred to together as
_b_oLh_ or £1. or they should be connected significantly by gift}; or while. Compromises are
attempts to console the partner. Compromises are usually Suggestions.

Example: We can go to the Bahamas and then we can go to Paris.
Well, do you want to do both?
We'll go there on our next vacation.
They have a nice golf course there. (Responds to partner who likes to golf
after he/she says he/she does not want to go to Disneyworld).

Since Compromises are usually also Suggestions, they are usually also coded as Suggests r.-
Action or Location in the MOVE section. : .

2. AGREES/ACCEPTS”NCORPORATES OTHER'S IDEA is accepting a J
Suggestion. Agreements represent a ”yes-vote" to the partner's proposal of a location or
activity.

Example: Yeah.
Good idea
That'll be fun.
We'll have a great time there
I've always wanted to see the Florida Keys (in a response to a Suggestion
of going to Florida).

3. DISAGREES/CHALLENGES OTHER'S IDEAS function primarily to prevent
acceptance of a Suggestion. Most Disagreements are reasons or arguments for not
accepting a particular Suggestion. These range from the fact that the partner doesn't want
to go to a Suggested place or doesn't want to engage in a Suggested activity.
Disagreements are reactions to previous Suggestions, factual statement, or Requests for
Action.

There are two kinds of Disagreements:
Disagreements/Challenges Other's Idea Directly are distinguished by use of the
first person singular pronoun 1.

Example: I don't want to go to Egypt.
No.
I think that would be boring

Disagreements/Challenges Other's Idea Indirectly can be difficult to distinguish
from Relevant Comments or simple Requests for Information. It is important to feel that
the speaker is already assuming an answer to the question. This assumed answer must be a
reason not to accept the Suggestion such that makes the Suggestion unfeasible. If the two
are incompatible, then a Disagreement is involved.

Example: We're already doing too much.
Do you think two weeks will be long enough for all that?
Isn't it awfully far?

143

4. ANSWERS REQUEST FOR INFORMATION/VALIDATION include
appropriate responses to Requests for Information, and, in some cases, they are responses
to Requests for Validation. Usually Answers are factual statements.

Examples: It's not far.
It won't take more than four hours.
Yes.
About two hundred miles
Spain, is south of France.
No, it wouldn't

The major problem in coding Answers is to be sure that they should not be coded higher.
They could also be Agreements or Disagreements.

5. COMPLIES WITH REQUEST FOR ACTION include an appropriate response
to a Request for Action. It indicates performance of the action which has been Requested.
Compliance requires a previous Request for Action. The most common way of indicating
Compliance is the particle %.

Examples: Got it. (response to a request to write something down).
0

Let's see, First day: fly to Rio. (response to request to read back the plan).

6. ACKNOWLEDGMENT functions to affirm speaker's participation in the
interaction. They respond positively to a speaker and to his or her ideas, but do not
actually incorporate or Agree with those ideas. The simplest form is affirming that the
message has been received. Acknowledgments may be responscs to any kind of remark.
They include positive particles like y;ea_h, uh huh, right, and M. repetition of an
utterance, and requests f or inf orrnation about what a person said, meant, thinks, feels, or
wants.

 

Examples: That's an idea.
You can say where you want to go.
I get your point.
Rome (repeating).
Skiing? (repeating with question)
What do you think?
So you think we should spend three days there.

7. NO CLEAR RESPONSE FUNCTION included items with no understandable
content or incomplete content due to interruption or trailing off and are coded in the No
Clear Move Function. Items with MOVE function only are coded as No Clear Response
Function. Many particle such as hey, _we_ll, fl, and you know are coded in this category if
they are uttered separately from any more meaningful remarks and they have no clear
function of their own.

Examples: Well, uh...
of course, I'm also
Let's see now.
Tahiti. (where Tahiti has the purely MOVE function of a Suggestion)
This is fun (where the chunk is being coded as Relevant Comment)

A _ .

 

144
OTHER FUNCTION

 

1. RELEVANT COMMENT are usually factual statements, but their contents may
vary due to the way they function in the Complex Response Convention. They are never
questions, commands, or Suggestions. Relevant Comments are determined acmrding to
the functions they do not have. If a remark does not exhibit one of the MOVE or
RESPONSE functions we have identified, yet it is still understandable and relevant to the
task, then it is coded Relevant Comment

Examples: This is fun.
Let's see, Egypt, we can probably make it in a few hours flying.
We'd probably fly American.

2. MINDREADING/STATES OTHER'S FEELINGS involves one person
speaking for another person in the presence of that person and attributes to the other person
ideas, desires, wishes, opinions, needs, choices, or suggestions which the person has not
stated previously in the interaction. The expression should contain a proper name or a
pronoun that refers to another person. (i.e., fl, yfl).

 

Examples: We want to spend at least three days there.
You don't want to go antique shopping.
You think it will be boring.

145
SPECIAL RESPONSES

1. Acknowledging Question

Some Requests of Information have RESPONSE function which are validating or
Acknowledging an individual. Example: Mt do you think, Bob? This is coded as
Acknowledgment in the RESPONSE Portion and Requests Inf ormation/Validation in the
MOVE Portion.

2. Answering Questions with Questions

Sometimes questions are responded to with questions that repeat or paraphrase the
question asked. These questions are considered to be Acknowledging Questions. Some
examples in response to the Request for Information Where do you want to go are:

Where? /Spain/

Where do I want to go?/Spain.?
In the above examples, the first chunk of each is coded as Requests Information/Validation
in the MOVE section and as Acknowledgment in the Response. The second chunk in each
is coded as Suggests Action or Location Indirectly in the MOVE section and Answers
Request for Inf ormation/Validation in the RESPONSE section.

3. Challenges

Challenges are tests of an idea, usually in question form, that are meant to
undermine the feasibility or adequacy of the idea in light of other facts. The major problem
is distinguishing Challenges from Requests for Information or Relevant Comments that do
not have the Disagreeing function. Do not code a Challenge unless you are reasonably
certain that it is one.

4. Checking Questions

Checking questions are repetitions of what a person has said with question
intonation. They function to make certain that a message was received correctly, and,
therefore, they are Acknowledging the message.

5. False Starts
Always use the last start.

6. Fragments

Fragments can often accomplish functions even thought the utterances are
incomplete. An example might be so, after the Bahamas, we go to.... In this case, the
fragment functions as a Request for Information. We would not code as that unless we can
tell whether a fragment has accomplished some function if it elicits the appropriate response
for that function. Fragments are usually, however, coded as No Clear Response Function
and No Clear Move Function.

7. OK and Other Particles
OK and other particles (yeah, uh huh, 39, and sure) have many functions.
Examine them in the appropriate context

146

8. Repeated Suggestions

Repeated Suggestions may serve a variety of function and to keep matters simple,
we code most of these as Relevant Comment. An example includes So, we're going to
New York where it has just been decided that the couple will go to New York. Sometimes
Suggestions are repeated with question intonation, in which case they are considered to be
Checking Questions or Challenges if the tone is sarcastic.

The kinds of repeated Suggestions that are coded as Suggests Action or Location
include:
1) Suggestion may be repeated several times by the speaker. As long as the Suggestion is
repeated prior to Agreement, it can be coded as Suggests Action or Location.

2) Suggestions may be repeated as part of an elaboration of the initial Suggestion. They are
coded as Suggests Action or Location.

3) Some couple may make major decisions before writing them down, then the writing
process repeats the Suggestions. When a Suggestion is repeated in this situation, it may be "
concretized by being ordered somewhere of by specifying the time during which the
Suggestion will be carried out These are coded as Suggests Action or Location. ‘
Example: Third day, fly to Switzerland
Then go to Germany

9. Responses to Relevant Comments
Code them as Acknowledgments.

10. You know

When you know occurs at the end of a statement, it is usually a Request for
Validation. If you know occurs at the beginning or in the middle of a statement, we do not
chunk and code it separately as a Request for Validation, since, in those positions, it
usually serves a different function.

147

 

APPENDIX L
Sample Coding Sheet
(adapted from Condon, Cooper, and Grotevant, 1984)
INTERACTION CODE SHEET page of ID
Date Rated Rater

 

 

Chunk12345678910111213
1. Move
irrelevant comment
sug act or loc directly
sug act or loc indirectly
requests info/validation
requests action

no clear move function

II. Response

initiates compromise
agrees-accepts-incorporate
disagrees directly
disagrees indirectly
answers req for info/val
complies w/req for action
acknowledgment

no clear reactive function

III. Other
relevant comment

mind reads/dictates feeling

 

148
APPENDIX M

Sample Coding Sheet - 2
(adapted from Condon, Cooper, and Grotevant, 1984)

 

 

INTERACTION SHEET

Date scored ID. Number
Esmde

I. Move

suggests action or location directly
suggests action or location indirectly
requests inf ormation/validation
requests action

irrelevant comment

no clear move function

II. Response

agrees, accepts, incorporates ideas
disagrees, conflicts directly
disagrees, conflicts indirectly
initiates compromise

answers request for inf ormation/ val
complies with request for action
acknowledgment

no clear reactive function

111. Other

relevant comment

mind reads/states other's feeling

Male

 

 

149
APPENDIX N

ID. Number

Problem Solving Communication Totals

Senarateness

direct disagreement
indirect disagreement
request for action
irrelevant comment
Total
Permeability
agreement
acknowledgment
requests for information
relevant comment
compliance with request for action
Total
Mutuality
initiation of compromise
statement of other's feelings
answers to requests
indirect suggestion
Total
Self-Assertion
direct suggestion

Total

Male

 

Female

150
APPENDIX 0

ID. Number

 

 

Couple Problem Solving Rating Score

Thank you for your participation in this research project. Please respond to the following
questions. Circle the correct response.

Circle the correct response.

1. Your sex: (1) male (2) female
2. How successful do you believe you and your partner were in completing this task? I!
1 2 3 4 5 6 ,
Not at all Very i L
successful successful a
3. How stressful did you find this task?
1 2 3 4 5 6 7
Not at all Very

stressful stressful

151
APPENDIX P

ID. Number

 

 

Couple Problem Solving Rating Score (RATER)

1 How successful was this couple in completing this task.
1 2 3 4 5 6 7
Not at all Very
successful successful

 

 

REFERENCES

152

References

Ainsworth, M.D.S. (1967). Infancy in Uganda: Infant care and mggrowth of love.
Baltimore: Johns Hopkins University Press.
Ainsworth, M.D.S. (1982). Attachment: Retrospect and prospect. In C.M. Parkes

& J. Stevenson-Hinde, The place of aflchment in human behavior (pp. 1-64). New York:

 

Basic Books.

Ainsworth, M.D.S., Blehar, M.C., Waters, E., & Wall, S.(1978). Patterns of
attachment: A psychological study of the strange situation. I-Iillsdale, NJ: Lawrence
Erlbaum.

Ainsworth, M.D.S., & Bowlby, J. (1991). An ethological approach to personality
development. American Psychologist, 46, 333-341.

Ballard-Reish, D., & Elton, M. (1992). Gender orientation and the Bern Sex Role
Inventory: A psychological construct revisited. Sex Roles. 27. 291-306.

Baldwin, M.W., & Fehr, B. (1992). Instability of attachment style ratings:
Unreliability of measurement or variability of construct? Unpublished manuscript,
University of Winnipeg.

Bartholomew, K. (1990). Avoidance of intimacy: An attachment perspective.
Journal of Socigand Personal Relationships, 7, 141-178.

Bartholomew, K., & Horowitz, L.M. (1991). Attachment styles among young
adults: A test of a f our-category model. Journal of Personality and Social Psychology,61_,
226-244.

Bern, S.L. (1974). The measurement of psychological androgyny. Journal of

Consulting and Clinical Psychology, 42, 155-162.

 

 

153
Bem, S.L. (1977). On the utility of alternative procedures for assessing

psychological androgyny. Journal of Consulting and Clinical Psychology, 45, 196-205.

Bem, S.L. (1984). Androgyny and gender schema theory: A conceptual and
empirical integration. In T.B. Sonderegger (Ed.), Nebraska synapgsium on motivation (pp.
179-226). Lincoln: University of Nebraska Press.

Bieger, GR. (1985). Bern Sex-Role Inventory. In DJ. Keyser & R.C. Sweetland
(Eds), Test critiques, vol. II] (pp. 51-57). Kansas City, MO: Test Corporation of
America.

Billings, A. (1979). Conflict resolution in distressed and nondistressed married
couples. Journal of Consulting and Clinical Psychology, 47, 368-376.

Bowen, M. (1978). Family therapy in clinical practice. New York: Aronson.

Bowlby, J. (1969/1982). Attachment and loss: Vol.1. Attachment (2nd ed.). New
York: Basic Books.

Bowlby, J. (1973). mohment and loss: Vol. 2. Separation: Anxiety and anger.
New York: Basic Books.

Bowlby, J. (1979/1992). The making and breaking of affectiong bond; London:
Tavistock/Routledge.

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

Bowlby, J. (1988). A secure @eiment—child attachment afl healthy hugaa
development. New York: Basic Books, Inc.

Brennan, K., Shaver, P.R., & Tobey, A.E. (1991). Attachment styles, gender,
and parental drinking. Journal of Sgaizard Personal Relationships, 8, 451-466.

Bretherton, I. (1992). The origins of attachment theory: John Bowlby and Mary
Ainsworth. Developmentg Psychology,_2_8_, 759-775.

Byng-Hall, J. (1995). Creating a secure family base: Some implications of
attachment theory for family therapy. family Process, 34. 45—58.

154
Chodorow, N. (1978). The reproduction of mothering. University of California

Press.

Cohn, D.A., Silver, D.H., Cowan, C.P., Cowan, P.A., & Pearson, J. (1992).
Working models of childhood attachment and couple relationships. Journal of Family
Issues 13, 432-449.

 

Collins, N.L., & Read, SJ. (1990). Adult attachment, working models, and
relationship quality in dating couples. Journal of Personality Modal Psychology, 58,
644-664.

Condon, S.M., Cooper, C.R., & Grotevant, H.D. (1984). Manual for the analyses
of family discourse, Psychological Documents. 14. (Ms. No. 2616).

Cooper, C.R., Grotevant, H.D., & Condon, SM. (1982). Methodological
challenges of selectivity in f arnily interaction: Assessing temporal patterns of individuation.
Journal of Mar—rriagQand The Family, 44, 749-754.

Deaux, K. (1984). From individual differences to social categories: Analysis of a
decade's research on gender. American Psycholoigst. 39, 105-116.

Duggan, E.S., & Brennan, K.A. (1994). Social avoidance and its relations to
Bartholomew's adult attachment typology. Journal of Social_ and Personal Relationship_s,
1_1, 147-153.

Feeney, J.A. (1994). Attachment style, communication patterns, and satisfaction
across the life cycle of marriage. Personal Relationships, 1, 333-348.

Feeney, J .A., & Noller, P. (1990). Attachment style as a predictor of adult
romantic relationships. Journal of Pererity and Social Psychology,_2_, 281-291.

Feeney, J .A., & Noller, P. (1991). Attachment style and verbal descriptions of
romantic partners. Jourflof SociaLand Personal Relationships, 8, 187-215.

Feeney, J .A., & Noller, P. (1992). Attachment style and romantic love:

Relationship dissolution. Australian Joumal of Psychology, 44, 69-74.

 

 

155
Feeney, J.A., Noller, P., & Hanrahan, M. (1992). Assessing adult attachment:

Developments in the conceptualization of security and insecurity. In M. B. Sperling &
W.H. Berman (Eds), Attachment in adults: Theory. asses_s_ment and treatment (pp. 128-
152). New York: Guilford Press

Feeney, J.A., Noller, P., & Patty, J. (1993). Adolescents' interactions with the
opposite sex: Influence of attachment style and gender. Journal of Adolescence, 16. 169-
186.

Feeney, J.A., Noller, P., & Callan, V.J. (1994). Attachment style, communication
and satisfaction in the early years of marriage. Advances in Personal Relationships, 5, 269-
308.

Fiala, K.B., & Pietromonaco, PR. (1991, June). Thinking about a romantic
relationship: Attachment style and gender influence emotional reactions and perceptions.
Paper presented at the annual meeting of the American Psychological Society, Washington,
DC.

George, C., Kaplan, N. & Main, M. (1985). The Berkeley Adult Attachment
Interview. Unpublished protocol, Department of Psychology, University of California,
Berkeley.

Gilbert, LA. (1987). Female and male emotional dependency implications for the
therapist-client relationship. Professional Psychology: Theogy, Research and Practice. 18.
555-561.

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

Griffin, D., & Bartholomew, K. (1994). Models of the self and other“. Fundamental
dimensions underlying measures of adult attachment. Jorml of Personalng and Social
Psychology, 67, 430-445.

156
Grotevant, H.D., & Cooper, CR. (1983). The role of family communication

patterns in adolescent identity and role taking. Paper presented at the biennial meeting of the
Society for Research in Child Development, Detroit, Michigan.

Grotevant, H.D., & Cooper CR. (1985). Patterns of interaction in family
relationships and the development of identity exploration in adolescence. 9114
Development. 56, 415-428.

Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment
process. Journal of Persona—limind Social Psycholgyhsg, 511-524.

Hazan, C., & Shaver, P. (1990). Love and work: An attachment-theoretical

. __“-.. _ ._
m "
h.—
z

perspective. Journal of Personality and Social Psychology, 59, 270-280.

Hazan, C., & Shaver, PR. (1994). Attachment as an organizational framework for
research on close relationships. Psychological Inguifl, 5, 1-22.

Heavey, C.L., Layne, C., & Christensen, A. (1993). Gender and conflict structure
in marital interaction: A replication and extension. Journal of Consulting and Clinical
Psychology, 61, 16-27.

Hendrick, C., & Hendrick, S. (1992). Attachment theory and close adult
relationships. Psychological Inquiry, 5, 38-41.

Hiller, D.V., & Philliber, W.M. (1985). Internal consistency of the Bem Sex-Role
Inventory, Social Psychology Quarterly, 48, 373-380.

Holmes, J.G., & Boon, SD. (1990). Developments in the field of close
relationships: Creating foundations for intervention strategies. Personality and Social
Psychology Bulletin, 16, 23-41.

Kaplan, AG. (1979). Toward an analysis of sex-role related issues in the
therapeutic relationship. Psychhatry. 43. 112-120.

Karen, R. (1994). Becoming attached: Unfolding the mystegy of the inf ant-mother
bond and its immct on later life. New York: Warner Books, Inc.

 

 

157
Kirkpatrick, L.A., & Davis, KB. (1994). Attachment style, gender, and

relationship stability: A longitudinal analysis. Journal of Personality and Social
Psychology, 66, 502-512.

Kobak, R.R., & Hazan, C.(l991). Attachment in marriage: Effects of security and
accuracy of working models. Journal of Personality and Social Psychology, _6_0_, 861-869.

Kobak, R.R., & Sceery, A. (1988). Attachment in late adolescence: Working
models, affect regulation, and representations of self and others. Child Development, 59,
135- 146.

Koren, P., & Carlton, K. (1980). Marital conflict: Relations among behaviors,
outcomes, and distress. Joumal of Conaultingand Clinical Psychology,i8, 460-468.

Kunce, L.J., & Shaver, PR. (1994). An attachment-theoretical approach to
caregiving in romantic relationships. In K. Bartholomew & D. Perlman (Eds), Advances
in personal relationships. Vol. 5 (pp. 205-237). London: Jessica Kingsley Publishers, Ltd.

Lang-Takac E. & Osterweil, Z. (1992). Separateness and connectedness:
Differences between the genders. Sex Roles. 27. 277-289.

Lamke, L. K., Sollie, D.L., Durbin, R.G., & Fithatrick, J.A. (1994).
Masculinity, femininity and relationship satisfaction: The mediating role of interpersonal
competence. Joumalfiof Socifial_ and Personal Relationshig, 11, 535-554.

Lee, J.A. (1977). The colors of love. Englewood Cliffs, NJ: Prentice-Hall.

Levy. MB. (1989). Integration of lovestyles and attachment styles: Cross-Ma
influences and a clarification of concepts, measurement, and conceptualization.
Unpublished doctoral dissertation, University of Cincinnati, Ohio.

Levy, M.B., & Davis, K.E.(1988). Lovestyles and attachment styles compared:

Their relations to each other and to various relationship characteristics. Journal of Social
and Person;al Relationships, 5, 439-471.

Lopez, F.G. (1993). Cognitive processes in close relationships: Recent findings

and implications for counseling. Journal of Counseling and Development. 71. 310-315.

158
Lopez, F.G., Gover, M.R., Leskela, J., Sauer, E., Schirmer, L., & Wyssmann,

J. (in press). Attachment styles, shame, guilt, and collaborative problem-solving
orientations. Personal Relationship_s.

Maccoby, EB. (1990). Gender and relationships: A developmental account.
American Psvfichologist. 45. 513-520.

Main, M., Kaplan, N., & Cassidy, J. (1985) Security in infancy, childhood, and
adulthood: A move to the level of representation. Monographs of the Society for Resear__c_h
in Child Development. 50. 66-104.

Margolin, G., & Wampold, BE. (1981). Sequential analysis of conflict and accord
in distressed and nondistressed marital partners. Journal of Consultingand Clinical
Psychology, 49, 554—567.

Markman, H.J., Silvem, L., Clements, M., & Kraft-Hanak, S. (1993). Men and
women dealing with conflict in heterosexual relationships. Journal, of Social Issues, 49.
107- 125.

Mayseless, O. (1991). Adult attachment patterns and courtship violence. F_am_ily
Relation_s. 40. 21-28.

Mikulincer, M., & Erev, I. (1991). Attachment style and the structure of romantic
love. British Journal of Social Psychologyfl, 273-291.

Mikulincer, M., Florian, V., & Tolmacz, R. (1990). Attachment styles and fear of
personal death: A case study of affect regulation. Journal of Personalig and Social
Psychology, 58, 273-280.

Mikulincer, M., & Nachshon, D. (1991). Attachment styles and patterns of self-
disclosure. Journflf Personaljtaflnd Social Psychology,_61, 321-331.

Miller, J. B. (1991). The construction of anger in women and men. In J.V. Jordan,
A.G. Kaplan, J. B. Miller, I.P. Stiver, & J.L. Surrey, Women's growth in connection.
(pp. 181-196). New York: The Guilford Press.

159
Mintz, L.B., & O'Neil, J.M. (1990). Gender roles, sex, and the process of

psychotherapy: Many questions and few answers. J_oar;nal of Coaaselinwd
Development, 68, 381-387.

Parkes, CM. (1972) Bereavement. New York: International Universities Press.

Pietromonaco, P.R., & Carnelley, KB. (1994). Gender and working models of
attachment: Consequences for perception of self and romantic relationships. MOD—81
Relationship. 1. 3-26.

Pistole, MC. (1989). Attachment in adult romantic relationships: Styles of conflict
resolution and relationship satisfaction. Journal of Social_ and Persomj Relationships, 6,
505-510.

Pistole, MC. (1993). Attachment relationships: Self -disclosure and trust mug
of Mental He_alth Coun_selingy1_5, 94-106.

Pistole, MC. (1994). Adult attachment styles: Some thoughts on closeness-
distance struggles. F_amily Process, 33, 147-159.

Pistole, MC. (1994, August). Ended love relationships: Differences in how it
hurts. Paper presented at the annual meeting of the American Psychological Association,
Los Angeles.

Pistole, C.M., Clark, E.M., & Tubbs, A.L. (1995). Love relationships:
Attachment style and the investment model. Journal of Mental Health Counseling, 17, 199-
209.

Rothbard, J.C., Roberts, L.J., Leonard, K.E., & Eiden, RD. (1993, August).
Attachment styles and marital interaction behavior. Paper presented at the annual meeting of
the American Psychological Association, Toronto.

Rusbult, C.E., Johnson, D.J., & Morrow, GD. (1986). Impact of couple patterns
of problem solving on distress and nondistress in dating relationships. Journal of

Personality and Social Psychology, 50, 744-753.

160
Rusbult, C.E., Verette, J., Whitney, G.A., Slovik, L.F., & Lipkus, I. (1991).

Accommodation processes in close relationships: Theory and preliminary empirical
evidence. Journal of Personality and Social Psychologyfifl, 53-78.

Satterfield, W.A., & Lyddon, W.J. (1995). Client attachment and perception of the
working alliance with counselor trainees. Journal of Counseling Psychology,_zlg, 187-189.

Sayers, S.L., & Baucom, D. H. (1991). Role of femininity and masculinity in
distressed couples' communication. Journal of Pemoaality and SocMsycholo 61,
641-647.

Scharfe, E, & Bartholomew, K. (1994a). Reliability and stability of adult
attachment patterns. Personal Relationships, 1, 23-43.

Scharfe, E, & Bartholomew, K. (1994b, August) Stability of adult attachment
representations: A two year follow-up. Paper presented at the annual meeting of the
American Psychological Association, Los Angeles.

Scharfe, E, & Bartholomew, K. (1995). Accommodation and attachment
representations in young couples. Journal of Social and Personal Relationships, 12, 389-
401.

Senchak, M., & Leonard, KE. (1992). Attachment styles and marital adjustment
among newlywed couples. Journal of Social_ and Persopal Relationships, 9, 51—64.

Shaver, P., & Hazan, C. (1987). Being lonely, feeling in love: Perspectives from
attachment theory. low of Socia_l_B__eaavior and Personmg, 105-124.

Shaver, P., & Hazan, C. (1988). A biased overview of the study of love. Jamal
of Socia_l_and Personal Relationships, 5, 473-501.

Shaver, P., & Hazan, C. (1993). Adult romantic attachment; Theory and evidence.
In D. Perlman & W. Jones (Eds), Advances in parsonal relationships, 4, (pp. 29-70).
London: Jessica Kingsley Publishers, Ltd.

 

 

161
Shaver, P., Hazan, C., & Bradshaw, D. (1988). Love as attachment: The

integration of three behavioral systems. In R. J. Stemberg & M.L. Barnes (Eds), Tag
Psychology of Love, (pp. 68-99). New Haven: Yale University Press.

Simpson, J .A. (1990). Influence of attachment styles on romantic relationships.
Journal of Personality and Social Pavchology, _5_9, 971-980.

Simpson, J.A., Rholes, W.S., & Nelligan, J.S. (1992). Support-seeking and
support-giving within couples in an anxiety-provoking situation: The role of attachment
styles. Jo_umal of PersonaligLand Social Psychology, Q 434-446.

Spence, .I.T., & Helmreich, RH. (1981). Androgyny versus gender schema: A
comment on Bem's gender schema theory. Psychological Review, _§8_, 365-368.

Watzlawick, P. (1966). A structured family interview. Family Process. 5. 256-
271.

Weiss, R.S. (1982) Attachment in adult life. In C.M. Parkes & J. Stevenson-Hinde
(Eds), 111apl_ace iattachment in human behavior. (pp. 171-184). New York: Basic
Books.

Weitzman, L. (1985). The divorce revolt_rtion. New York: The Free Press.

White, BB. (1989). Gender differences in marital communication patterns. PM!
Process. 28. 89-106.

Winstead, B.A., & Derlega, V.J. (1993). Gender and close relationships: An

introduction. Journal of Social Issues, 49. 1-9.

   

"IIIIIIIIIIIIIIII“