Citation
The influence of ritual use, institutional support, and conflict resolution style on lesbian couples' relationship satisfaction

Material Information

Title:
The influence of ritual use, institutional support, and conflict resolution style on lesbian couples' relationship satisfaction
Creator:
Pace, Caroline
Publication Date:
Language:
English
Physical Description:
vii, 219 leaves : ; 29 cm.

Subjects

Subjects / Keywords:
Cohabitation ( jstor )
Conflict resolution ( jstor )
Dyadics ( jstor )
Female homosexuality ( jstor )
Marital satisfaction ( jstor )
Problem solving ( jstor )
Rituals ( jstor )
Sexual orientation ( jstor )
Social psychology ( jstor )
Women ( jstor )
Counselor Education thesis, Ph.D ( lcsh )
Dissertations, Academic -- Counselor Education -- UF ( lcsh )
Alachua County ( local )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1999.
Bibliography:
Includes bibliographical references (leaves 195-218).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Caroline Pace.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
030366361 ( ALEPH )
42843960 ( OCLC )

Downloads

This item has the following downloads:


Full Text











THE INFLUENCE OF RITUAL USE, INSTITUTIONAL SUPPORT, AND CONFLICT RESOLUTION STYLE ON LESBIAN COUPLES'
RELATIONSHIP SATISFACTION












By

CAROLINE PACE


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1999















ACKNOWLEDGMENTS


I would like to extend my deepest thanks to those family members, friends, and colleagues who have given me guidance, support, and encouragement throughout the process of conceptualizing, designing, implementing, and writing this dissertation. It was my incredible good fortune that Dr. Ellen Amatea was "randomly" assigned to be my advisor when I began my studies at the University of Florida. She has served as advisor, teacher, supervisor, and chair of my dissertation committee. I am so appreciative of the rigorous precision with which she thinks and writes as well as her endless patience, prompt feedback, and scholarly brilliance. Dr. David Miller, who speaks the language of statistics with exquisite grace, has been a veritable fountain of information and a delightful human being. Dr. Phyllis Meek-a wisewoman, mentor, and feminist role model-has believed in me with unflinching certainty and impeccably modeled for me the skills of critical thinking and feminist analysis. I am grateful to Dr. Mary HowardHamilton for the opportunities she has given me to collaborate with her and for the vote of confidence.

Words cannot adequately express the deep gratitude I feel for my colleagues at the Gainesville Family Institute. The mentoring, training, and supervision I received from the GFI faculty and students have been nothing short of transformative-both personally and professionally. I am privileged to know Andres Nazario-a wonderfully feisty and compassionate person. He has offered me numerous wonderful opportunities








to challenge myself and to do so in the warmth of his support. I amn also grateful to my friend, Lynne Barolet-Fogarty, for her consistent encouragement and the opportunity to learn from her spirit. I am appreciative of Herb Steier, an extraordinary role model of irreverence.

Without my community of lesbian friends and family this project would not have been possible. Participation in a study that requires one to acknowledge one's lesbian identity and to examine one's relationship is potentially challenging and risky. I would like to thank the 90 couples who took that risk. Also, I am indebted to Woody Blue (who encouraged me to "exploit" her outgoing personality in the search for potential participants) for her help with data collection. My lesbian family-Mickey Barron, Suzanne Miller, and Lorelei Esser-kept me nurtured and encouraged.

My family of origin has offered inspiration and invaluable support. I am grateful to my mom for her unconditional love and absolute faith in me. The memory of my father-a scholar and a gentle person-has been my constant companion on this j ourney. Additionally, I appreciate the example set by my Aunt Caroline, the first woman in the family to get a Ph.D.

Last, but certainly not least, I am indebted to my partner, Jude, for her steadfast love and unquestioning faith in me. Her patience and calm have sustained our relationship. Her intelligence and willingness to challenge me have inspired my creativity.


iii















TABLE OF CONTENTS

Page

ACKNOWLEDGMENTS ..................................................... ii

ABSTRACT .................................................................. vi

CHAPTERS

1 INTRODUCTION ..........................................................1I

Theoretical Framework ................................................ 3
Scope of the Problem................................................... 7
Need for the Study..................................................... 22
Purpose ................................................******"*** 25
Research Questions..................................................... 25
Definition of Termis.................................................... 26
Organization of the Study ............................................. 30

2 REVIEW OF THE LITERATURE........................................ 32

Introduction........................................................... 32
Lesbian Couples: A Contextual View.................................. 32
Research on Lesbian Relationship Satisfaction .......................... 33
Research on Social Support Influences ................................. 53
Research on Ritual ..................................................... 74
Research on Conflict Resolution ....................................... 96
Summary .............................................................111I

3 METHODOLOGY ....................................................... 113

Statement of Purpose ................................................ 113
Hypotheses........................................................... 113
Delineation of Relevant Variables..................................... 114
Data Analysis ........................................................ 117
Description of the Population ........................................ 118
Sampling Procedures.................................................. 118
Subjects.............................................................. 121


iv









Data Collection....................................................... 126
Instrumentation....................................................... 127

4 DATA ANALYSIS AND RESULTS..................................... 144

Analysis Procedures .................................................. 144
Analysis Results...................................................... 150
Hypothesis Testing ................................................... 158
Summary............................................................. 162

5 DISCUSSION............................................................ 163

Overview of the Study................................................. 163
Research Sample ...................................................... 164
Association Between Couple Relationship Satisfaction and Partner
Agreement Regarding Conflict Resolution Style Utilized .............165
Association Between Couple Relationship Satisfaction and Conflict
Resolution Style Utilized............................................. 168
Association Between Couple Relationship Satisfaction and Degree of
Relationship Ritualization ..........................................171
Association Between Couple Relationship Satisfaction and Ritual
Meaningfulness...................................................... 175
Association Between Couple Relationship Satisfaction and Institutional
Support for the Relationship ........................................ 178
Recommendations ..................................................... 179
Summary .................................................................... 187

APPENDICES

A LETTER TO PARTICIPANTS ...............................188

B PARTICIPANT SELECTION PROTOCOL.................... 191

C DEMOGRAPHIC DATA SHEET ............................. 191

D INSTITUTIONAL SUPPORT SCALE......................... 193

REFERENCES............................................................... 195

BIOGRAPHICAL SKETCH ................................................. 219


v















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE INFLUENCE OF RITUAL USE, INSTITUTIONAL SUPPORT, AND CONFLICT RESOLUTION STYLE ON LESBIAN COUPLES' RELATIONSHIP SATISFACTION

By

Caroline Pace

May, 1999

Chairperson: Ellen Amatea, Ph.D.
Major Department: Counselor Education

The purpose of this study was to examine the influence of five variables on the

relationship satisfaction of lesbian couples. Four of these variables focused on aspects of partner interaction: (a) the extent of ritualization in couple life, (b) the meaningfulness of ritualization in couple life, (c) the type of conflict resolution style utilized by couples, and

(d) the agreement between partners regarding conflict resolution style. The first two of these variables were measured by the Family Ritual Questionnaire while the latter two variables were measured by the Conflict Resolution Style Inventory. The fifth variable-perceived institutional support-focused on the couples' participation in the larger social context as measured by the Institutional Support Scale. The sample for this study consisted of 180 women (90 couples) who self-identified as lesbians and had lived together in a committed monogamous relationship for at least one year. Bisexual women


vi








and transgendered women were excluded from the study as were women living with children.

Regression analyses were used to evaluate the contribution of the five variables in predicting couple relationship satisfaction as measured by averaging the Dyadic Adjustment Scale scores of both partners in a relationship. Statistically significant (p<.05) associations were found between couple relationship satisfaction and two of four conflict resolution styles-positive problem-solving and conflict engagement styles. No significant associations were found between couple relationship satisfaction and partner conflict resolution style agreement, extent of relationship ritualization, ritual meaningfulness, and perceived institutional support. However, post hoc analyses revealed indirect effects of relationship ritualization and ritual meaningfulness on relationship satisfaction when conflict resolution style was controlled. Discussion of these results, the study's limitations, and suggestions for future research with this population were then presented.


vii















CHAPTER 1
INTRODUCTION


Love relationships play a significant role in the lives of most human beings-including lesbians. Traditionally, researchers interested in determining what factors contribute to stable and satisfying intimate relationships have focused primarily on heterosexual married couples (Actelli, 1988; Davis & Oathout, 1987; Gottman, Coan, Carrere, & Swanson, 1998; Jacobson & Addis, 1993; Rusbult, 1983). Only recently have researchers begun to examine the love relationships of "nontraditional" couples such as cohabiting heterosexuals and gays and lesbians (Cardell, Finn, & Maracek, 198 1; Dailey, 1979; Duffy & Rusbult, 1986; Kurdek, 1988a, 1992b, 1998; Kurdek & Schmidt. 1986a,

1 986b, 1 986c). However, these researchers have tended to look only at the internal processes of couple life--how couples report making decisions, resolving conflicts, or structuring their roles. There has been scant attention given to couples' perceptions of the larger social context and its influence on couple relational processes.

Because of the negative cultural attitudes towards homosexuality, the consequences of partners choosing to acknowledge publicly their lesbian love relationship can present significant problems that would not be encountered by heterosexual partners. For example, researchers (Carter & McGoldrick, 1989; ImberBlack & Roberts, 1992; Roth, 1985; Slater, 1995) have identified that the lack of marker events for lesbian couples--such as weddings or the birth of a child--often creates


I





2


difficulties for couples trying to define their change in status from single to couplehood. Additionally, life cycle transitions such as retirement, serious illness, or death can be problematic for lesbian couples. Disqualification of the couple relationship by the larger society may lead to public ceremonies in which partners have to disguise their feelings and their relationship, thus intensifying their problems rather than promoting a sense of connectedness and continuity.

The lesbian couple's acknowledgement of their relationship in the larger

community is very likely to be influenced by the extent to which either or both partners perceive it is "safe" to acknowledge their relationship. According to FBI statistics, hate crimes targeting gays and lesbians rose 42% from 1991 to 1995 (Mills, 1997). In Florida, the number of hate crimes committed against members of sexual orientation minorities has nearly tripled in the last 2 years ("Hate Crimes on the Rise," 1998). Couples who feel they cannot risk being open about their relationship may feel isolated from their work, school, or social community. On the other hand, those couples who choose to risk acknowledging their relationship may form connections within a community of likeminded lesbian couples but feel isolated from the larger heterosexual community and its cultural institutions. Lesbian partners often deal with a loss of status in the larger community if they are open about their relationship or with a sense of alienation and oppression if they maintain secrecy about their relationship (Carter & McGoldrick, 1989; Roth, 1985). Consequently, lesbian couples often must find ways to develop their love relationships in a cultural context that, at best, marginalizes or regard these relationships as invisible and, at worst, subjects them to overt hostility and violence.






3


Although the larger social context appears to have a significant influence on the lives of lesbian couples, no studies to date have examined the relative effect of institutional support on a couple's relationship satisfaction. Instead, most researchers have looked exclusively at the internal dynamics of the couple. This study sought to assess the relative influence on couple relationship satisfaction of support from the larger context as well as the influence of three aspects of couple interaction.

Theoretical Framework

To understand better the factors influencing lesbian relationship satisfaction, a theoretical framework is needed which addresses the contributions of internal couple dynamics and the external dynamics of the larger cultural context in which the couple is embedded. No existing theory integrates both perspectives. Consequently, this study was based upon two theoretical traditions: Gottman's (1994) balance theory and Kurdek's relationship appraisal model.

Gottman's (1994) balance theory has been used to conceptualize the interactional processes by which heterosexual married couples sustain satisfying relationships. This theory is based on multimethod assessments of significant aspects of couple interaction (e.g.. partner's problem solving, power, affect, distance regulation, and physiology) in an 8-year longitudinal investigation of 79 couples as well as laboratory observations and ongoing assessments of four cohorts of couples in different stages of the family life cycle. Over the past two decades, Gottman and his colleagues conducted empirical research that ultimately led to the formulation of balance theory. According to this theory, satisfied couples maintain a balance in their ratio of positive to negative interactions.





4


Two theoretical premises underlie balance theory. First, satisfied couples develop a ratio of positive to negative affective interactions as well as the ratio of positive to negative problem solving interactions. Over time, the negative aspects are "balanced" by positive aspects. For satisfied couples, this balance is "tilted" toward the positive. Gottman (1994) described balance theory as "an ecology of marital behaviors in which a ration of positivity to negativity that is highly tilted toward positivity needs to be maintained" (p. 380).

The second theoretical premise underlying balance theory is the notion that partners in satisfied couples develop a "fit" regarding preferred interactional style. Gottman (1994) postulated three types of "stable" couples: volatile, avoiding, and validating. Each of these couple types results in a "stable," satisfied relationship in which positive and negative interactions are balanced. Volatile relationships balance a high level of argument and conflict with significant amounts of laughter and passionate romance. In other words, high volumes of positive interactions balance high volumes of negative interactions. On the other hand, high levels of neutral affect characterize avoiding types of couple relationships. The avoiding couple type "appears to involve a minimization of the importance of disagreement. It results in a good deal of calm interaction, but pays the price with emotional distance" (Gottman, 1994, p. 182). In avoiding couple types, a low level of negative interactions is balanced by a low level of positive interactions. Finally, the validation type of couple balances careful selection of when to confront conflict with warmth and "we-ness." Partners in these relationships demonstrate little intensity regarding either disagreement or passion. A moderate amount of negativity is balanced by a moderate amount of positivity in validating couples.





5


Gottman (1994) addressed cultural variation in the distribution of couple types, but only from an ethnic point of view. He postulated that while one couple type may be more prevalent in a specific ethnic culture, the eroding effect of an unbalanced (e.g., high ratio of negativity to positivity) relationship would remain the same. Gottman has not addressed the applicability of the balance theory to lesbian couples although he reported a study-yet to be published-regarding lesbian and gay couples' conflict resolution strategies (cited in Gottman, 1994).

Although Gottman' s balance theory (Gottman, 1994) incorporates behavior,

cognition, and physiology, it is primarily an interactional model which does not account for contextual forces that may influence couples. His theory is based on the study of heterosexual marital couples--couples that benefit from the social sanctions given their relationships by the larger cultural context. Members of marginalized and stigmatized groups, such as lesbians, often do not receive the institutional support for their relationships afforded heterosexual couples. Brown (I1989a) described three aspects of lesbian couples' cultural context that cut across all demographic characteristics: marginality, biculturalism, and normative creativity. Marginality is expressed as a feeling of "otherness" and a cognitive recognition of minority status. According to Brown (1 989b, 1991), partners in lesbian relationships may experience this feeling of "otherness" as nontraditional gender role identity because they do not receive the traditional recognition afforded a female via her socially proscribed association with a male. For some lesbians this may mean feeling as if they do not belong to their gender; however, this feeling of gender inappropriateness often changes as women are able to





6


reevaluate belief systems and validate the importance of nontraditional gender roles in couple relationships (Vargo, 1987).

Unlike other minority groups that share a common experience, lesbians come

from a variety of religions, racial and ethnic cultures, age groups, or socioeconomic status (Brown, 1989b). In fact, there is no singular lesbian couple reality. Lesbian partners are members of both the heterosexual and homosexual cultures; they are, in essence, bicultural. Almost all lesbians have lived at least the early part of their lives under the influence of assumed heterosexuality, and even after affirming a lesbian identity, many women continue to pass as heterosexual in certain circumstances (Brown, 1 989b). Although this bicultural experience is not unique to lesbian couples, unlike their counterparts in other oppressed minority groups, they do not share their minority status with members of their own families of origin (Brown, 1 989b).

According to Brown (I1989b), as a result of their minority status in a group for which previously established norms do not exist, lesbian couples are afforded the opportunity to create their own norms. Accordingly, lesbians may enjoy tremendous freedom to invent their relationships and roles in a creative manner (Grahn, 1984). One of the most important ways that families create for their members a sense of belonging is through the establishment of rituals. Lesbian families have not been afforded many of the traditional rituals afforded their heterosexual counterparts such as wedding showers, marriage ceremonies, and anniversary celebrations, yet they may utilize what Brown (1 989b) terms normative creativity to devise rituals of their own.

Building on Brown's contextual model and Gottman's balance conceptualization, Kurdek (1998) hypothesized a model of relationship satisfaction applicable to both






7


heterosexual married and homosexual cohabiting couples. Kurdek (1998) posited that relationship satisfaction was related to both internal forces within the relationshipy--such as shared agreements about problem solving, autonomy, equality, and intimacy-and external forces outside the relationship-such as institutional supports. Kurdek (1998) found that institutional supports act as barriers to leaving the relationship, particularly for heterosexual married couples. The inclusion of external forces as significant influences on relationship satisfaction gives Kurdek's (1998) model a contextual focus that Gottman's (1994) balance theory does not address.

The present study utilized both Gottman's (1994) balance theory and Kurdek's (1998) model of internal and external forces in conceptualizing potential sources of influence on relationship satisfaction. Specifically, in this study the assessment of the balance in relationship positivity and negativity were represented by measurement of the variables of degree of ritualization and conflict resolution style respectively. In addition to these internal factors, the influence of external factors-institutional support for the relationship-were assessed as well.

Scope of the Problem

While it is difficult to "measure" just how many women are involved in lesbian love relationships, estimates suggest that over 70% of lesbians are in committed relationships ("The Gay Market", 1992; Gonsiorek & Weinrich, 1991). No census data exist regarding the number of lesbians in the United States although estimates range from 2% to 26% of the population, with most estimates in the 2% to 10% range (Brown, 199 1; Paul & Weinrich, 1982). Yet, the importance of lesbian relationships is not limited to the women in them. In fact, members of lesbian couples live in a greater family context as





8


daughters, sisters, aunts, cousins, and parents. It has been estimated that one in every five families has a gay or lesbian child (Dahiheimer & Feigal, 1991) and between 6 million and 14 million children have at least one lesbian or gay parent (Kantrowitz, 1996). The extensive impact of this sexual orientation minority on such a great number of people supports the need for further research in this area.

Lesbian communities have been shown to significantly value therapy (Morgan, 1992). In fact, lesbians comprise a substantial amount of the women seeking psychological services (Morgan, 1992). While many therapist training programs have begun to include coursework regarding the needs of clients from culturally diverse backgrounds, a focus limited to racial, ethnic, and/or religious minorities may result in the exclusion of lesbian populations from these classes (Markowitz, 1991). Martin (1982) has suggested that therapists operate from a perspective negatively biased as a result of having been socialized in a culture that simultaneously stigmatizes and renders invisible gays and lesbians. In order to work effectively with lesbian clients, therapists must re-educate themselves, yet a dearth of empirical research exists to help clinicians understand this population.

Research on Lesbian Couple Relationships

Historically, minority issues as a whole have been excluded from academic study and literature; consequently, until very recently homosexuality has been omitted as a focus of research inquiry. The topic of lesbian relationship issues-essentially a minority issue within a minority issue-has received even less attention. Essentially, women in lesbian relationships have suffered a greater oppression due to a double minority status-as a sexual orientation minority and a gender minority. This double minority status





9


means that documentation of lesbian couple issues have seldom been noted in the literature, even by couples and family therapists (Brown, 1991; Clark & Serovich, 1997). In their review of literature on lesbian, gay, and bisexual issues in couples and family therapy journals during the period 1975-1995, Clark and Serovich (1997) found that lesbian couples in particular and homosexual families in general were rarely discussed. Of the 13,217 articles reviewed from 17 couples and family therapy j ournals, only 77 (.006%) focused on lesbian, gay, or bisexual issues (Clark & Serovich, 1997). Of these 77, only 15 (19.5%) specifically addressed lesbian couple relationships. Thus, articles regarding lesbian couples accounted for .00 1% of the total number of articles published in the 20-year span covered in the analysis. Further research is needed to understand this important segment of society.

What research has been conducted seems to have followed three traditions. Initially, scholars were interested in determining what made lesbians and gay men different from their heterosexual counterparts. This led to a narrow focus on the psychological makeup of the individual rather than on lesbians and gay men as participants in larger relationship systems (e.g., as partners in couple relationships or members of families). With the shift in psychology away from individual pathology and toward family systems theories, comparative studies sought similarities between lesbian and gay couples and families and their heterosexual counterparts. Most recently, through the larger lens of contextual conceptualizations, researchers have begun to investigate the concurrent processes by which lesbians and gay men are both oppressed by and members of multiple systems including familial, work, educational, legal, and medical institutions.





10


The contributions from each of these three traditions is reviewed in the following sections.

The search for differences: Individual identity models. Initial investigations

regarding homosexuality were influenced by the growth of psychiatry and sexology in the late 19th century, as were the efforts to pathologize gays and lesbians as deviant individuals (Faderman, 1991). This early literature decontextualized partners in lesbian and gay relationships by viewing them as individuals participating in sinful behavior. In part, it was this shift away from viewing lesbians and gays as sinners engaged in an "unnatural" behavior and toward regarding them as individuals with deviant personality types that allowed the conceptualization of homosexual identity development to arise. Although the concept of homosexual identity development perpetuated an individual focus in lesbian and gay research, it had a profound effect on the depathologizing of sexual orientation minorities.

Lesbian identity development, the process by which a lesbian identity is formed and managed, is considered one of the most important factors that affects the lives of lesbians individually and in couples (Cass, 1979; Chapman & Brannock, 1987; Coleman, 1982; DeMonteflores & Schultz, 1978; Minton & McDonald, 1984; Silberkraus, 1995; Sophie, 1986; Troiden, 1979). Unlike their heterosexual counterparts, lesbians must develop and maintain their identities within a larger cultural context structure that continues to control, legislate, oppress, and stigmatize any kind of sexual variance. In the

1 950s several significant studies supported a shift in the conceptualization of homosexuality from deviant to a normal variant of sexuality. One of the first to do this was a study by Kinsey and colleagues (Kinsey, Pomeroy, Martin, & Gebhard, 1953)




'1'~'~


which documented that lesbian feelings and experiences were common to a widespread group of women rather than an isolated few as had previously been thought. Later investigations would focus on the effect of lesbian identity development on relationship quality (Jordan, 1995; McGuire, 1995; Silberkraus, 1995; Wendell, 1986).

By the 1960s the feminist movement emerged. Women united in the call for the deconstruction of patriarchal power structures. Concurrently, lesbians and gay men began rallying against institutional heterosexism. In contrast to prior research--which focused on the etiology of the "deviant" sexual behavior and its cure--new research focused on the "normal" personality development of lesbian and gay individuals (Armon, 1960; Dank, 1971; Evans, 1970; Freedman, 1971; Green, 1972; Hopkins, 1969; Loney, 1971; Saghir & Robbins; 1973; Siegelman, 1972; Thompson, McCandless, & Strickland, 1971; Wilson & Green, 1971). Although this research was born out of a political context, investigations regarding lesbian and gay relationships were ignored in favor of a continued focus on the individual.

It was Cass' (1979) groundbreaking work on a five-stage model of homosexual identity formation that influenced researchers to shift to conceptualizing homosexuality from the broader framework of identity rather than simply from a (sexual) behavioral perspective. Subsequently, other models of lesbian and gay identity development have been proposed (e.g., Bell, Weinberg, & Hammersmith, 198 1; Chapman & Brannock, 1987; Coleman, 1982; deMonteflores & Schultz, 1978; Minton & McDonald, 1984; Silberkraus, 1995; Sophie, 1986; Troiden, 1979). It was the influence of homosexual identity development literature that spurred the eventual declassification of homosexuality as a mental illness. In 1973 the American Psychiatric Association voted






12


to remove homosexuality from its official list of mental disorders. Subsequently, in 1975 the American Psychological Association elected to support homosexuality as a legitimate lifestyle. It was not until 1991 that the American Association of Marriage and Family Therapy addressed the issue of sexual orientation minorities by including a nondiscrimination clause in its code of ethics (American Association of Marriage and Family Therapy, 199 1). Furthermore, it was only last year that the Commission of the Accreditation for Marriage and Family Therapy and Education adopted a similar policy (Clark & Serovich, 1997).

The search for similarities: Lesbian relationships and families. As researchers searched for differences between lesbians and gay men and their heterosexual counterparts, they began to focus on the lesbian family. Initially, individuals were once again the center of this research. Investigators sought to determine whether or not differences existed between lesbian and heterosexual mothers and the children raised in these households (Bozett, 1987; Chesler, 1986; Cramer, 1986; Gibbs, 1989; Gottman, 1990; Green, Mandel, Hotvedt, Gray, & Smith, 1986; Hoeffer, 1981; Kirkpatrick, Smith, & Roy, 1981; Mackin, 1980; Pagelow, 1980). For instance, more similarities than differences were found regarding the psychological well being, gender identity, sex role socialization, and availability of male roles models for children raised by lesbian mothers and children raised by heterosexual mothers (Cramer, 1986; Gibbs, 1989; Gottman, 1990; Green et al., 1986; Hoeffer, 198 1; Kirkpatrick et al., 198 1).

Lesbian couples and families have only recently been publicly recognized as

legitimate family forms. Although many studies have attempted to delineate the factors that contribute to relationship satisfaction in heterosexual couples (see review by Lewis





13


& Spanier, 1979), until recently very few empirical studies have sought to determine the factors that influence relationship satisfaction in lesbian and gay couples (Bell & Weinberg, 1978; Kurdek, 1998; Peplau, Padesky, & Hamilton, 1982). Only in the last 20 years have empirical studies compared similarities among lesbian, gay, and heterosexual couples with regard to the degree and correlates of relationship satisfaction (Blumstein & Schwartz, 1983; Cardell et al., 198 1; Duffy & Rosbult, 1986; Kurdek, 198 8b, 1992a, 1992b, 1998; Kurdek & Schmitt, 1986a, 1986b, 1986c, 1987b; Peplau & Cochran, 1990; Schneider, 1986). For instance, in one of the classic investigations to compare lesbian, gay, and heterosexual couples, Blumstein and Schwartz (1983) studied decision making regarding money, work, and sex. They examined the interpersonal dynamics within four types of relationships: cohabiting lesbian, cohabiting gay, cohabiting heterosexual, and married heterosexual. Blumstein and Schwartz (1983) interviewed 300 couples and administered questionnaires to more than 3,500 couples; the focus of this large body of work was on the internal "workings" of the relationship. While informative, such comparative studies have decontextualized lesbian relationships by ignoring significant cultural influences such as sexism and heterosexism.

The contextual view: Lesbian couples as a minority family form. Although

research on heterosexual couples provides some insights into the relationships of lesbian and gay couples, the larger social constructs of gender and sex-role socialization prevent a parallel application of such research to members of same-gender couples (Morin, 1977). Because the socialization of males and females differs in this culture, it is important to investigate the unique contribution each gender brings to ways of relating in couples (Vargo, 1987). Because lesbian and gay relationships are comprised of two people of the





14


same gender, the couple inherently gets a double measure of the positive and the negative aspects of their gender socialization. On the other hand, in heterosexual couples the positive and negative aspects of female cultural socialization as well as the positive and negative aspects of male cultural socialization are combined. Thus, studies based exclusively on heterosexual couples may not be generalizable to lesbian and gay couples.

According to many researchers (Murphy, 1992), lesbian couples differ from other couple types (including other homosexual couples) because they are "doubly devalued and oppressed" as women in a culture that is sexist and misogynist and as homosexuals in society that is homophobic and heterosexist. Furthermore, lesbian couples face other stresses that may affect their relationships including socially imposed restrictions on the public display of affection and the lack of visible role models, culture, and history (Riddle & Sang, 1978; Sang, 1978). Additionally, lesbian couples must find ways to thrive in a culture that invalidates and even overtly opposes the legitimization of this family form (Maggiore, 1988). Several researchers have emphasized the need for studies that focus exclusively on lesbians and lesbian relationships (Burch, 1986; Peplau, 1982; Peplau, Padesky, & Hamilton, 1982).

Social Support and Lesbian Relationship Satisfaction

Very few researchers have focused on examining the social support networks of lesbian couples. No research, to date, has examined the relationship between institutional support and relationship satisfaction in lesbian couples. Instead, most researchers have investigated the dimension, types, and degrees to which individuals who are lesbian receive social support from family of origin, partners, gay-lesbian friends, heterosexual friends, and health professionals (Albro & Tully, 1979; Aura, 1985; Chafetz, Sampson,





15


Beck, & West, 1974; Grossman & Kerner, 1998; Jordan, 1995; Jordan & Deluty, 1998; Kurdek, 1988a; Kurdek & Schmitt, 1987a, 1987b; Levy, 1992; Meyer, 1989; Olson, 1988; Walsh, 1995; Wayment & Peplau, 1995; Wood, 1983). One of the first studies of lesbian social support found that lesbians received most support from lesbian and gay friends rather than from heterosexual friends and family (Chafetz et al., 1974). In a later investigation, Albro and Tully (1979) discovered that lesbians relied on members of the homosexual micro-culture for support rather than members of the heterosexual macroculture. Furthermore, researchers support the notion that social support enhances lesbians' personal well being (Meyer, 1989; Olson, 1988; Walsh, 1995; Wayment & Peplau, 1995). Most models of homosexual identity development acknowledge the importance of social support as a factor in the development of a positive lesbian or gay identity (Cass, 1979; Chapman & Brannock, 1987; Coleman, 1982; Troiden, 1979; Walsh, 1995). Although these studies have been informative, most of these investigations have examined social support reported by individuals. It is unclear whether social support of the couple contributes to relationship satisfaction in lesbian couple relationships.

Leavy and Adams (1986) determined that lesbians involved in committed

relationships scored significantly higher on measures of social support and self-esteem than those not in relationships. Similarly, Walsh (1995) found that lesbians who reported the highest social support were those with high self-esteem, low internalized homophobia, and involved in committed relationships. Even among studies with samples composed of both lesbian couples and gay couples, social support systems have seldom been assessed. In a study of lesbian and gay couples, Kurdek (1988a) found that friends





16


were the primary sources of social support. He determined that psychological well being was not influenced by lack of family support. Kurdek (I1988a) hypothesized that lesbians and gay men have compensated for family support lost due to the stigma associated with a homosexual identity by creating an extended family of friends.

The three contextual elements of lesbian couples' experience described by

Brown-marginality, biculturalism, and normative creativity-illustrate key factors that may serve to unify lesbian communities. Members of minority groups frequently connect with each other via those aspects of their culture that differentiate them from members of the majority group. These differentiating experiences, characteristics, beliefs, and customs enhance the sense of belonging and affiliation. Empirical research has validated Brown's (I1989b) notion that many lesbians receive support from other members of the lesbian community that they do not receive from heterosexual women or family of origin members (Cass, 1979; Chafetz et al., 1974; Kurdek, 1988a; Walsh, 1995).

Studies of exclusively lesbian, exclusively gay, and combined lesbian and gay samples have focused on interpersonal social supports provided couples. However, few researchers have focused on examining the influence of other sources of support on these love relationships. Institutional supports confer status and privileges upon members of a couple in regard to financial, legal, religious, familial, and social welfare domains. Adams and Jones (1997) proposed that institutional barriers are needed to sustain heterosexual marriages. In one of the few studies of institutional supports given lesbian, gay, and heterosexual couples, Kurdek (1998) hypothesized that barriers to leaving a relationship would apply pressure to couples to stay together in spite of personal unhappiness. Specifically, he postulated that the individuals in socially sanctioned






17


relationships (e.g., heterosexual married couples) would report significantly greater barriers to leaving the relationship than their counterparts in socially stigmatized relationships (e.g., cohabiting gay or lesbian couples). As expected, he found that partners in lesbian and gay couples reported fewer barriers to leaving the relationship than did married heterosexual partners (Kurdek, 1998). Furthermore, he determined that barriers to leaving the relationship effectively predicted couple dissolution for lesbian couples and heterosexual married couples. Although this study suggests that lesbian and gay couples perceive fewer institutional supports for their relationships than do heterosexual married couples, there is little information regarding the extent to which institutional supports influence relationship satisfaction. This study addressed this limitation.

Couple and Family Ritual Use and Relationship Satisfaction

Lesbian couples are often afforded the opportunity to create their own norms as a result of minority status in a group for which previously established norms do not exist. Consequently, lesbians may enjoy tremendous freedom to invent their relationships and roles in a creative manner (Brown, 1989b; Gralin, 1984). One of the most important ways that couples and families create a sense of belonging for their members is through the establishment of rituals.

Researchers have begun to investigate the phenomenon of family rituals as attention in the literature has shifted from a focus on family pathology to resiliency. However, this research focused exclusively on heterosexual families. Furthermore, most of the literature regarding the influence of ritual in families defined family in a traditional manner-a constellation of parents, children, and first-order extended family relations





18


such as grandparents, aunts, uncles, and cousins. Very little research has focused exclusively on the couple subsystem.

Family rituals have been recognized as substantial forces that organize behavior in family systems (Bossard & Boll, 1950; Fiese, 1992; Fiese, Hooker, Kotary, & Schwagler, 1993; Sameroff & Fiese, 1992; Wolin & Bennett, 1984). According to Fiese (1992) "family rituals appear to exert influence on family life by pairing meaning and affect with patterned interaction" (p. 634). Rituals have been shown to play an important role in the development and maintenance of family satisfaction, identity, support, strength, and cohesiveness (Bossard & Boll, 1950; Curran, 1983; Fiese, 1992; Jenson, James, Boyce, & Hartnett, 1983; Meredith, 1985; Meredith, Abbott, Lamanna, & Sanders, 1989; Schvaneveldt & Lee, 1983; Wise, 1986; Wolin & Bennett, 1984). Empirical research has demonstrated that rituals have a positive and even a protective effect on families. Investigations have documented that rituals help families create, maintain, and transmit to future generations a paradigm of family life and shared belief systems (Bossard & Boll, 1950; Reiss, 1982), enhance family members' sense of belonging (Wolin, Bennett, Noonan, & Tietelbaum, 1980), promote bonding and intimacy among family members (Meredith, 1985), preserve a sense of meaningfulness (Schvaneveldt & Lee, 1983), and create and maintain family cohesion (Wolin et al., 1980).

The positive aspects of family rituals comprise much of the research to date; however, some evidence has been documented regarding potential negative aspects of rituals. Meredith (1985) and Wolin and Bennett (1984) noted that rituals may negatively impact families if they are perceived as meaningless, superficial, or boring. Additionally,





19


rituals may be imposed on family members in an unhealthy manner, or they may be used to control and manipulate inappropriately (Bossard & Boll, 1950; Meredith, 1985). Furthermore, rituals may require too much effort or may result in family conflict that is disruptive to members (Meredith, 1985). Finally, rituals may be used to perpetuate physical, sexual, and emotional abuse.

Almost all of the empirical research regarding rituals has focused on their place in family life rather than in couple life. Furthermore, researchers have only studied heterosexual families. Bruess (1994) conducted one of the few empirical studies to date describing the types of rituals used by intimate partners. Analyzing the rituals reported by marital partners and members of friendship dyads, Bruess (1994) identified seven major types of rituals. These consisted of (a) "couple time" rituals which are comprised of enjoyable activities, escape episodes, and togetherness rituals, (b) "symbolic" rituals which consist of private codes, play, and celebration rituals, (c) "spiritual" rituals, (d) communicationo" rituals, (e) "daily routines," (f) "intimacy expression" rituals, and (g) "patterns-habits-mannerisms" (Bruess, 1994).

Fiese and her colleagues compared 115 married heterosexual couples making the transition to parenthood (Fiese et al., 1993). They found that those couples reporting high levels of meaningfulness associated with their rituals were more satisfied with their marriages than their counterparts who reported relatively hollow rituals.

There has been no empirical research conducted concerning lesbian couples' use of ritual. Although anecdotal accounts have supported the notion that ritual plays an important role in lesbian couple life (Laird, 1988; Slater, 1995), Slater (1995) identified ritual as a method by which lesbian couples may "compensate for the usually absent





20


social recognition of their family status" (p. 69). She posited that the heterosexist nature of society explicitly excludes lesbian couples from the definition of family and thereby bars them from participation in traditional rituals that affirm family membership and facilitate developmental transitions (Slater, 1995). The present investigation extended the research on family ritual to include lesbian partners. Conflict Resolution and Relationship Satisfaction

Conflict resolution has been referred to as the pivotal dynamic in understanding couples' interactions (Braiker & Kelly. 1979). Researchers have suggested that couple satisfaction is mediated by conflict resolution strategies utilized by partners (Bowman, 1990; Christensen & Heavey, 1990; Gottman, 1994; Gottman & Krokoff, 1989; Markman, Floyd, Stanley, & Storaasli, 1988). Furthermore, investigations relying on samples of heterosexual couples consistently demonstrated that couples who are able to resolve conflicts mutually are more satisfied than couples who do not (Christensen & Shenk, 1991; Epstein, DeGiovanni, & Jayne-Lazarus, 1978; Gottman & Levenson, 1992; Markman et al., 1988; O'Leary & Smith, 1991).

To date, four studies have investigated the contribution of conflict resolution strategies to relationship satisfaction among lesbian and gay couples. Most recently, Kurdek (199 8) developed a multidimensional model of relationship quality that he studied in a sample of heterosexual married, cohabiting lesbian, and cohabiting gay couples. He hypothesized that couples benefit from both internal and external forces that sustain the relationship. Specifically, Kurdek (1998) postulated that gender-linked forces within the relationship would facilitate happiness in each partner while institutional forces outside the relationship would promote the continuation of the relationship in spite






21


of personal dissatisfaction. He proposed that it is the combination of internal and external forces that significantly contributes to relationship quality. Kurdek (1998) investigated four internal forces (constructive problem solving, intimacy, autonomy, and equality) and one external force (barriers to leaving the relationship). Kudek (1998) found that constructive problem-solving appraisals by both partners were significantly associated with relationship satisfaction.

In a prior study, Kurdek (1988) investigated the conflict resolution strategies of lesbian and gay couples. He found that partners in gay relationships did not differ significantly from partners in lesbian relationships regarding the types of conflict resolution strategies they utilized to resolve relationship disputes. In an evaluation of the relationship happiness of these couples, Kurdek (I 988b) identified "positive" problemsolving strategies such as infrequent withdrawal, infrequent negative initiatives, and frequent positive efforts significantly influenced relationship satisfaction among both lesbian and gay couples. These results supported those of Falbo and Peplau (1980) who compared lesbian and gay couples' conflict resolution styles in an assessment of power strategies. They found no significant difference between gay partners and lesbian partners.

Metz, Rosser, and Strapko (1994) compared 36 heterosexual, 36 lesbian, and 36 gay couples matched for age and length of relationship. Overall, they found that couple type did not differentiate conflict resolution style utilized. The differences they did find seemed linked to gender rather than sexual orientation. They did note that lesbian partners reported greater satisfaction and more hopefulness in addition to greater use of positive conflict resolution styles (Metz et al., 1994).





22


Need for the Study

Only recently have researchers begun to examine the factors that contribute to the maintenance or enhancement of lesbian relationships in the face of institutionalized oppression on the basis of sexual orientation. This study extended this research by exploring possible influences on lesbian relationship satisfaction. Four potential influences were examined: perceived institutional support for the relationship, conflict resolution style, degree of couple ritualization, and ritual meaningfulness.

A significant body of research has explored the social support networks of lesbians as individuals. However, little research has focused on support for lesbian couples. Furthermore, many of the investigations of lesbian couples' support systems have relied on data drawn from only one member of the couple, thereby excluding potentially significant dyadic differences. Whiile investigations regarding the supports utilized by individual women-even those in committed relationships-provide useful information, there is a need to examine the influence of institutional supports on lesbian couples' satisfaction.

Conflict resolution styles have been shown to affect significantly relationship satisfaction in heterosexual samples (Christensen & Shenk, 1991; Epstein et al., 1978; Gottman & Levenson, 1992; Markman et al., 1988; O'Leary & Smith, 1991). However, there is limited empirical investigation of the influence of conflict resolution style on lesbian relationship satisfaction. Moreover, results from existing studies assessing the influence of lesbian partners' conflict resolution on relationship satisfaction have been inconsistent (Kurdek, 1998; Metz et al., 1994; Falbo & Peplau, 1988). Consequently,





23


additional research is needed to clarify the role of conflict resolution style on the lesbian couples' relationship satisfaction.

Most intimate relationships are sustained via a dynamic and complex network of intentional and less-than-intentional strategies. Lesbian couples, unlike their heterosexual counterparts, often must sustain their intimate relationships in the face of adversity or even overt hostility from family, "friends," and society at large. Rituals have been shown to contribute to the stability and cohesion of family life in healthy families (Wolin et al., 1980) and in those families facing stresses such as severe illness (Fiese, 1997) or alcoholism (Wolin & Bennett, 1984; Wolin et al., 1980). In addition, rituals have been linked with greater relationship satisfaction among married couples, especially those facing stressful life transitions (Fiese et al., 1993). No research to date has assessed the contribution of ritual to relationship satisfaction among lesbian couples. Consequently, a need exists to assess the contribution of ritualization to relationship satisfaction among lesbian couples.

Much of the research on lesbian couples has been plagued with methodological difficulties. First, as with any stigmatized or marginalized population, it has been difficult for researchers to obtain a truly representative sample. Many studies have relied on network or snowball sample selection techniques. Particularly absent from many of these studies have been lesbians of color and of lower socioeconomic status. Another problem with the methodology of a number of studies of lesbian couples has been the reliance on only one member of the couple from which to gather data (e.g., Cardell et al., 198 1; Duffy & Rusbult, 1986; Falbo & Peplau, 1980). Of course, it is unlikely that both members of a couple will view the relationship identically. Critical





24


information may be overlooked when only one partner is studied. Furthermore, even when both partners have been included in the study, some researchers have chosen to treat each individual as a separate unit of analysis (e.g., Howard, Blumstein, & Schwartz, 1986; Kurdek & Schmitt, 1986a; Metz et al., 1994). As Kurdek (1998) pointed out, "because partners' scores share common variability-as is usually the case with dyadic data--the test statistic associated with type-of-couple differences is too large, resulting in inaccurate estimates of effect size and p values that are smaller than they should be" (p. 554).

Another limitation of a majority of past lesbian couple research has been the focus on intrapartner links rather than on cross-partner linkages. Intrapartner links are those relationships between an individual's perception of relationship satisfaction and predictor variables while cross-partner links are those relationships between one partner's appraisal of satisfaction and the predictor variable scores for the other partner. The importance of assessing both types of links is based on the theory that an individual's experience of hisher relationship is simultaneously influenced by both his-her own appraisal of the relationship as well as by his-her partner's appraisal of the relationship (Kenny, 1996; Kurdek, 1998).

A final problem with some of the past lesbian couple research is the reliance on measurements with unknown psychometric properties. A number of investigations have utilized assessment instruments that have no documented reliability or validity (e.g., Blumstein & Schwartz, 1983). Furthermore, some researchers have significantly altered assessment measures originally created and normed on heterosexual or, in some cases, all





25


male samples in order to adapt them for use with a lesbian sample (e.g., Kus, 1985). Assumptions about validity simply cannot be maintained under these circumstances.

Purpose

The purpose of this study was to examine the influence of five variables on the relationship satisfaction among lesbian couples. Four of these variables focused on aspects of partner interaction: (a) the extent of ritualization in couple life, (b) the meaningfulness of ritualization in couple life, (c) the type of conflict resolution style utilized by couples, and (d) the agreement between partners regarding conflict resolution style. The fifth variable-perceived institutional support-focused on the couples' participation in the larger social context. The sample for this study of lesbian couples included women who self-identified as lesbians. Bisexual women and transgendered women were excluded from the study. Furthermore, this sample consisted only of lesbian couples who lived together and had been in a committed monogamous relationship for at least 1 year.

Research Questions

Limited empirical research has been conducted on lesbian relationships. The research that does exist has failed to consider the joint contribution to relationship satisfaction of the internal dynamics between partners and the external influence of the cultural context in which these relationships are sustained. Because the larger cultural context exerts a powerful influence on lesbian couple life, examining the processes by which such couples construct an identity as a couple and create shared rituals together appears to be an appropriate target for research. The goal of this study was to expand the body of knowledge used to understand lesbian couple satisfaction by examining potential





26


forces inside the relationship--such as ritualization and conflict resolution style-as well as forces outside the relationship-such as institutional supports-which may contribute to relationship satisfaction. To this end, the following research questions were posed:

1. What style of conflict resolution do lesbian partners develop and how does a

couple's conflict resolution style contribute to their relationship satisfaction?

2. What shared rituals are developed by lesbian couples and how do these

contribute to relationship satisfaction?

3. How do lesbian couples develop an identity as a couple in the face of a larger

social context that may be critical of their relationship? What types of

supports and barriers are perceived by the couple and how do these contribute

to relationship satisfaction?

Definition of Terms

Conflict resolution style. The manner in which couples resolve disputes is the conflict resolution style. Four styles of conflict resolution were assessed in this investigation: positive problem solving, conflict engagement, withdrawal, and compliance.

Homosexual. Homosexual is a term that refers both to lesbians and gay men. Generally, this term is considered to be a clinical diagnostic label and, therefore, a pathology-based descriptor.

Institutional support. There are a number of societal institutions that support and validate heterosexual couples. For instance, religious sects sanction heterosexual marriage. Additionally, these marriages receive other forms of institutional support including legal and financial supports such as public recognition and even tax benefits.





27


Institutional supports such as these are not easily accessible to lesbian couples due to the marginalized position that lesbians occupy in a society that stigmatizes both the lesbian identity and lesbian relationships. For the purpose of this investigation, institutional support includes aid, recognition, and status afforded by the following institutions: religious, educational, social (including family, friends, peers, and lesbian communityy, financial, work, social welfare, medical, and legal.

Lesbian. After nearly two decades of research on lesbian identity, there still is no consensus about the definition of this concept (Cass, 1984; Falco, 1991). The range of meanings spans the spectrum from defining oneself as lesbian to consistent lesbianrelated behavior (Falco. 1991). A distinction must be made between the term "homosexual identity" and "lesbian identity." Homosexual identity, a narrower term than lesbian identity, refers strictly to sexual behavior (Levine & Evans, 1991). On the other hand, lesbian identity encompasses the global experience of being lesbian. Additionally, the term "homosexual" often is found offensive by lesbians as it has historically been used to pathologize and diagnose what is now considered a healthy variation of human sexuality. In contrast, the term lesbian connotes a positive description of the emotional, political, and lifestyle (including but not limited to sexual) aspects of life (Levine & Evans, 1991). Finally, a differentiation between lesbian identity and sexual orientation must be addressed. Sexual orientation, which may have biological components, is a behavioral construct whereas lesbian identity is both a social and a cognitive construct (Falco, 1991). Although there is much debate regarding possible genetic influences, most recent theories concur that an individual is not born with a fully constructed lesbian identity; rather, this identity is constructed and maintained through interaction with others





28


(Levine & Evans, 1991; Richardson, 1981). Lesbian identity development is not synonymous with lesbian behavior or same-sex attachments. Although same-sex sexual behavior may indeed be a precursor to the development of a gay or lesbian identity, individuals frequently engage in same-sex sexual behavior without defining themselves as lesbian (Cass, 1984; Elliott, 1985, 1992; Levine & Evans, 1991). Thus, behavior alone in insufficient to define a person as lesbian.

Lesbian (or gay) identity. Lesbian identity refers to the total experience of being a lesbian; the core of a woman's essence as a lesbian. This includes but is not limited to sexual behavior. Emotional, affectional, political, and lifestyle aspects of lesbian experience also comprise lesbian identity (Levine & Evans, 1991). According to Brown (1995), lesbian is "primarily a self-ascribed definition held by a woman over time and across situations as having primary sexual, affectional, and relational ties to other women" (p. 4). This definition does not include women who are bisexual or transgendered. Furthermore, individual women may engage in same-sex sexual behavior that is exploratory or transitory without defining themselves as lesbians. These women would not be considered lesbians for the purpose of this study.

Lesbian couple-lesbian family. In this study, a lesbian couple was operationalized as two lesbians who are currently committed to a monogamous relationship with each other and who have been together for at least 1 year. According to Slater (1995), "lesbian couples constitute complete family units in and of themselves, the terms couple, family, and family of creation (can) be used interchangeably" (p. 32). These terms were used synonymously in this study.





29


Ritual. Bossard and Boll (1950), in their groundbreaking study of 400 families, identified ritual as "a prescribed procedure, arising out of family interaction, involving a pattern of defined behavior, which is directed toward some specific end or purpose, and acquires rigidity and a sense of rightness as a result of its continuing history" (p. 29). Later, Wolin and Bennett (1984) defined ritual as symbolic communication systematically repeated across time because of its perceived enhancement of family satisfaction. Wolin and Bennet (1984) identified three types of family rituals: patterned interactions, family celebrations, and family traditions. Patterned interactions are rituals which occur with the greatest frequency in family life but are the least planned. Dinner time and bedtime routines as well as weekend activities are examples of patterned interactions. Family celebrations include holidays and special occasions celebrated by many members of a culture. Religious holidays, weddings, and Independence Day celebrations are examples of family celebrations. Family traditions are less culturespecific and more family-specific. Family reunions, birthdays, and anniversaries are examples of family traditions.

Dyadic ritual. Dyadic rituals are defined for the purpose of this study as the

symbolic, patterned interactions shared by both members of a couple. Although members of other dyadic relationships such as friends or coworkers may also share interpersonal rituals, the focus of this study was on those rituals shared by committed couples, specifically lesbian couples.

Ritual dimensions. Ritual dimensions include occurrence, roles, routines, attendance, affect, symbolic significance, continuation, and deliberateness (Fiese & Kline, 1993). Rituals are performed through roles and routines. Families may have





30


expectations about who is to attend rituals as well as who is to be excluded. Rituals have affective meaning for the family as well as a symbolic interpretation. Rituals are passed down through generations and may require advance planning and preparation.

Ritual settings. Ritual setting are those circumstances and situations in which rituals are performed such as dinner time, weekends, vacations, annual celebrations, special celebrations, religious holidays, and cultural and ethnic traditions (Fiese & Kline, 1993).

Ritualization. A term used by Betcher (1987), ritualization refers to the ongoing and patterned interactions between members of a couple which are endowed with relational significance. The couple may rely on a variety of types of ritualization to maintain and enhance the relationship. For instance, daily rituals such as sharing dinner every night may serve to connect the couple while "renewal" rituals such as revisiting the site of their first date may serve to maintain the relationship over time. Furthermore, commemorativev" rituals such as sending holiday greeting cards together and annual rituals like attending family reunions also contribute to the maintenance of the relationship (Betcher, 1987).

Organization of the Study

Chapter 2 of this investigation is a review of the related literature. Following

Chapter 2 is a delineation of the methodology in Chapter 3, including a statement of the purpose of the study, hypotheses, description of relevant variables, data analyses, description of the population, subjects, sampling procedures, data collection, and instrumentation. Chapter 4 is a presentation of the results of the statistical analyses of






31


the data. Chapter 5 includes a discussion of the results, implications for therapy, limitations of the study, and suggestions for further investigation.















CHAPTER 2
REVIEW OF THE LITERATURE

Introduction

Lesbian couples have not been studied extensively by social scientists. In this chapter the literature on social support factors affecting lesbian satisfaction is reviewed. In addition, literature on couple conflict resolution style and the use of ritual in couple and family life is discussed.

Lesbian Couples: A Contextual View

Brown (1 989b) described three aspects of lesbian couples' cultural context that cut across all demographic characteristics: marginality, biculturalism, and normative creativity. Marginality is expressed as a feeling of "otherness" and a cognitive recognition of minority status. According to Brown (Il989b, 199 1), partners in lesbian relationships may experience this feeling of "otherness" as nontraditional gender role identity because they do not receive the traditional recognition afforded a female via her socially proscribed association with a male. For some lesbians this may mean feeling as if they do not belong to their gender; however, this feeling of gender inappropriateness often changes as women are able to reevaluate belief systems and validate the importance of nontraditional gender roles in couple relationships (Vargo, 1987).

Unlike other minority groups that share a common experience, lesbians come

from a variety of religions, racial and ethnic cultures, age groups, or socioeconomic status (Brown, 1 989b). In fact, there is no singular lesbian couple reality. Lesbian partners are


32






33


members of both the heterosexual and homosexual cultures; they are, in essence, bicultural. Almost all lesbians have lived at least the early part of their lives under the influence of assumed heterosexuality, and even after affirming a lesbian identity many women continue to pass as heterosexual in certain circumstances (Brown, 1 989b). Although this bicultural experience is not unique to lesbian couples, unlike their counterparts in other oppressed minority groups, they do not share their minority status with members of their own families of origin (Brown, 1 989b). According to Brown (I1989b), as a result of their minority status in a group for which previously established norms do not exist, lesbian couples are afforded the opportunity to create their own norms. Accordingly, lesbians may enjoy tremendous freedom to invent their relationships and roles in a creative manner (Grahn, 1984). One of the most important ways that families create for their members a sense of belonging is through the establishment of rituals. Lesbian families have not been afforded many of the traditional rituals afforded their heterosexual counterparts such as wedding showers, marriage ceremonies, and anniversary celebrations, yet they may utilize what Brown (1 989b) terms normative creativity to devise rituals of their own.

Research on Lesbian Relationship Satisfaction

Researchers have shown that lesbians highly value permanent love relationships. In one of the first surveys of lesbian partners, Bell and Weinberg (1978) found that 25% of 293 respondents reported having a permanent relationship with one partner to be the mostt important thing in life," while another 35% said it was "very important." Further, Peplau and Cochran (1990) estimated that 45% to 80% of lesbians are involved in committed relationships. Additionally, many lesbians develop lifelong partnerships.





3 4


More than 90% of the 706 lesbian couples surveyed by Bryant and Demnian (1994) reported that they were committed to their relationships for life. While some investigations have found similar levels of relationship satisfaction among lesbian, gay, and heterosexual couples (Blumstein & Schwartz, 1983; Duffy & Rusbult, 1986; Kurdek & Schmitt, 1 986c), other studies have noted higher reports of satisfaction among lesbian couples (Bryant & Demian, 1994; Kurdek, 1988b; Zacks, Green, & Marrow, 1988).

Research regarding relationship satisfaction among lesbian couples has followed three traditions: (a) comparative studies involving heterosexual couples and homosexual couples (lesbian and/or gay), (b) comparative studies of different types of homosexual relationships (lesbian, gay, and/or bisexual), and (c) descriptive studies involving samples comprised exclusively of lesbian couples. Studies comparing relationship satisfaction among various heterosexual and homosexual couple types have found very few differences (Blumstein & Schwartz, 1983; Cardell et al., Duffy & Rusbult, 1986; Green, Bettinger, & Zacks, 1996; Kurdek, 1994c, 1998; Kurdek & Schmitt, 1986a, 1986b, 1986c, 1987a; Lange, 1989; Norton, 1995; Oberstone & Sukonek, 1976; Peplau & Cochran, 1990). Although comparisons between heterosexual couples and lesbian or gay couples are informative, lesbian and gay couples face some unique stressors-such as homophobia--that may affect their relationships (Falco, 1991, Murphy, 1992). Further, while studies comparing lesbian couples with gay couples yield interesting information, these investigations may overlook the distinct factors that may influence couples in which both partners are women (Falco, 1991, Murphy, 1992). Therefore, studies of lesbian couples are warranted. Research on lesbian relationships have examined a number of correlates of relationship satisfaction including disclosure of lesbian identity (Green et






35


al., 1996; Jordan, 1995; Kurdek, 1988a; 1995b; Oberstone & Sukoneck, 1976) and satisfaction with support (Kurdek & Schmidt, 1 986a, 1 986b, 1 986c, 1 988a; Oberstone & Sukoneck, 1976).

Research on Lesbian. Gay. and Heterosexual Couple's Satisfaction

Relationship satisfaction is one of the most frequently studied variables in

investigations of heterosexual as well as lesbian and gay couples. Early comparative studies sought to determine similarities and differences between lesbian relationships and heterosexual relationships. More similarities than differences were found in general measures of relationship satisfaction. Consequently, more recent investigations have focused on possible correlates of relationship satisfaction in these different couple types.

Research comparing levels of relationship satisfaction. Morin (1977) used the term "heterosexual bias" to refer to the pathologizing of lesbian relationships. Heterosexual bias refers to the tendency of research to reflect societal beliefs that heterosexual relationships set the normative standards against which all other relationships are to be measured; in essence, heterosexual relationships are conceptualized as "normal" and "healthy." Consequently, early studies of lesbian and gay relationships focused on the relative "adjustment" these couples reported in comparison with heterosexual couples. Cardell et al. (1981) found no difference between 10 lesbian, 5 gay, and 10 heterosexual couples in measures of "adjustment." Lesbian and gay couples were just as likely as heterosexual couples to score in the "well-adj usted" range on traditional measures of marital satisfaction. In a similar investigation, Peplau (1982) compared matched samples of lesbian, gay, and heterosexual couples on measures of love and liking their partner. No significant differences were found. The findings of






36


these and similar studies suggest that lesbian and gay relationships can be fully as functioning and healthy as heterosexual relationships.

Several investigations have compared overall relationship satisfaction among heterosexual, gay, and lesbian couples. One of the largest of such studies compared married heterosexual, cohabiting heterosexual, cohabiting lesbian, and cohabiting gay couples (Blumstein & Schwartz, 1983). Across the four couple types, reports of relationship satisfaction were similar. Very few couple type differences were found. Furthermore, these researchers reported that emotional, sexual, and financial commitment predicted relationship endurance.

In a series of comparative investigations, Kurdek and Schmitt (1 986a, 1 986b)

surveyed a sample comprised of four types of couples: 56 cohabiting lesbian couples, 50 cohabiting gay couples, 35 cohabiting heterosexual couples, and 44 married heterosexual couples. Overall, they found fewer differences than similarities among couple types in each of the investigations. Kurdek and Schmitt (I1986c) found no significant differences between these four types of couples in regard to level of relationship satisfaction. Relying on data from self-report measures, they found that dyadic attachment levels were related to relationship satisfaction, love for partner, and liking of partner. Further, they found that cohabiting heterosexual couples reported the lowest levels of dyadic attachment of the four couple types. These investigations support the notion that lesbian and gay partners are just as satisfied with their relationships as are their heterosexual counterparts.

Kurdek (1998) investigated five predictors of relationship dissolution and change in relationship satisfaction over time in a sample of 236 married heterosexual, 51 lesbian






37


couples, and 66 gay cohabiting couples. He found no differences based on couple type on measures of initial relationship satisfaction. Similarly, heterosexual and lesbian and gay couples did not differ on the trajectory of change over a 5-year time period although each couple showed a decline in relationship satisfaction.

Research on the applicability of heterosexual models of relationships. A number of studies have investigated the applicability of traditionally heterosexual models of relationships to lesbian and gay couples. In one such study, the investment model (Rusbult, 1980, 1983) of relationship maintenance and appraisal was examined. The investment model conceptualizes relationship satisfaction as an ongoing process in which partners evaluate and re-evaluate the relative rewards and costs of continuing the relationship. Duffy and Rusbult (1986) compared 25 lesbians, 25 gay men, 25 heterosexual women, and 25 heterosexual men matched for age and education in an investigation of relationship satisfaction and commitment. They used a series of selfreport measures designed to assess both the specific and the global aspects of relationship costs, relationship rewards, alternatives to the relationship, and investment in the relationship. Global measures of satisfaction and commitment were also collected. These researchers found more similarities than differences among the groups. No significant differences among the groups were found regarding reports of relationship costs, relationship rewards, alternatives to the relationship, and investment. Differences on the basis of gender were identified; specifically, women reported higher levels of investment and greater degrees of commitment. Heterosexuals reported greater costs and marginally greater investment than did either lesbians or gays. For all couples types, greater satisfaction was associated with higher levels of rewards and lower levels of






38


costs. Likewise, no differences by couple type were reported in predictors of commitment. These researchers concluded that gender rather than sexual orientation accounted for differences.

Hatton (1995) surveyed 103 lesbians in cohabiting relationships. Results of this study offer support for Rusbult's (Duffy & Rusbult, 1986) investment model. Hatton (1995) found positive relationships between relationship satisfaction, commitment, and personal power; however, the relationship power model was not fully supported. Although relationship satisfaction was the focus of this study, only individual data were analyzed rather than couple data. These studies provide support for the applicability of heterosexual models of relationship development and appraisal to lesbian and gay couples.

In another study of traditional heterosexual models of relationship development, Kurdek and Schmitt (1986a) used a subsample from previous research comprised of 33 married heterosexual couples, 34 cohabiting heterosexual couples, 41 gay couples, and 47 lesbian couples in an investigation focused on stages of relationship development. They were particularly interested in indicators of relationship quality such as agreement, satisfaction with affection and sex, beliefs about sex, cohesion, and tension. They identified similarities regarding relationship development among these types of couples; specifically, couples in the first year of relationship (blending stage) reported relatively greater satisfaction while couples in the second and third (nesting stage) years of relationship development reported relatively greater levels of stress and disillusionment. Couples in the fourth and fifth (maintaining stage) years were also assessed. Results from this investigation supported the predicted curvilinear relationship between






39


relationship quality and stage of relationship such that higher quality would be reported at blending and maintaining stages. This study supports the belief that lesbian and gay relationships follow developmental patterns similar to those found in heterosexual couples. One of the few significant differences identified among the four couple types was in the area of support: married couples reported greater support from family than did lesbian and gay couples. In fact, lesbian couples reported receiving less family support than did either gay couples or heterosexual couples. However, these differences did not influence their reports of relationship satisfaction. This study supports the use of developmental models of relationships originally theorized regarding heterosexual relationships with lesbian and gay couples.

Kurdek (I 992b, 1 994a) continued to investigate the applicability of heterosexual models of relationships with lesbian and gay couples in a 4-year longitudinal study. He surveyed 39 married heterosexual couples, 39 lesbian cohabiting couples, and 39 gay cohabiting couples, matched on length of time living together, regarding relative levels of relationship satisfaction. This study combined variables from the contextual model (e.g., satisfaction with social support, dysfunctional beliefs regarding relationships) and the interdependence model (e.g., perceived rewards from the relationship, perceived costs to the relationship, relationship satisfaction). Unlike most previous studies that used the individual as the unit of analysis, this investigation used the couple as the unit of analysis.

Kurdek (I 992b) found evidence to support the connection between relationship satisfaction and the two models in both lesbian and gay couples. He found that lesbian couples and heterosexual couples did not differ on level of satisfaction, although gay couples reported lower satisfaction. Kurdek (1 992b) cautioned that this may be due to a






40


sampling artifact in the selection of gay couples. He found that lesbian partners and wives did not differ on a set of individual difference scores--dysfunctional relationship beliefs, social support, expressiveness, and negative affectivity--while husbands and gay partners did differ on these comparisons. In terms of interdependence scores, gay couples reported fewer relationship rewards and more relationship costs than did either lesbian couples or heterosexual couples. Further, relationship satisfaction was positively related to perceived relationship rewards and emotional investment and negatively related to perceived relationship costs in each of the three groups. Finally, this examination highlighted the mediation effect of interdependence variables on individual difference variables in the prediction of relationship satisfaction.

Schneider (1986) also investigated interdependence variables in a study that

compared lesbian and heterosexual couples on measures of relationship satisfaction as defined by interdependence, durability, and equality. Overall, results supported more similarities than differences between lesbian and heterosexual couples. Although both types of couples reported a high degree of optimism regarding the continuance of their relationships, this construct was statistically significant only for heterosexual couples. Furthermore, analyses of the interdependence factor data revealed that lesbian and heterosexual couples differed on only 3 of 13 subfactors: wills, banking practices, and life insurance policies. This finding is not surprising when one considers the lack of legal sanctions afforded lesbian couples. Finally, analyses of equality data showed that lesbian couples differed from heterosexual couples only on the issue of household chores. Lesbian couples reported greater equality regarding household chores than did heterosexual couples. Further, lesbians reported using personal preference rather than






41


traditional gender roles in decision making regarding division of chores. This investigation adds support to the use of heterosexual models of relationships with lesbian and gay partners.

Zacks et al. (1988) used the Circumflex Model (Olson, 1986) as the basis for an investigation of the levels of cohesion, adaptability, and relationship satisfaction in a sample of 52 lesbian couples and 1140 heterosexual couples. According to this model, balanced levels of adaptability and cohesion are characteristic of healthy heterosexual couples and families. Data analyses revealed that lesbians scored significantly higher than heterosexual couples on measures of cohesion. These researchers explained that higher cohesion scores could be attributed to the tendency for lesbian partners to turn to each other for support and validation in the face of societal rejection. Further, these scholars noted that the findings of this study indicate a cultural--heterosexual--bias that exists regarding the term "enmeshment." Lesbian couples also scored higher than heterosexual couples on measures of adaptability. Zacks and colleagues (1988) suggested that lesbian couples must create their own relationship roles, rules, and norms in the absence of societal models. Thus, high adaptability scores represent positive qualities of a relationship. Finally, lesbian couples reported greater relationship satisfaction than did heterosexual couples. This study supports the notion that qualities that are functional and desirable in lesbian relationships such as emotional closeness and role flexibility may differ in degree from those in heterosexual relationships.

In a related study, Green and colleagues (1996) investigated relationship quality, disclosure of sexual orientation, and relationship dissolution in a sample of 52 lesbian couples, 50 gay couples, and 1358 heterosexual married couples. Data for the lesbian





42


and gay couples were collected in 1986 and compared with data from two different national samples in 1983 and 1976 (Green et al., 1996). These researchers found that lesbian and gay couples reported greater cohesion and flexibility in their relationships than did heterosexual couples. Further, lesbian couples reported more cohesion and flexibility than did gay couples. Both cohesion and flexibility predict relationship satisfaction in lesbian and gay couples and relationship dissolution in lesbian couples. For lesbian couples, degree of disclosure of lesbian sexual orientation family members was not related to relationship satisfaction.

Heterosexual models of relationships conceptualize partner homogamy as one influence in the appraisal of relationship satisfaction (Murstein, 1970). This theory of partner selection suggests that "like is attracted to like." Kurdek and Schmitt (1 987a) assessed partner similarity measured by demographic characteristics and relationship quality in a sample of 45 married heterosexual, 35 heterosexual cohabiting, 50 gay cohabiting, and 56 lesbian cohabiting couples. Partner scores were correlated on measures of relationship satisfaction. Lesbian and gay couples' scores on love for partner were correlated, but this was not true for either heterosexual cohabiting or married couples. Partner homogamy was most pervasive in lesbian couples. These researchers found similarities between partner ages for each couple type. Heterosexual cohabiting couples had similar income, education, and job prestige. Gay couples had the greatest disparity in age, income, and education.

In another study of partner similarities, Kurdek and Schmitt (1 986b) compared 370 lesbian, gay, and heterosexual couples on measures of sex-role self-concept and relationship satisfaction. Bern's (1974) theory of sex-role self-concept was used in the






43


study. Consequently, four types of sex-role self-concepts were assessed: androgynous, feminine, masculine, and undifferentiated. Individual partner analyses revealed that relationship quality differed by sex-role self-concept rather than by couple type. Androgynous and feminine partners reported higher relationship quality than did masculine and undifferentiated partners. Couples in which one or both partners were feminine or androgynous reported the highest relationship satisfaction.

Research comparing correlates of relationship satisfaction. Intimacy has been correlated with relationship satisfaction in a number of studies of heterosexual and lesbian and gay couples (Anderson, 1990; Eldridge & Gilbert, 1990; Snyder, 1991; Zacks et al., 1988). Anderson (1990) compared intimacy, fusion, and relationship quality of women in lesbian and heterosexual relationships. She found that lesbians scored higher on self-report measures of intimacy than did their heterosexual counterparts. Further, no differences between lesbians and heterosexual women were found on measures of fusion or measures of relationship satisfaction. Intimacy and relationship satisfaction were correlated for lesbian couples but not for heterosexual couples. This study refutes suggestions that lesbian couples tend to have higher levels of fusion than do other couples types.

In a recent examination of relationship satisfaction and intimacy, Norton (1995) surveyed 20 lesbian, 11I gay, and 31 heterosexual cohabiting individuals. She found similarities between couple types regarding levels of intimacy and relationship satisfaction. Further, in all three couples types intimacy predicted relationship satisfaction. Although this investigation focused on couple relationships, only one






44


member of each couple was assessed. Consequently, partner differences that may influence relationship satisfaction were not taken into account.

Another correlate of relationship satisfaction often investigated in samples of

heterosexual couples is power. Lange (1989) examined the relationship between power, nurturance, and trust in predicting relationship satisfaction and commitment among cohabiting lesbian and heterosexual couples. She found many similarities between the two couple types investigated. Specifically, both lesbian couples and heterosexual couples reported a balanced allocation of power was strongly associated with relationship satisfaction. However, there were also differences noted across couple types. For example, a significant association between a balanced allocation of power was unrelated to commitment for lesbian couples; however, heterosexual women reported higher levels of commitment and equal balance of power. Receiving communal nurturance (the feminine prototype of support) predicted relationship satisfaction and commitment in both couple types while receiving agentic nurturance (the masculine prototype of support) was negatively predictive of satisfaction. Further, trust was positively related to satisfaction and commitment for both couple types. Research Comparing Relationship Satisfaction of Lesbian and Gay Male Couples

While studies comparing heterosexual, lesbian, and gay couples offer valuable information, these investigations may overlook unique influences on relationship satisfaction shared by lesbian and gay couples. As members of a stigmatized and marginalized group, lesbian and gay partners face stressors that many partners in heterosexual couples do not. Consequently, studies comparing lesbian and gay couples






45


provide unique insights into the couple life of partners who must live without the social sanctions afforded other types of couples.

In his 1978 study, Tuller compared 5 lesbian and 10 gay couples. He stated that lesbian couples reported a greater need for sexual fidelity and a broader network of friends than did gay couples. Later, in an investigation of the importance of fidelity for married, cohabiting, and lesbian women, Blumstein and Schwartz (1983) reported no significant difference among these couple types. It remains unclear how fidelity influences relationship satisfaction in lesbian couples.

In one of the few longitudinal investigations of relationship quality among lesbian and gay couples, Kurdek (198 8b) surveyed a sample of 47 lesbian and 65 gay cohabiting couples. Analyses of initial relationship satisfaction revealed that lesbian couples reported higher relationship satisfaction, higher liking of partner, more motivation to stay in the relationship, more trust, and more shared decision making than gay couples. More than half of the gay couples in the survey reported nonmonogamous relationships although this did not affect relationship quality measures. In a one-year follow-up study, Kurdek (1989) sampled 45 lesbian and 74 gay couples. Compared to gay couples, lesbian couples demonstrated an overall pattern of enhanced relationship quality over time. Greater relationship quality was reported by lesbian couples at both times of assessment. Further, compared to gay couples, lesbian couples reported stronger liking of partner, stronger motivation for being in the relationship, greater trust, and more frequent shared decision making at both times of assessment. Kurdek (1989) found that satisfaction with social support, high expressiveness, and few beliefs that disagreement is






46


destructive to the relationship predicted later relationship quality in gay couples but not in lesbian couples.

Kurdek (199 1) compared 75 gay couples and 51 lesbian couples on self-report measures of relationship satisfaction. This study investigated variables related to contextual, investment, and problem-solving models of intimate relationships. Overall, he found more similarities than differences between lesbian and gay couples on mean levels of each model variable. Similarly, very few differences were found between these two couple types in terms of the strength of the correlates of relationship satisfaction. Variables from each of the three models were significantly related to relationship satisfaction in both lesbian and gay couples. In addition to support for correlations between each of the models and relationship satisfaction independently, Kurdek (199 1) found support for a mediational model that proposed ordered linkages among variables from the three models. Specifically, the combination of contextual, investment, and problem-solving model variables accounted for nearly 50% of the variance in relationship satisfaction.

In a later study, Kurdek (1995a) investigated developmental changes in

relationship quality in 42 lesbian and 61 gay cohabiting couples. He analyzed survey data from three annual assessments of current levels of attachment, autonomy, and equality in the relationship, importance of attachment, autonomy, and equality in an ideal relationship, and relationship commitment. Lesbian partners reported ideal equality as more important than did gay partners. The importance of ideal equality to lesbian partners changed curvilinearly overtime, but no such change was found for gay partners. Lesbian and gay partners showed decreases in the importance of ideal attachment over






47


time. For both lesbian partners and gay partners, changes in the difference between current and ideal levels of equality predicted changes in relationship commitment over time. Similarly, current levels of attachment and current levels of autonomy predicted relationship commitment for both lesbian partners and gay partners.

In one of the largest national surveys of lesbian and gay couples, Bryant and

Demnian (1994) investigated correlates of relationship satisfaction in a sample of 560 gay couples and 706 lesbian couples. They found that high quality of sexual interaction, high commitment to the relationship, low frequency of abusive behavior, low frequency of serious arguments, high joint income, and legal preparedness were significantly associated with relationship satisfaction among lesbian couples and gay couples. The authors of this study caution that in only 483 instances did both members of a couple complete the one-page survey. Consequently, statistical techniques were used in which responses from individuals whose partners also participated in the survey were assigned a weight of .50.

Research on Relationship Satisfaction in Lesbian Couples

While comparative studies of lesbian, gay, and heterosexual couples are

informative, studies focused exclusively on lesbian relationships often highlight the importance of variables that are unique to lesbian couples. Several studies sought to determine the factors that may affect relationship satisfaction reported by lesbian couples. Specifically, the influence on relationship satisfaction of disclosure of sexual orientation and social support has been investigated.

Disclosure of sexual orientation and relationship satisfaction. Anecdotal literature (Cabaj, 1988; Murphy, 1992; Roth, 1985) has focused on the potential influence of






48


comingg out" on relationship satisfaction. The decision about whether or not to be open about the relationship may present stressful challenges to lesbian couples, particularly if there is disagreement regarding how, when, to whom, and in what situations to disclose (Falco, 1991). This stress may be expressed in the couple relationship, thereby decreasing the satisfaction level (Brooks, 198 1). Some empirical evidence suggests that relationship satisfaction is negatively impacted by partner discrepancy in degree of disclosure of sexual orientation and positively impacted by partner's disclosure of sexual orientation. However, other investigations have found no relationship between partners' disclosure of sexual orientation and relationship satisfaction reports.

Berger (1990) surveyed 114 gay men and 29 lesbians regarding the effects of passingn" as heterosexual on the quality of relationship satisfaction. Respondents who were known as gay or lesbian to significant others such as family and close friends reported higher levels of relationship satisfaction than those respondents who had not acknowledged their sexual orientation to those close to them. The extent of passing with other groups such as distant friends and acquaintances did not appear to be related to the quality of the relationship. The disproportionate number of gay men in this study may limit its generalizability to lesbians.

Romano (1990) investigated the relationship between homophobia and lesbian relationship satisfaction. A sample of 102 partnered lesbians was surveyed using selfreport measures. Homophobia from family, society, and self were assessed. Romano (1990) found a negative correlation between relationship satisfaction and each type of homophobia. When combined, the three different homophobia variables predicted relationship satisfaction. However, this study relied on only one partner from each






49


couple. While this method of data collection is statistically less complex, potentially important data are missing.

Melamed (1992) investigated the effects of internalized homophobia on

relationship satisfaction and commitment in a national sample of 233 lesbian couples. Unlike the couples in many other studies, these partners had been together for a significant length of time (mean = 7.5 years). She found that couples' levels of internalized homophobia among couples varied inversely with relationship satisfaction and commitment. Further, in an analysis of individual data Melamned (1992) found that lesbians with lower levels of homophobia reported higher levels of self-esteem.

In a related study utilizing both individual and couple data, McGuire (1995)

surveyed 44 lesbian couples who had been in relationships for at least 1 year. She used a measure of "homophobia" which assessed subjects' attitudes toward the fact of their lesbianism being known by others as well as their attitudes regarding their own homosexuality and homosexuality in general. She found that partner difference in level of "homophobia" was predictive of decreased relationship satisfaction although individual levels of "homophobia" were not significantly related to reported relationship satisfaction levels. It should be noted, however, that the instrument used to assess "homophobia" in this study was normed on gay men and was modified for use with lesbians. Additionally, this study did not attempt to investigate the effect of disclosing one's status as a member of a committed lesbian couple; rather, an individual focus was utilized in which each woman's "homophobia" was assessed.

In a similar study, Jordan (1995) found that individual women who more widely disclosed their sexual orientation reported greater relationship satisfaction than their






50


counterparts who were more "closeted." Additionally, the women in this study reported lower levels of relationship satisfaction when a greater discrepancy between the partners existed regarding disclosure of sexual orientation. However, it is unclear how many of the 499 women in committed relationships were included in couple data. Furthermore, Jordan (1995) assessed the degree to which individual women disclosed their sexual identity not the degree to which they identified themselves to others as part of a committed lesbian couple.

Caron and Ulin (1997) investigated the relationship between couple satisfaction and "closeting"--the degree to which the couple disclosed their relationship. They surveyed a sample of 124 lesbians involved in a "committed" relationship. They analyzed four potential sources of closeting: family, extended kin, friends, and work associates. The strongest correlations among these sources of closeting and relationship satisfaction occurred in the realm of family. Further, these researchers found that relationship closeting (compared to personal closeting) had the highest correlation with satisfaction. Consequently, lesbian couples who were able to be more open about their relationships reported higher levels of satisfaction. Although these authors reported that the lesbians in their sample represent women in committed relationships, partners involved in relationships as short as two month's duration were included in the study. Furthermore, both members of a couple were not assessed. Consequently, important information regarding dyadic differences may have been overlooked.

Several studies have found no relationship between a woman's willingness to disclose individually her sexual orientation and reported relationship satisfaction (Eldridge & Gilbert, 1990; Peplau et al., 1982). These researchers expected that women






51


who were able to acknowledge their sexual orientation to others in the family, at work, or in social arenas would report higher levels of relationship satisfaction; however, none was found (Eldridge & Gilbert, 1990; Peplau et al., 1982).

Peplau et al. (1982) and later Caldwell and Peplau (1984, 1992) surveyed 127 lesbians regarding relationship satisfaction, self-disclosure, and personal power. They found that lesbians reported high levels of relationship satisfaction and egalitarian power distribution as well as equal degree of commitment and involvement in the relationship. However, they found no relationship between disclosure of lesbian identity and relationship satisfaction. They had hypothesized that individual partner characteristics such as age, student-work status, and religion would predict relationship satisfaction but this hypothesis was not confirmned--a finding previously been noted by Cotton (1975). However, only one partner from each couple participated in the study and more extensive variables to assess similarity (e.g., race, degree of self-disclosure) were not used.

Eldridge and Gilbert (1990) conducted a study in which correlates of relationship satisfaction were assessed utilizing both partners of 275 lesbian couples. Degree of disclosure of lesbian identity was not significantly correlated with relationship satisfaction. Disclosure of lesbian couple status was not assessed. However, many of the psychological variables assessed did correlate significantly with relationship satisfaction. For instance, higher levels of personal autonomy were associated with lower levels of relationship satisfaction. Further, self-esteem and satisfaction with life were significantly positively associated with relationship satisfaction. However, role conflicts (e.g., partner versus self, professional versus self) were negatively correlated with relationship satisfaction.






52


Thus, it is unclear whether or not disclosure of individual sexual orientation is

related to relationship satisfaction in lesbian women. Further research is needed in order to understand this relationship.

Social support and relationship satisfaction. Murphy (1989) examined the role of parental support on lesbian couple satisfaction. She found that perceptions of parents' attitudes toward the couple, the partner, and the daughter's lesbian sexual orientation do affect relationship quality. This study is one of the few to examine support for the lesbian couple relationship rather than simply support for the lesbian individual.

In a study of 173 lesbian couples, Donaldson (1993) examined the relationships between autonomy-mutuality, social support, and relationship satisfaction. Using selfreport measures, she found that perceived mutuality was positively correlated with, and accounted for the largest percent of explained variance in, relationship satisfaction. Additionally, dyadic attachment and perceived social support positively correlated with relationship satisfaction. Specifically, perceived support from partner and friends was higher than that from family. Further, she found limited support for a negative correlation between mutuality and relationship satisfaction.

Jenkins (1996) examined the relationship between differentiation from family of origin and relationship satisfaction in a sample of lesbian couples. She found that intimacy within the relationship was strongly associated with relationship satisfaction. Level of individuation within the relationship was shown to have a positive relationship with level of relationship satisfaction; however, differentiation in general and differentiation from family was not related to reports of relationship satisfaction among the lesbian couples in the sample.






53


In summary, research on relationship satisfaction has followed three traditions: (a) comparative studies of heterosexual, lesbian, and gay couples, (b) comparative studies of lesbian and gay couples, and (c) descriptive studies of lesbian couples. Overall, studies of relationship satisfaction among these three types of couples have revealed more similarities than differences. Furthermore, investigations based on models developed in study of heterosexual couples (e.g., the investment, problem solving, interdependence, contextual models) have provided support for the use of these models with lesbian and gay couples. Studies of lesbian couples have focused on the influence of disclosure of sexual identity and social support on relationship satisfaction. Most studies have focused on disclosure of individual sexual orientation to family and friends. Few studies have examined the influence on relationship satisfaction of the disclosure of lesbian couple identity. Likewise, investigations of social support have focused on support for the individual rather than for the couple.

Research on Social Support Influences

Social support has been viewed as an important asset to individuals and couples alike. It has been depicted as a buffer against stress among heterosexual couples (e.g., Cobb, 1976; Cohen & Hoberman, 1983; Cutrona, 1984; Cutrona & Russell, 1987), but it is unclear whether social support provides a similar protective function in lesbians. Similarly, while social support has been conceptualized as a main effect in studies of heterosexual couples (Cohen & Hoberman, 1983; House, 198 1; Lin, Ensel, Simeone, & Kuo, 1979), few similar studies have been conducted on samples of lesbian couples (Kurdek, 1998; Kurdek & Schmitt, 1986a, 1986b, 1986c; Meyer, 1989). Most researchers agree that social support plays an important role in the lives of human beings,






54


yet there is little agreement regarding its definition (Cobb, 1976; Cooke, Rossmann, McCubbin, & Patterson, 1988; Fischer, 1982; Kahn & Antonucci, 1980; Mitchell, Billings, & Moos, 1982; Pilisuk, 1982; Procidano & Heller, 1983; Schaefer, Coyne, & Lazarus, 198 1; Weiss, 1974). Similarly, much disagreement exists concerning the best way to operationalize the concept (Cohen & Hoberman, 1983; Coyne & DeLongis, 1986; Cummins, 1988; Heller, 1979; Lakey & Heller, 1988; Procidano & Heller, 1983; Stemp, Turner, & Noh, 1986; Turner, Frankel, & Levin, 1983).

Social support and its measurement have been conceptualized from a variety of

perspectives. Theorists have debated whether received or perceived social support should be assessed (Cummins, 1988; Lakey & Heller, 1988). This debate is tied to the question regarding the process by which social support is believed to function, either as a main effect or a buffer. The conceptualization of social support is further complicated by the discourses regarding the multiple provisions of support, numerous dimensions of support, and various mechanisms by which support is delivered. Regardless of the conceptual framework from which social support has been measured, much of the research regarding social support in lesbian populations has been limited to the support provided to individuals by individuals. Rarely have investigations assessed either the impact of institutional support or the support given lesbian couples (Kurdek, 1998). Definitions

Social support is often conceptualized as a multidimensional construct. These definitions have been based on the belief that human beings need to form meaningful relationships with each other (Stemp, Turner, & Noh, 1986). Moreover, these definitions focus on the specific classes of information (Cobb, 1976), exchanges (Fischer, 1982), or






55


provisions (Weiss, 1974) which comprise social support. For example, Cobb (1976) defined social support as information of three types: information which contributes to an individual's belief that he or she is loved and cared for, information which supports an individual's feeling that she or he is valued, and information that contributes to an individual's sense of belonging and mutual involvement in a network. Further, Cobb (1976) hypothesized that social support could take three forms: emotion, esteem, or network support. In contrast with several other theorists, Cobb's definition is limited to the provision of nonmaterial supports.

On the other hand, Weiss (1974) conceptualized social support from the

perspective of "relationship provisions." He posited that individuals have a "fund of sociability" which is a specified amount of interpersonal interaction that he or she needs. These needs are potentially met in a variety of ways and via a variety of relationship sources. Those relationships that provide social support are specialized in regard to the "relationship provisions" each offers (Weiss, 1974). Weiss (1974) outlined six relationship provisions: attachment, social integration, reassurance of worth, nurturance, reliable alliance, and guidance.

Like Cobb's conception of "emotional" support, Weiss' attachment provision represents a sense of belonging in the individual. Likewise, a parallel exists between Cobb's "network support" and Weiss' social integration provision. Both of these social supports provide companionship and a network for mutual exchange. Reassurance of worth, similar to Cobb's "esteem support," offers the individual affirmation of his or her competence. Weiss' opportunity for nurturance provision is unlike any of Cobb's social supports. This provision occurs only in parent-child relationships and provides parents






56


with a feeling of being needed and a sense of meaning in their lives. Family relationships, including the primary relationship between the members of a couple, are the domain from which a sense of reliable alliance is provided. This provision engenders in the individual a feeling of unconditional love and support. Finally, Weiss' guidance provision assists individuals with decision-making in stressful situations via the sharing of information. This provision is unlike any of Cobb's social supports. However, like Cobb, Weiss does not include the provision of material goods and services in his conceptualization of social support.

Material goods and services are included among the social supports

conceptualized by several theorists. Pilisuk (1982) identified three components of social support: "material help," "emotional assurance," and "communication of concern and caring". Similarly, Kahn and Antonucci (1980) defined social support as interpersonal interactions that include "aid," "affect," and/or "affirmation." Likewise, Schaefer et al. (1981) classified social supports into three types: tangible support, emotional support, and informational support. House (198 1) analyzed a number of conceptual frameworks concerning social support and attempted to integrate these into what he considers to be an inclusive definition. He suggested that social support is "a flow of emotional concern, instrumental aid, information, and/or appraisal (information relevant to self-evaluation) between people" (p. 26).

Similar to the authors previously cited, Cooke and colleagues (198 8) have

conceptualized social support in terms of provisions. These authors' conceptualization parallels provisions identified by other theorists by including the importance of network, emotional, esteem, and appraisal supports. However, they also add altruistici" support as






57


an equally important provision. Moreover, they note that reciprocal exchange is a salient provision of support as well.

While some theorists have conceptualized social support in terms of provisions, others have proposed that social support is best understood in terms of the mechanisms by which it is believed to affect well-being. Mitchell and colleagues (1982) identified three categories of social support mechanisms: direct, indirect, and interactive. They proposed that "direct" effects are those benefits to the global well-being of an individual as a result of social support. These direct effects may occur when an individual is in a situation of stress or crisis or even when life circumstances are relatively uneventful. On the other hand, indirect effects on well-being may be accomplished by social supports which have the effect of reducing the intensity with which the individual perceives the stress. Finally, interaction effects refer to the buffering effect that social support may have in times of stress. Henderson (1987) reviewed studies of the impact of socially supportive interactions. He found that when support is lacking, physiological and psychiatric symptoms increased. Similarly, Turner, Frankel, and Levin (1983) reviewed social support literature and concluded that mental health and physical well-being were significantly positively related to social support. Assessment of Social Support

In light of the disagreement among researchers and theorists regarding the

definition of social support, it is not surprising that a similar conflict exists concerning the best method of measuring social support. Two opposing theories have been proposed. Some researchers advocate assessing perceived social support while others believe that social network, or embeddedness, must be studied. Furthermore, this debate






58


has not been limited to the efficacy of these opposing methods; it has also included a discussion of the strength of the relationship between them.

A comprehensive systematic analysis of the individual's total social field is

generally required for a social network analysis (Turner et al., 1983). This approach has a number of limitations, particularly when applied to individuals with minority status. First, it is based on the assumption that all members of an individual's social network make positive contributions to the individual's social support. In the case of lesbians or members of other socially stigmatized groups, this may not be the case; in fact, the opposite may be true. It is quite possible that some members of a lesbian's social network are unaware of her orientation and, therefore, contribute no support--at least in this important arena. Other members of a lesbian's social network may adamantly oppose her lifestyle, thus contributing negatively to her social support and adding to her stress. Thus, measures based on the density or frequency of social contacts may offer minimal meaningful information. Additionally, social network analyses often ignore the complexity of the interaction between the stressful and the stress-buffering aspects of relationships (Coyne & DeLongis, 1986). In a review of social network research, Coyne and DeLongis (1986) found that a number of investigations identified strong relationships among the conflictual or upsetting aspects of social interactions, signs of depression, lack of well-being, and distress and weak relationships among sources of positive support, well-being, and lack of distress.

It is not clear whether beneficial effects of social support are the results of supportive behaviors on the part of network members--as is the case in received support--or the result of the belief that support is available--as is the case in perceived






59


support (Lakey & Heller, 1988). This might not be a concern if perceived support is a reflection of received support. However, investigations regarding the link between these two facets of support have shown a weak or nonexistent relationship between measures of received support and measures of perceived support and satisfaction with support (Lakey & Heller, 1988). In fact, researchers have found that perceived availability of support may either wholly or partly protect individuals from negative effects of high stress levels (Cohen & Hoberman, 1983).

Based on these findings, it appears that perceived support is more accurately assessed than received social support. Consequently, an instrument developed for this investigation to assess perceived support was utilized to assess the support lesbian partners perceive they have available to them from a variety of social institutions.

Another consideration regarding the measurement of social support is whether social support acts as a main effect or a buffer effect. Some researchers have found that social support processes function as a main effect in that the support itself is beneficial regardless of the presence of stress (Cohen & Hoberman, 1983; House, 198 1; Lin et al., 1979). On the other hand, researchers have found that the interaction between social support and stress may function in such a way that support is more effective in situations of relatively high stress rather than in situations of relatively low stress. In this case, social support would function as a buffer in which support mediates the psychological and/or physiological effects of stress on the individual (Heller & Swindle, 1983; Levy, 1983; Mitchell et al., 1982).

Scholars suggest that the buffering effect of social support is cognitively mediated in that support is thought to influence the individual's appraisal of stressful situations






60


(Cohen, Mermeistein, Kamarck, & Hoberman, 1985; Heller & Swindle, 1983). The individual is thought to weigh the resources for support available against the risks and difficulty faced (Lazarus & Launier, 1978). Consequently, the perceived availability of social support may result in the reappraisal of a threatening situation as less so (Wilcox & Vernberg, 1985).

Research on Lesbians and Social Support

Research on lesbian individuals. In one of the first studies to investigate the social support of lesbian individuals, Chafetz and colleagues (1974) interviewed 51 lesbians in Houston, Texas. They found that the women in this study reported a lack of both general and specific support from a variety of social institutions such as family, friends, employers, and helping professionals. Of these women, 50% reported receiving little or no support from family of origin. More than 40% of the lesbians in the sample reported that the majority of their close friends are lesbian or gay. These researchers did not distinguish between women in committed relationships and those who were single.

Albro and Tully (1979) expanded the study initiated by Chafetz and colleagues (1974). Questionnaires completed by a sample of 91 lesbians revealed that these women felt isolated from the heterosexual "macro-culture." They reported that the women in their sample received emotional support from the lesbian "micro-culture." Both of these early studies support the notion that lesbians compensate for lack of support from the dominant culture or family of origin by constructing support networks in the lesbian community. However, these studies did not address the potential impact of this strategy on the lesbian couple relationship.






61


In another investigation of the various sources of social support available to

lesbian individuals, Jay and Young (1977) surveyed 900 lesbians regarding the types of people who know of their sexual orientation and the perceived reactions of these people to this knowledge. They found that more lesbians told their straight friends about their sexual orientation than told their family members. Employers and employees were least likely to know of a woman's lesbianism. In regard to the perceived "reactions" of significant people who knew of their lesbianism, individuals in the sample cited mothers as least supportive and gay or lesbian friends as most supportive. No data regarding couple status were reported.

Leavy and Adams (1986) determined that relationship status and age affect social support and self-esteem among lesbian individuals. The lesbians in this sample who were involved in committed relationships scored significantly higher on measures of social support and self-esteem than those not in relationships. Although relationship status was assessed in this study, the data were treated as individual data rather than couple data. Consequently, the influence of social support on the relationships of lesbian couples in this sample is unclear.

D'Augelli, Collins, and Hart (1987) examined the social support pattern of 34 lesbians living in a rural setting. Members of an organized network of lesbians completed questionnaires. Of the lesbians in the sample, 91% reported seeking help from women, 71% reported seeking help from partners, and 68% reported seeking help from heterosexual friends who knew of the woman's sexual orientation. Further, a majority of the women had not told their families of their sexual orientation; more than half expected






62


rejection from their fathers and nearly half expected the same from their mothers. This study did not assess relationship status.

More recently, Walsh (1995) expanded studies of social support by examining five provisions of social support (attachment, reliable alliance, guidance, reassurance of worth, and social integration) in a sample of 209 lesbians with a "positive sexual identity." She found that lesbians with high self-esteem and low internalized homophobia reported the most social support. She determined that different provisions of social support did not come from different sources. Overall, social support was provided by a combination of friends--lesbian, gay, and heterosexual. Lesbians in a relationship had greater social support than did single lesbians. This study did not attempt to investigate the influence of social support on the couple relationship although the sample included women in committed relationships.

Wayment and Peplau (1995) also investigated five types of social support that lesbian and heterosexual women receive and the relationship between these types of support and psychological well-being. They compared 75 single lesbians, 75 coupled lesbians, 75 single heterosexual women, and 75 coupled heterosexual women using data from self-report measures originally collected from a sample of 600 women during the

1 980s (Aura, 1985). The groups were matched for race, employment, and parent status. They found that overall support reported by lesbian and heterosexual women was equivalent. Women in relationships reported greater well-being than did single women regardless of sexual orientation. A strong relationship between psychological well-being and support was found for all women. Specifically, lesbian women's well-being was associated with reassurance-of-worth support while heterosexual women's well-being






63


was associated with guidance support. The focus of this study was on lesbians as individuals rather than on couples.

Jordan (1995) surveyed 499 lesbians regarding relationship satisfaction, degree of disclosure of sexual orientation, and measures of adjustment including social support. She found that social support was related to degree of disclosure of sexual orientation with those women who more widely disclosed their lesbianism reporting greater social support. Jordan identified a hierarchy of disclosure in which lesbians were most likely to disclose their sexual orientation to lesbian and gay friends, followed by straight friends, family, and co-workers. Furthermore, in an analysis of couple data Jordan (1995) found that disclosure of sexual orientation was significantly related to relationship satisfaction. However, it is unclear how many of the 499 women in committed relationships were included in couple data.

In a recent comparative study of the self-reported stress levels and social support of 29 lesbians and 77 heterosexual women, Tait (1997) found no difference regarding stress experienced by these groups. In terms of social support, she identified differences between sources of support for lesbians and heterosexual women. Specifically, she found that lesbians turned to their partners first and then to mental health professionals such as psychologists or psychiatrists while heterosexual women turned to a variety of helpers including family, psychologists, psychiatrists, and partners with no one type being dominant. While Tait (1997) expanded the assessment of social support to include sources outside family and friendship network, she did not gather information regarding the influence of support on the couple system.






64


In a study of lesbian, gay, and bisexual youth, Gilmore (1996) investigated the relationships among social support, general stress, gay-related stress, anxiety, comfort with sexual orientation, depression, substance use, and self-esteem. He surveyed 105 individuals ages 14 years to 20 years using a variety of self-report instruments. Results of this investigation suggest that low social support and general stress both predicted depression and substance use. Other predictors of depression included gay-related stress, younger age, and discomfort with sexual orientation. Couples were not included in this study, in part due to the age of the sample.

Grossman and Kerner (1998) studied the social support and self-esteem of 32 lesbian and 58 gay youth ages 14 years to 2lyears. Unlike most other samples, the individuals in this study were predominantly Black or Latino/a. Using self-report measures, these researchers found that self-esteem was a moderate predictor of emotional distress in the total sample and a strong predictor in the lesbian subsample. Satisfaction with support was not significantly related to emotional distress that the authors suggest may be due to the fact that a global assessment of support was used rather than one in which specific support sources were assessed.

In a comparative study of heterosexual, lesbian, gay, and bisexual adults, O'Neill (1998) surveyed 201 individuals to determine potential predictors of depression including lack of social support, dysfunctional thinking, stressful life events, autonomy, and sociotropy. He found no differences between heterosexual and combined gay-lesbianbisexual groups in regard to relationships between current depression and risk factors. Nor did he find that the gay-lesbian-bisexual group reported greater levels of each risk






65


factor although this group did report greater levels of current depression. Relationship status was not considered in this investigation.

Jordan and Deluty (1998) examined the relationship between lesbian individuals' disclosure of sexual orientation and four indices of psychological adjustment: anxiety, self-esteem, positive affectivity, and social support. The 499 lesbians in the sample responded to a set of questionnaires. They found that the more widely a lesbian disclosed her sexual orientation, the greater her level of social support. Further, these researchers identified that disclosure of sexual orientation was positively associated with less anxiety, more positive affectivity, and greater self-esteem. Jordan and Deluty (1998) assessed disclosure to family, gay and nongay friends, and coworkers. They did not assess disclosure to members of other potential support sources such as helping professionals and clergy.

Research on lesbian couples. In one of the first studies to compare social support systems of different types of couples, Wood (1983) compared 20 lesbian couples with 20 cohabiting heterosexual couples and 20 married heterosexual couples. She also examined the relative functioning of each couple type on three indicators: psychosocial maturity, intra-couple functioning, and couple-community functioning. Contrary to her expectation, Wood found significant differences between the groups. Compared to married heterosexual couples and cohabiting heterosexual couples the lesbian couples received social support from friends rather than from family. Lesbian couples in this sample had more functional adaptability patterns and were more psychosocially mature than their cohabiting heterosexual counterparts.






66


Wood (1983) used a snowball sampling technique, which necessarily calls into question the generalizability of the sample. Many researchers must resort to snowball sampling when attempting to locate participants from minority groups such as lesbians. However, this technique is particularly problematic because this study investigated social support. It is likely that lesbians identified via some sort of network would have support from that network. Although membership in such a network does not equate with high levels of support, it does suggest that the women in this sample have a network of people who are aware of their lesbianism. Thus, women who are relatively closeted may be underrepresented in this sample. Wood (1983) acknowledged other limits regarding the generalizability of her results based on the demographic similarity of the sample. Additionally, Wood (1983) found order effects on the social support scores, which suggests that the instrument she devised to measure this construct may not be valid. Finally, the return rates for each of the groups differed significantly. It is possible, then, that the groups may be less similar than Wood posits.

Aura (1985) conducted another comparative study to describe similarities and

differences between the social support systems of lesbian and heterosexual women. She compared a sample of 50 lesbians with a sample of 50 heterosexual women regarding the sources, amount, and value of social support for these women. The samples were matched on age, education, employment, number of children, and length of relationship. Although she was primarily interested in the role of partners, friends, and relatives in social support, she developed a taxonomy of 20 types of support divided into four categories: personal, social, accomplishment, and instrumental. The results of this investigation revealed that lesbians and their heterosexual counterparts did not differ on





67


the value placed on social support. These two samples did not differ on the total amount of social support perceived. Aura (1985) did find that the groups differed on the sources from which they perceived support. Heterosexual women reported receiving a greater amount of total social support from relatives than did lesbians. Lesbians reported that support lacking from family of origin was obtained via additional support from their partners. Heterosexual women perceived less total social support from their partners than did lesbians.

While Aura (1985) did not claim to have obtained a representative sample of lesbians, she did match the lesbian and heterosexual women on key demographic characteristics. However, the results of the study may be limited due to the original technique of obtaining the lesbian sample from lesbian and feminist organizations. Additionally, the small sample size may compromise the degree to which the results of the study can be applied. Further, this investigation did not focus on support given the couple although couples were included in the sample. Finally, this study was limited to support from friends and family and, therefore, did not assess other potential sources of support such as institutional supports.

Meyer (1989) examined the relationship between social support and relationship closeness among 60 lesbian couples in varying stages of relationship development. Using self-report measures to assess both emotional closeness and support, she found that social support and emotional closeness were significantly related for the sample of lesbians. Further, she identified a negative relationship between perceived support from friends and couple closeness but a positive relationship between perceived support from





68


family and couple closeness. Meyer (1989) did not assess forms of support other than that provided from family and friends.

Olson (198 8) examined the relationship between social support and psychological well-being in a sample of 65 lesbian couples using several self-report measures. She found a significant positive correlation between the quality of social support and the number of close relationships and psychological well-being. Further, she identified that cohesion, adaptability, and homosexual identity formation were also positively correlated with psychological well-being in this sample of lesbian couples. Although Olson (1988) assessed couples, the predictor variable she used was an individual construct (psychological well-being) rather than a couple construct (e.g., relationship satisfaction).

In another study of support from family and friends, Markowitz (1993) compared 500 gay men and lesbians with 500 heterosexuals. Using self-report measures, she found that lesbians and gay men were no more estranged from their families of origin than heterosexual women and men. She suggested that lesbians and gay men utilize friendship networks and create families of choice to supplement rather than replace support from relatives.

In a series of studies, Kurdek and Schmitt (1986a, 1986b, 1986c, 1987b) expanded the comparative study of women's social support to include cohabiting heterosexual women as well as married heterosexual women and cohabiting lesbians. Additionally, they included a focus on the relationship quality as well as on social support. These researchers hypothesized that social support would act as a buffer to stress and thereby enhance relationship satisfaction. Both partners in 44 married, 35 heterosexual cohabiting, 56 lesbian, and 50 gay couples completed questionnaires.





69


Lesbians reported receiving less social support from family than did married women and cohabiting women. In a separate analysis using the same data set, Kurdek and Schmitt (1 987b) found that regardless of whether individual or couples scores were used, lesbian couples reported receiving more social support from friends than from family. Contrary to expectation, they found the reverse true for cohabiting heterosexual women. These women, like their married counterparts, reported family more than friends as their primary source of social support.

Kurdek and Schmitt (1986a) suggested that the results of these studies be

interpreted with caution. They pointed out that the sample size was small and that the return rate for the questionnaires was low. Furthermore, they noted that because stage of relationship was not factored into the design of the investigation a potentially important mediating variable may have been excluded.

In a later study, Kurdek (1988a) compared the perception of social support among 50 lesbian and 69 gay couples. In this study, he assessed the relationships between social support and both relationship quality and psychological adjustment using self-report measures. The results of this investigation once again demonstrated that lesbians (and gay men) received social support more from friends than from relatives. Additionally, psychological adjustment in this sample was related to satisfaction with social support, large number of providers of support, and the perception of support from friends and partners rather than family. Interestingly, Kurdek (I1988a) also found that a significant discrepancy between partners regarding familial support was related to high relationship satisfaction. He concluded that a high degree of support from one partner's family may






70


compensate for a lack of support from the other partner's family and that support from one family may be sufficient to enhance relationship quality.

In his most recent comparative analysis of relationship quality among married

heterosexual, cohabiting lesbian, and cohabiting gay couples Kurdek (1998) investigated five dimensions of relationship quality: intimacy, autonomy, equality, constructive problem solving, and barriers to leaving. In data collected from 236 married heterosexual, 51 lesbian cohabiting, and 66 gay cohabiting couples, two relationship outcomes were assessed: trajectory of change in relationship satisfaction and relationship dissolution over 5 years. The lesbian couples reported more intimacy, more autonomy, more equality, and fewer barriers to leaving the relationship than did their married heterosexual counterparts. However, lesbian partners also reported a higher rate of relationship dissolution than did the married heterosexual and gay cohabiting couples.

Kurdek (1998) cautioned that these results may not be generalizable due to the recruitment strategy used to obtain the lesbian couples. Furthermore, demographic data suggest that the sample was highly educated. The equivalence of couples on demographic variables was not obtained by matching; instead, it was approximated using statistical corrections. Additionally, he noted that although the assessment of relationship quality seems both valid and reliable, it may not be comprehensive. Finally, because all measures relied on self-report, response bias may be a factor in the interpretation of the data.

Research on gy men. Although gender socialization differences make parallels between lesbians and gay men impossible, studies based solely on samples of gay men lend support to the research regarding lesbian social support. Berger and Mallon (1993)





71


investigated several types of social support received by a sample of single and coupled gay men. They found that the most frequent source of social support for gay men was a friend, specifically, a close male friend. This supports Kurdek's (1988a) notion regarding the creation of extended families composed of friends to compensate for lost family support. Berger and Mallon (1993) determined that the most common type of support experienced by the gay men in the sample was emotional support when compared to financial, practical, advice, guidance, and social supports.

In a study of coupled and single gay men, Sommers (1982) discovered that men in couples experienced more support than did their single counterparts. This finding is similar to that reported by Leavy and Adams (1986) in their sample of lesbians. Furthermore, Sommers (1982) reported that involvement in the gay subculture was related to positive gay identity development for the men in the sample, regardless of relationship status.

Research on Institutional Support

According to Heller and Swindle (1983), the degree to which an individual has social support is a function of the interaction between the availability of "supportive structures in the environment" and the capacity with which the individual elicits and maintains supportive relationships that "present themselves in the environment."9 Consequently, social support is not a unidirectional process. Lesbian couples may have limited access to social support structures due to exclusion from socially sanctioned supports afforded heterosexual couples. Institutional supports for lesbians such as schools, religious assemblies, and legal organizations may be a source of rejection rather than support (Kurdek, 1998; Walsh, 1995).






72


In one of the first studies to investigate institutional support, Chafetz and colleagues (1974) interviewed 51 lesbians in Houston, Texas. They found that the women in this study reported a lack of both general and specific support from a variety of social institutions. Of these women, 50% reported receiving little or no support from family of origin. Additionally, 90% of the women in the sample reported feeling a lack of support from religious institutions, and 66% said they did not feel support from economic institutions. Further, 40% of the women in the sample said they had been harassed by police officers, and 90% reported feeling lesbians were ignored by the media. The primary sources of support for these women were lesbian friends and the lesbian community. These authors focused exclusively on the support given lesbians as individuals and did not address the support lesbians received for their relationships.

In a survey of 91 lesbians Albro and Tully (1979) examined the ways that these women functioned in the homosexual "micro-culture" as well as in the heterosexual "macro-culture." They discovered that more than half of the women in the sample did not share their lesbian identities with members of the heterosexual "macro-culture" including family of origin or coworkers. These researchers also found a positive correlation between degree of social interaction in the lesbian "micro-culture" and willingness to be open regarding sexual orientation. Conversely, they discovered a negative correlation between age and disclosure of sexual orientation.

Bryant and Demian (1994) surveyed a national sample of 706 lesbian couples and 560 gay couples. Respondents rated the degree of support for each of 12 sources including gay and heterosexual friends, family, gay and heterosexual church, gay organizations, coworkers and boss. No significant differences were found between gay





73


couples and lesbian couples. Couples reported low levels of support from all types of relatives. Overall, couples ranked coworkers as more supportive than family members. Churches were the only sources rated as more "hostile" than supportive overall. One quarter of those surveyed reported that nearly half of the potential sources of support were totally unavailable because these sources were unaware of the couple's relationship.

In Kurdek's (1998) most recent study described above, barriers to leaving the

relationship was one of five indices of relationship quality (intimacy, autonomy, equality, constructive problem solving) assessed in a sample of 236 married heterosexual, 51 lesbian cohabiting, and 66 gay cohabiting couples. In this study barriers to leaving the relationship were defined as forces external to the relationship that would keep couples in the relationship regardless of personal happiness. Five barriers were assessed: religious beliefs, pressure from family and friends and community, financial difficulties, personal obligation, and responsibility for dependents. Two relationship outcomes were assessed: trajectory of change in relationship satisfaction and relationship dissolution over 5 years. Kurdek (1998) found that lesbian couples reported fewer barriers to leaving the relationship than did heterosexual couples. Further, he found that of the five indices of relationship quality assessed, only barriers to leaving the relationship uniquely predicted relationship dissolution.

Kurdek's (1998) investigation extends the literature regarding social support for lesbian relationships because it considers the influence of cultural institutions rarely studied such as religious and financial institutions in addition to those frequently studied in past research (e.g., family, fiends). However, Kurdek conceptualizes the influence of cultural institutions on lesbian relationships from a negative point of view in that he






74


considers these institutional forces as constraints to leaving a relationship despite personal unhappiness rather than as supportive of the relationship. Furthermore, the cultural institutions included in Kurdek' s conceptualization of barriers to leaving the relationship are not comprehensive. Additional cultural institutions may influence lesbian relationships (e.g., legal, healthcare, educational).

In summary, there is limited agreement regarding the conceptualization and

measurement of social support. However, some general conclusions can be drawn from the literature. First, some researchers have conceptualized social support as a main effect while others view it as having a buffering effect on the psychological functioning of individuals. Second, perceived support is more accurately assessed than received social support. Consequently, an instrument developed for this investigation to assess perceived support was utilized to assess the support lesbian partners perceive they have available to them from a variety of social institutions. Third, models of social support have been developed for individual rather than couple or family unit. Fourth, while lesbian couples report they receive support primarily from friends rather than from family, no investigations to date have assessed other institutional sources of support.

Research on Ritual

A preponderance of the recent literature regarding secular ritual has focused on the use of ritual in family systems and has been based on foundational concepts from anthropology, sociology, and most recently, psychology. Much of the literature regarding family ritual has focused on the types and functions of rituals in family systems (Bennett, Wolin, & McAvity, 1988; Bennett, Wolin, Reiss, & Teitelbaum, 1987; Bossard & Boll, 1950; Boyce, Jenson, James, & Peacock, 1983; Fiese, 1992; Fiese & Kline, 1993;





75


Imber-Black & Roberts, 1992; Kline, 1992; Mccubbin & McCubbin, 1987; Meredith, 1985; Meredith et al., 1989; Mize, 1988; Myerhoff, 1977; Reiss, 1981; Reiss & Oliveri, 1991; Roberts, 1988; Schvaneveldt & Lee, 1983; Wolin & Bennett, 1984; Wolin et al., 1980). The latest research has begun to focus on the salience of rituals for family members (Fiese, 1992; Fiese & Kline, 1993; Fiese et al., 1993; January, 1996). Only recently have researchers begun to investigate the importance of rituals in the life of marital partners or friendship dyads (Bruess, 1994; January, 1996). While the empirical literature regarding family rituals is limited and research regarding interpersonal rituals among couples is sparse, the empirical research regarding rituals in lesbian couples is nonexistent. To date, no clinical studies have focused on the use and functions of ritual in this family form. The following sections review both the anecdotal and empirical literature on the definitions, types, and functions of rituals in families, heterosexual couples, and lesbian couples.

Definitions

Historically, the study of ritual was the exclusive domain of anthropologists and often centered on religious and cultural rites (Durkheim, 1965). Implicit in this connotation of ritual are the concepts of magic, myth, and mysticism (Bossard & Boll, 1950), yet rituals are more often commonplace occurrences that routinely take place within the context of interpersonal relationships (Bossard & Boll, 1950; Fiese, 1992; Fiese & Kline, 1993; Wolin & Bennett, 1984). Family rituals range from extremely circumscribed religious ceremonies such as baptisms and bar mitzvahs to commonplace daily interaction patterns such as the way that parents and children say goodnight to each other.






76


The concept of ritual remains enigmatic in practice and in research. In fact, the multiplicity of definitions of this concept bears witness to its intangible nature. Describing family ritual, Wolin and Bennett (1984) explained, "Ritual itself is an elusive concept, on the one hand transparent and conspicuous in its enactment, on the other, subtle and mysterious in its boundaries and effects on participants" (p. 401). Much of the current literature attempting to define ritual focuses on family ritual. In these instances family is defined as parents and children and/or extended kin such as grandparents, aunts, uncles, and cousins. Bossard and Boll (1950) collected data from 400 family case records spanning a 60-year period for their historic study of family ritual. They defined ritual as a purposeful, prescribed procedure that takes on a "sense of rightness as a result of its continuing history" (p. 29) in family interactions. They identified that rituals take on many forms and arise from virtually all aspects of family life. Bossard and Boll (1950) argued that rituals are the "core" of family life. In their work with families of alcoholics, Wolin and Bennett (1984) defined ritual as a form of symbolic communication repeated over time due to their contribution to family satisfaction. Schvaneveldt and Lee (1983) identified two types of family rituals: rituals associated with tradition such as holidays and religious ceremonies and rituals that spontaneously erupt out of family interactions. These researchers defined ritual as a prescribed pattern of behavior that is purposeful in nature (Schvaneveldt & Lee, 1983). Ritual versus Routine

It is important to make the distinction between ritual and routine for it is this distinction that is at the heart of this investigation. Routines have been defined as predictable activities enacted within the family context while rituals involve routines






77


which have a symbolic meaning (Wise, 1986). Routines are characterized by their functional aspect, such as eating dinner or shopping for groceries weekly, whereas rituals have a symbolic meaning that transcends the activity itself, such as eating out on weekends or taking a family vacation. Routines can become rituals if those enacting them perceive them to be relationally or symbolically significant (Fiese, 1992; Wise, 1986). For example, getting children ready for bed is not simply a matter of changing clothes, brushing teeth, turning out lights but is a cherished time to interact and share about the day's activities. Essentially, any routine can become a ritual once it acquires a symbolic status. It is the symbolic quality of ritual that instills meaning in what would ordinarily be repetitive or habitual experiences. The shared meaning of a family ritual endows a routine task or mundane experience with salience and potency (Myerhoff, 1977).

Types of Rituals

Schvaneveldt and Lee (1983) categorized rituals into two very general classes: traditional and spontaneous. More recent authors have offered other typologies. Wolin and Bennett (1984) divided rituals into three types: traditions, patterned family interactions, and celebrations. They defined each of these categories according to the setting and degree to which each is linked to cultural customs. Family traditions are unique to families and less influenced by the cultural context in which the family lives. Traditions include family reunions, birthdays, and vacations. Family traditions may be developed in order to serve a particular family need (i.e., family meetings). Patterned family interactions are spontaneous rituals that organize daily life. They involve little planning and occur with great frequency. Examples of patterned family interactions





78


include dinner time and bedtime rituals. Family celebrations are culturally influenced rituals that families devise in accordance with holidays or developmental transitions such as Thanksgiving and weddings.

Fiese and Kline (1993) developed a seven-part classification of family ritual

settings. Dinnertime rituals are defined as those activities and interactions involved in a shared family meal. Weekend rituals involve recreational and planned events that occur on nonworking days. Vacation rituals are the leisure activities surrounding a family vacation. Annual celebrations include yearly commemorations of family birthdays, anniversaries, and the first day of school. Special celebrations include weddings, graduations, and family reunions as well as other celebrations that occur regardless of culture or religion. Religious holidays include rituals that honor Christmas and Chanukah. Cultural and ethnic traditions include naming ceremonies, funerals, and cooking ethnic foods.

Janine Roberts (1988) developed a categorization of family ritual styles. Strict rules for role performance and unforgiving attendance policies for all family members characterize the first style--the rigidly ritualized families. The second type are underritualized families who infrequently share rituals and usually ignore significant family milestones such as birthdays. The third type of families have a skewed ritualization focus in which they center their rituals exclusively on one aspect of family life such as religion or the needs of a single family member. Families with hollow ritualization, the fourth type, focus on the routine aspect of family rituals to the exclusion of almost all else, thereby enacting rituals with very little meaning and symbolic attachment. Families with interrupted ritualization, the fifth type, often deal with





79


unexpected family crises such as serious injury or death of a member. The sixth type, families with flexible ritualization style, consistently practice meaningful rituals, despite transitional periods, due to their ability to adapt roles and routines to the changing developmental needs of family members.

In a similar typology, Imber-Black and Roberts (1992) identified five ritual styles: minimized, rigid, obligatory, imbalanced, and rigid. A minimized ritual style is characterized by haphazard or absent rituals. In families with a minimized ritual style, important events are ignored or barely acknowledged and other aspects of life (e.g., work, telephone calls) are allowed to interrupt rituals. Unvaried and highly prescribed rituals that often leave family members feeling confined or constrained characterize the rigid ritual style. These families are unable to allow rituals to develop as the family changes and often reflect the family's narrow prescription of roles for its members. A lack of meaning for the ritual and an absence of spontaneity or playfulness characterize the obligatory ritual style. Families with an obligatory ritual style often feel anxiety or guilt at the thought of trying to adapt these rituals to meet the changing needs of family members. The imbalanced ritual style often occurs as a result of significant differences within the family. For instance, families in which members come from different cultures, religions, or races may embrace rituals from one heritage to the exclusion of the other. Lesbian and gay couples often have an imbalanced ritual style if one partner's family is accepting of the couple and the other partner's family is not. The flexible ritual style is characterized by the simultaneous ability to offer a sense of continuity and connectedness over time while validating developmental changes. The flexible ritual style allows for





80


growth and change among family members and relationships (Imber-Black & Roberts, 1992).

Bruess (1994) conducted one of few empirical studies to date of the types of

interpersonal rituals. She analyzed the rituals used by marital partners and members of friendship dyads. Seven major types of rituals were identified (as well as seven subtypes) in marital dyads. Couple time rituals are comprised of enjoyable activities, escape episodes, and togetherness rituals while symbolic rituals consist of private codes, play, and celebration rituals. Couples also engage in spiritual rituals, communication rituals, daily routine rituals, rituals of intimacy expression, and ritualized patterns-habitsmannerisms (Bruess, 1994). Members of friendship dyads also shared many of these ritual types.

Conceptualization of Rituals

Researchers in the fields of anthropology, psychology, and sociology have investigated the central dimensions of family ritual. Cultural anthropologists have suggested that symbols are the smallest unit of ritual (Turner, 1967, 1969).

Wolin and Bennett (1984) identified two dimensions of rituals: commitment and flexibility. They described the commitment dimension as the family's commitment to ritualization. Commitment to ritualization is the degree to which a family embraces and encourages rituals.. The ability of a family to adapt ritual to the changing needs of members and the family as a whole comprises the flexibility dimension. Wolin and Bennett (1984) argued a strong commitment to ritualization and the capacity to successfully adapt family rituals engenders in family members a sense of meaning and family identity.





81


Fiese and Kline (1993) identified eight dimensions of rituals: occurrence, roles, routines, attendance, affect, symbolic significance, continuation, and deliberateness. The frequency with which rituals are performed constitutes its occurrence. Family members adopt specified roles and perform designated routines in their enactment of family rituals (Bossard & Boll, 1950). Often, family rules govern who is expected to attend and absences become significant (Bennett et al., 1988). The emotional investment in the ritual constitutes its affect dimension while the attachment of meaning to the activity is reflected in the symbolic significance dimension. Continuation reflects the degree to which a ritual is passed down through the generations (Reiss, 1981). Finally, the planning and intentional execution of a family ritual constitutes its deliberateness dimension.

The Meanine Factor and the Routine Factor of Ritual

Fiese and Kline (1993) found that ritual dimensions could be compressed into two distinct factors: meaning and routines. The meaning factor describes the symbolic component of family rituals including the way that members make sense of family activities and interactions. The dimensions of deliberateness, symbolic significance, affect, attendance, and occurrence contribute to the meaning component. Consequently, meaningfulness is a cluster of associated factors involving emotional investment, expression of meaningfu~lness or specialness, a commitment to occurrence, and deliberateness in the planning and implementation of family rituals. The meaning factor of family ritual has been positively associated with adolescent self-esteem (Fiese, 1992) and marital satisfaction (Fiese et al., 1993) as well as negatively associated with anxiety health symptoms in adolescent raised by an alcoholic parent (Fiese & Kline, 1993).






82


Fiese and Kline (1993) investigated the relationship between family ritual

involvement and self-esteem in 241 college undergraduates. They found that students who reported having been raised in an alcoholic family reported significantly lower meaning in family rituals than did their counterparts raised in nonalcoholic families. Furthermore, those students from alcoholic families who also reported high levels of meaningfulness of rituals demonstrated fewer physical symptoms of anxiety than did students reporting lower meaningfulness of rituals. Regardless of the type of home in which the students were raised, meaningfulness of family ritual was related to high selfesteem levels.

In a study of family rituals and family functioning among 90 nonclinical Catholic families in New Mexico, January (1996) found a significant positive association between healthy family functioning and meaningful rituals. Using the Family Ritual Questionnaire to assess the degree and meaning of family ritual, she determined that families from different ethnic traditions differed in the degree of ritualization utilized over time. Specifically, Hispanic families experienced fewer declines in family ritual as children aged than did Anglo families. Families in this investigation that scored relatively higher on indices of healthy family functioning also reported higher levels of meaning in their family rituals (January, 1996).

The routine factor represents the degree to which roles are rigidly proscribed and the flexibility with which the family practices rituals. The second component, routines, is comprised of the dimensions of roles and routines. The routine factor, as defined by Fiese (1992), has not been empirically investigated.





83


Research on the Benefits of Rituals

Much of the research regarding the use of rituals has centered on the various

benefits that rituals produce for families rather than for dyadic relationships. Although a substantial amount of this literature is anecdotal in nature, some empirical investigations have undertaken the task of exploring the multifaceted purposes of rituals.

Enhancing family identity. Bossard and Boll's (1950) groundbreaking

investigation concluded that family ritual provides a sense of family culture by promoting the intergenerational transmission of a family's beliefs, values, and goals. Bossard and Boll (1950), writing from a sociological perspective, characterized ritual as "the core of family culture." As such, it provides a sense of continuity between generations as well as opportunities for learning and interaction. Psychologists Sameroff and Fiese (1992) described rituals as a significant aspect of the "family code," a regulatory system between individual and culture. Researchers have found that family rituals contribute to the establishment and maintenance of a family's identity (Reiss, 198 1) and relay family goals and values between and within generations (Wolin & Bennett, 1984).

In their landmark study of the role of ritual in families of alcoholics, Wolin and Bennett (1984) determined that rituals assist in the establishment of family identity. Their studies of the intergenerational transmission of alcoholism found that rituals seem to serve a "protective" function. Family rituals contribute to a sense of belonging by maintaining the roles, rules, and behavioral expectations. Although many rituals are culturally patterned, idiosyncratic family rituals enhance members' sense of a shared family identity thereby encouraging cohesiveness among members (Wolin & Bennett, 1984).






84


Rosenthal and Marshall (1988) investigated the intergenerational transmission of family ritual and its function of providing continuity between generations. They interviewed a random sample of 112 adults, stratified by age and sex, from Hamilton, Canada. These researchers found that several factors affected the degree of change in family ritual over time. Specifically, they identified decrease in family finances, increase in women working outside the home, and increase in travel as factors contributing to a decline in family ritualization over time. Furthermore, they identified specific rituals that seemed relatively impermeable to change: Christmas celebrations and Sunday family times. Finally, Rosenthal and Marshall (1988) found that parents raised in homes with a high level of ritualization as children tended to continue this tradition with their own children.

Creating a family paradigm. Models of family functioning have begun to include the role of ritual in the culture of the family. Reiss (198 1) posited a cognitive model of family functioning based in social construction theory. A culmination of 15 years of research involving interviews with more than 400 families, Reiss' (198 1) exploration of family rituals is rich with qualitative information. He posited that a family's reality is socially constructed in such a manner that members share a set of organizing beliefs and assumptions which structure patterns of daily living.

Enhancing streng-th and stability. Meredith (1985) found that rituals strengthened family bonds. In a subsequent study, Meredith et al. (1989) surveyed three generations of family members in an effort to determine the relationship between ritual and family strengths. They assessed family strength using measures of family loyalty, pride, trust, and effective conflict resolution. These researchers determined that a significant positive





85


relationship existed between family strength and family rituals. Additionally, they uncovered gender and generational differences in family members' perceptions of the importance of rituals. Women more than men, regardless of generation, viewed rituals as important contributors to family strength. Furthermore, members of elder generations identified rituals as making greater contributions to family strength than did younger generations (Meredith et al., 1989).

Enhancing belonging. Family rituals are thought to protect against loneliness and uncertainty by structuring certain family activities and clearly defining behavioral expectations and roles (Wolin & Bennett, 1984). Meredith (1985) found that rituals strengthened family bonds by encouraging meaningful and often symbolic contact among family members. Kline (1992) investigated the possible influence of family rituals on individual psychological organization, particularly attachment. In a study of college students, she found that those young adults raised in family environments rich in ritual had a stronger sense of "belonging-in-the-world" than those young adults raised in homes with relatively fewer rituals. In a second investigation, Kline (1992) identified that female students from families with relatively high levels of ritualization had an easier separation from parents and a better adjustment to college than did their counterparts from families with relatively lower levels of ritualization.

Strengthening couple identity. Bruess (1994) identified functions of rituals in interpersonal dyads such as marital relationships and friendships. She found that in couples "rituals enact 'masonry' functions by building and cementing the relational bond and commitment" (p. 186). Additionally, marital partners utilize rituals to contribute to the culture and the maintenance of the relationship by providing opportunities to learn





86


more about each other, promoting communication, and offering occasions for bonding, thus enhancing the happiness and longevity of the relationship. Couples also engage in rituals in order to define marital roles and promote continuity and organization of daily life. Bruess' (1994) investigation demonstrated that marital partners engage in rituals in order to create a sense of couple identity. Rituals also provide a sense of predictability and security. Finally, rituals enhance excitement via spontaneous flexibility among marital partners.

In a qualitative study of 26 unmarried Texas women, Mize (198 8) investigated the relationships among family of origin rituals, gender, and meaning. Using in-depth interviews, she identified levels of meaning and levels of ritualization. Mize (1988) found that women with relatively high levels of ritualization also reported relatively high levels of meaning compared to women who reported low levels of ritualization. She noted that ritual was considered a "female-centered experience" and that it afforded women the opportunity for affiliation.

Buffers ag-ainst stress. Models of family functioning have begun to include the role of ritual in the culture of the family. Reiss (198 1) posited a cognitive model of family functioning in which a family's reality is socially constructed such that members share a set of organizing beliefs and assumptions which structure patterns of daily living. This structure is challenged at times of family stress (e.g., developmental transitions) or crisis (e.g., death, acute illness, delinquency, teen pregnancy). The family may experience a reduction or even total loss of "implicit" functioning thereby rendering its members unable to attribute meaning to events. No longer a source of rejuvenation and resilience, events such as birthdays or anniversaries become a burden. In essence, the






87


family disconnects from its past. Families that successfully manage the stressor or crisis create a new "family paradigmn" which is viewed by members as a family accomplishment and restores "implicit" functioning. According to Reiss (198 1) the family rituals are the "repository" of the family paradigm--they permit the ongoing experience of the transformed values and assumptions. In this way, family rituals simultaneously perpetuate and embody the family's approach to the larger systems in which it is embedded.

The community system in which the family lives imposes meaning to stressful events and crises (Reiss & Oliveri, 1991). According to Reiss and Oliveri (1991) two dimensions of stress are "accountability" and "competence." Accountability refers to the degree to which the family is viewed as having created the stress while competence refers to the family's ability to cope with and protect itself from the stress. These researchers predicted that families that perceive themselves as having a sense of mastery over the social context of which it is a part would emphasize rituals that served to connect them to this social world. On the other hand, families that do not perceive themselves as able to influence their cultural context would emphasize rituals that isolate them from it. Reiss and Oliveri (1991) studied 45 middle class, White parents involved in PTAs in Washington, DC, using a Q-sort methodology. Their results were as predicted but are necessarily limited by the nature of the sample.

Boyce and colleagues (1983) theorized that rituals are an integral component of healthy family functioning. They suggested that social experiences such as family rituals and family routines facilitate a sense of continuity and stability that serve a protective function for families during times of stress. These authors argued that during times of





88


crisis, rituals perform important functions in family life--provision of social support and security. In this way, rituals foster feelings of protection and stability for family members challenged with stressful situations (Boyce et al., 1983).

Family rituals were described as an "index of family integration" by McCubbin and McCubbin (1987). As such, they assist the family in effectively dealing with stress, problems, and crisis. Family rituals are a resource that contributes to the family's ability to organize and give meaning to shared experiences in times of hardship. The resiliency model of family functioning was developed in order to expand prior family stress theory which had been based exclusively on middle class, White, heterosexual families (McCubbin, McCubbin, Thompson, & Thompson, 1995). This model acknowledges the significant contribution of family celebrations, traditions, routines, and rituals as resources for effectively coping with family stress and crisis. These resources are seen as a source of stability and support for families in times of need--in essence, a buffer against stress.

Studies of alcoholic families were among the first to test this "buffer theory." Using semi-structured interviews, the ritual patterns of 25 families of alcoholics (including adult children) were assessed (Wolin et al., 1980). These investigators predicted that families in which rituals were chronically disrupted by alcoholic member's behavior would be more likely to have children that later developed alcoholism or had alcoholic spouses. On the other hand, these researchers hypothesized, families that were able to maintain ritual practices in spite of alcoholic members would be more likely to break the cycle of alcoholism. The results of the investigation supported both hypotheses.





89


In a subsequent study of 68 couples raised in alcoholic families, the development of "family heritage" or ritual practices was explored (Bennett et al., 1987). Interviews of these couples identified the careful planning and implementation of family heritage practices as a protective factor in the intergenerational transmission of alcoholism. Couples described as "high deliberate" employed explicit strategies regarding their family ritual early in their marriages. These couples developed ritual patterns based on the spouse's nonalcoholic family of origin or other models of sharing traditions without alcohol.

In a continued effort to explore the concept of "deliberateness" in family ritual

patterns, 37 alcoholic families were compared with a matched sample of 45 nonalcoholic families (Bennett, Wolin, & Reiss, 1988). The emotional, behavioral, and cognitive functioning of the children in each family were assessed. Children from "high deliberate" families, families able to consistently plan and implement family rituals, functioned at higher levels emotionally and behaviorally than their counterparts in "low deliberate" families.

These pioneering studies of the influence of ritual in family functioning lend credibility to the concept of ritual as a protective factor in the face of family stress. Additional information regarding the cultural and ethnic backgrounds of participants might have been helpful in understanding the general izability of the results. Of course, the small sample size indicates a need for replication with larger numbers of families. Additionally, the self-identified sample may limit generalizability particularly in families in which denial is the predominant coping strategy (Jacob, Favorini, Meisel. & Anderson, 1978). Wolin and Bennett (1984) used a sample of families that had already identified





90


alcohol-related problems and therefore this sample may have been skewed with a greater proportion of families that had begun to take steps to solve family problems.

Fiese and Kline (1993) further studied the protective function of family rituals

using a sample of 241 undergraduates in New York and one of each student's parents. A preponderance of the sample was Whiite, female, age 17-21, and middle class. In this study only those rituals occurring in the dinnertime, vacation, and weekend settings were assessed. They found that undergraduates who were raised in a family in which one parent was alcoholic reported less meaning associated with family rituals. This trend was also reported for fathers but not for mothers. Furthermore, the students from alcoholic homes who reported highly meaningful family rituals scored significantly lower on a measure of the physical symptoms of emotional disturbance than did their counterparts who reported family rituals relatively lacking in meaning. Differences in family members' perceptions of rituals were found as well, suggesting that assessment of all family members is an important consideration in conducting research regarding ritual (Fiese & Kline, 1993). A possible response bias may exist due to the fact that students were asked to identify problematic drinking among their parents. Students raised in an alcoholic home may view family life in a more negative light than their counterparts in nonalcoholic families.

Families facing the stress of normative developmental transitions may also reap the protective benefits of rituals. In a study of 115 married couples making the transition to parenthood, Fiese and her colleagues (Fiese et al., 1993) found that couples with preschool-age children reported more meaningfulness in family rituals than did couples whose first child was still an infant. Couples who scored higher on the meaning factor of





91


family ritual reported greater marital satisfaction than their counterparts who reported relatively hollow family rituals. It is possible that meaningful rituals protect the couple from the developmental stress of first-time parenting and that these rituals serve to renew the relationship after an intense period of transition (Fiese et al., 1993).

In a comparative study of 43 families with a developmentally delayed child and 43 families with children developing "normally," Bucy (1995) investigated the role of ritual and family stress. She found no differences between family types on the basis of degree of ritualization or ritual meaningfulness. Regression analysis revealed that the interaction of meaningfulness and ritualization to be the best predictor of parenting stress. Bucy (1995) concluded that couples adapting to parenting a disabled child may increase their resilience to stress by creating a predictable and orderly family life characterized by meaningful family rituals.

Research on Dyadic Rituals

While family ritual literature is abundant, the literature regarding rituals shared by members of dyadic relationships is relatively scarce. Furthermore, what literature exists in this area is primarily anecdotal rather than empirical in nature (Imber-Black, 1 988a, 1988b; Oring, 1984).

In one of the few empirical studies of couple rituals, Fiese and her colleagues (Fiese et al., 1993) compared 115 couples making the transition to parenthood. These parents were predominantly middle class, White, and in their early 30s. These researchers used the Dyadic Adjustment Scale to assess marital satisfaction and the Family Ritual Questionnaire to assess the relative degree of meaning associated with the couples' rituals. She found that those couples who reported relatively greater degree of





92)


meaning associated with their rituals were more satisfied with their marriages than their counterparts who reported relatively hollow rituals.

In a qualitative investigation of possible types and functions of dyadic rituals, Bruess (1994) interviewed 20 married couples and collected data from an additional 79 couples. Bruess (1994) identified a typology of rituals in two kinds of dyadic relationships: marriages and friendship dyads. While the ritual typology for marital dyads had some overlap with that for friendship dyads, significant differences were found. This suggests that rituals may be relationship-specific. Bruess' (1994) investigation yielded seven types of rituals among couples and six ritual types among friendship pairs. In a subsequent study of the functions of ritual in marital and friendship dyads, Bruess (1994) sampled 494 married partners. She reported that couples in this study described benefits from ritual use similar to those reported by families. Bruess (1994) concluded that for married couples rituals fostered a sense of belonging and couple identity, and facilitated opportunities for positive interaction.

Valenti (1997) investigated the effects on relationship satisfaction and intimacy of educating heterosexual couples about the use of positive rituals in their relationships. A total of 45 White couples participated in a lecture regarding the positive use of rituals in relationships. Measures of intimacy and relationship satisfaction as well as the Family Ritual Questionnaire were administered before and after the lecture. Valenti (1997) found no statistically significant results. Valenti (1997) discussed the results of this study as due in part to the homogeneity of the sample, small sample size, and lack of sensitivity of the instruments used to assess couple rituals.





93


Rituals in Lesbian Relationships

To date, no empirical investigations have been conducted on the influence of ritual in lesbian relationships. However, in a national survey of 706 lesbian couples, Bryant and Demian (1994) found that 60% of lesbians had symbolized their commitment with a relationship ritual. Anecdotal literature points out both similarities and differences between the ways that lesbian couples and heterosexual couples use ritual in their relationships (Butler, 1991; Carter & McGoldrick, 1989; Imber-Black, 1988a; ImberBlack & Roberts, 1992; Roth, 1985; Slater, 1995). This literature is reviewed in the following section.

Several authors suggest (Butler, 1991; Imber-Black, 1 988a; Imber-Black &

Roberts, 1992; Slater, 1995) that one of the ways in which lesbian couples cope with the absence of public acknowledgement and social approval for their relationships is by developing rituals to validate their unions. According to Slater (1995), lesbian couples live in a greater cultural context in which particular family types--namely, traditional heterosexual families with married parents--are "invited" to be recognized through ritual marking important milestones such as marriages, deaths, births, retirements, and graduations. In contrast, lesbian couples and families are not invited. Consequently, lesbian couples create unique rituals to assist them with basic needs faced by most couples. Slater (1995) noted, "Boundaries around their relationship that are socially recognized and respected, social validation that the family unit is a legitimate and viable grouping, and a sense of membership in a group of similar families that make up the surrounding social community" (p. 63). She identified three types of rituals commonly utilized by lesbian couples: daily rituals and family traditions, celebration and holiday




Full Text

PAGE 1

THE INFLUENCE OF RITUAL USE, INSTITUTIONAL SUPPORT, AND CONFLICT RESOLUTION STYLE ON LESBIAN COUPLES' RELATIONSHIP SATISFACTION By CAROLINE PACE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY V, UNIVERSITY OF FLORIDA 1999

PAGE 2

ACKNOWLEDGMENTS I would like to extend my deepest thanks to those family members, friends, and colleagues who have given me guidance, support, and encouragement throughout the process of conceptuaUzing, designing, implementing, and writing this dissertation. It was my incredible good fortune that Dr. Ellen Amatea was "randomly" assigned to be my advisor when I began my studies at the University of Florida. She has served as advisor, teacher, supervisor, and chair of my dissertation committee. I am so appreciative of the rigorous precision with which she thinks and writes as well as her endless patience, prompt feedback, and scholarly brilliance. Dr. David Miller, who speaks the language of statistics with exquisite grace, has been a veritable fountain of information and a delightful human being. Dr. Phyllis Meek— a wisewoman, mentor, and feminist role model — has believed in me with unflinching certainty and impeccably modeled for me the skills of critical thinking and feminist analysis. I am grateful to Dr. Mary HowardHamilton for the opportunities she has given me to collaborate with her and for the vote of confidence. Words cannot adequately express the deep gratitude I feel for my colleagues at the Gainesville Family Institute. The mentoring, training, and supervision I received from the GFI faculty and students have been nothing short of transformative— both personally and professionally. I am privileged to know Andres Nazario— a wonderfully feisty and compassionate person. He has offered me numerous wonderful opportunities

PAGE 3

to challenge myself and to do so in the warmth of his support. I am also grateful to my friend, Lynne Barolet-Fogarty, for her consistent encouragement and the opportunity to learn from her spirit. I am appreciative of Herb Steier, an extraordinary role model of irreverence. Without my community of lesbian friends and family this project would not have been possible. Participation in a study that requires one to acknowledge one's lesbian identity and to examine one's relationship is potentially challenging and risky. I would like to thank the 90 couples who took that risk. Also, I am indebted to Woody Blue (who encouraged me to "exploit" her outgoing personality in the search for potential participants) for her help with data collection. My lesbian family — Mickey Barron, Suzanne Miller, and Lorelei Esser — kept me nurtured and encouraged. My family of origin has offered inspiration and invaluable support. I am grateful to my mom for her unconditional love and absolute faith in me. The memory of my father — a scholar and a gentle person — ^has been my constant companion on this journey. Additionally, I appreciate the example set by my Aunt Caroline, the first woman in the family to get a Ph.D. ,„ Last, but certainly not least, I am indebted to my partner, Jude, for her steadfast love and unquestioning faith in me. Her patience and calm have sustained our relationship. Her intelligence and willingness to challenge me have inspired my creativity. iii

PAGE 4

I TABLE OF CONTENTS ^ ' . u..: -:,. V, M . Page ACKNOWLEDGMENTS ii ABSTRACT vi CHAPTERS 1 INTRODUCTION 1 Theoretical Framework 3 Scope of the Problem 7 Need for the Study 22 Purpose 25 Research Questions 25 Definition of Terms 26 Organization of the Study 30 2 REVIEW OF THE LITERATURE 32 Introduction '\ 32 Lesbian Couples: A Contextual View 32 Research on Lesbian Relationship Satisfaction 33 Research on Social Support Influences 53 Research on Ritual 74 Research on Conflict Resolution 95 Summary 211 3 METHODOLOGY 113 Statement of Purpose 1 1 3 Hypotheses 113 V Delineation of Relevant Variables II4 Data Analysis Hy Description of the Population 1 1 g Sampling Procedures Hg Subjects 221 iv

PAGE 5

Data Collection 126 Instrumentation 127 4 DATA ANALYSIS AND RESULTS 144 Analysis Procedures 1 44 Analysis Results 150 Hypothesis Testing 158 Summary 162 5 DISCUSSION 163 Overview of the Study 163 Research Sample 164 Association Between Couple Relationship Satisfaction and Partner Agreement Regarding Conflict Resolution Style Utilized 165 Association Between Couple Relationship Satisfaction and Conflict Resolution Style Utilized 168 Association Between Couple Relationship Satisfaction and Degree of Relationship Ritualization 171 Association Between Couple Relationship Satisfaction and Ritual Meaningfulness 175 Association Between Couple Relationship Satisfaction and Institutional Support for the Relationship 1 78 Recommendations 179 Summary 187 APPENDICES : , . ^ A LETTER TO PARTICIPANTS 188 B PARTICIPANT SELECTION PROTOCOL 191 C DEMOGRAPHIC DATA SHEET 191 D INSTITUTIONAL SUPPORT SCALE 193 REFERENCES 195 BIOGRAPHICAL SKETCH 219 V

PAGE 6

Abstract of Dissertation Presented to the Graduate Scliool of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE INFLUENCE OF RITUAL USE, INSTITUTIONAL SUPPORT, AND CONFLICT RESOLUTION STYLE ON LESBIAN COUPLES' RELATIONSHIP SATISFACTION By Caroline Pace May, 1999 Chairperson: Ellen Amatea, Ph.D. Major Department: Counselor Education The purpose of this study was to examine the influence of five variables on the relationship satisfaction of lesbian couples. Four of these variables focused on aspects of partner interaction: (a) the extent of ritualization in couple life, (b) the meaningfulness of ritualization in couple life, (c) the type of conflict resolution style utilized by couples, and (d) the agreement between partners regarding conflict resolution style. The first two of these variables were measured by the Family Ritual Questionnaire while the latter two variables were measured by the Conflict Resolution Style Inventory. The fifth variable— perceived institutional support— focused on the couples' participation in the larger social context as measured by the Institutional Support Scale. The sample for this study consisted of 180 women (90 couples) who self-identified as lesbians and had lived together in a committed monogamous relationship for at least one year. Bisexual women vi

PAGE 7

and transgendered women were excluded from the study as were women living with children. Regression analyses were used to evaluate the contribution of the five variables in predicting couple relationship satisfaction as measured by averaging the Dyadic Adjustment Scale scores of both partners in a relationship. Statistically significant (£<.05) associations were found between couple relationship satisfaction and two of four conflict resolution styles — positive problem-solving and conflict engagement styles. No significant associations were found between couple relationship satisfaction and partner conflict resolution style agreement, extent of relationship ritualization, ritual meaningfubess, and perceived institutional support. However, post hoc analyses revealed indirect effects of relationship ritualization and ritual meaningfubess on relationship satisfaction when conflict resolution style was controlled. Discussion of these results, the study's limitations, and suggestions for future research with this population were then presented. , f f * . ,«' vii

PAGE 8

CHAPTER 1 ^ S INTRODUCTION ' ^ Love relationships play a significant role in the lives of most human beingsincluding lesbians. Traditionally, researchers interested in determining what factors contribute to stable and satisfying intimate relationships have focused primarily on heterosexual married couples (ActeUi, 1988; Davis & Oathout, 1987; Gottman, Coan, Carrere, & Swanson, 1998; Jacobson & Addis, 1993; Rusbult, 1983). Only recently have researchers begun to examine the love relationships of "nontraditional" couples such as cohabiting heterosexuals and gays and lesbians (Cardell, Finn, & Maracek. 1981; Dailey, 1979; Duffy & Rusbult, 1986; Kurdek, 1988a, 1992b, 1998; Kurdek & Schmidt, 1986a, 1986b, 1986c). However, these researchers have tended to look only at the internal processes of couple Iife--how couples report making decisions, resolving conflicts, or structuring their roles. There has been scant attention given to couples' perceptions of the larger social context and its influence on couple relational processes. ; Because of the negative cultural attitudes towards homosexualify, the consequences of partners choosing to acknowledge publicly their lesbian love relationship can present significant problems that would not be encountered by heterosexual partners. For example, researchers (Carter & McGoldrick, 1989; ImberBlack & Roberts, 1992; Roth, 1985; Slater, 1995) have identified that the lack of marker events for lesbian couples-such as weddings or the birth of a child-often creates

PAGE 9

difficulties for couples trying to define their change in status fi-om single to couplehood. Additionally, life cycle transitions such as retirement, serious illness, or death can be problematic for lesbian couples. Disqualification of the couple relationship by the larger society may lead to public ceremonies in which partners have to disguise their feelings and their relationship, thus intensifying their problems rather than promoting a sense of connectedness and continuity, ' The lesbian couple's acknowledgement of their relationship in the larger community is very likely to be influenced by the extent to which either or both partners perceive it is "safe" to acknowledge their relationship. According to FBI statistics, hate crimes targeting gays and lesbians rose 42% from 1991 to 1995 (Mills, 1997). In Florida, the number of hate crimes committed against members of sexual orientation minorities has nearly tripled in the last 2 years ("Hate Crimes on the Rise," 1998). Couples who feel they cannot risk being open about their relationship may feel isolated from their work, school, or social community. On the other hand, those couples who choose to risk acknowledging their relationship may form connections within a community of likeminded lesbian couples but feel isolated from the larger heterosexual community and its cultural instimtions. Lesbian partners often deal with a loss of status in the larger community if they are open about their relationship or with a sense of alienation and oppression if they maintain secrecy about their relationship (Carter & McGoldrick, 1989; Roth, 1 985). Consequently, lesbian couples often must find ways to develop their love relationships in a cultural context that, at best, marginalizes or regard these relationships as invisible and, at worst, subjects them to overt hostility and violence.

PAGE 10

3 Although the larger social context appears to have a significant influence on the lives of lesbian couples, no studies to date have examined the relative effect of institutional support on a couple's relationship satisfaction. Instead, most researchers have looked exclusively at the internal dynamics of the couple. This study sought to assess the relative influence on couple relationship satisfaction of support from the larger context as well as the influence of three aspects of couple interaction. Theoretical Framework To understand better the factors influencing lesbian relationship satisfaction, a theoretical framework is needed which addresses the contributions of internal couple dynamics and the external dynamics of the larger cultural context in which the couple is embedded. No existing theory integrates both perspectives. Consequently, this study was based upon two theoretical traditions: Gottman's (1994) balance theory and , Kurdek's relationship appraisal model. % , ' Gottman's (1994) balance theory has been used to conceptualize the interactional processes by which heterosexual married couples sustain satisfying relationships. This theory is based on muhimethod assessments of significant aspects of couple interaction (e.g.. partner's problem solving, power, affect, distance regulation, and physiology) in an 8-year longitudinal investigation of 79 couples as well as laboratory observations and ongoing assessments of four cohorts of couples in different stages of the family life cycle. Over the past two decades, Gottman and his colleagues conducted empirical research that ultimately led to the formulation of balance theory. According to this theory, satisfied couples maintain a balance in their ratio of positive to negative interactions. " ' -v" '' ,>.''."

PAGE 11

Two theoretical premises underlie balance theory. First, satisfied couples develop a ratio of positive to negative affective interactions as well as the ratio of positive to negative problem solving interactions. Over time, the negative aspects are "balanced" by positive aspects. For satisfied couples, this balance is "tilted" toward the positive. Gottman (1994) described balance theory as "an ecology of marital behaviors in which a ration of positivity to negativity that is highly tilted toward positivity needs to be maintained" (p. 380). r, -.^ The second theoretical premise underlying balance theory is the notion that partners in satisfied couples develop a "fit" regarding preferred interactional style. Gottman (1994) postulated three types of "stable" couples: volatile, avoiding, and validating. Each of these couple types results in a "stable," satisfied relationship in which positive and negative interactions are balanced. Volatile relationships balance a high level of argument and conflict with significant amounts of laughter and passionate romance. In other words, high volumes of positive interactions balance high volumes of negative interactions. On the other hand, high levels of neutral affect characterize avoiding types of couple relationships. The avoiding couple type "appears to involve a minimization of the importance of disagreement. It results in a good deal of calm interaction, but pays the price with emotional distance" (Gottman, 1994, p. 182). In avoiding couple types, a low level of negative interactions is balanced by a low level of positive interactions. Finally, the validation type of couple balances careful selection of when to confi-ont conflict with warmth and "we-ness." Partners in these relationships demonstrate little intensity regarding either disagreement or passion. A moderate amount of negativity is balanced by a moderate amount of positivity in validating couples.

PAGE 12

5 Gottman (1994) addressed cultural variation in the distribution of couple types, but only from an ethnic point of view. He postulated that while one couple type may be more prevalent in a specific ethnic culture, the eroding effect of an unbalanced (e.g., high ratio of negativity to positivity) relationship would remain the same. Gottman has not addressed the applicability of the balance theory to lesbian couples although he reported a study — ^yet to be published — regarding lesbian and gay couples' conflict resolution strategies (cited in Gottman, 1 994). Although Gottman's balance theory (Gottman, 1994) incorporates behavior, cognition, and physiology, it is primarily an interactional model which does not account for contextual forces that may influence couples. His theory is based on the study of heterosexual marital couples — couples that benefit from the social sanctions given their relationships by the larger cultural context. Members of marginalized and stigmatized groups, such as lesbians, often do not receive the institutional support for thenrelationships afforded heterosexual couples. Brown (1989a) described three aspects of lesbian couples' cultural context that cut across all demographic characteristics: marginality, biculturalism, and normative creativity. Marginality is expressed as a feeling of "otherness" and a cognitive recognition of minority status. According to Brown (1989b, 1991), partners in lesbian relationships may experience this feeling of "otherness" as nontraditional gender role identity because they do not receive the traditional recognition afforded a female via her socially proscribed association with a male. For some lesbians this may mean feeling as if they do not belong to their gender; however, this feeling of gender inappropriateness often changes as women are able to

PAGE 13

reevaluate belief systems and validate the importance of nontraditional gender roles in couple relationships (Vargo, 1987). Unlike other minority groups that share a common experience, lesbians come from a variety of religions, racial and ethnic cultures, age groups, or socioeconomic status (Brown, 1989b). In fact, there is no singular lesbian couple reality. Lesbian partners are members of both the heterosexual and homosexual cultures; they are, in essence, bicultural. Almost all lesbians have lived at least the early part of their lives under the influence of assumed heterosexuality, and even after affirming a lesbian identity, many women continue to pass as heterosexual in certain circumstances (Brown, 1 989b). Although this bicultural experience is not unique to lesbian couples, unlike their .:' counterparts in other oppressed minority groups, they do not share their minority status with members of their own families of origin (Brown, 1 989b). According to Brown (1 989b), as a result of their minority status in a group for which previously established norms do not exist, lesbian couples are afforded the opportunity to create their ovra norms. Accordingly, lesbians may enjoy tremendous freedom to invent their relationships and roles in a creative manner (Grahn, 1984). One of the most important ways that families create for their members a sense of belonging is through the establishment of riUials. Lesbian families have not been afforded many of the traditional rituals afforded their heterosexual counterparts such as wedding showers, marriage ceremonies, and anniversary celebrations, yet they may utilize what Brown : (1 989b) terms normative creativity to devise rituals of their own. Building on Brown's contextual model and Gottman's balance conceptualization, Kurdek (1998) hypothesized a model of relationship satisfaction applicable to both

PAGE 14

heterosexual married and homosexual cohabiting couples. Kurdek (1998) posited that relationship satisfaction was related to both internal forces within the relationship — such as shared agreements about problem solving, autonomy, equality, and intimacy — and external forces outside the relationship — such as institutional supports. Kurdek (1998) found that institutional supports act as barriers to leaving the relationship, particularly for heterosexual married couples. The inclusion of external forces as significant influences on relationship satisfaction gives Kurdek's (1998) model a contextual focus that Gottman's (1994) balance theory does not address. The present study utilized both Gottman's (1994) balance theory and Kurdek's (1998) model of internal and external forces in conceptualizing potential sources of influence on relationship satisfaction. Specifically, in this study the assessment of the balance in relationship positivity and negativity were represented by measurement of the variables of degree of ritualization and conflict resolution style respectively. In addition to these internal factors, the influence of external factors— institutional support for the relationship — were assessed as well. : , ' ' Scope of the Problem While it is difficult to "measure" just how many women are involved in lesbian love relationships, estimates suggest that over 70% of lesbians are in committed relationships ("The Gay Market", 1992; Gonsiorek & Weinrich, 1991). No census data exist regarding the number of lesbians in the United States although estimates range from 2% to 26% of the population, with most estimates in the 2% to 10% range (Brown, 1991; Paul & Weinrich, 1982). Yet, the importance of lesbian relationships is not limited to the women in them. In fact, members of lesbian couples live in a greater family context ; as

PAGE 15

daughters, sisters, aunts, cousins, and parents. It has been estimated that one in every five famiUes has a gay or lesbian child (Dahlheimer & Feigal, 1991) and between 6 million and 14 million children have at least one lesbian or gay parent (Kantrowitz, 1996). The extensive impact of this sexual orientation minority on such a great number of people supports the need for further research in this area. • > Lesbian communities have been shown to significantly value therapy (Morgan, 1992). In fact, lesbians comprise a substantial amount of the women seeking psychological services (Morgan, 1992). While many therapist training programs have begun to include coursework regarding the needs of clients from culturally diverse backgrounds, a focus limited to racial, ethnic, and/or religious minorities may result in the exclusion of lesbian populations from these classes (Markowitz, 1991). Martin (1982) has suggested that therapists operate from a perspective negatively biased as a result of having been socialized m a culture that simultaneously stigmatizes and renders invisible gays and lesbians. In order to work effectively with lesbian clients, therapists must re-educate themselves, yet a dearth of empirical research exists to help clinicians understand this population. Research on Lesbian Couple Relationship s Historically, minority issues as a whole have been excluded from academic study and literature; consequently, until very recently homosexuality has been omitted as a focus of research inquiry. The topic of lesbian relationship issues— essentially a minority issue within a minority issue— has received even less attention. Essentially, women in lesbian relationships have suffered a greater oppression due to a double minority status-as a sexual orientation minority and a gender minority. This double minority status

PAGE 16

9 means that documentation of lesbian couple issues have seldom been noted in the literature, even by couples and family therapists (Brown, 1991; Clark & Serovich, 1997). In their review of literature on lesbian, gay, and bisexual issues in couples and family therapy journals during the period 1975-1995, Clark and Serovich (1997) found that lesbian couples in particular and homosexual families in general were rarely discussed. Of the 13,217 articles reviewed from 17 couples and family therapy journals, only 77 (.006%) focused on lesbian, gay, or bisexual issues (Clark & Serovich, 1997). Of these 77, only 15 (19.5%) specifically addressed lesbian couple relationships. Thus, articles regarding lesbian couples accounted for .001% of the total number of articles published in the 20-year span covered in the analysis. Further research is needed to understand this important segment of society. What research has been conducted seems to have followed three traditions. Initially, scholars were interested in determining what made lesbians and gay men different from their heterosexual counterparts. This led to a narrow focus on the psychological makeup of the individual rather than on lesbians and gay men as participants in larger relationship systems (e.g., as partners in couple relationships or members of families). With the shift in psychology away from individual pathology and toward family systems theories, comparative studies sought similarities between lesbian and gay couples and families and their heterosexual counterparts. Most recently, through the larger lens of contextual conceptualizations, researchers have begun to investigate the concurrent processes by which lesbians and gay men are both oppressed by and members of multiple systems including familial, work, educational, legal, and medical institutions.

PAGE 17

10 The contributions from each of these three traditions is reviewed in the following sections. The search for differences: Individual identity models . Initial investigations regarding homosexuality were influenced by the grov^ of psychiatry and sexology in the late 19th century, as were the efforts to pathologize gays and lesbians as deviant individuals (Faderman, 1991). This early literature decontextualized partners in lesbian and gay relationships by viewing them as individuals participating in sinful behavior. In part, it was this shift away from viewing lesbians and gays as sinners engaged in an "unnatural" behavior and toward regarding them as individuals with deviant personality types that allowed the conceptualization of homosexual identity development to arise. Although the concept of homosexual identity development perpetuated an individual focus in lesbian and gay research, it had a profound effect on the depathologizing of sexual orientation minorities. Lesbian identity development, the process by which a lesbian identity is formed and managed, is considered one of the most important factors that affects the lives of lesbians individually and in couples (Cass, 1979; Chapman &, Brannock, 1987; Coleman, 1982; DeMonteflores & Schultz, 1978; Minton & McDonald, 1984; Silberkraus, 1995; Sophie, 1 986; Troiden, 1 979). Unlike their heterosexual counterparts, lesbians must develop and maintain their identities within a larger cultural context structure that continues to control, legislate, oppress, and stigmatize any kind of sexual variance. In the 1 950s several significant studies supported a shift in the conceptualization of homosexuality from deviant to a normal variant of sexuality. One of the first to do this was a study by Kinsey and colleagues (Kinsey, Pomeroy, Martin, & Gebhard, 1953)

PAGE 18

11 which documented that lesbian feelings and experiences were common to a widespread group of women rather than an isolated few as had previously been thought. Later investigations would focus on the effect of lesbian identity development on relationship quality (Jordan, 1995; McGuire, 1995; Silberkraus, 1995; Wendell, 1986). By the 1960s the feminist movement emerged. Women united in the call for the deconstruction of patriarchal power structures. Concurrently, lesbians and gay men began rallying against institutional heterosexism. In contrast to prior research-which focused on the etiology of the "deviant" sexual behavior and its cure-new research focused on the "normal" personality development of lesbian and gay individuals (Armon, 1960; Dank, 1971; Evans, 1970; Freedman, 1971; Green, 1972; Hopkins, 1969; Loney, 1971; Saghir &, Robbins; 1973; Siegelman, 1972; Thompson, McCandless, & Strickland, 1971; Wilson & Green, 1971). Although this research was bom out of a political context, investigations regarding lesbian and gay relationships were ignored in favor of a ' continued focus on the individual. It was Cass' ( 1 979) groundbreaking work on a five-stage model of homosexual identity formation that influenced researchers to shift to conceptualizing homosexuality from the broader framework of identity rather than simply from a (sexual) behavioral perspective. Subsequently, other models of lesbian and gay identity development have been proposed (e.g.. Bell, Weinberg, & Hammersmith, 1981; Chapman & Brannock, 1987; Coleman, 1982; deMonteflores & Schultz, 1978; Minton & McDonald, 1984; Silberkraus, 1995; Sophie, 1986; Troiden, 1979). It was the influence of homosexual identity development literature that spurred the eventual declassification of homosexuality as a mental illness. In 1973 the American Psychiatric Association voted

PAGE 19

to remove homosexuality from its official list of mental disorders. Subsequently, in 1975 the American Psychological Association elected to support homosexuality as a legitimate lifestyle. It was not until 1991 that the American Association of Marriage and Family Therapy addressed the issue of sexual orientation minorities by including a nondiscrimination clause m its code of ethics (American Association of Marriage and Family Therapy, 1991). Furthermore, it was only last year that the Commission of the Accreditation for Marriage and Family Therapy and Education adopted a similar policy (Clark & Serovich, 1997). ^ ' , ' • The search for similarities: Lesbian relationships and families . As researchers searched for differences between lesbians and gay men and their heterosexual counterparts, they began to focus on the lesbian family. Initially, individuals were once again the center of this research. Investigators sought to determine whether or not differences existed between lesbian and heterosexual mothers and the children raised in these households (Bozett, 1987; Chesler, 1986; Cramer, 1986; Gibbs, 1989; Gottman, 1990; Green, Mandel, Hotvedt, Gray, &. Smith, 1986; Hoeffer, 1981; Kirkpatrick, Smith, & Roy, 1981; Macklin, 1980; Pagelow, 1980). For instance, more similarities than differences were found regarding the psychological well being, gender identity, sex role socialization, and availability of male roles models for children raised by lesbian mothers and children raised by heterosexual mothers (Cramer, 1986; Gibbs, 1989; Gottman, 1990; Green et al., 1986; Hoeffer, 1981; Kirkpatrick et al., 1981). Lesbian couples and families have only recently been publicly recognized as legitimate family forms. Although many studies have attempted to delineate the factors that contribute to relationship satisfaction in heterosexual couples (see review by Lewis

PAGE 20

13 & Spanier, 1979), until recently very few empirical studies have sought to determine the factors that influence relationship satisfaction in lesbian and gay couples (Bell & Weinberg, 1978; Kurdek, 1998; Peplau, Padesky, & Hamilton, 1982). Only in the last 20 years have empirical studies compared similarities among lesbian, gay, and heterosexual couples with regard to the degree and correlates of relationship satisfaction (Blumstein & Schwartz, 1983; Cardell et al., 1981; Duffy & Rosbult, 1986; Kurdek, 1988b, 1992a, 1992b, 1998; Kurdek & Schmitt, 1986a, 1986b, 1986c, 1987b; Peplau & Cochran, 1990; Schneider, 1986). For instance, in one of the classic investigations to compare lesbian, gay, and heterosexual couples, Blumstein and Schwartz (1983) studied decision making regarding money, work, and sex. They examined the interpersonal dynamics within four types of relationships: cohabiting lesbian, cohabiting gay, cohabiting heterosexual, and married heterosexual. Blumstein and Schwartz (1983) interviewed 300 couples and administered questionnaires to more than 3,500 couples; the focus of this large body of work was on the internal "workings" of the relationship. While informative, such comparative studies have decontextualized lesbian relationships by ignoring significant cultural influences such as sexism and heterosexism. The contextual view: T .esbian couples as a minoritv familv form . Although research on heterosexual couples provides some insights into the relationships of lesbian and gay couples, the larger social constructs of gender and sex-role socialization prevent a parallel application of such research to members of same-gender couples (Morin, 1977). Because the socialization of males and females differs in this culture, it is important to investigate the unique contribution each gender brings to ways of relating in couples (Vargo, 1987). Because lesbian and gay relationships are comprised of two people of the

PAGE 21

same gender, the couple inherently gets a double measure of the positive and the negative aspects of their gender socialization. On the other hand, in heterosexual couples the positive and negative aspects of female cultural socialization as well as the poshive and negative aspects of male cultural socialization are combined. Thus, studies based exclusively on heterosexual couples may not be generalizable to lesbian and gay couples. According to many researchers (Murphy, 1992), lesbian couples differ from other couple types (including other homosexual couples) because they are "doubly devalued and oppressed" as women in a culture that is sexist and misogynist and as homosexuals in society that is homophobic and heterosexist. Furthermore, lesbian couples face other stresses that may affect their relationships including socially imposed restrictions on the public display of affection and the lack of visible role models, culture, and history (Riddle & Sang, 1978; Sang, 1978). Additionally, lesbian couples must find ways to thrive in a culture that invalidates and even overtly opposes the legitimization of this family form (Maggiore, 1988). Several researchers have emphasized the need for studies that focus exclusively on lesbians and lesbian relationships (Burch, 1986; Peplau, 1982; Peplau, Padesky, & Hamilton, 1982). Social Su pport and Lesbian Relationship Satisfaction Very few researchers have focused on examining the social support networks of lesbian couples. No research, to date, has examined the relationship between institutional support and relationship satisfaction in lesbian couples. Instead, most researchers have investigated the dimension, types, and degrees to which individuals who are lesbian receive social support from family of origin, partners, gay-lesbian friends, heterosexual friends, and health professionals (Albro & Tully, 1979; Aura, 1985; Chafetz, Sampson,

PAGE 22

Beck, & West, 1974; Grossman & Kemer, 1998; Jordan, 1995; Jordan & Deluty, 1998; Kurdek, 1988a; Kurdek & Schmitt, 1987a, 1987b; Levy, 1992; Meyer, 1989; Olson, 1988; Walsh, 1995; Wayment & Peplau, 1995; Wood, 1983). One of the first studies of lesbian social support found that lesbians received most support from lesbian and gay friends rather than from heterosexual friends and family (Chafetz et al, 1974). In a later investigation, Albro and Tully (1979) discovered that lesbians relied on members of the homosexual micro-culture for support rather than members of the heterosexual macroculture. Furthermore, researchers support the notion that social support enhances lesbians' personal well being (Meyer, 1989; Olson, 1988; Walsh, 1995; Wayment & Peplau, 1995). Most models of homosexual identity development acknowledge the importance of social support as a factor in the development of a positive lesbian or gay identity (Cass, 1979; Chapman & Brannock, 1987; Coleman, 1982; Troiden, 1979; Walsh, 1995). Although these studies have been informative, most of these investigations have examined social support reported by individuals. It is unclear whether social support of the couple contributes to relationship satisfaction in lesbian couple relationships. Leavy and Adams (1986) determined that lesbians involved in committed relationships scored significantly higher on measures of social support and self-esteem than those not in relationships. Similarly, Walsh (1995) found that lesbians who reported the highest social support were those with high self-esteem, low internalized homophobia, and involved in committed relationships. Even among studies with samples composed of both lesbian couples and gay couples, social support systems have seldom been assessed. In a study of lesbian and gay couples, Kurdek (1988a) found that friends

PAGE 23

16 were the primary sources of social support. He determined that psychological well being was not influenced by lack of family support. Kurdek (1988a) hypothesized that lesbians and gay men have compensated for family support lost due to the stigma associated with a homosexual identity by creating an extended family of friends. The three contextual elements of lesbian couples' experience described by Brown — ^marginality, biculturalism, and normative creativity — illustrate key factors that may serve to unify lesbian communities. Members of minority groups frequently connect with each other via those aspects of their culture that differentiate them from members of the majority group. These differentiating experiences, characteristics, beliefs, and customs enhance the sense of belonging and affiliation. Empirical research has validated Brown's (1989b) notion that many lesbians receive support from other members of the lesbian community that they do not receive from heterosexual women or family of origin members (Cass, 1979; Chafetz et al., 1974; Kurdek, 1988a; Walsh, 1995). _ Studies of exclusively lesbian, exclusively gay, and combined lesbian and gay samples have focused on interpersonal social supports provided couples. However, few researchers have focused on examining the influence of other sources of support on these love relationships. Institutional supports confer status and privileges upon members of a couple in regard to financial, legal, religious, familial, and social welfare domains. Adams and Jones (1997) proposed that institutional barriers are needed to sustain heterosexual marriages. In one of the few studies of institutional supports given lesbian, gay, and heterosexual couples, Kurdek (1998) hypothesized that barriers to leaving a relationship would apply pressure to couples to stay together in spite of personal unhappiness. Specifically, he postulated that the individuals in socially sanctioned

PAGE 24

relationships (e.g., heterosexual married couples) would report significantly greater barriers to leaving the relationship than their counterparts in socially stigmatized relationships (e.g., cohabiting gay or lesbian couples). As expected, he found that partners in lesbian and gay couples reported fewer barriers to leaving the relationship than did married heterosexual partners (Kurdek, 1998). Furthermore, he determined that barriers to leaving the relationship effectively predicted couple dissolution for lesbian couples and heterosexual married couples. Although this study suggests that lesbian and gay couples perceive fewer institutional supports for their relationships than do heterosexual married couples, there is little information regarding the extent to which institutional supports influence relationship satisfaction. This study addressed this limitation. • ^ Couple and Family Ritual Use and Relationship Satisfaction Lesbian couples are often afforded the opportunity to create their own norms as a resuh of minority status m a group for which previously established norms do not exist. Consequently, lesbians may enjoy tremendous freedom to invent their relationships and roles in a creative manner (Brown, 1989b; Grahn, 1984). One of the most important ways that couples and families create a sense of belonging for their members is through the estabhshment of rituals. ' Researchers have begun to investigate the phenomenon of family rituals as attention in the literature has shifted from a focus on family pathology to resiliency. However, this research focused exclusively on heterosexual families. Furthermore, most of the literature regarding the influence of rimal in families defined family in a traditional manner— a constellation of parents, children, and first-order extended family relations , .

PAGE 25

18 such as grandparents, aunts, uncles, and cousins. Very little research has focused exclusively on the couple subsystem. Family rituals have been recognized as substantial forces that organize behavior in family systems (Bossard & Boll, 1950; Fiese, 1992; Fiese, Hooker, Kotary, & Schwagler, 1993; Sameroff & Fiese, 1992; Wolin & Bennett, 1984). According to Fiese (1992) "family rituals appear to exert influence on family life by pairing meaning and affect with patterned interaction" (p. 634). Rituals have been shown to play an important role in the development and maintenance of family satisfaction, identity, support, strength, and cohesiveness (Bossard & Boll, 1950; Curran, 1983; Fiese, 1992; Jenson, James, Boyce, & Hartnett, 1983; Meredith, 1985; Meredith, Abbott, Lamanna, & Sanders, 1989; Schvaneveldt & Lee, 1983; Wise, 1986; Wolin & Bennett, 1984). Empirical research has demonstrated that rituals have a positive and even a protective effect on families. Investigations have documented that rituals help families create, maintain, and transmit to future generations a paradigm of family life and shared belief systems (Bossard & Boll, 1 950; Reiss, 1 982), enhance family members' sense of belonging (Wolin, Bennett, Noonan, & Tietelbaum, 1980), promote bonding and intimacy among family members (Meredith, 1985), preserve a sense of meaningfulness (Schvaneveldt & Lee, 1983), and create and maintain family cohesion (Wolin et al., 1980). The positive aspects of family rituals comprise much of the research to date; however, some evidence has been documented regarding potential negative aspects of rituals. Meredith (1985) and Wolin and Bennett (1984) noted that rituals may negatively impact families if they are perceived as meaningless, superficial, or boring. Additionally,

PAGE 26

19 rituals may be imposed on family members in an unhealthy manner, or they may be used to control and manipulate inappropriately (Bossard & Boll, 1950; Meredith, 1985). Furthermore, rituals may require too much effort or may result in family conflict that is disruptive to members (Meredith, 1985). Finally, rituals may be used to perpetuate physical, sexiial, and emotional abuse. Ahnost all of the empirical research regarding rituals has focused on their place in family life rather than in couple life. Furthermore, researchers have only studied heterosexual families. Bruess (1994) conducted one of the fev^ empirical studies to date describing the types of rituals used by intimate partners. Analyzing the rituals reported by marital partners and members of friendship dyads, Bruess (1994) identified seven major types of rituals. These consisted of (a) "couple time" rituals which are comprised of enjoyable activities, escape episodes, and togetherness rituals, (b) "symbolic" rituals which consist of private codes, play, and celebration rituals, (c) "spiritual" rituals, (d) ' "communication" rituals, (e) "daily routines," (f) "intimacy expression" rituals, and (g) "pattems-habits-mannerisms" (Bruess, 1994). Fiese and her colleagues compared 1 1 5 married heterosexual couples making the transition to parenthood (Fiese et al., 1993). They found that those couples reporting high levels of meaningfulness associated with their rituals were more satisfied with their marriages than their counterparts who reported relatively hollow rituals. ; ' There has been no empirical research conducted concerning lesbian couples' use of ritual. Although anecdotal accounts have supported the notion that ritual plays an important role in lesbian couple life (Laird, 1988; Slater, 1995), Slater (1995) identified ritual as a method by which lesbian couples may "compensate for the usually absent '"4";,.-^ ' V -

PAGE 27

20 social recognition of their family status" (p. 69). She posited that the heterosexist nature of society explicitly excludes lesbian couples from the definition of family and thereby bars them from participation in traditional rituals that affirm family membership and facilitate developmental transitions (Slater, 1995). The present investigation extended the research on family ritual to include lesbian partners. Conflict Resolution and Relationship Satisfaction Conflict resolution has been referred to as the pivotal dynamic in understanding couples' interactions (Braiker & Kelly, 1979). Researchers have suggested that couple satisfaction is mediated by conflict resolution strategies utilized by partners (Bowman, 1 990; Christensen & Heavey, 1 990; Gottman, 1 994; Gottman & Krokoff, 1989; Markman, Floyd, Stanley, & Storaasli, 1988). Furthermore, investigations relying on samples of heterosexual couples consistently demonstrated that couples who are able to resolve conflicts mutually are more satisfied than couples who do not (Christensen &. Shenk, 1991; Epstein, DeGiovanni, & Jayne-Lazarus, 1978; Gottman & Levenson, 1992; Markman et al., 1988; O'Leary «& Smith, 1991). To date, four studies have investigated the contribution of conflict resolution " strategies to relationship satisfaction among lesbian and gay couples. Most recently, Kurdek ( 1 998) developed a multidimensional model of relationship quality that he studied in a sample of heterosexual married, cohabiting lesbian, and cohabiting gay couples. He hypothesized that couples benefit from both internal and external forces that sustain the relationship. Specifically, Kurdek (1998) postulated that gender-linked forces within the relationship would facilitate happiness in each partner while institutional forces outside the relationship would promote the continuation of the relationship in spite

PAGE 28

21 of personal dissatisfaction. He proposed that it is the combination of internal and external forces that significantly contributes to relationship quality. Kurdek (1998) investigated four internal forces (constructive problem solving, intimacy, autonomy, and equality) and one external force (barriers to leaving the relationship). Kudek ( 1 998) found that constructive problemsolving appraisals by both partners were significantly associated with relationship satisfaction. In a prior study, Kurdek (1988) investigated the conflict resolution strategies of lesbian and gay couples. He found that partners in gay relationships did not differ significantly from partners in lesbian relationships regarding the types of conflict resolution strategies they utilized to resolve relationship disputes. In an evaluation of the relationship happmess of these couples, Kurdek (1988b) identified "positive" problemsolving strategies such as infrequent withdrawal, infrequent negative initiatives, and fi-equent positive efforts significantly influenced relationship satisfaction among both ' ' lesbian and gay couples. These results supported those of Falbo and Peplau ( 1 980) who compared lesbian and gay couples' conflict resolution styles in an assessment of power strategies. They found no significant difference between gay partners and lesbian partners. ' ' .. vMetz, Rosser, and Strapko (1994) compared 36 heterosexual, 36 lesbian, and 36 gay couples matched for age and length of relationship. Overall, they found that couple type did not differentiate conflict resolution style utilized. The differences they did find seemed linked to gender rather than sexual orientation. They did note that lesbian partners reported greater satisfaction and more hopefulness in addition to greater use of positive conflict resolution styles (Metz et al., 1994). Y ' " : . r-^

PAGE 29

Need for the Study Only recently have researchers begun to examine the factors that contribute to the maintenance or enhancement of lesbian relationships in the face of institutionalized oppression on the basis of sexual orientation. This study extended this research by exploring possible influences on lesbian relationship satisfaction. Four potential influences were examined: perceived institutional support for the relationship, conflict resolution style, degree of couple ritualization, and ritual meaningfulness. A significant body of research has explored the social support networks of lesbians as individuals. However, little research has focused on support for lesbian couples. Furthermore, many of the investigations of lesbian couples' support systems have relied on data drawn from only one member of the couple, thereby excluding potentially significant dyadic differences. While investigations regarding the supports utilized by individual women— even those in committed relationships— provide useful information, there is a need to examine the influence of institutional supports on lesbian couples' satisfaction. ^ r * ' Conflict resolution styles have been shown to affect significantly relationship satisfaction in heterosexual samples (Christensen & Shenk, 1991; Epstein et al., 1978; Gottman & Levenson, 1992; Markman et al., 1988; O'Leary & Smith, 1991). However, there is limited empirical investigation of the influence of conflict resolution style on lesbian relationship satisfaction. Moreover, results from existing studies assessing the influence of lesbian partners' conflict resolution on relationship satisfaction have been inconsistent (Kurdek, 1998; Metz et al., 1994; Falbo & Peplau, 1988). Consequently,

PAGE 30

23 additional research is needed to clarify the role of conflict resolution style on the lesbian couples' relationship satisfaction. Most intimate relationships are sustained via a dynamic and complex network of intentional and less-than-intentional strategies. Lesbian couples, unlike their heterosexual counterparts, often must sustain their intimate relationships in the face of adversity or even overt hostility from family, "fHends," and society at large. Rituals have been shown to contribute to the stability and cohesion of family life in healthy families (Wolin et al., 1980) and in those families facing stresses such as severe illness (Fiese, 1997) or alcoholism (Wolin & Bennett, 1984; Wolin et al., 1980). In addition, rituals have been linked with greater relationship satisfaction among married couples, especially those facing stressful life transitions (Fiese et al., 1993). No research to date has assessed the contribution of ritual to relationship satisfaction among lesbian couples. Consequently, a need exists to assess the contribution of ritualization to relationship satisfaction among lesbian couples. Much of the research on lesbian couples has been plagued with methodological difficulties. First, as with any stigmatized or marginalized population, it has been difficult for researchers to obtain a truly representative sample. Many studies have relied on network or snowball sample selection techniques. Particularly absent from many of these studies have been lesbians of color and of lower socioeconomic status. Another problem with the methodology of a number of studies of lesbian couples has been the reliance on only one member of the couple from which to gather data (e.g.. Garden et al., 1981; Duffy & Rusbult, 1986; Falbo & Peplau, 1980). Of course, it is unlikely that both members of a couple will view the relationship identically. Critical

PAGE 31

information may be overlooked wlien only one partner is studied. Furthermore, even when both partners have been included in the study, some researchers have chosen to treat each individual as a separate unit of analysis (e.g., Howard, Blumstein, & Schwartz, 1986; Kurdek & Schmitt, 1986a; Metz et al, 1994). As Kurdek (1998) pointed out, "because partners' scores share common variability — as is usually the case with dyadic data~the test statistic associated with type-of-couple differences is too large, resulting in inaccurate estimates of effect size and p values that are smaller than they should be" (p. 554). Another limitation of a majority of past lesbian couple research has been the focus on intrapartner links rather than on cross-partner linkages. Intrapartner links are those relationships between an individual's perception of relationship satisfaction and predictor variables while cross-partner links are those relationships between one partner's appraisal of satisfaction and the predictor variable scores for the other partner. The importance of assessing both types of links is based on the theory that an individual's experience of hisher relationship is simultaneously influenced by both his-her own appraisal of the relationship as well as by his-her partner's appraisal of the relationship (Kenny, 1996; Kurdek, 1998). A final problem with some of the past lesbian couple research is the reliance on measurements with unknown psychometric properties. A number of investigations have utilized assessment instruments that have no documented reliability or validity (e.g., Blumstein & Schwartz, 1983). Furthermore, some researchers have significantly altered assessment measures originally created and normed on heterosexual or, in some cases, all

PAGE 32

male samples in order to adapt them for use with a lesbian sample (e.g., Kus, 1 985). Assumptions about validity simply cannot be maintained under these circumstances. Purpose The purpose of this study was to examine the influence of five variables on the relationship satisfaction among lesbian couples. Four of these variables focused on aspects of partner interaction: (a) the extent of ritualization in couple life, (b) the meaningfulness of ritualization in couple life, (c) the type of conflict resolution style utilized by couples, and (d) the agreement between partners regarding conflict resolution style. The fifth variable — ^perceived institutional support — focused on the couples' participation in the larger social context. The sample for this study of lesbian couples included women who self-identified as lesbians. Bisexual women and transgendered women were excluded from the study. Furthermore, this sample consisted only of lesbian couples who lived together and had been in a committed monogamous relationship for at least 1 year. .. Research Questions Limited empirical research has been conducted on lesbian relationships. The research that does exist has failed to consider the joint contribution to relationship ; : , satisfaction of the internal dynamics between partners and the external influence of the cultural context in which these relationships are sustained. Because the larger cultural context exerts a powerful influence on lesbian couple life, examining the processes by which such couples construct an identity as a couple and create shared rituals together appears to be an appropriate target for research. The goal of this study was to expand the body of knowledge used to understand lesbian couple satisfaction by examining potential

PAGE 33

forces inside the relationship — such as ritualization and conflict resolution style — as well as forces outside the relationship — such as institutional supports — which may contribute to relationship satisfaction. To this end, the following research questions were posed: . 1 . What style of conflict resolution do lesbian partners develop and how does a couple's conflict resolution style contribute to their relationship satisfaction? 2. What shared rituals are developed by lesbian couples and how do these contribute to relationship satisfaction? 3. How do lesbian couples develop an identity as a couple in the face of a larger social context that may be critical of their relationship? What types of supports and barriers are perceived by the couple and how do these contribute to relationship satisfaction? Definition of Terms Conflict re solution stvle . The manner in which couples resolve disputes is the conflict resolution style. Four styles of conflict resolution were assessed in this investigation: positive problem solving, conflict engagement, withdrawal, and compliance. Homosexual. Homosexual is a term that refers both to lesbians and gay men. Generally, this term is considered to be a clinical diagnostic label and, therefore, a pathology-based descriptor. Institutional support. There are a number of societal institutions that support and validate heterosexual couples. For instance, religious sects sanction heterosexual marriage. Additionally, these marriages receive other forms of institutional support including legal and financial supports such as public recognition and even tax benefits.

PAGE 34

27 Institutional supports such as these are not easily accessible to lesbian couples due to the marginalized position that lesbians occupy in a society that stigmatizes both the lesbian identity and lesbian relationships. For the purpose of this investigation, institutional support includes aid, recognition, and status afforded by the following institutions: religious, educational, social (including family, friends, peers, and lesbian community), financial, work, social welfare, medical, and legal. Lesbian. After nearly two decades of research on lesbian identity, there still is no consensus about the definition of this concept (Cass, 1984; Falco, 1991). The range of meanings spans the spectrum from defining oneself as lesbian to consistent lesbianrelated behavior (Falco. 1991). A distinction must be made between the term "homosexual identity" and "lesbian identity." Homosexual identity, a narrower term than lesbian identity, refers strictly to sexual behavior (Levine & Evans, 1991). On the other hand, lesbian identity encompasses the global experience of being lesbian. Additionally, the term "homosexual" often is found offensive by lesbians as it has historically been used to pathologize and diagnose what is now considered a healthy variation of human sexuality. In contrast, the term lesbian connotes a positive description of the emotional, political, and lifestyle (including but not limited to sexual) aspects of life (Levine & Evans, 1991). Finally, a differentiation between lesbian identity and sexual orientation must be addressed. Sexual orientation, which may have biological components, is a behavioral construct whereas lesbian identity is both a social and a cognitive construct (Falco, 1991). Although there is much debate regarding possible genetic influences, most recent theories concur that an individual is not bom with a fiilly constructed lesbian identity; rather, this identity is constructed and maintained through interaction with others

PAGE 35

(Levine & Evans, 1991; Richardson, 1981). Lesbian identity development is not synonymous with lesbian behavior or same-sex attachments. Although same-sex sexual behavior may indeed be a precursor to the development of a gay or lesbian identity, individuals frequently engage in same-sex sexual behavior without defining themselves as lesbian (Cass, 1984; Elliott, 1985, 1992; Levine & Evans, 1991). Thus, behavior alone in insufficient to define a person as lesbian. . Lesbian Cor gavl identity . Lesbian identity refers to the total experience of being a lesbian; the core of a woman's essence as a lesbian. This includes but is not limited to sexual behavior. Emotional, affectional, political, and lifestyle aspects of lesbian experience also comprise lesbian identity (Levine & Evans, 1991). According to Brown (1995), lesbian is "primarily a self-ascribed definition held by a woman over time and across situations as having primary sexual, affectional, and relational ties to other women" (p. 4). This defmition does not include women who are bisexual or transgendered. Furthermore, individual women may engage in same-sex sexual behavior that is exploratory or transitory without defining themselves as lesbians. These women would not be considered lesbians for the purpose of this study. Lesbian couple-lesbian family. In this study, a lesbian couple was operationalized as two lesbians who are currently committed to a monogamous relationship with each other and who have been together for at least 1 year. According to Slater (1995), "lesbian couples constitute complete family units in and of themselves, the terms couple, family, md family of creation (can) be used interchangeably" (p. 32). These terms were used synonymously in this study.

PAGE 36

Ritual . Bossard and Boll (1950), in their groundbreaking study of 400 families, identified ritual as "a prescribed procedure, arising out of family interaction, involving a pattern of defined behavior, which is directed toward some specific end or purpose, and acquires rigidity and a sense of rightness as a result of its continuing history" (p. 29). Later, Wolin and Bennett (1984) defined ritual as symbolic communication systematically repeated across time because of its perceived enhancement of family satisfaction. Wolin and Bennet (1984) identified three types of family rituals: patterned interactions, family celebrations, and family traditions. Patterned interactions are rituals which occur with the greatest frequency in family life but are the least planned. Dinner time and bedtime routines as well as weekend activities are examples of patterned interactions. Family celebrations include holidays and special occasions celebrated by many members of a culture. Religious holidays, weddings, and Independence Day celebrations are examples of family celebrations. Family traditions are less culturespecific and more family-specific. Family reunions, birthdays, and anniversaries are examples of family traditions. Dyadic ritual. Dyadic rituals are defined for the purpose of this study as the symbolic, patterned interactions shared by both members of a couple. Although members of other dyadic relationships such as friends or coworkers may also share interpersonal rituals, the focus of this study was on those rituals shared by committed couples, specifically lesbian couples. Ritual dimensions. Rimal dimensions include occurrence, roles, routines, attendance, affect, symbolic significance, continuation, and deliberateness (Fiese & Kline, 1993). Rituals are performed through roles and routines. Families may have

PAGE 37

30 expectations about who is to attend rituals as well as who is to be excluded. Rituals have affective meaning for the family as well as a symbolic interpretation. Rituals are passed down through generations and may require advance plarming and preparation. Ritual settings . Ritual setting are those circumstances and situations in which rituals are performed such as dinner time, weekends, vacations, annual celebrations, special celebrations, religious holidays, and cultural and ethnic traditions (Fiese & Kline, 1993). Ritualization . A term used by Betcher (1987), ritualization refers to the ongoing and patterned interactions between members of a couple which are endowed with relational significance. The couple may rely on a variety of types of ritualization to maintain and enhance the relationship. For instance, daily rituals such as sharing dinner every night may serve to connect the couple while "renewal" rituals such as revisiting the site of their first date may serve to maintain the relationship over time. Furthermore, "commemorative" rituals such as sending holiday greeting cards together and annual rituals like attending family reunions also contribute to the maintenance of the relationship (Betcher, 1987). Organization of the Study ' , Chapter 2 of this investigation is a review of the related literature. Following Chapter 2 is a delineation of the methodology in Chapter 3, including a statement of the purpose of the study, hypotheses, description of relevant variables, data analyses, description of the population, subjects, sampling procedures, data collection, and instrumentation. Chapter 4 is a presentation of the results of the statistical analyses of

PAGE 38

31 the data. Chapter 5 includes a discussion of the results, implications for therapy, limitations of the study, and suggestions for further investigation.

PAGE 39

A J,f H ^^ CHAPTER 2 , : . . REVIEW OF THE LITERATURE ' . ; . \ • . Introduction Lesbian couples have not been studied extensively by social scientists. In this chapter the literature on social support factors affecting lesbian satisfaction is reviewed. In addition, literature on couple conflict resolution style and the use of ritual in couple and family life is discussed. Lesbian Couples: A Contextual View Brown (1989b) described three aspects of lesbian couples' cultural context that cut across all demographic characteristics: marginality, biculturalism, and normative " creativity. Marginality is expressed as a feeling of "otherness" and a cognitive recognition of minority status. According to Brown (1989b, 1991), partners in lesbian relationships may experience this feeling of "otherness" as nontraditional gender role identity because they do not receive the traditional recognition afforded a female via her socially proscribed association with a male. For some lesbians this may mean feeling as if they do not belong to their gender; however, this feeling of gender inappropriateness often changes as women are able to reevaluate belief systems and validate the importance ofnontraditional gender roles in couple relationships (Vargo, 1987). Unlike other minority groups that share a common experience, lesbians come from a variety of religions, racial and ethnic cultures, age groups, or socioeconomic status (Brown, 1989b). In fact, there is no singular lesbian couple reality. Lesbian partners are 32

PAGE 40

members of both the heterosexual and homosexual cultures; they are, in essence, bicultural. Almost all lesbians have lived at least the early part of their lives under the influence of assumed heterosexuality, and even after affirming a lesbian identity many women contmue to pass as heterosexual in certain circumstances (Brown, 1989b). Although this bicultural experience is not unique to lesbian couples, unlike their counterparts in other oppressed minority groups, they do not share their minority status wdth members of their own families of origin (Brown, 1989b). According to Brown (1989b), as a result of their minority status in a group for which previously established norms do not exist, lesbian couples are afforded the opportimity to create their own norms. Accordingly, lesbians may enjoy tremendous fi-eedom to invent their relationships and roles in a creative manner (Grahn, 1984). One of the most important ways that families create for their members a sense of belonging is through the establishment of rituals. Lesbian families have not been afforded many of the traditional rituals afforded their heterosexual counterparts such as wedding showers, marriage ceremonies, and anniversary celebrations, yet they may utilize what Brown (1989b) terms normative creativity to devise rituals of their own. < jt" Research on Lesbian Relationship Satisfaction Researchers have shown that lesbians highly value permanent love relationships. In one of the first surveys of lesbian partners, Bell and Weinberg (1978) found that 25% of 293 respondents reported having a permanent relationship with one partner to be the "most important thing in life," while another 35% said it was "very important." Further, Peplau and Cochran (1990) estimated that 45% to 80% of lesbians are involved in committed relationships. Additionally, many lesbians develop lifelong partnerships.

PAGE 41

More than 90% of the 706 lesbian couples surveyed by Bryant and Demian (1994) reported that they were committed to their relationships for life. While some investigations have found similar levels of relationship satisfaction among lesbian, gay, and heterosexual couples (Blumstein & Schw^artz, 1983; Duffy & Rusbult, 1986; Kurdek & Schmitt, 1986c), other studies have noted higher reports of satisfaction among lesbian couples (Bryant & Demian, 1994; Kurdek, 1988b; Zacks, Green, & Marrow^, 1988). Research regarding relationship satisfaction among lesbian couples has followed three traditions: (a) comparative studies involving heterosexual couples and homosexual couples (lesbian and/or gay), (b) comparative studies of different types of homosexual relationships (lesbian, gay, and/or bisexual), and (c) descriptive studies involving samples comprised exclusively of lesbian couples. Studies comparing relationship satisfaction among various heterosexual and homosexual couple types have found very few differences (Blumstein &. Schwartz, 1983; Cardell et al., Duffy «fe Rusbuh, 1986; Green, Bettinger, & Zacks, 1996; Kurdek, 1994c, 1998; Kurdek &, Schmitt, 1986a, 1986b, 1986c, 1987a; Lange, 1989; Norton, 1995; Oberstone & Sukonek, 1976; Peplau & Cochran, 1990). Although comparisons between heterosexual couples and lesbian or gay couples are informative, lesbian and gay couples face some unique stressors-such as homophobia--that may affect their relationships (Falco, 1991, Murphy, 1992). Further, while studies comparing lesbian couples with gay couples yield interesting information, these investigations may overlook the distinct factors that may influence couples in which both partners are women (Falco, 1991, Murphy, 1992). Therefore, studies of lesbian couples are warranted. Research on lesbian relationships have examined a number of correlates of relationship satisfaction including disclosure of lesbian identity (Green et

PAGE 42

35 al., 1996; Jordan, 1995; Kurdek, 1988a; 1995b; Oberstone & Sukoneck, 1976) and satisfaction with support (Kurdek & Schmidt, 1986a, 1986b, 1986c, 1988a; Oberstone & Sukoneck, 1976). " Research on Lesbian, Gay, and Heterosexual Couple's Satisfaction Relationship satisfaction is one of the most frequently studied variables in ., investigations of heterosexual as well as lesbian and gay couples. Early comparative studies sought to determine similarities and differences between lesbian relationships and v heterosexual relationships. More similarities than differences were found in general measures of relationship satisfaction. Consequently, more recent investigations have focused on possible correlates of relationship satisfaction in these different couple types. Research c omparing levels of relationship satisfaction . Morin (1977) used the term "heterosexual bias" to refer to the pathologizing of lesbian relationships. Heterosexual bias refers to the tendency of research to reflect societal beliefs that heterosexual relationships set the normative standards against which all other relationships are to be measured; in essence, heterosexual relationships are conceptualized as "normal" and "healthy." Consequently, early studies of lesbian and gay relationships focused on the relative "adjustment" these couples reported in comparison with heterosexual couples. Cardell et al. (1981) found no difference between 10 lesbian, 5 gay, and 10 heterosexual couples in measures of "adjustment." Lesbian and gay couples were just as likely as heterosexual couples to score in the "well-adjusted" range on traditional measures of marital satisfaction. In a similar investigation, Peplau (1982) compared matched samples of lesbian, gay, and heterosexual couples on measures of love and liking their partner. No significant differences were found. The findings of

PAGE 43

36 these and similar studies suggest that lesbian and gay relationships can be fully as functioning and healthy as heterosexual relationships. Several investigations have compared overall relationship satisfaction among heterosexual, gay, and lesbian couples. One of the largest of such studies compared married heterosexual, cohabiting heterosexual, cohabiting lesbian, and cohabiting gay couples (Blumstein & Schwartz, 1983). Across the four couple types, reports of relationship satisfaction were similar. Very few couple type differences were found. Furthermore, these researchers reported that emotional, sexual, and financial commitment predicted relationship endurance. In a series of comparative investigations, Kurdek and Schmitt ( 1 986a, 1 986b) surveyed a sample comprised of four types of couples: 56 cohabiting lesbian couples, 50 cohabiting gay couples, 35 cohabiting heterosexual couples, and 44 married heterosexual couples. Overall, they found fewer differences than similarities among couple types in each of the investigations. Kurdek and Schmitt (1986c) found no significant differences between these four types of couples in regard to level of relationship satisfaction. Relying on data from self-report measures, they found that dyadic attachment levels were related to relationship satisfaction, love for partner, and liking of partner. Further, they found that cohabiting heterosexual couples reported the lowest levels of dyadic attachment of the four couple types. These investigations support the notion that lesbian and gay partners are just as satisfied with their relationships as are their heterosexual coimterparts. Kurdek (1998) investigated five predictors of relationship dissolution and change in relationship satisfaction over time in a sample of 236 married heterosexual, 51 lesbian

PAGE 44

couples, and 66 gay cohabiting couples. He found no differences based on couple type on measures of initial relationship satisfaction. Similarly, heterosexual and lesbian and gay couples did not differ on the trajectory of change over a 5-year time period although each couple showed a decline in relationship satisfaction. Research on the applicability of heterosexual models of relationships . A number of studies have investigated the applicability of traditionally heterosexual models of relationships to lesbian and gay couples. In one such study, the investment model (Rusbult, 1980, 1983) of relationship maintenance and appraisal was examined. The investment model conceptualizes relationship satisfaction as an ongoing process in which partners evaluate and re-evaluate the relative rewards and costs of continuing the relationship. Duffy and Rusbult (1986) compared 25 lesbians, 25 gay men, 25 heterosexual women, and 25 heterosexual men matched for age and education in an investigation of relationship satisfaction and commitment. They used a series of selfreport measures designed to assess both the specific and the global aspects of relationship costs, relationship rewards, alternatives to the relationship, and investment in the relationship. Global measures of satisfaction and commitment were also collected. These researchers found more similarities than differences among the groups. No significant differences among the groups were found regarding reports of relationship costs, relationship rewards, alternatives to the relationship, and investment. Differences on the basis of gender were identified; specifically, women reported higher levels of investment and greater degrees of commitment. Heterosexuals reported greater costs and marginally greater investment than did either lesbians or gays. For all couples types, greater satisfaction was associated with higher levels of rewards and lower levels of

PAGE 45

costs. Likewise, no differences by couple type were reported in predictors of commitment. These researchers concluded that gender rather than sexual orientation accounted for differences. Hatton (1995) surveyed 103 lesbians in cohabiting relationships. Results of this study offer support for Rusbult's (Duffy & Rusbult, 1986) investment model. Hatton (1995) found positive relationships between relationship satisfaction, commitment, and personal power; however, the relationship power model was not fully supported. Although relationship satisfaction was the focus of this study, only individual data were analyzed rather than couple data. These studies provide support for the applicability of heterosexual models of relationship development and appraisal to lesbian and gay couples. In another study of traditional heterosexual models of relationship development, Kurdek and Schmitt (1986a) used a subsample from previous research comprised of 33 married heterosexual couples, 34 cohabiting heterosexual couples, 41 gay couples, and 47 lesbian couples in an investigation focused on stages of relationship development. They were particularly interested in indicators of relationship quality such as agreement, satisfaction with affection and sex, beliefs about sex, cohesion, and tension. They identified similarities regarding relationship development among these types of couples; specifically, couples in the first year of relationship (blending stage) reported relatively greater satisfaction while couples in the second and third (nesting stage) years of relationship development reported relatively greater levels of stress and disillusionment. Couples in the fourth and fifth (maintaining stage) years were also assessed. Results from this investigation supported the predicted curvilinear relationship between

PAGE 46

39 relationship quality and stage of relationship such that higher quality would be reported at blending and maintaining stages. This study supports the belief that lesbian and gay relationships follow developmental patterns similar to those found in heterosexual couples. One of the few significant differences identified among the four couple types . was in the area of support: married couples reported greater support from family than did lesbian and gay couples. In fact, lesbian couples reported receiving less family support than did either gay couples or heterosexual couples. However, these differences did not influence their reports of relationship satisfaction. This study supports the use of 5 . developmental models of relationships originally theorized regarding heterosexual relationships with lesbian and gay couples. " Kurdek (1992b, 1994a) continued to investigate the applicability of heterosexual * models of relationships with lesbian and gay couples in a 4-year longitudinal study. He surveyed 39 married heterosexual couples, 39 lesbian cohabiting couples, and 39 gay cohabiting couples, matched on length of time living together, regarding relative levels of J. .. relationship satisfaction. This study combined variables from the contextual model (e.g., satisfaction with social support, dysfunctional beliefs regarding relationships) and the interdependence model (e.g., perceived rewards from the relationship, perceived costs to the relationship, relationship satisfaction). Unlike most previous studies that used the individual as the unit of analysis, this investigation used the couple as the unit of analysis. Kurdek ( 1 992b) found evidence to support the connection between relationship satisfaction and the two models in both lesbian and gay couples. He found that lesbian couples and heterosexual couples did not differ on level of satisfaction, although gay couples reported lower satisfaction. Kurdek (1992b) cautioned that this may be due to a

PAGE 47

40 sampling artifact in the selection of gay couples. He found that lesbian partners and wives did not differ on a set of individual difference scores-dysfunctional relationship beliefs, social support, expressiveness, and negative affectivity--while husbands and gay partners did differ on these comparisons. In terms of interdependence scores, gay couples reported fewer relationship rewards and more relationship costs than did either lesbian couples or heterosexual couples. Further, relationship satisfaction was positively related to perceived relationship rewards and emotional investment and negatively related to perceived relationship costs in each of the three groups. Finally, this examination highlighted the mediation effect of interdependence variables on individual difference variables in the prediction of relationship satisfaction. Schneider (1986) also investigated interdependence variables in a study that compared lesbian and heterosexual couples on measures of relationship satisfaction as defined by interdependence, durability, and equality. Overall, results supported more similarities than differences between lesbian and heterosexual couples. Although both types of couples reported a high degree of optimism regarding the continuance of their relationships, this construct was statistically significant only for heterosexual couples. Furthermore, analyses of the interdependence factor data revealed that lesbian and heterosexual couples differed on only 3 of 13 subfactors: wills, banking practices, and life insurance policies. This finding is not surprising when one considers the lack of legal sanctions afforded lesbian couples. Finally, analyses of equality data showed that lesbian couples differed from heterosexual couples only on the issue of household chores. Lesbian couples reported greater equality regarding household chores than did heterosexual couples. Further, lesbians reported using personal preference rather than

PAGE 48

traditional gender roles in decision making regarding division of chores. This investigation adds support to the use of heterosexual models of relationships with lesbian and gay partners. " Zacks et al. ( 1 988) used the Circumflex Model (Olson, 1 986) as the basis for an investigation of the levels of cohesion, adaptability, and relationship satisfaction in a sample of 52 lesbian couples and 11 40 heterosexual couples. According to this model, balanced levels of adaptability and cohesion are characteristic of healthy heterosexual couples and families. Data analyses revealed that lesbians scored significantly higher than heterosexual couples on measures of cohesion. These researchers explained that higher cohesion scores could be attributed to the tendency for lesbian partners to turn to each other for support and validation in the face of societal rejection. Further, these scholars noted that the findings of this study indicate a cultural--heterosexual--bias that exists regarding the term "enmeshment." Lesbian couples also scored higher than heterosexual couples on measures of adaptability. Zacks and colleagues (1988) suggested that lesbian couples must create their own relationship roles, rules, and norms in the absence of societal models. Thus, high adaptability scores represent positive qualities of a relationship. Finally, lesbian couples reported greater relationship satisfaction than did heterosexual couples. This study supports the notion that qualities that are fimctional and desirable in lesbian relationships such as emotional closeness and role flexibility may differ in degree from those in heterosexual relationships. jj'". In a related study. Green and colleagues ( 1 996) investigated relationship quality, disclosure of sexual orientation, and relationship dissolution in a sample of 52 lesbian couples, 50 gay couples, and 1358 heterosexual married couples. Data for the lesbian

PAGE 49

and gay couples were collected in 1986 and compared with data from two different national samples in 1983 and 1976 (Green et al., 1996). These researchers found that lesbian and gay couples reported greater cohesion and flexibility in their relationships than did heterosexual couples. Further, lesbian couples reported more cohesion and flexibility than did gay couples. Both cohesion and flexibility predict relationship satisfaction in lesbian and gay couples and relationship dissolution in lesbian couples. For lesbian couples, degree of disclosure of lesbian sexual orientation family members was not related to relationship satisfaction. Heterosexual models of relationships conceptualize partner homogamy as one influence in the appraisal of relationship satisfaction (Murstein, 1970). This theory of partner selection suggests that "like is attracted to like." Kurdek and Schmitt ( 1 987a) assessed partner similarity measured by demographic characteristics and relationship quality in a sample of 45 married heterosexual, 35 heterosexual cohabiting, 50 gay cohabiting, and 56 lesbian cohabiting couples. Partner scores were correlated on measures of relationship satisfaction. Lesbian and gay couples' scores on love for partner were correlated, but this was not true for either heterosexual cohabiting or married couples. Partner homogamy was most pervasive in lesbian couples. These researchers found similarities between partner ages for each couple type. Heterosexual cohabiting couples had similar income, education, and job prestige. Gay couples had the greatest disparity in age, income, and education. In another study of partner similarities, Kurdek and Schmitt ( 1 986b) compared 370 lesbian, gay, and heterosexual couples on measures of sex-role self-concept and relationship satisfaction. Bem's (1974) theory of sex-role self-concept was used in the

PAGE 50

43 study. Consequently, four types of sex-role self-concepts were assessed: androgynous, feminine, masculine, and undifferentiated. Individual partner analyses revealed that relationship quality differed by sex-role self-concept rather than by couple type. Androgynous and feminine partners reported higher relationship quality than did masculine and undifferentiated partners. Couples in which one or both partners were feminine or androgynous reported the highest relationship satisfaction. *^ Research comparing correlates of relationship satisfaction . Intimacy has been ' correlated with relationship satisfaction in a number of studies of heterosexual and lesbian and gay couples (Anderson, 1990; Eldridge & Gilbert, 1990; Snyder, 1991; Zacks et al., 1988). Anderson (1990) compared intimacy, fusion, and relationship quality of women in lesbian and heterosexual relationships. She found that lesbians scored higher on self-report measures of intimacy than did their heterosexual counterparts. Further, no differences between lesbians and heterosexual women were found on measures of fusion or measures of relationship satisfaction. Intimacy and relationship satisfaction were correlated for lesbian couples but not for heterosexual couples. This study refutes suggestions that lesbian couples tend to have higher levels of fusion than do other couples types. • In a recent examination of relationship satisfaction and intimacy, Norton (1995) surveyed 20 lesbian, 1 1 gay, and 31 heterosexual cohabiting individuals. She found similarities between couple types regarding levels of intimacy and relationship satisfaction. Further, in all three couples types intimacy predicted relationship satisfaction. Although this investigation focused on couple relationships, only one

PAGE 51

. ; 44 member of each couple was assessed. Consequently, partner differences that may ; influence relationship satisfaction were not taken into account. Another correlate of relationship satisfaction often investigated in samples of heterosexual couples is power. Lange (1989) examined the relationship between power, nurturance, and trust m predicting relationship satisfaction and commitment among cohabiting lesbian and heterosexual couples. She found many similarities between the two couple types investigated. Specifically, both lesbian couples and heterosexual couples reported a balanced allocation of power was strongly associated with relationship satisfaction. However, there were also differences noted across couple types. For example, a significant association between a balanced allocation of power was unrelated to commitment for lesbian couples; however, heterosexual women reported higher levels ofcommitment and equal balance of power. Receiving communal nurturance (the feminine prototype of support) predicted relationship satisfaction and commitment in both couple types while receiving agentic nurturance (the masculine prototype of support) was negatively predictive of satisfaction. Further, trust was positively related to satisfaction and commitment for both couple types. Research Comparing R elationship Satisfaction of Lesbian and Gav Male Coup les While studies comparing heterosexual, lesbian, and gay couples offer valuable information, these investigations may overlook unique influences on relationship satisfaction shared by lesbian and gay couples. As members of a stigmatized and marginalized group, lesbian and gay partners face stressors that many partners in heterosexual couples do not. Consequently, studies comparing lesbian and gay couples

PAGE 52

provide unique insights into the couple Ufe of partners who must live without the social sanctions afforded other types of couples. In his 1978 study, Tuller compared 5 lesbian and 10 gay couples. He stated that lesbian couples reported a greater need for sexual fidelity and a broader network of friends than did gay couples. Later, in an investigation of the importance of fidelity for married, cohabiting, and lesbian women, Blumstein and Schwartz (1983) reported no " significant difference among these couple types. It remains unclear how fidelity influences relationship satisfaction in lesbian couples. In one of the few longitudinal investigations of relationship quality among lesbian and gay couples, Kurdek (1988b) surveyed a sample of 47 lesbian and 65 gay cohabiting couples. Analyses of initial relationship satisfaction revealed that lesbian couples reported higher relationship satisfaction, higher liking of partner, more motivation to stay in the relationship, more trust, and more shared decision making than gay couples. More than half of the gay couples in the survey reported nonmonogamous relationships although this did not affect relationship quality measures. In a one-year follow-up study, Kurdek (1989) sampled 45 lesbian and 74 gay couples. Compared to gay couples, lesbian couples demonstrated an overall pattern of enhanced relationship quality over time. Greater relationship quality was reported by lesbian couples at both times of assessment. Further, compared to gay couples, lesbian couples reported stronger liking of partner, stronger motivation for being in the relationship, greater trust, and more frequent shared decision making at both times of assessment. Kurdek (1 989) found that satisfaction with social support, high expressiveness, and few beliefs that disagreemem is

PAGE 53

destructive to the relationship predicted later relationship quality in gay couples but not in lesbian couples. Kurdek (1991) compared 75 gay couples and 5 1 lesbian couples on self-report measures of relationship satisfaction. This study investigated variables related to contextual, investment, and problem-solving models of intimate relationships. Overall, he found more similarities than differences between lesbian and gay couples on mean levels of each model variable. Similarly, very fev/ differences were found between these two couple types in terms of the strength of the correlates of relationship satisfaction. Variables from each of the three models were significantly related to relationship satisfaction in both lesbian and gay couples. In addition to support for correlations between each of the models and relationship satisfaction independently, Kurdek (1991) found support for a mediational model that proposed ordered linkages among variables from the three models. Specifically, the combination of contextual, investment, and problem-solving model variables accounted for nearly 50% of the variance in relationship satisfaction. . In a later study, Kurdek (1995a) investigated developmental changes in " relationship quality in 42 lesbian and 61 gay cohabiting couples. He analyzed survey data from three annual assessments of current levels of attachment, autonomy, and equality in the relationship, importance of attachment, autonomy, and equality in an ideal relationship, and relationship commitment. Lesbian partners reported ideal equality as more important than did gay partners. The importance of ideal equality to lesbian partners changed curvilinearly overtime, but no such change was found for gay partners. Lesbian and gay partners showed decreases in the importance of ideal attachment over

PAGE 54

47 time. For both lesbian partners and gay partners, changes in the difference between current and ideal levels of equality predicted changes in relationship commitment over time. Similarly, current levels of attachment and current levels of autonomy predicted relationship commitment for both lesbian partners and gay partners. In one of the largest national surveys of lesbian and gay couples, Bryant and Demian (1994) investigated correlates of relationship satisfaction in a sample of 560 gay couples and 706 lesbian couples. They found that high quality of sexual interaction, high commitment to the relationship, low frequency of abusive behavior, low frequency of serious arguments, high joint income, and legal preparedness were significantly associated with relationship satisfaction among lesbian couples and gay couples. The authors of this study caution that in only 483 instances did both members of a couple complete the one-page survey. Consequently, statistical techniques were used in which responses from individuals whose partners also participated in the survey were assigned a weight of .50. Research on Relationship Satisfaction in Lesbian Couples While comparative studies of lesbian, gay, and heterosexual couples are informative, studies focused exclusively on lesbian relationships often highlight the importance of variables that are unique to lesbian couples. Several studies sought to determine the factors that may affect relationship satisfaction reported by lesbian couples. Specifically, the influence on relationship satisfaction of disclosure of sexual orientation and social support has been investigated. Disclosure of sexual orientation and rela t ionship satisfaction Anecdotal literature (Cabaj, 1988; Murphy, 1992; Roth, 1985) has focused on the potential influence of

PAGE 55

"coming out" on relationship satisfaction. The decision about whether or not to be open about the relationship may present stressful challenges to lesbian couples, particularly if there is disagreement regarding how, when, to whom, and in what situations to disclose (Falco, 1991). This stress may be expressed in the couple relationship, thereby decreasing the satisfaction level (Brooks, 1981). Some empirical evidence suggests that relationship satisfaction is negatively impacted by partner discrepancy in degree of disclosure of sexual orientation and positively impacted by partner's disclosure of sexual orientation. However, other investigations have found no relationship between partners' disclosure of sexual orientation and relationship satisfaction reports. Berger (1990) surveyed 1 14 gay men and 29 lesbians regarding the effects of "passing" as heterosexual on the quality of relationship satisfaction. Respondents who were known as gay or lesbian to significant others such as family and close friends reported higher levels of relationship satisfaction than those respondents who had not acknowledged their sexual orientation to those close to them. The extent of passing with other groups such as distant friends and acquaintances did not appear to be related to the quality of the relationship. The disproportionate number of gay men in this study may limit its generalizability to lesbians. Romano (1990) investigated the relationship between homophobia and lesbian relationship satisfaction. A sample of 1 02 partnered lesbians was surveyed using selfreport measures. Homophobia from family, society, and self were assessed. Romano ( 1 990) found a negative correlation between relationship satisfaction and each type of homophobia. When combined, the three different homophobia variables predicted relationship satisfaction. However, this study relied on only one partner from each

PAGE 56

couple. While this method of data collection is statistically less complex, potentially important data are missing. Melamed (1992) investigated the effects of internalized homophobia on relationship satisfaction and commitment in a national sample of 233 lesbian couples. Unlike the couples in many other studies, these partners had been together for a significant length of time (mean = 7.5 years). She found that couples' levels of internalized homophobia among couples varied inversely with relationship satisfaction and commitment. Further, in an analysis of individual data Melamed (1992) found that lesbians with lower levels of homophobia reported higher levels of self-esteem. In a related study utilizing both individual and couple data, McGuire ( 1 995) surveyed 44 lesbian couples who had been in relationships for at least 1 year. She used a measure of "homophobia" which assessed subjects' attitudes toward the fact of their lesbianism being known by others as well as their attitudes regarding their own homosexuality and homosexuality in general. She found that partner difference in level of "homophobia" was predictive of decreased relationship satisfaction although individual levels of "homophobia" were not significantly related to reported relationship satisfaction levels. It should be noted, however, that the instrument used to assess "homophobia" in this study was normed on gay men and was modified for use with lesbians. Additionally, this study did not attempt to investigate the effect of disclosing one's status as a member of a committed lesbian couple; rather, an individual focus was utilized in which each woman's "homophobia" was assessed. In a similar study, Jordan (1995) found that individual women who more widely disclosed their sexual orientation reported greater relationship satisfaction than their

PAGE 57

counterparts who were more "closeted." Additionally, the women in this study reported lower levels of relationship satisfaction when a greater discrepancy between the partners existed regarding disclosiire of sexual orientation. However, it is unclear how many of the 499 women in committed relationships were included in couple data. Furthermore, Jordan (1995) assessed the degree to which individual women disclosed their sexual identity not the degree to which they identified themselves to others as part of a , committed lesbian couple. Caron and Ulin (1997) investigated the relationship between couple satisfaction and "closeting"--the degree to which the couple disclosed their relationship. They surveyed a sample of 124 lesbians involved in a "committed" relationship. They . analyzed four potential sources of closeting: family, extended kin, friends, and work associates. The strongest correlations among these sources of closeting and relationship satisfaction occurred in the realm of family. Further, these researchers found that relationship closeting (compared to personal closeting) had the highest correlation with satisfaction. Consequently, lesbian couples who were able to be more open about their relationships reported higher levels of satisfaction. Although these authors reported that the lesbians in their sample represent women in committed relationships, partners involved in relationships as short as two month's duration were included in the study. Furthermore, both members of a couple were not assessed. Consequently, important information regarding dyadic differences may have been overlooked. Several studies have found no relationship between a woman's willingness to disclose individually her sexual orientation and reported relationship satisfaction (Eldridge & Gilbert, 1990; Peplau et al., 1982). These researchers expected that women

PAGE 58

who were able to acknowledge their sexual orientation to others in the family, at work, or in social arenas would report higher levels of relationship satisfaction; however, none was found (Eldridge & Gilbert, 1 990; Peplau et al, 1 982). Peplau et al. ( 1 982) and later Caldwell and Peplau ( 1 984, 1 992) surveyed 1 27 lesbians regarding relationship satisfaction, self-disclosure, and personal power. They found that lesbians reported high levels of relationship satisfaction and egalitarian power distribution as well as equal degree of commitment and involvement in the relationship. However, they found no relationship between disclosure of lesbian identity and relationship satisfaction. They had hypothesized that individual partner characteristics such as age, student-work status, and religion would predict relationship satisfaction but this hypothesis was not confumed~a finding previously been noted by Cotton (1975). However, only one partner from each couple participated in the study and more extensive variables to assess similarity (e.g., race, degree of self-disclosure) were not used. Eldridge and Gilbert (1990) conducted a study in which correlates of relationship satisfaction were assessed utilizing both partners of 275 lesbian couples. Degree of disclosure of lesbian identity was not significantly correlated with relationship satisfaction. Disclosure of lesbian couple status was not assessed. However, many of the psychological variables assessed did correlate significantly with relationship satisfaction. For instance, higher levels of personal autonomy were associated with lower levels of relationship satisfaction. Further, self-esteem and satisfaction with life were significantly positively associated with relationship satisfaction. However, role conflicts (e.g., partner versus self, professional versus self) were negatively correlated with relationship satisfaction.

PAGE 59

52 Thus, it is unclear whether or not disclosure of individual sexual orientation is related to relationship satisfaction in lesbian women. Further research is needed in order to understand this relationship. Social support and relationship satisfaction . Murphy (1989) examined the role of parental support on lesbian couple satisfaction. She found that perceptions of parents' attitudes toward the couple, the partner, and the daughter's lesbian sexual orientation do affect relationship quality. This study is one of the few to examine support for the lesbian couple relationship rather than simply support for the lesbian individual. In a study of 1 73 lesbian couples, Donaldson ( 1 993) examined the relationships between autonomy-mutuality, social support, and relationship satisfaction. Using selfreport measures, she found that perceived mutuality was positively correlated with, and accounted for the largest percent of explained variance m, relationship satisfaction. Additionally, dyadic attachment and perceived social support positively correlated with relationship satisfaction. Specifically, perceived support from partner and friends was higher than that from family. Further, she found limited support for a negative correlation between mutuality and relationship satisfaction. Jenkins (1996) examined the relationship between differentiation from family of origin and relationship satisfaction in a sample of lesbian couples. She found that intimacy within the relationship was strongly associated with relationship satisfaction. Level of individuation within the relationship was shown to have a positive relationship with level of relationship satisfaction; however, differentiation in general and differentiation from family was not related to reports of relationship satisfaction among the lesbian couples in the sample. .

PAGE 60

• 53 In summary, research on relationship satisfaction has followed three traditions: (a) comparative studies of heterosexual, lesbian, and gay couples, (b) comparative studies of lesbian and gay couples, and (c) descriptive studies of lesbian couples. Overall, studies of relationship satisfaction among these three types of couples have revealed more similarities than differences. Furthermore, investigations based on models developed in study of heterosexual couples (e.g., the investment, problem solving, interdependence, contextual models) have provided support for the use of these models with lesbian and gay couples. Studies of lesbian couples have focused on the influence of disclosure of sexual identity and social support on relationship satisfaction. Most studies have focused on disclosure of individual sexual orientation to family and friends. Few studies have examined the influence on relationship satisfaction of the disclosure of lesbian couple identity. Likewise, investigations of social support have focused on support for the ' individual rather than for the couple. ' r: 'Research on Social Support Influences ,. Social support has been viewed as an important asset to individuals and couples alike. It has been depicted as a buffer against stress among heterosexual couples (e.g., Cobb, 1976; Cohen & Hoberman, 1983; Cutrona, 1984; Cutrona & Russell, 1987), but it is unclear whether social support provides a similar protective function in lesbians. Similarly, while social support has been conceptualized as a main effect in studies of heterosexual couples (Cohen & Hoberman, 1983; House, 1981; Lin, Ensel, Simeone, & Kuo, 1 979), few similar smdies have been conducted on samples of lesbian couples (Kurdek, 1998; Kurdek & Schmitt, 1986a, 1986b, 1986c; Meyer, 1989). Most researchers agree that social support plays an important role in the lives of human beings,

PAGE 61

54 yet there is little agreement regarding its definition (Cobb, 1976; Cooke, Rossmann, McCubbin, & Patterson, 1988; Fischer, 1 982; Kahn & Antonucci, 1 980; Mitchell, Billings, & Moos, 1982; Pilisuk, 1982; Procidano & Heller, 1983; Schaefer, Coyne, & Lazarus, 1981; Weiss, 1974). Similarly, much disagreement exists concerning the best way to operationalize the concept (Cohen & Hoberman, 1983; Coyne & DeLongis, 1986; Cummins, 1988; Heller, 1979; Lakey & Heller, 1988; Procidano & Heller, 1983; Stemp, Turner, & Noh, 1 986; Turner, Frankel, «& Levin, 1 983). Social support and its measurement have been conceptualized from a variety of perspectives. Theorists have debated whether received or perceived social support should be assessed (Cummins, 1988; Lakey & Heller, 1988). This debate is tied to the question regarding the process by which social support is believed to function, either as a main effect or a buffer. The conceptualization of social support is further complicated by the discourses regarding the multiple provisions of support, numerous dimensions of support, and various mechanisms by which support is delivered. Regardless of the conceptual framework from which social support has been measured, much of the research regarding social support in lesbian populations has been limited to the support provided to individuals by individuals. Rarely have investigations assessed either the impact of institutional support or the support given lesbian couples (Kurdek, 1998). Definitions Social support is often conceptualized as a multidimensional construct. These definitions have been based on the belief that human beings need to form meaningful relationships with each other (Stemp, Turner, & Noh, 1986). Moreover, these definitions focus on the specific classes of information (Cobb, 1976), exchanges (Fischer, 1982), or

PAGE 62

provisions (Weiss, 1974) which comprise social support. For example, Cobb (1976) defined social support as information of three types: information which contributes to an individual's belief that he or she is loved and cared for, information which supports an individual's feeling that she or he is valued, and information that contributes to an individual's sense of belonging and mutual involvement in a network. Further, Cobb (1976) hypothesized that social support could take three forms: emotion, esteem, or network support. In contrast with several other theorists, Cobb's definition is limited to the provision of nonmaterial supports. On the other hand, Weiss ( 1 974) conceptualized social support from the . • perspective of "relationship provisions." He posited that individuals have a "fund of sociability" which is a specified amount of interpersonal interaction that he or she needs. These needs are potentially met in a variety of ways and via a variety of relationship sources. Those relationships that provide social support are specialized in regard to the "relationship provisions" each offers (Weiss, 1974). Weiss (1974) outlined six relationship provisions: attachment, social integration, reassurance of worth, nurturance, reliable alliance, and guidance. ' Like Cobb's conception of "emotional" support, Weiss' attachment provision represents a sense of belonging in the individual. Likewise, a parallel exists between Cobb's "network support" and Weiss' social integration provision. Both of these social supports provide companionship and a network for mutual exchange. Reassurance of worth, similar to Cobb's "esteem support," offers the individual affirmation of his or her competence. Weiss' opportunity for nurturance provision is unlike any of Cobb's social supports. This provision occurs only in parent-child relationships and provides parents

PAGE 63

with a feeling of being needed and a sense of meaning in their lives. Family relationships, including the primary relationship between the members of a couple, are the domain from which a sense of reliable alliance is provided. This provision engenders in the individual a feeling of unconditional love and support. Finally, Weiss' guidance provision assists individuals wdth decision-making in stressfiil situations via the sharing of information. This provision is unlike any of Cobb's social supports. However, like Cobb, Weiss does not include the provision of material goods and services in his conceptualization of social support. Material goods and services are included among the social supports conceptualized by several theorists. Pilisuk (1982) identified three components of social support: "material help," "emotional assurance," and "communication of concern and caring". Similarly, Kahn and Antonucci (1980) defined social support as interpersonal interactions that include "aid," "affect," and/or "affirmation." Likewise, Schaefer et al. (1981) classified social supports into three types: tangible support, emotional support, and informational support. House (1981) analyzed a number of conceptual frameworks concerning social support and attempted to integrate these into what he considers to be an inclusive definition. He suggested that social support is "a flow of emotional concern, instrumental aid, information, and/or appraisal (information relevant to self-evaluation) between people" (p. 26). Similar to the authors previously cited, Cooke and colleagues (1988) have conceptualized social support in terms of provisions. These authors' conceptualization parallels provisions identified by other theorists by including the importance of network, emotional, esteem, and appraisal supports. However, they also add "altruistic" support as

PAGE 64

an equally important provision. Moreover, they note that reciprocal exchange is a salient provision of support as well. While some theorists have conceptualized social support in terms of provisions, others have proposed that social support is best understood in terms of the mechanisms by which it is believed to affect well-being. Mitchell and colleagues (1982) identified three categories of social support mechanisms: direct, indirect, and interactive. They proposed that "direct" effects are those benefits to the global well-being of an individual as a result of social support. These direct effects may occur when an individual is in a situation of stress or crisis or even when life circumstances are relatively uneventful. On the other hand, indirect effects on well-being may be accomplished by social supports which have the effect of reducing the intensity with which the individual perceives the stress. Finally, interaction effects refer to the buffering effect that social support may have in times of stress. Henderson (1987) reviewed studies of the impact of socially supportive interactions. He found that when support is lacking, physiological and psychiatric symptoms increased. Similarly, Turner, Frankel, and Levin (1983) reviewed social support literature and concluded that mental health and physical well-being were significantly positively related to social support. Assessment of Social Support " . ' ' f In light of the disagreement among researchers and theorists regarding the defmition of social support, it is not surprising that a similar conflict exists concerning the best method of measuring social support. Two opposing theories have been proposed. Some researchers advocate assessing perceived social support while others believe that social network, or embeddedness, must be studied. Furthermore, this debate

PAGE 65

58 has not been limited to the efficacy of these opposing methods; it has also included a discussion of the strength of the relationship between them. A comprehensive systematic analysis of the individual's total social field is generally required for a social network analysis (Turner et al., 1983). This approach has a number of limitations, particularly when applied to mdividuals with minority status. First, it is based on the assumption that all members of an individual's social network make positive contributions to the individual's social support. In the case of lesbians or members of other socially stigmatized groups, this may not be the case; in fact, the opposite may be true. It is quite possible that some members of a lesbian's social network are unaware of her orientation and, therefore, contribute no support--at least in this important arena. Other members of a lesbian's social network may adamantly oppose her lifestyle, thus contributing negatively to her social support and adding to her stress. Thus, measures based on the density or frequency of social contacts may offer minimal meaningful information. Additionally, social network analyses often ignore the complexity of the interaction between the stressfial and the stress-buffering aspects of relationships (Coyne 8c DeLongis, 1 986). In a review of social network research, Coyne and DeLongis (1986) found that a number of investigations identified strong relationships among the conflictual or upsetting aspects of social interactions, signs of depression, lack of well-being, and distress and weak relationships among sources of positive support, well-being, and lack of distress. ' It is not clear whether beneficial effects of social support are the results of supportive behaviors on the part of network members~as is the case in received support~or the resuh of the belief that support is available-as is the case in perceived

PAGE 66

support (Lakey & Heller, 1988). This might not be a concern if perceived support is a reflection of received support. However, investigations regarding the link between these two facets of support have shown a weak or nonexistent relationship between measures of received support and measures of perceived support and satisfaction with support (Lakey &. Heller, 1988). In fact, researchers have found that perceived availability of support may either wholly or partly protect individuals from negative effects of high stress levels (Cohen & Hoberman, 1983). Based on these findings, it appears that perceived support is more accurately assessed than received social support. Consequently, an instrument developed for this investigation to assess perceived support was utilized to assess the support lesbian partners perceive they have available to them from a variety of social institutions. Another consideration regarding the measurement of social support is whether social support acts as a main effect or a buffer effect. Some researchers have found that social support processes function as a main effect in that the support itself is beneficial regardless of the presence of stress (Cohen & Hoberman, 1983; House, 1981; Lin et al., 1 979). On the other hand, researchers have found that the interaction between social support and stress may fiinction in such a way that support is more effective in situations of relatively high stress rather than in situations of relatively low stress. In this case, social support would function as a buffer in which support mediates the psychological and/or physiological effects of stress on the individual (Heller & Swindle, 1983; Levy, 1983; Mitchell etal., 1982). Scholars suggest that the buffering effect of social support is cognitively mediated in that support is thought to influence the individual's appraisal of stressful situations

PAGE 67

60 (Cohen, Mermelstein, Kamarck, & Hoberman, 1985; Heller & Swindle, 1983). The individual is thought to weigh the resources for support available against the risks and difficulty faced (Lazarus & Launier, 1978). Consequently, the perceived availability of social support may result m the reappraisal of a threatening situation as less so (Wilcox & Vemberg, 1985). Research on Lesbians and Social Support Research on lesbian individuals . In one of the first studies to investigate the social support of lesbian individuals, Chafetz and colleagues (1974) interviewed 51 lesbians in Houston, Texas. They found that the women in this study reported a lack of both general and specific support from a variety of social institutions such as family, Mends, employers, and helping professionals. Of these women, 50% reported receiving little or no support fi-om family of origin. More than 40% of the lesbians in the sample reported that the majority of their close fi-iends are lesbian or gay. These researchers did not distinguish between women in committed relationships and those who were single. Albro and Tully (1979) expanded the study initiated by Chafetz and colleagues (1974). Questionnaires completed by a sample of 91 lesbians revealed that these women feh isolated from the heterosexual "macro-culture." They reported that the women in their sample received emotional support from the lesbian "micro-culture." Both of these early studies support the notion that lesbians compensate for lack of support from the dominant culture or family of origin by constructing support networks in the lesbian community. However, these studies did not address the potential impact of this strategy on the lesbian couple relationship. •

PAGE 68

61 In another investigation of the various sources of social support available to lesbian individuals, Jay and Young (1977) surveyed 900 lesbians regarding the types of people who know of their sexual orientation and the perceived reactions of these people to this knowledge. They found that more lesbians told their straight friends about their sexual orientation than told their family members. Employers and employees were least likely to know of a woman's lesbianism. In regard to the perceived "reactions" of significant people who knew of their lesbianism, individuals in the sample cited mothers as least supportive and gay or lesbian friends as most supportive. No data regarding couple status were reported. . ' Leavy and Adams (1 986) determined that relationship status and age affect social support and self-esteem among lesbian individuals. The lesbians in this sample who were involved in committed relationships scored significantly higher on measures of social support and self-esteem than those not in relationships. Although relationship status was assessed in this study, the data were treated as individual data rather than couple data. Consequently, the influence of social support on the relationships of lesbian couples in this sample is unclear. D' Augelli, Collins, and Hart (1 987) examined the social support pattern of 34 lesbians livmg in a rural setting. Members of an organized network of lesbians completed questionnaires. Of the lesbians in the sample, 91% reported seeking help from women, 71% reported seeking help from partners, and 68% reported seeking help from heterosexual friends who knew of the woman's sexual orientation. Further, a majority of the women had not told their families of their sexual orientation; more than half expected

PAGE 69

62 rejection from their fathers and nearly half expected the same from their mothers. This study did not assess relationship status. More recently, Walsh (1995) expanded studies of social support by examining five provisions of social support (attachment, reliable alliance, guidance, reassurance of , . worth, and social integration) in a sample of 209 lesbians with a "positive sexual ^ :% identity." She found that lesbians with high self-esteem and low internalized homophobia reported the most social support. She determined that different provisions of social support did not come from different sources. Overall, social support was provided by a combination of friends-lesbian, gay, and heterosexual. Lesbians in a relationship had greater social support than did single lesbians. This study did not attempt to investigate the influence of social support on the couple relationship although the sample included women in committed relationships. • . , Wayment and Peplau ( 1 995) also investigated five types of social support that lesbian and heterosexual women receive and the relationship between these types of support and psychological well-being. They compared 75 single lesbians, 75 coupled lesbians, 75 single heterosexual women, and 75 coupled heterosexual women using data from self-report measures originally collected from a sample of 600 women during the 1980s (Aura, 1985). The groups were matched for race, employment, and parent status. They found that overall support reported by lesbian and heterosexual women was equivalent. Women in relationships reported greater well-being than did single women regardless of sexual orientation. A strong relationship between psychological well-being and support was found for all women. Specifically, lesbian women's well-being was ' • associated with reassurance-of-worth support while heterosexual women's well-being

PAGE 70

63 was associated with guidance support. The focus of this study was on lesbians as individuals rather than on couples. Jordan (1995) surveyed 499 lesbians regarding relationship satisfaction, degree of disclosure of sexual orientation, and measures of adjustment including social support. She found that social support was related to degree of disclosure of sexual orientation with those women who more widely disclosed their lesbianism reporting greater social support. Jordan identified a hierarchy of disclosure in which lesbians were most likely to disclose their sexual orientation to lesbian and gay friends, followed by straight friends, family, and co-workers. Furthermore, in an analysis of couple data Jordan (1995) found that disclosure of sexual orientation was significantly related to relationship satisfaction. However, it is unclear how many of the 499 women in committed relationships were included in couple data. In a recent comparative study of the self-reported stress levels and social support of 29 lesbians and 77 heterosexual women, Tait (1997) found no difference regarding stress experienced by these groups. In terms of social support, she identified differences between sources of support for lesbians and heterosexual women. Specifically, she found that lesbians turned to their partners first and then to mental health professionals such as psychologists or psychiatrists while heterosexual women turned to a variety of helpers including family, psychologists, psychiatrists, and partners with no one type being dominant. While Tait (1997) expanded the assessment of social support to include sources outside family and friendship network, she did not gather information regarding the influence of support on the couple system. -

PAGE 71

In a study of lesbian, gay, and bisexual youth, Gilmore (1996) investigated the relationships among social support, general stress, gay-related stress, anxiety, comfort with sexual orientation, depression, substance use, and self-esteem. He surveyed 105 individuals ages 14 years to 20 years using a variety of self-report instruments. Results of this investigation suggest that low social support and general stress both predicted depression and substance use. Other predictors of depression included gay-related stress, younger age, and discomfort with sexual orientation. Couples were not included in this study, in part due to the age of the sample. Grossman and Kemer (1998) studied the social support and self-esteem of 32 lesbian and 58 gay youth ages 14 years to 21 years. Unlike most other samples, the individuals in this study were predominantly Black or Latino/a. Using self-report measures, these researchers found that self-esteem was a moderate predictor of emotional distress in the total sample and a strong predictor in the lesbian subsample. Satisfaction with support was not significantly related to emotional distress that the authors suggest may be due to the fact that a global assessment of support was used rather than one in which specific support sources were assessed. In a comparative study of heterosexual, lesbian, gay, and bisexual adults, O'Neill (1998) surveyed 201 individuals to determine potential predictors of depression including lack of social support, dysftmctional thinking, stressful life events, autonomy, and sociotropy. He found no differences between heterosexual and combined gay-lesbianbisexual groups in regard to relationships between current depression and risk factors. Nor did he fmd that the gay-lesbian-bisexual group reported greater levels of each risk

PAGE 72

factor although this group did report greater levels of current depression. Relationship status was not considered in this investigation. . Jordan and Deluty (1998) examined the relationship between lesbian individuals' disclosure of sexual orientation and four indices of psychological adjustment: anxiety, self-esteem, positive affectivity, and social support. The 499 lesbians in the sample responded to a set of questionnaires. They found that the more widely a lesbian disclosed her sexual orientation, the greater her level of social support. Further, these researchers identified that disclosure of sexual orientation was positively associated with less anxiety, more positive affectivity, and greater self-esteem. Jordan and Deluty (1998) assessed disclosure to family, gay and nongay friends, and coworkers. They did not assess disclosure to members of other potential support sources such as helping professionals and clergy. Research on lesbian couples . In one of the first studies to compare social support systems of different types of couples. Wood (1983) compared 20 lesbian couples with 20 cohabiting heterosexual couples and 20 married heterosexual couples. She also examined the relative functioning of each couple type on three indicators: psychosocial maturity, intra-couple ftinctioning, and couple-community functioning. Contrary to her expectation. Wood found significant differences between the groups. Compared to married heterosexual couples and cohabiting heterosexual couples the lesbian couples received social support from friends rather than from family. Lesbian couples in this sample had more fimctional adaptability patterns and were more psychosocially mature than their cohabiting heterosexual counterparts.

PAGE 73

66 Wood (1983) used a snowball sampling technique, which necessarily calls into question the generalizability of the sample. Many researchers must resort to snowball sampling when attempting to locate participants from minority groups such as lesbians. However, this technique is particularly problematic because this study investigated social support. It is likely that lesbians identified via some sort of network would have support from that network. Although membership in such a network does not equate with high levels of support, it does suggest that the women in this sample have a network of people who are aware of their lesbianism. Thus, women who are relatively closeted may be underrepresented in this sample. Wood (1983) acknowledged other limits regarding the generalizability of her results based on the demographic similarity of the sample. Additionally, Wood (1983) found order effects on the social support scores, which suggests that the instrument she devised to measure this construct may not be valid. Finally, the return rates for each of the groups differed significantly. It is possible, then, that the groups may be less similar than Wood posits. Aura (1 985) conducted another comparative study to describe similarities and differences between the social support systems of lesbian and heterosexual women. She compared a sample of 50 lesbians with a sample of 50 heterosexual women regarding the sources, amount, and value of social support for these women. The samples were matched on age, education, employment, number of children, and length of relationship. Although she was primarily interested in the role of partaers, friends, and relatives in social support, she developed a taxonomy of 20 types of support divided into four categories: personal, social, accompUshment, and instrumental. The results of this investigation revealed that lesbians and their heterosexual counterparts did not differ on

PAGE 74

the value placed on social support. These two samples did not differ on the total amount of social support perceived. Aura (1985) did find that the groups differed on the sources from which they perceived support. Heterosexual women reported receiving a greater amount of total social support from relatives than did lesbians. Lesbians reported that support lacking from family of origin was obtained via additional support from their partners. Heterosexual women perceived less total social support from their partners than did lesbians. While Aura ( 1 985) did not claim to have obtained a representative sample of lesbians, she did match the lesbian and heterosexual women on key demographic characteristics. However, the results of the study may be limited due to the original technique of obtaining the lesbian sample from lesbian and feminist organizations. Additionally, the small sample size may compromise the degree to which the results of the study can be applied. Further, this investigation did not focus on support given the couple although couples were included in the sample. Finally, this study was limited to support from friends and family and, therefore, did not assess other potential sources of support such as institutional supports. Meyer (1989) examined the relationship between social support and relationship closeness among 60 lesbian couples in varying stages of relationship development. Using self-report measures to assess both emotional closeness and support, she found that social support and emotional closeness were significantly related for the sample of lesbians. Further, she identified a negative relationship between perceived support from friends and couple closeness but a positive relationship between perceived support from

PAGE 75

68 family and couple closeness. Meyer (1989) did not assess forms of support other than that provided from family and friends. Olson (1988) examined the relationship between social support and psychological well-being in a sample of 65 lesbian couples using several self-report measures. She found a significant positive correlation between the quality of social support and the number of close relationships and psychological well-being. Further, she identified that cohesion, adaptability, and homosexual identity formation were also positively correlated with psychological well-being in this sample of lesbian couples. Although Olson (1988) assessed couples, the predictor variable she used was an individual construct (psychological well-being) rather than a couple construct (e.g., relationship satisfaction). In another study of support from family and friends, Markowitz (1993) compared 500 gay men and lesbians with 500 heterosexuals. Using self-report measures, she found that lesbians and gay men were no more estranged from their families of origin than heterosexual women and men. She suggested that lesbians and gay men utilize friendship networks and create families of choice to supplement rather than replace support from relatives. In a series of studies, Kurdek and Schmitt (1986a, 1986b, 1986c, 1987b) expanded the comparative study of women's social support to include cohabiting heterosexual women as well as married heterosexual women and cohabiting lesbians. Additionally, they included a focus on the relationship quality as well as on social support. These researchers hypothesized that social support would act as a buffer to ' stress and thereby enhance relationship satisfaction. Both partners in 44 married, 35 heterosexual cohabiting, 56 lesbian, and 50 gay couples completed questionnaires.

PAGE 76

Lesbians reported receiving less social support from family than did married women and cohabiting women. In a separate analysis using the same data set, Kurdek and Schmitt (1987b) found that regardless of whether individual or couples scores were used, lesbian couples reported receiving more social support from friends than from family. Contrary to expectation, they found the reverse true for cohabiting heterosexual women. These women, like their married counterparts, reported family more than friends as their primary source of social support. ' Kurdek and Schmitt ( 1 986a) suggested that the results of these studies be interpreted with caution. They pointed out that the sample size was small and that the return rate for the questionnau-es was low. Furthermore, they noted that because stage of relationship was not factored into the design of the investigation a potentially important mediating variable may have been excluded. In a later study, Kurdek ( 1 988a) compared the perception of social support among 50 lesbian and 69 gay couples. In this study, he assessed the relationships between social support and both relationship quality and psychological adjustment using self-report measures. The results of this investigation once again demonstrated that lesbians (and gay men) received social support more from friends than from relatives. Additionally, psychological adjustment in this sample was related to satisfaction with social support, large number of providers of support, and the perception of support from friends and partners rather than family. Interestingly, Kurdek (1988a) also found that a significant discrepancy between partners regarding familial support was related to high relationship satisfaction. He concluded that a high degree of support from one partner's family may

PAGE 77

compensate for a lack of support from the other partner's family and that support from one family may be sufficient to enhance relationship quality. In his most recent comparative analysis of relationship quality among married heterosexual, cohabiting lesbian, and cohabiting gay couples Kurdek (1998) investigated five dimensions of relationship quality: intimacy, autonomy, equality, constructive problem solving, and barriers to leaving. In data collected from 236 married heterosexual, 51 lesbian cohabiting, and 66 gay cohabiting couples, two relationship outcomes were assessed: trajectory of change in relationship satisfaction and relationship dissolution over 5 years. The lesbian couples reported more intimacy, more autonomy, more equality, and fewer barriers to leaving the relationship than did their married heterosexual counterparts. However, lesbian partners also reported a higher rate of relationship dissolution than did the married heterosexual and gay cohabiting couples. Kurdek (1998) cautioned that these results may not be generalizable due to the K ' ^ h ; i. ; r r ' ' recruitment strategy used to obtain the lesbian couples. Furthermore, demographic data suggest that the sample was highly educated. The equivalence of couples on demographic variables was not obtained by matching; instead, it was approximated using statistical corrections. Additionally, he noted that although the assessment of relationship quality seems both valid and reliable, it may not be comprehensive. Finally, because all measures relied on self-report, response bias may be a factor in the interpretation of the data. Research on gay men. Although gender socialization differences make parallels between lesbians and gay men impossible, studies based solely on samples of gay men lend support to the research regarding lesbian social support. Berger and Mallon (1993)

PAGE 78

11 investigated several types of social support received by a sample of single and coupled gay men. They found that the most frequent source of social support for gay men was a friend, specifically, a close male friend. This supports Kurdek's (1988a) notion regarding the creation of extended families composed of friends to compensate for lost family support. Berger and Mallon (1993) determined that the most common type of support experienced by the gay men in the sample was emotional support when compared to financial, practical, advice, guidance, and social supports. In a study of coupled and single gay men, Sommers (1982) discovered that men in couples experienced more support than did their single counterparts. This finding is similar to that reported by Leavy and Adams ( 1 986) in their sample of lesbians. Furthermore, Sommers (1982) reported that involvement in the gay subculture was related to positive gay identity development for the men in the sample, regardless of relationship status. Research on Institutional Support ^ According to Heller and Swindle ( 1 983), the degree to which an individual has social support is a fimction of the interaction between the availability of "supportive structures in the environment" and the capacity with which the individual elicits and maintains supportive relationships that "present themselves in the environment." Consequently, social support is not a unidirectional process. Lesbian couples may have limited access to social support structures due to exclusion from socially sanctioned supports afforded heterosexual couples. Institutional supports for lesbians such as schools, religious assemblies, and legal organizations may be a source of rejection rather than support (Kurdek, 1998; Walsh, 1995).

PAGE 79

In one of the first studies to investigate institutional support, Chafetz and colleagues (1974) interviewed 51 lesbians in Houston, Texas. They found that the w^omen in this study reported a lack of both general and specific support from a variety of social institutions. Of these women, 50% reported receiving little or no support from family of origin. Additionally, 90% of the women in the sample reported feeling a lack of support from religious institutions, and 66% said they did not feel support from economic institutions. Further, 40% of the women in the sample said they had been harassed by police officers, and 90% reported feeling lesbians were ignored by the media. The primary sources of support for these women were lesbian friends and the lesbian community. These authors focused exclusively on the support given lesbians as individuals and did not address the support lesbians received for their relationships. ' , In a survey of 91 lesbians Albro and Tully (1979) examined the ways that these women functioned in the homosexual "micro-culture" as well as in the heterosexual "macro-culture." They discovered that more than half of the women in the sample did not share their lesbian identities with members of the heterosexual "macro-culture" including family of origin or coworkers. These researchers also found a positive correlation between degree of social interaction in the lesbian "micro-culture" and willingness to be open regarding sexual orientation. Conversely, they discovered a negative correlation between age and disclosure of sexual orientation. . . Bryant and Demian (1994) surveyed a national sample of 706 lesbian couples and 560 gay couples. Respondents rated the degree of support for each of 1 2 sources including gay and heterosexual fnends, family, gay and heterosexual church, gay organizations, coworkers and boss. No significant differences were found between gay

PAGE 80

couples and lesbian couples. Couples reported low levels of support from all types of relatives. Overall, couples ranked coworkers as more supportive than family members. Churches were the only sources rated as more "hostile" than supportive overall. One quarter of those surveyed reported that nearly half of the potential sources of support were totally unavailable because these sources were unaware of the couple's relationship. In Kurdek's (1998) most recent study described above, barriers to leaving the relationship was one of five indices of relationship quality (intimacy, autonomy, equality, constructive problem solving) assessed in a sample of 236 married heterosexual, 51 lesbian cohabiting, and 66 gay cohabiting couples. In this study barriers to leaving the relationship were defined as forces external to the relationship that would keep couples in the relationship regardless of personal happiness. Five barriers were assessed: religious beliefs, pressure from family and friends and community, financial difficulties, personal obligation, and responsibility for dependents. Two relationship outcomes were assessed: trajectory of change in relationship satisfaction and relationship dissolution over 5 years. Kurdek ( 1 998) found that lesbian couples reported fewer barriers to leaving the relationship than did heterosexual couples. Further, he found that of the five indices of relationship quality assessed, only barriers to leaving the relationship uniquely predicted relationship dissolution. Kurdek's (1998) investigation extends the literature regarding social support for lesbian relationships because it considers the influence of cultural institutions rarely studied such as religious and financial institutions in addition to those frequently studied in past research (e.g., family, friends). However, Kurdek conceptualizes the influence of cultural institutions on lesbian relationships from a negative point of view in that he

PAGE 81

considers these institutional forces as constraints to leaving a relationship despite personal unhappiness rather than as supportive of the relationship. Furthermore, the cultural institutions included in Kurdek's conceptualization of barriers to leaving the relationship are not comprehensive. Additional cultural institutions may influence lesbian relationships (e.g., legal, healthcare, educational). ' ^ In summary, there is limited agreement regarding the conceptualization and measurement of social support. However, some general conclusions can be drawn from the literature. First, some researchers have conceptualized social support as a main effect while others view it as having a buffering effect on the psychological functioning of ' individuals. Second, perceived support is more accurately assessed than received social support. Consequently, an instrument developed for this investigation to assess perceived support was utilized to assess the support lesbian partners perceive they have available to them from a variety of social institutions. Third, models of social support have been developed for individual rather than couple or family unit. Fourth, while lesbian couples report they receive support primarily from friends rather than from family, no • investigations to date have assessed other institutional sources of support. f Research on Ritual A preponderance of the recent literature regarding secular ritual has focused on the use of ritual in family systems and has been based on foundational concepts from anthropology, sociology, and most recently, psychology. Much of the literature regarding family ritual has focused on the types and ftinctions of rituals in family systems (Bennett, Wolin, & McAvity, 1988; Bennett, Wolin, Reiss, & Teitelbaum, 1987; Bossard & Boll, 1950; Boyce, Jenson, James, & Peacock, 1983; Fiese, 1992; Fiese & Kline, 1993;

PAGE 82

75 Imber-Black & Roberts, 1992; Kline, 1992; Mccubbin & McCubbin, 1987; Meredith, 1985; Meredith et al., 1989; Mize, 1988; Myerhoff, 1977; Reiss, 1981; Reiss & Oliveri, 1991; Roberts, 1988; Schvaneveldt & Lee, 1983; Wolin & Bennett, 1984; Wolin et al, 1980). The latest research has begun to focus on the salience of rituals for family members (Fiese, 1992; Fiese & Kline, 1993; Fiese et al., 1993; January, 1996). Only recently have researchers begun to investigate the importance of rituals in the life of marital partners or friendship dyads (Bruess, 1994; January, 1996). While the empirical literature regarding family rituals is limited and research regarding interpersonal rituals among couples is sparse, the empirical research regarding rituals in lesbian couples is nonexistent. To date, no clinical studies have focused on the use and functions of ritual in this family form. The following sections review both the anecdotal and empirical literature on the definitions, types, and functions of rituals in families, heterosexual couples, and lesbian couples, Definitions ^ ^ ='V ' Historically, the study of ritual was the exclusive domain of anthropologists and often centered on religious and cultural rites (Durkheim, 1965). Implicit in this connotation of ritual are the concepts of magic, myth, and mysticism (Bossard & Boll, 1950), yet rituals are more often commonplace occurrences that routinely take place within the context of interpersonal relationships (Bossard & Boll, 1950; Fiese, 1992; Fiese & Kline, 1993; Wolin & Bennett, 1984). Family rituals range from extremely circumscribed religious ceremonies such as baptisms and bar mitzvahs to commonplace daily interaction patterns such as the way that parents and children say goodnight to each other.

PAGE 83

76 The concept of ritual remains enigmatic in practice and in research. In fact, the multiplicity of defmitions of this concept bears witness to its intangible nature. Describing family ritual, Wolin and Bennett (1984) explained, "Ritual itself is an elusive concept, on the one hand transparent and conspicuous in its enactment, on the other, subtle and mysterious in its boundaries and effects on participants" (p. 401). Much of the current literature attempting to define ritual focuses on family ritual. In these instances family is defined as parents and children and/or extended kin such as grandparents, aunts, uncles, and cousins. Bossard and Boll (1 950) collected data from 400 family case records spanning a 60-year period for their historic study of family ritual. They defined i ritual as a purposeful, prescribed procedure that takes on a "sense of rightness as a result of its continuing history" (p. 29) in family interactions. They identified that rituals take on many forms and arise fi-om virtually all aspects of family life. Bossard and Boll (1950) argued that rituals are the "core" of family life. In their work with families of alcoholics, Wolin and Bennett (1 984) defined ritual as a form of symbolic communication repeated over time due to their contribution to family satisfaction. Schvaneveldt and Lee (1 983) identified two types of family rituals: rituals associated with tradition such as holidays and religious ceremonies and rituals that spontaneously erupt out of family interactions. These researchers defined ritual as a prescribed pattern of behavior that is purposeful in nature (Schvaneveldt & Lee, 1 983). Ritual versus Routine It is important to make the distinction between ritual and routine for it is this distinction that is at the heart of this investigation. Routines have been defined as predictable activities enacted within the family context while rituals involve routines

PAGE 84

which have a symbolic meaning (Wise, 1986). Routines are characterized by their functional aspect, such as eating dinner or shopping for groceries weekly, whereas rituals have a symbolic meaning that transcends the activity itself, such as eating out on weekends or taking a family vacation. Routines can become rituals if those enacting them perceive them to be relationally or symbolically significant (Fiese, 1992; Wise, 1986). For example, getting children ready for bed is not simply a matter of changing clothes, brushing teeth, turning out lights but is a cherished time to interact and share about the day's activities. Essentially, any routine can become a ritual once it acquires a symbolic status. It is the symbolic quality of ritual that instills meaning m what would ordinarily be repetitive or habitual experiences. The shared meaning of a family ritual endows a routine task or mundane experience with salience and potency (Myerhoff, 1977). Types of Rituals ^ " ' ^ • ' ' ' Schvaneveldt and Lee (1983) categorized rituals into two very general classes: traditional and spontaneous. More recent authors have offered other typologies. Wolin and Bennett (1984) divided rituals into three types: traditions, patterned family interactions, and celebrations. They defined each of these categories according to the setting and degree to which each is linked to cultural customs. Family tt-aditions are unique to families and less influenced by the cultural context in which the family lives. Traditions include family reunions, birthdays, and vacations. Family traditions may be developed in order to serve a particular family need (i.e., family meetings). Patterned family interactions are spontaneous rituals that organize daily life. They involve little planning and occur with great frequency. Examples of patterned family interactions

PAGE 85

include dinner time and bedtime rituals. Family celebrations are culturally influenced rituals that families devise in accordance with holidays or developmental transitions such as Thanksgiving and weddings. Fiese and Kline (1993) developed a seven-part classification of family ritual settings. Dinnertime rituals are defined as those activities and interactions involved in a shared family meal. Weekend rituals involve recreational and planned events that occur on nonworking days. Vacation rituals are the leisure activities surrounding a family vacation. Annual celebrations include yearly commemorations of family birthdays, anniversaries, and the first day of school. Special celebrations include weddings, graduations, and family reunions as well as other celebrations that occur regardless of culture or religion. Religious holidays include rituals that honor Christmas and Chanukah. Cultural and ethnic traditions include naming ceremonies, funerals, and cooking ethnic foods. Janine Roberts (1 988) developed a categorization of family ritual styles. Strict ^ rules for role performance and unforgiving attendance policies for all family members characterize the first style--the rigidly ritualized families. The second type are underritualized families who infrequently share rituals and usually ignore significant family milestones such as birthdays. The third type of famiUes have a skewed ritualization focus in which they center their rituals exclusively on one aspect of family life such as reUgion or the needs of a single family member. Families with hollow ritualization, the fourth type, focus on the routine aspect of family rituals to the exclusion of almost all else, thereby enacting rituals with very little meaning and symbolic attachmem. Families with interrupted ritualization, the fifth type, often deal with

PAGE 86

unexpected family crises such as serious injury or death of a member. The sixth type, famines with flexible ritualization style, consistently practice meaningful rittials, despite transitional periods, due to their ability to adapt roles and routines to the changing developmental needs of family members. "* " ' In a similar typology, Imber-Black and Roberts (1992) identified five ritual styles: minimized, rigid, obligatory, imbalanced, and rigid. A minimized ritual style is characterized by haphazard or absent rituals. In families with a minimized ritual style, important events are ignored or barely acknowledged and other aspects of life (e.g., work, telephone calls) are allowed to interrupt rituals. Unvaried and highly prescribed rituals that often leave family members feeling confined or constrained characterize the rigid ritual style. These families are unable to allow rituals to develop as the family changes and often reflect the family's narrow prescription of roles for its members. A lack of meaning for the ritual and an absence of spontaneity or playfiilness characterize the obligatory ritual style. Families with an obligatory ritual style often feel anxiety or guilt at the thought of trying to adapt these rituals to meet the changing needs of family members. The imbalanced ritual style often occurs as a result of significant differences within the family. For instance, families in which members come fi-om different cultures, religions, or races may embrace rituals from one heritage to the exclusion of the other. Lesbian and gay couples often have an imbalanced ritual style if one partner's family is accepting of the couple and the other partner's family is not. The flexible ritual style is Characterized by the simultaneous ability to offer a sense of continuity and connectedness over time while validating developmental changes. The flexible ritual style allows for

PAGE 87

growth and change among family members and relationships (Imber-Black & Roberts, 1992). Bruess (1994) conducted one of few empirical studies to date of the types of interpersonal rituals. She analyzed the rituals used by marital partners and members of friendship dyads. Seven major types of rituals were identified (as well as seven subtypes) in marital dyads. Couple time rituals are comprised of enjoyable activities, escape episodes, and togetherness rituals while symbolic rituals consist of private codes, play, and celebration rituals. Couples also engage in spiritual rituals, communication rituals, daily routine rituals, rituals of intimacy expression, and ritualized pattems-habitsmannerisms (Bruess, 1994). Members of friendship dyads also shared many of these ritual types. Conceptualization of Rituals Researchers in the fields of anthropology, psychology, and sociology have investigated the central dimensions of family ritual. Cultural anthropologists have suggested that symbols are the smallest unit of ritual (Turner, 1967, 1969). Wolin and Bennett (1984) identified two dimensions of rituals: commitment and flexibility. They described the commitment dimension as the family's commitment to ritualization. Commitment to ritualization is the degree to which a family embraces and encourages rituals The ability of a family to adapt ritual to the changing needs of members and the family as a whole comprises the flexibility dimension. Wolin and Bennett (1 984) argued a strong commitment to ritualization and the capacity to ; successfiilly adapt family rituals engenders in family members a sense of meaning and family identity.

PAGE 88

Fiese and Kline (1993) identified eight dimensions of rituals: occurrence, roles, routines, attendance, affect, symbolic significance, continuation, and deliberateness. The frequency with which rituals are performed constitutes its occurrence. Family members adopt specified roles and perform designated routmes in their enactment of family rituals (Bossard & Boll, 1950). Often, family rules govern who is expected to attend and absences become significant (Bennett et al., 1988). The emotional investment in the ; ritual constitutes its affect dimension while the attachment of meaning to the activity is reflected in the symbolic significance dimension. Continuation reflects the degree to which a ritual is passed down through the generations (Reiss, 1981). Finally, the planning and intentional execution of a family ritual constitutes its deliberateness dimension. The Meaning Factor and the Routine Factor of Ritual '^^^^^^ Fiese and Kline (1 993) found that ritual dimensions could be compressed into two distinct factors: meaning and routines. The meaning factor describes the symbolic component of family rituals including the way that members make sense of family : activities and interactions. The dimensions of deliberateness, symbolic significance, affect, attendance, and occurrence contribute to the meaning component. Consequently, meaningfulness is a cluster of associated factors involving emotional investment, expression of meaningftilness or specialness, a commitment to occurrence, and deliberateness in the planning and implementation of family rituals. The meaning factor of family ritual has been positively associated with adolescent self-esteem (Fiese, 1992) and marital satisfaction (Fiese et al., 1993) as well as negatively associated with anxiety health symptoms in adolescent raised by an alcoholic parent (Fiese & Kline, 1993).

PAGE 89

Fiese and Kline (1993) investigated the relationship between family ritual involvement and self-esteem in 241 college undergraduates. They found that students who reported having been raised in an alcoholic family reported significantly lower meaning in family rituals than did their counterparts raised in nonalcoholic families. Ftirthermore, those students from alcoholic families who also reported high levels of meaningfubess of rituals demonstrated fewer physical symptoms of anxiety than did students reporting lower meaningfulness of rituals. Regardless of the type of home in which the students were raised, meaningftilness of family ritual was related to high selfesteem levels. In a study of family rituals and family functioning among 90 nonclinical Catholic families in New Mexico, January (1996) found a significant positive association between healthy family functioning and meaningful rituals. Using the Family Ritual Questionnaire to assess the degree and meaning of family ritual, she determined that families from different ethnic traditions differed in the degree of ritualization utilized over time. Specifically, Hispanic families experienced fewer declines in family ritual as children aged than did Anglo families. Families in this investigation that scored relatively higher on indices of healthy family fiinctioning also reported higher levels of meaning in their family rituals (January, 1 996). The routine factor represents the degree to which roles are rigidly proscribed and the flexibility with which the family practices rituals. The second component, routines, is comprised of the dimensions of roles and routines. The routine factor, as defined by Fiese (1992), has not been empirically investigated. • I V

PAGE 90

83 Research on the Benefits of Rituals Much of the research regarding the use of rituals has centered on the various benefits that rituals produce for families rather than for dyadic relationships. Although a substantial amount of this literature is anecdotal in nature, some empirical investigations have imdertaken the task of exploring the multifaceted purposes of rituals. . Enhancing family identity . Bossard and Boll's (1950) groundbreaking ' ^ investigation concluded that family ritual provides a sense of family culture by promoting the intergenerational transmission of a family's beliefs, values, and goals. Bossard and Boll (1950), writing from a sociological perspective, characterized ritual as "the core of family culture." As such, it provides a sense of continuity between generations as well as opportunities for learning and interaction. Psychologists Sameroff and Fiese (1992) described rituals as a significant aspect of the "family code," a regulatory system between individual and culture. Researchers have found that family rituals contribute to the establishment and maintenance of a family's identity (Reiss, 1981) and relay family goals and values between and within generations (Wolin & Bennett, 1 984). In their landmark study of the role of ritual in families of alcoholics, Wolin and Bennett (1984) determined that rituals assist in the establishment of family identity. Their studies of the intergenerational transmission of alcoholism found that rituals seem to serve a "protective" function. Family rituals contribute to a sense of belonging by maintaining the roles, rules, and behavioral expectations. Although many rituals are culturally patterned, idiosyncratic family rituals enhance members' sense of a shared family identity thereby encouraging cohesiveness among members (Wolin & Bennett, 1984).

PAGE 91

Rosenthal and Marshall (1988) investigated the intergenerational transmission of family ritual and its function of providing continuity between generations. They interviewed a random sample of 1 12 adults, stratified by age and sex, from Hamilton, Canada. These researchers found that several factors affected the degree of change in family ritual over time. Specifically, they identified decrease in family finances, increase in women working outside the home, and increase in travel as factors contributing to a decline in family ritualization over time. Furthermore, they identified specific rituals that seemed relatively impermeable to change: Christmas celebrations and Sunday family times. Finally, Rosenthal and Marshall (1988) found that parents raised in homes with a high level of ritualization as children tended to continue this tradition whh their own children. Creating a family paradigm. Models of family fiinctioning have begun to include the role of ritual in the culture of the family. Reiss (1981) posited a cognitive model of family fimctioning based in social construction theory. A culmination of 15 years of research involving interviews with more than 400 families, Reiss' (1981) exploration of family rituals is rich with qualitative information. He posited that a family's reality is socially constructed in such a manner that members share a set of organizing beliefs and assumptions which structure patterns of daily living. Enhancing strength and stability. Meredith (1985) found that rituals strengthened family bonds. In a subsequent study, Meredith et al. (1989) surveyed three generations of family members in an effort to determine the relationship between ritual and family strengths. They assessed family strength using measures of family loyalty, pride, trust, and effective conflict resolution. These researchers determined that a significant positive

PAGE 92

relationship existed between family strength and family rituals. Additionally, they uncovered gender and generational differences in family members' perceptions of the importance of rituals. Women more than men, regardless of generation, viewed rituals as important contributors to family strength. Furthermore, members of elder generations identified rituals as making greater contributions to family strength than did younger generations (Meredith etal, 1989). Enhancing belonging. Family rituals are thought to protect against loneliness and uncertainty by structuring certain family activities and clearly defining behavioral expectations and roles (Wolin & Bennett, 1 984). Meredith (1 985) found that rituals strengthened family bonds by encouraging meaningful and often symbolic contact among family members. Kline ( 1 992) investigated the possible influence of family rituals on individual psychological organization, particularly attachment. In a study of college students, she found that those young adults raised in family environments rich in ritual had a stronger sense of "belonging-in-theworld" than those young adults raised in homes with relatively fewer rituals. In a second investigation, Kline (1992) identified that •. female students fi-om families with relatively high levels of ritualization had an easier separation fi-om parents and a better adjustment to college than did their counterparts from families with relatively lower levels of ritualization. Strengthening couple identity . Bruess (1994) identified fiinctions of rituals in interpersonal dyads such as marital relationships and friendships. She found that in couples "rituals enact 'masonry' functions by building and cementing the relational bond and commitment" (p. 186). Additionally, marital partners utilize rituals to contribute to the culture and the maintenance of the relationship by providing opportunities to learn '

PAGE 93

more about each other, promoting communication, and offering occasions for bonding, thus enhancing the happiness and longevity of the relationship. Couples also engage in rituals in order to define marital roles and promote continuity and organization of daily life. Bruess' (1994) investigation demonstrated that marital partners engage in rituals in order to create a sense of couple identity. Rituals also provide a sense of predictability and security. Finally, rituals enhance excitement via spontaneous flexibility among marital partners. In a qualitative study of 26 unmarried Texas women, Mize (1988) investigated the relationships among family of origin rituals, gender, and meaning. Using in-depth ' interviews, she identified levels of meaning and levels of ritualization. Mize (1988) found that women with relatively high levels of ritualization also reported relatively high levels of meaning compared to women who reported low levels of ritualization. She noted that ritual was considered a "female-centered experience" and that it afforded women the opportunity for affiliation. Buffers against stress. Models of family fimctioning have begun to include the role of ritual in the culture of the family. Reiss (1981) posited a cognitive model of family fimctioning in which a family's reality is socially constructed such that members share a set of organizing beliefs and assumptions which structure patterns of daily living. This structure is challenged at times of family stress (e.g., developmental transitions) or crisis (e.g., death, acute ilhiess, delinquency, teen pregnancy). The family may experience a reduction or even total loss of "implicit" fimctioning thereby rendering its ' members unable to attribute meaning to events. No longer a source of rejuvenation and resilience, events such as birthdays or anniversaries become a burden. In essence, the

PAGE 94

87 family disconnects from its past. Families that successfully manage the stressor or crisis create a new "family paradigm" which is viewed by members as a family accomplishment and restores "implicit" functioning. According to Reiss (1981) the family rituals are the "repository" of the family paradigm-they permit the ongoing experience of the transformed values and assumptions. In this way, family rituals simultaneously perpetuate and embody the family's approach to the larger systems in which it is embedded. The community system in which the family lives imposes meaning to stressful events and crises (Reiss & Oliveri, 1991). According to Reiss and Oliveri (1991) two dimensions of stress are "accountability" and "competence." Accountability refers to the degree to which the family is viewed as having created the stress while competence refers to the family's ability to cope with and protect itself from the stress. These researchers predicted that families that perceive themselves as having a sense of mastery over the social context of which it is a part would emphasize rituals that served to connect them to this social world. On the other hand, families that do not perceive themselves as able to influence their cultural context would emphasize rituals that isolate them from it. Reiss and Oliveri ( 1 99 1 ) studied 45 middle class. White parents involved in PTAs in Washington, DC, using a Q-sort methodology. Their results were as predicted but are necessarily limited by the nature of the sample. Boyce and colleagues (1983) theorized that rituals are an integral component of healthy family fimctioning. They suggested that social experiences such as family rituals and family routines facilitate a sense of continuity and stability that serve a protective function for families during times of stress. These authors argued that during times of

PAGE 95

crisis, rituals perform important functions in family life-provision of social support and security. In this way, rituals foster feelings of protection and stability for family members challenged with stressful situations (Boyce et al., 1983). Family rituals were described as an "index of family integration" by McCubbin and McCubbin (1987). As such, they assist the family in effectively dealing with stress, problems, and crisis. Family rituals are a resource that contributes to the family's ability to organize and give meaning to shared experiences in times of hardship. The resiliency model of family functioning was developed in order to expand prior family stress theory which had been based exclusively on middle class. White, heterosexual families (McCubbin, McCubbin, Thompson, & Thompson, 1995). This model acknowledges the significant contribution of family celebrations, traditions, routines, and rituals as resources for effectively coping with family stress and crisis. These resources are seen as a source of stabihty and support for families in times of need-in essence, a buffer against stress. Studies of alcoholic families were among the first to test this "buffer theory." Using semi-structured interviews, the ritual patterns of 25 families of alcoholics (including adult children) were assessed (Wolin et al., 1980). These investigators predicted that families in which rituals were chronically disrupted by alcoholic member's behavior would be more likely to have children that later developed alcoholism or had . alcoholic spouses. On the other hand, these researchers hypothesized, families that were able to maintain ritual practices m spite of alcoholic members would be more likely to break the cycle of alcoholism. The results of the investigation supported both ' hypotheses. ,. /

PAGE 96

In a subsequent study of 68 couples raised in alcoholic families, the development of "family heritage" or ritual practices was explored (Bennett et al., 1987). Interviews of these couples identified the careful plarming and implementation of family heritage practices as a protective factor in the intergenerational transmission of alcoholism. Couples described as "high deliberate" employed explicit strategies regarding their family ritual early in their marriages. These couples developed ritual patterns based on the spouse's nonalcoholic family of origin or other models of sharing traditions without alcohol. In a continued effort to explore the concept of "deliberateness" in family ritual patterns, 37 alcoholic families were compared with a matched sample of 45 nonalcoholic families (Bennett, Wolin, & Reiss, 1988). The emotional, behavioral, and cognitive functioning of the children in each family were assessed. Children from "high deliberate" families, families able to consistently plan and implement family rituals, functioned at higher levels emotionally and behaviorally than their counterparts in "low deliberate" families. These pioneering studies of the influence of ritual in family functioning lend credibility to the concept of ritual as a protective factor in the face of family stress. Additional information regarding the cultural and ethnic backgrounds of participants might have been helpful in understandmg the generalizability of the results. Of course, the small sample size indicates a need for replication with larger numbers of families. Additionally, the self-identified sample may limit generalizability particularly in families in which denial is the predominant coping strategy (Jacob, Favorini, Meisel, 8c Anderson, 1978). Wolin and Bennett (1984) used a sample of families that had already identified

PAGE 97

90 alcohol-related problems and therefore this sample may have been skewed with a greater proportion of families that had begun to take steps to solve family problems. Fiese and Kline (1993) further studied the protective function of family rituals using a sample of 241 undergraduates in New York and one of each student's parents. A preponderance of the sample was White, female, age 17-21, and middle class. In this study only those rituals occurring in the dinnertime, vacation, and weekend settings were assessed. They found that undergraduates who were raised in a family in which one parent was alcoholic reported less meaning associated with family rituals. This trend was also reported for fathers but not for mothers. Furthermore, the students from alcoholic homes who reported highly meaningful family rituals scored significantly lower on a measure of the physical symptoms of emotional disturbance than did their counterparts who reported family rituals relatively lacking in meaning. Differences in family members' perceptions of rituals were found as well, suggesting that assessment of all family members is an important consideration in conducting research regarding ritual (Fiese & Kline, 1993). A possible response bias may exist due to the fact that students were asked to identify problematic drinking among their parents. Students raised in an alcoholic home may view family life in a more negative light than their counterparts in nonalcoholic families. Families facing the stress of normative developmental transitions may also reap the protective benefits of rituals. In a study of 1 15 married couples making the transition to parenthood, Fiese and her colleagues (Fiese et al., 1993) found that couples with preschool-age children reported more meaningfulness in family rituals than did couples whose first child was still an infant. Couples who scored higher on the meaning factor of

PAGE 98

91 family ritual reported greater marital satisfaction than their counterparts who reported V: relatively hollow family rituals. It is possible that meaningful rituals protect the couple from the developmental stress of first-time parenting and that these rituals serve to renew the relationship after an intense period of transition (Fiese et al., 1 993). In a comparative study of 43 families with a developmentally delayed child and 43 families with children developing "normally," Bucy (1 995) investigated the role of ritual and family stress. She found no differences between family types on the basis of degree of ritualization or ritual meaningftilness. Regression analysis revealed that the interaction of meaningfulness and ritualization to be the best predictor of parenting stress. Bucy (1995) concluded that couples adapting to parenting a disabled child may increase their resilience to stress by creating a predictable and orderly family life characterized by meaningful family rituals. Research on Dyadic Rituals * While family ritual literature is abundant, the literature regarding rituals shared by members of dyadic relationships is relatively scarce. Furthermore, what literature exists in this area is primarily anecdotal rather than empirical in nature (Imber-Black, 1 988a, 1988b; Oring, 1984). |In one of the few empirical studies of couple rituals, Fiese and her colleagues (Fiese et al., 1993) compared 115 couples making the transition to parenthood. These parents were predominantly middle class. White, and in their early 30s. These researchers used the Dyadic Adjustment Scale to assess marital satisfaction and the Family Ritual Questionnaire to assess the relative degree of meaning associated with the couples' rituals. She found that those couples who reported relatively greater degree of

PAGE 99

meaning associated with their rituals were more satisfied with their marriages than their counterparts who reported relatively hollow rituals. In a qualitative investigation of possible types and functions of dyadic rituals, Bruess (1994) interviewed 20 married couples and collected data from an additional 79 couples. Bruess (1 994) identified a typology of rituals in two kinds of dyadic relationships: marriages and fiiendship dyads. While the ritual typology for marital dyads had some overlap with that for friendship dyads, significant differences were found. This suggests that rituals may be relationship-specific. Bruess' (1994) investigation yielded seven types of rituals among couples and six ritual types among friendship pairs. In a subsequent study of the functions of ritual in marital and fi-iendship dyads, Bruess ( 1 994) sampled 494 married partners. She reported that couples in this study described benefits from ritual use similar to those reported by families. Bruess ( 1 994) concluded that for married couples rituals fostered a sense of belonging and couple identity, and facilitated opportunities for positive interaction. Valenti (1 997) investigated the effects on relationship satisfaction and intimacy of educating heterosexual couples about the use of positive rituals in their relationships. A total of 45 White couples participated in a lecture regarding the positive use of rituals in relationships. Measures of intimacy and relationship satisfaction as well as the Family Ritual Questionnaire were administered before and after the lecture. Valenti (1997) found no statistically significant results. Valenti (1997) discussed the results of this study as due in part to the homogeneity of the sample, small sample size, and lack of sensitivity ofthe instruments used to assess couple rituals. ,

PAGE 100

Rituals in Lesbian Relationships af : To date, no empirical investigations have been conducted on the influence of ritual in lesbian relationships. However, in a national survey of 706 lesbian couples, Bryant and Demian (1994) found that 60% of lesbians had symbolized their commitment with a relationship ritual. Anecdotal literature points out both similarities and differences between the ways that lesbian couples and heterosexual couples use ritual in their ^. relationships (Butler, 1991; Carter & McGoldrick, 1989; Imber-Black, 1988a; ImberBlack & Roberts, 1992; Roth, 1985; Slater, 1995). This literature is reviewed in the following section. Several authors suggest (Butler, 1991; Imber-Black, 1988a; Imber-Black &, Roberts, 1992; Slater, 1995) that one of the ways in which lesbian couples cope with the absence of public acknowledgement and social approval for their relationships is by developing rituals to validate their unions. According to Slater (1995), lesbian couples live in a greater cultural context in which particular family types-namely, traditional heterosexual families with married parents-are "invited" to be recognized through ritual marking important milestones such as marriages, deaths, births, retirements, and graduations. In contrast, lesbian couples and families are not invited. Consequently, lesbian couples create unique rituals to assist them with basic needs faced by most couples. Slater (1995) noted, "Boundaries around their relationship that are socially recognized and respected, social validation that the family unit is a legitimate and viable grouping, and a sense of membership in a group of similar families that make up the surrounding social community" (p. 63). She identified three types of rituals commonly utilized by lesbian couples: daily rituals and family traditions, celebration and holiday

PAGE 101

rituals, and family life cycle transition rituals. According to Slater (1995), lesbian couples commonly rely on daily rituals and family traditions to validate their relationships and provide a sense of belonging. This is because these rituals require no participation from anyone outside the family. For instance, the way that the couple cooks dinner together or shares events of the day may provide a sense of continuity and stability for the partners yet this acknowledgement of their relationship does not extend beyond the doors of their home. Such daily rituals and family traditions affirm to the couple that they are not simply "roommates" and can be particularly powerful for those couples who fear retribution from family or society on the basis of their sexual orientation (Slater, 1995). Slater (1995) noted that lesbian couples often choose to mark more events than heterosexual couples. For instance, lesbian couples frequently celebrate a number of anniversaries: the day they met, the day they moved in together, the day they bought rings together. In this way, lesbian partners create a family story together in which past, present, and implied fiiture are united. . While daily rituals and family traditions serve to provide a sense of continuity and validation of the relationship, family life cycle transition rituals mark developmental changes and milestones in the lesbian relationship (Slater, 1 995). Slater ( 1 995) noted, "These rituals acknowledge changes within each individual partner, within the family of creation, in the couple's relationship to their families of origin, and in the relationship between the family and the community at large" (p. 74). Lesbian couples' exclusion from social sanction in celebrations of commitment, anniversaries, or the birth of a child may encourage these couples to create rituals with symbols and traditions that validate their unique position in the larger cultural context rather than relying on those

PAGE 102

95 traditionally found in heterosexual celebrations (Butler, 1991; Imber-Black, 1988a; Slater, 1995). Additionally, lesbians may create rituals to mark events not experienced by heterosexual families such as "coming out" (Gargaetas, 1995; Neu, 1995). Neu (1995) conceptualized the process of coming out as a rite-of-passage and explored the use or ritual to commemorate this transition. She described the use of feminist rituals to "empower women for personal and social transformation." She noted that lesbians may create rituals in which women's ways of thinking, feeling, and living become normative. The commemoration of holidays or religious celebrations can be especially public events as they often include extended family or even members of the community at large. For example, lesbian partners may not be invited by their families of origin to share in Thanksgiving, Christmas, or Passover gatherings. Furthermore, lesbian partners may not feel that they can attend public celebrations (e.g.. Mass, holiday office parties) of these events as a couple (Carter & McGoldrick, 1989; Slater, 1995). Consequently, lesbian couples are faced with the task of negotiating a way to "mark the event and remain an unmarked family" (Slater, 1995). Lesbian couples may choose to create elaborate celebrations of idiosyncratic family holidays (e.g., birthdays or anniversaries) that are not imbued with inherent social expectations from family of origin or society at large (Slater, 1995). Additionally, lesbian couples may create rituals to commemorate events unique to the lesbian community (Slater, 1995). Imber-Black and Roberts (1992) identify five ritual styles: minimized, rigid, obligatory, imbalanced, and flexible. Lesbian couples may suffer the difficulties of an imbalanced ritual style (Imber-Black & Roberts, 1992) particularly if the family of origin of one partner is more accepting than that of the other partner. The rituals of the

PAGE 103

accepting family may be adopted to the exclusion of the rituals of the rejecting family because most holidays and celebrations can be shared with the former but not the latter. While the couple benefits from a sense of belonging, the partner whose family is rejecting suffers the double loss of family relationships and familiar family rituals (Imber-Black & Roberts, 1992). This situation may also preclude the couple devising their own unique rituals. Carter and McGoldrick (1989) addressed the potential for rigid, closed family structures among lesbian couples because these couples lack "marker events" to commemorate important transitions or change in family status (Roth, 1985). ImberBlack and Roberts (1992) noted that lesbian couples may feel shame or guilt from the absence of family or community supports for rites-of-passage such as unions. Fear of openly marking a life cycle transition may prevent lesbian couples from developing rituals. These authors pointed out that rituals to "celebrate the uncelebrated" offer lesbian couples a way to have significant life transitions witnessed and validated. In summary, both anecdotal and empirical literature regarding the types and fimctions of rituals suggest that rituals play an important role in the lives of most heterosexual family members. However, no empirical study has been conducted to determine what types of rituals lesbian couples use. Furthermore, although anecdotal literature supports the existence and value of ritual in lesbian couple relationships, no empirical validation of specific benefits on relationship satisfaction exists. Research on Conflict Resolution . Research regarding heterosexual couples consistently demonstrates that couples who are able to resolve conflicts mutually are more satisfied than couples who do not

PAGE 104

(Christensen & Shenk, 1991; Epstein et al., 1978; Gottman, 1994; Gottman & Levenson, 1992; Markman, Floyd, & Stanley, 1988; Markman, Renick, Floyd, Stanley, & Clements, 1993; O'Leary & Smith, 1991). Although the bulk of empirical research regarding couple's conflict resolution has been conducted on samples of heterosexual partners, recent studies suggest that conflict resolution is also an important construct in the relationships of lesbian and gay couples. Conflict Resolution Theory v To understand better the factors influencing lesbian relationship satisfaction, a theoretical framework is needed which addresses the contributions of internal couple dynamics and the external dynamics of the larger cultural context in which the couple is embedded. No existing theory integrates both perspectives. Consequently, this study , was based upon two theoretical traditions: Gottman's (1 994) balance theory and Kurdek's relationship appraisal model. > . " # : : Gottman's (1 994) balance theory has been used to conceptualize the interactional processes by which heterosexual married couples sustain satisfying relationships. This theory is based on multimethod assessments of significant aspects of couple interaction (e.g., partner's problem solving, power, affect, distance regulation, and physiology) in an 8-year longitudinal investigation of 79 couples as well as laboratory observations and ongoing assessments of four cohorts of couples in different stages of the family life cycle. Over the past two decades, Gottman and his colleagues conducted empirical research that ultimately led to the formulation of balance theory. According to this theory, satisfied couples maintain a balance in their ratio of positive to negative interactions.

PAGE 105

Two theoretical premises underlie balance theory. First, satisfied couples develop a ratio of positive to negative affective interactions as well as the ratio of positive to negative problem solving interactions. Over time, the negative aspects are "balanced" by positive aspects. For satisfied couples, this balance is "tilted" toward the positive. Gottman (1994) described balance theory as "an ecology of marital behaviors" in which the ratio of positive interactions to negative interactions is "tilted" in the direction of positive interactions. Thus, successful and stable couples sustain their relationships by ensuring that negative interactions are "balanced" with a significant amount of positive interactions. ? . • The second theoretical premise underlying balance theory is the notion that partners in satisfied couples develop a "fit" regarding preferred interactional style. Gottman (1994) postulated three types of "stable" couples: volatile, avoiding, and validating. Each of these couple types results in a "stable," satisfied relationship in which positive and negative interactions are balanced. Volatile relationships balance a high level of argument and conflict with significant amounts of laughter and passionate romance. In other words, high volumes of positive interactions balance high volumes of negative interactions. On the other hand, high levels of neutral affect characterize avoiding types of couple relationships. The avoiding couple type "appears to involve a minimization of the importance of disagreement. It results in a good deal of calm interaction, but pays the price with emotional distance" (Gottman, 1994, p. 182). In avoiding couple types, a low level of negative interactions is balanced by a low level of positive interactions. Finally, the validation type of couple balances careful selection of when to confi-ont conflict with warmth and "we-ness." Partners in these relationships

PAGE 106

demonstrate little intensity regarding either disagreement or passion. A moderate amount of negativity is balanced by a moderate amount of positivity in validating couples. Gottman (1994) addressed cultural variation in the distribution of couple types, but only from an ethnic point of view. He postulated that while one couple type may be more prevalent in a specific ethnic culture, the eroding effect of an unbalanced (e.g., high ratio of negativity to positivity) relationship will remain the same. Gottman has not addressed the applicability of the balance theory to lesbian couples although he reports a study--yet to be published-regarding lesbian and gay couples' conflict resolution strategies (cited in Gottman, 1994). Although Gottman' s balance theory (1994) incorporates behavior, cognition, and physiology, it is primarily an interactional model which does not account for contextual forces that may influence couples. His theory is based on the study of heterosexual marital couples-couples that benefit from the social sanctions given their relationships by the larger cultural context. Members of marginalized and stigmatized groups, such as lesbians, often do not receive the institutional support for their relationships afforded heterosexual couples. Research o n Conflict Resolution Among Heterosexual Coup les Gottman (1998) reviewed research regarding conflict resolution in marital partners. He noted that observational studies have identified consistently characteristic patterns in conflictual marriages. Specifically, negative affect reciprocity and the demand-withdraw pattern. Additionally, conflictual marriages are characterized by a greater amount of negative than positive behaviors as well as the presence of specific types of negativity such as contempt, criticism, and stonewalling.

PAGE 107

• 100 Gottman and colleagues (1998) investigated seven marital interaction processes in a study of marital satisfaction among 130 newlywed couples. Divorce was predicted by four of the seven models. Specifically, the interaction pattern in which husbands reject their wives influence was predictive of divorce. Negative start-up by the wife and husbands inability to de-escalation the wife's low intensity negative affect were both predictive of divorce. Finally, lack of "physiological soothing" of husbands was also predictive of divorce. These authors concluded that these findings provide support for the balance model of positive-to-negative affect. In a study of 1 56 long-term marriages, Carstensen, Gottman, and Levenson (1995) used an observational-coding system to identify emotional responses in marital problem discussions. They found that older couples conflict resolution was more affectionate and less negative than middle-aged couples. Further, wives expressed more negative affect than did husbands. In contrast, husbands expressed more defensiveness than wives. Couples who reported lower levels of relationship satisfaction also demonstrated greater exchange of negative affect than did couples who reported higher levels of relationship satisfaction. In a recent study of heterosexual couples, Kurdek ( 1 995b) examined the link between husbands' and wives' conflict resolution styles. Specifically, he investigated three types of conflict resolution styles: conflict engagement, withdrawal, and compliance. This study sought to determine potential changes in marital satisfaction (over a 2-year period) associated with conflict resolution style. The sample consisted of 155 married heterosexual couples. Results support the notion that husbands' marital satisfaction is more often affected by the way that wives resolved conflict than wives'

PAGE 108

101 marital satisfaction is affected by their husbands' conflict resolution style. Specifically, Kurdek (1995b) found that initial conflict resolution styles predicted change in husbands' marital satisfaction but not in wives' marital satisfaction. The frequency with which wives used conflict engagement was significantly related to decrease in marital satisfaction while the frequency with which husbands used withdrawal was significantly related to decrease in marital satisfaction. Hojjat (1997) investigated types of conflict resolution and relationship satisfaction in heterosexual couples. He categorized conflict resolution style according the dimensions of activity and valence. Consequently, four types of conflict resolution styles were assessed: positive-passive, negative-passive, positive-active, and negative-active. Analyses of data revealed that agreement between partners' perceptions of conflict resolution styles was associated with greater relationship satisfaction. Similarly, individuals' accurate perceptions of their partners' conflict resolution style related positively to relationship satisfaction. Furthermore, Hojjat (1 997) found that women perceived themselves to use the negative-active style more than did men. In contrast, men perceived themselves to use the positive-passive conflict resolution style more than did women. The relationship between conflict resolution style and attachment style was examined by Fischer (1993). She surveyed 140 couples upon completion of a proble solving task that involved the possibility of conflict. Five conflict resolution style assessed: integrating, compromising, avoiding, dominating, and obliging. Three attachment styles were assessed: secure, anxious-ambivalent, and avoidant. Analyses of data revealed that individuals with secure attachment styles used imegrating or lemles were

PAGE 109

102 compromising conflict resolution styles. In contrast, individuals with avoidant attachment styles used dommating or avoiding conflict resolution styles. Further, securely attached individuals reported the highest relative relationship satisfaction followed by anxiously-ambivalently attached individuals and, in turn, avoidantly attached individuals. Solheim (1990) investigated conflict resolution styles and family work roles in a sample of 161 married couples over a 6-year time period. She used both qualitative and quantitative design elements. She found that both husbands and wives who reported greater use of withdrawal and control styles of conflict resolution also reported higher levels of family conflict. In contrast, partners who used negotiation reported less conflict regarding family work roles. Research on Lesbian Partners' Interactions ' . ' Bryant and Demian ( 1 994) surveyed a national sample of 706 lesbian couples and 560 gay couples. Respondents were asked to categorize the two greatest challenges to their relationship. Over 40% of lesbian couples named communication as an important relationship challenge. One of the most important aspects of couple communication is " the manner in which partners resolve conflicts in their relationships. The interaction of lesbian partners has frequently been depicted as pathological (Morin, 1 977). Communication and interaction patterns of lesbian couples have been characterized by negative terms such as "merger" or "fusion" (Krestan & Bepko, 1980; Lindenbaum, . . 1985; Pearlman, 1988; Roth, 1985). Krestan and Bepko ( 1 980) were the first to depict lesbian relationships as "fused/merged." They defined "fiision/merger" as a lack of differentiation between

PAGE 110

103 partners in lesbian couples. Citing examples from clinical settings, they suggested that regardless of presenting concerns, lesbian couples experience intense anxiety regarding a desire for autonomy within the relationship. They hypothesized that "fiision/merger" was a coping response to a hostile environment that ignored or attempted to disrupt the boundaries of the lesbian relationship. They identified six potential environmental factors that may contribute to "fusion/merger" in lesbian relationships: (a) societal assumptions of heterosexuality for all women, (b) sexism, (c) loss of heterosexual social role and status, (d) rejection from family, (e) jealousy from heterosexual women, and (f) influences from the lesbian community. Burch (1986) used the term "psychological merger" to describe "fusion/merger" and identified two factors that contribute to this phenomenon. First, systematic elements in the environment such as societal disregard and disrespect for lesbian relationships lead to a loss of status and support for the couples. Second, psychosexual development of women is such that ego boundaries are more permeable thereby affording women greater relational capacity, more difficulty with separateness, and greater tendency towards merger. Roth (1 985) added the socialization of women as a potential factor in the development of "fiision/merger." These theories regarding the structure and ftmctioning of lesbian interaction patterns were developed in the context of clinical populations. However, they have been generalized to describe all lesbian relationships. Furthermore, no empirical validation of these theories has been conducted either on clinical or nonclinical samples of lesbian couples. This tendency toward generalization of pathology is a common occurrence in research regarding lesbians and gays and reflects "heterosexual bias" (Morin, 1977)

PAGE 111

10^ Whereas heterosexual relationships reflect a blending of male socialization and female socialization, lesbian relationships inherently embody an intensification of the effects of gender-related phenomena. Consequently, the interaction and communication styles of lesbian partners will likely reflect gendered-pattems. Mencher (1984) ; interviewed a small, nonclinical sample of stable and well-functioning lesbian couples. She found that the intense closeness between the partners and centrality of the ' relationship in their lives were cited by all the women as accounting for the success of their relationships. While "fusion/merger" is often cited as inhibhing individual growth, this study supports the conceptualization of "fusion/merger" as a facilitative of an atmosphere of trust and safety conducive to risk-taking and self-actualization (Mencher, 1984). Slater ( 1 995) conceptualized "fiision/merger" as a protective strategy, without ' which lesbian couples may be at risk. She suggested that environmental pressures for the couple to behave as separate, disconnected individuals can be heightened if one or both members of the couple fear such an intense level of connectedness resulting in a failure of the couple to generate sufficient cohesiveness. In essence, "fusion/merger" may be fiinctional and necessary for lesbian couples developing relationships in a homophobic society (Slater, 1995). v . : ' Recent models of lesbian couple functioning have begun to focus not only on the devaluation lesbians experience m a society that is homophobic but also on the devaluation of women in a society that is sexist (Donaldson, 1993; Miller, 1984, 1986, 1988; Surrey, 1985). Contemporary models of women's development note primacy of relational processes in women's lives. Theorists at the Stone Center have begun to

PAGE 112

105 reframe relational patterns of lesbians as inherently healthy in terms of "fusion" (Donaldson, 1993; Miller, 1984, 1986, 1988; Surrey, 1985). According to these scholars, a major developmental goal for women is the creation, sustenance, and enlargement of mutually enhancmg relationships. Women's development in this framework rests neither on differentiation of self from other nor necessary disconnection from others for growth. Rather, the relational process itself becomes the hallmark for growth and development. Based on this theory of women's development, interaction patterns of healthy adult relationships must be reevaluated in terms of the quality of the relational process (Donaldson, 1993; Miller, 1984, 1986, 1988; Surrey, 1985). Mencher ( 1 990) reframed the construct of "fusion" from a pathological relationship function to a pattern of intimacy that allows all women, heterosexual and lesbian, to express relational strengths. She suggested renaming the term "embeddedness" in order to acknowledge the normative developmental needs and intimacy patterns of women. She supported removing the conceptualization of "fusion/merger" from the traditional, male-defined standard of autonomy and separation. In this way, lesbian relationship patterns may be viewed in a nonpathological framework. Consequently, research regarding lesbian relationships may be conducted without preconceived constructs for functioning and structure. Because lesbian and gay relationships are comprised of two people of the same gender, the couple inherently gets a double measure of the positive and the negative aspects of their gender socialization. On the other hand, in heterosexual couples the positive and negative aspects of female cultural socialization as well as the positive and negative aspects of male cultural

PAGE 113

106 socialization are combined. Consequently, the interaction and communication styles of lesbian partners will likely reflect gendered-pattems. Research on Lesbian and Gav Couple Conflict Resolution Kurdek (1994a) examined the areas in which couples experience conflict in a sample of 51 lesbian, 51 lesbian, and 108 heterosexual cohabiting couples. He investigated the frequency of conflict in 20 separate content areas and potential associations with relationship satisfaction. Six groups of conflict content areas were identified: (a) power, (b) social issues, (c) personal flaws, (d) intimacy, (e) distrust, and (f) personal distance. The order of frequency of conflict among the six areas was similar for lesbian, gay, and heterosexual couples. Greater frequency of power arguments and greater frequency of intimacy arguments were strongly negatively related to current relationship satisfaction levels in all couple types. Over time, increases in the frequency of power arguments were associated with decreases in relationship satisfaction for each couple type. -c ' . . Kurdek ( 1 99 1 ) compared 75 gay couples and 5 1 lesbian couples on self-report measures of relationship satisfaction. This study investigated variables related to contextual, investment, and problem-solving models of intimate relationships. Overall, he found more similarities than differences between lesbian and gay couples on mean levels of each model variable. Similarly, very few differences were found between these two couple types in terms of the strength of the correlates of relationship satisfaction. Variables from each of the three models were significantly related to reported relationship satisfaction in both lesbian and gay couples. ;

PAGE 114

Kurdek (1998) investigated five predictors of relationship dissolution and change in relationship satisfaction over time in a sample of 236 married heterosexual, 51 lesbian couples, and 66 gay cohabiting couples. This study was based in part on Gottman's balance theory of relationships; constructive problem solving was assessed from the perspective of self as well as the perspective of partner. Variables representative of the contextual model (e.g., barriers to leaving the relationship) as well as variables representative of gender-linked forces within the relationship (e.g., intimacy, autonomy, and equality) were assessed using self-report measures on an annual basis over a 5-year period. Analyses of data revealed no differences between heterosexual, lesbian, and gay couples in the strength vsath which the five variables assessed (constructive problemsolving, barriers to leaving the relationship, intimacy, autonomy, and equality) were related to relationship quality measures. One of the unique features of this investigation is the assessment of both self-appraisals and partner-appraisals of each of the five variables of relationship quality. Kurdek (1998) hypothesized that partners in both lesbian couples and gay couples would report greater constructive problem solving because gender-liked patterns of conflict resolution would be the same in these couple types but different in heterosexual couples. This hypothesis was not supported, however. Kurdek ( 1 998) explained this finding as a possible reflection of unidimensional assessment of constructive problem solving and suggests that fiirther research should include assessments of the investment that partners have in the issue at hand. Further, he found no differences based on couple type on measures of initial relationship satisfaction. Similarly, heterosexual and lesbian and gay couples did not differ on the trajectory of change over the 5-year time period although each couple showed a decline in relationship

PAGE 115

satisfaction. Although the five variables assessing relationship quality were linked to initial relationship satisfaction reports, they were not linked to the rate of change in relationship satisfaction levels over time. ' Metz and colleagues (1994) investigated differences in conflict resolution styles among 36 heterosexual, 36 lesbian and 36 gay couples matched for age and length of relationship. The study was based on an interactional process model of conflict resolution v^th a foundation in social learning theory. According to this model, conflict resolution styles can be organized into two dimensions: engagement versus avoidance and constructive versus destructive. Each of these dimensions can be further organized into a total of twelve styles. Analyses of data revealed few differences between the relationship satisfaction reports and the level of conflict among the couple types. However, lesbian partners reported significantly greater relationship satisfaction and greater hope for conflict resolution than did gay or heterosexual partners. Furthermore, lesbian partners endorsed more constructive conflict resolution styles than partners in either gay or heterosexual couples. The authors concluded that differences in conflict resolution styles appear to be correlated with gender role factors rather than with sexual orientation. Further, they acknowledged the importance of common lifestyle features, such as the presence of children in the home, as important influences in conflict resolution styles. In an examination of the relationships among conflict resolution style, sex-role orientation, and attachment styles, Maarse (1994) surveyed 1 10 gay men in committed relationships. Three types of conflict resolution style were assessed: integrating, avoiding, and dominating. Sex-role orientation was categorized as either masculine or

PAGE 116

feminine. He found that the dominating conflict resolution style was positively associated with masculine sex-role orientation. In contrast, the avoiding conflict resolution style was negatively associated with masculine sex-role orientation. On the other hand, feminine sex-role orientation was positively associated with the integrating conflict resolution style. Three quaUtative studies provide evidence of the importance of conflict resolution strategies in lesbian relationships. Park (1994) conducted a qualitative investigation of the normative developmental themes and processes in lesbian relationships in a sample of 18 couples who lived together from at least 1 year. She used a single-session interview with both partners in the relationship to gather information about the development of the relationship over time. Five salient themes were identified: conflict management, commitment, lesbian identity issues, sexuality, and attachment. Dom, (1 990) investigated factors that contributed to long-term lesbian relationships using a design that incorporated both quantitative and qualitative measures. A sample of nine couples who had lived together at least 8 years completed questionnaires and participated in both individual and joint interviews. Five factors were identified that contributed to the enduring quality of the long-term couple relationships: (a) good conflict resolution skills, (b) perceived equality in the relationship, (c) similarity between partners disclosure of sexual orientation, (d) fidelity, and (e) complementary personality styles. These studies support the importance of conflict resolution strategies in lesbian relationships. However, these studies do not address the specific conflict resolution

PAGE 117

styles used by these successful couples. No study has focused exclusively on the assessment of lesbian conflict resolution styles. •: . In a qualitative study of power strategies in intimate relationships, Falbo and Peplau (1980) investigated associations between gender, sexual orientation, and egalitarianism in a sample of 50 heterosexual female, 50 heterosexual male, 50 lesbian, and 50 gay college students. These researchers identified two dimensions of conflict resolution/power strategies: direct versus indirect and bilateral interactive versus unilateral interactive. Gender differences were found between male and female ., heterosexuals but not between gay and lesbian individuals. Male heterosexuals were more likely to use bilateral and direct strategies than were heterosexual females. No differences reflected the individuals' perception of their own power in the relationship. Individuals who perceived themselves as having more power in their relationship, such as heterosexual men, reported more bilateral and direct strategies. This study provides support for quantitative studies reviewed above regarding the salience of gender differences rather than sexual orientation differences in conflict resolution strategies t used. In summary, there is considerable empirical support for the balance theory of couple interaction proposed by Gottman. Although much of the research on couples' use of different types of conflict resolution styles has been conducted using samples of heterosexual couples, recent studies have begun to include lesbian couples as well. " Conflict resolution styles have been predictive of relationship satisfaction in heterosexual, lesbian, and gay couples. Few studies of conflict resolution styles have

PAGE 118

Ill focussed exclusively on lesbian couples. Furthermore, few studies of lesbian couple conflict resolution have used Gottman's balance theory as a foundation. Summary Brown's (1989b) conceptualization of marginality, biculturalism, and normative creativity provides a contextual framework from which to understand lesbian relationships. Initial studies of lesbian relationship satisfaction focused on comparisons with heterosexual couples or with gay couples. Overall, investigations of relationship satisfaction among these three types of couples have revealed more similarities than differences. Studies of lesbian couples have focused on the influence of disclosure of sexual identity and social support on relationship satisfaction. Most studies have focused on disclosure of individual sexual orientation to family and friends. Few studies have examined the influence on relationship satisfaction of the disclosure of lesbian couple identity. Likewise, investigations of social support have focused on support for the individual rather than for the couple. There is limited agreement regarding the conceptualization and measurement of social support. However, some general conclusions can be drawn from the literature. Perceived support is more accurately assessed than received social support. Models of social support have been developed for individual rather than couple or family units. While lesbian couples report they receive support primarily from friends rather than from family, no investigations to date have assessed other institutional sources of support. Empirical literature supports the relationship between couple satisfaction and social support in heterosexual couples but the influence of social and institutional supports on relationship satisfaction in lesbian couples is unclear.

PAGE 119

There is considerable empirical support for the balance theory of couple interaction; however, this model has rarely been used in studies of lesbian couples. Recently, studies have begun to focus on the conflict resolution styles of lesbian couples. These investigations suggest that conflict resolution styles are predictive of relationship satisfaction in heterosexual, lesbian, and gay couples. However, many of the studies to date have focused exclusively on self-appraisals of conflict resolution style ignoring the influence of partner-appraisals. Gottman's (1994) balance theory suggests that satisfied couples must balance V negative affect-such as that characteristic of conflict~with significant amounts of positive affect. Both anecdotal and empirical literature regarding the types and functions of rituals suggest that rituals are often a source of positive interaction in the lives of most heterosexual family members. However, no empirical study has been conducted to determine what types of rituals are used by lesbian couples. Furthermore, although anecdotal literature supports the existence and value of ritual in lesbian couple relationships, no empirical validation of specific benefits on relationship satisfaction exists. •>

PAGE 120

CHAPTERS METHODOLOGY Statement of Purpose The purpose of this study was to assess the influences of five variables on the relationship satisfaction reported by lesbian partners. The five variables examined were (a) the preferred conflict resolution style utilized by each partner, (b) the degree of agreement between partners' about each partner's preferred conflict resolution style, (c) the degree of relationship ritualization, (d) the level of ritual meaningfulness, and (e) the extent of perceived institutional support for the relationship. In this chapter the research hypotheses, relevant variables, data analysis, population, sample, and data collection procedures are described. Additionally, instrumentation and methodology are discussed. Hypotheses The following null hypotheses were evaluated in this study. Hi : There is no significant association between the level of relationship satisfaction reported by lesbian couples and the degree of partner agreement regarding the conflict resolution style utilized. H2: There is no significant association between the level of relationship satisfaction and the type of conflict resolution style utilized by each partner of a lesbian couple. >: 113

PAGE 121

H3: There is no significant association between the level of relationship satisfaction and the degree of relationship ritualization reported by lesbian couples. H4: There is no significant association between the level of relationship satisfaction and the extent of ritual meaningfulness reported by lesbian couples. H5: There is no significant association between the level of relationship satisfaction and the degree of perceived institutional support reported by lesbian couples. Delineation of Relevant Variables Dependent Variable Couple relationship satisfaction, the dependent variable m this study, is defined as the degree to which the couple reports feeling content with the quality of their adjustment to the relationship. This adjustment includes the extent to which the couple report experiencing consensus, cohesion, satisfaction, and affection in the relationship. Spanier's ( 1 989) Dyadic Adjustment Scale was used to assess these qualities. The Dyadic Satisfaction subscale of the Dyadic Adjustment Scale has been shown to adequately assess relationship satisfaction in heterosexual as well as lesbian and gay couples (Kurdek, 1992a; Hunsley, Pinsent, Lefebvre, James-Tanner, & Vito, 1995). This subscale is comprised of 10 items, whereas the entire Dyadic Adjustment Scale consists of 32 items. A couple score was calculated by averaging the scores of each partner in the couple. Independent Variables The following independent variables were assessed: personal conflict resolution style, partner agreement in conflict resolution style, relationship ritualization, ritual meaning, institutional support for the relationship, and demographic characteristics.

PAGE 122

" Its Personal conflict resolution style . Kurdek's (1994b) Conflict Resolution Style Inventory was used to measure the degree to which partners utilized four conflict resolution styles: positive problem solving, conflict engagement, withdrawal, and compliance. The CRSI assessment of conflict resolution style calculated each partner's relative endorsement of strategies from each of four conflict resolution styles. Therefore, conflict resolution style was seen as a continuous rather than a categorical variable. This means that partners received scores on each conflict resolution style in accordance with the degree to which each style is used in resolving relationship disputes. Specifically, the degree to which each partner in the couple utilized each of four conflict resolution styles was computed. Scores for each scale ranged from 4 points-which indicated very low endorsement of the conflict resolution style--to 20 points-which indicated very high endorsement of the conflict resolution style. Both members of the couple evaluated their own conflict resolution style (CRSI-Self scores) and their partner's conflict resolution r. • style (CRSI-Partner scores). Couple scores were calculated for each conflict resolution style by averaging both partners' CRSI-Self scores for each conflict resolution style. That is, the self-evaluations of both partners were averaged for each conflict resolution style to determine the degree to which the couple utilized each style. . v ' . Partner agreement in co nflict resolution stvle . Partner agreement regarding relative use of each of the four conflict resolution styles was assessed in two ways in this study. This allowed for multiple comparisons within the couple regarding the extent of agreement in conflict resolution style.

PAGE 123

116 First, scores from the CRSI-Self were used to determine the difference between partner's self-evaluations of conflict resolution style. That is, the CRSl-self scores for both partners were subtracted and the difference was used to determine the level of agreement between partners regarding each of the four conflict resolution styles. These differences ranged from 0 points-which indicated high levels of agreement--to 16 points which indicated low levels of agreement. Second, scores from the CRSI-Self and the CRSI-Partner were used to make comparisons between couple members regarding both self and partner evaluations of one member of the couple. This score was computed from the difference between each couple member's self and partner evaluations for each of the four conflict resolution styles. For instance, comparisons were made between couple member 1 's self-evaluation of her own conflict resolution style and couple member 2 ' s evaluation of her partner' s conflict resolution style. Like the partner agreement scores, these scores ranged from 0 points-which indicated high levels of agreement~to 16 points-which indicated low levels of agreement. Unlike the agreement scores, these scores incorporated the CRSIPartner evaluations for each couple member. Relationship ritualization. The degree of relationship ritualization in lesbian couples was assessed by Fiese and Kline's (1993) Family Ritual Questionnaire. The FRQ total score was used to represent the degree of relationship ritualization. This score was the sum of scores across seven settings: dinnertime, weekend, vacation, annual celebrations, special celebrations, religious holidays, and cultural and ethnic traditions. A couple score was calculated by averaging both partners' relationship ritualization scores.

PAGE 124

117 Higher scores indicated relatively higher relationship ritualization while lower scores indicated relatively lower relationship ritualization. Ritual meaningfulness . The FRO also assessed ritual meaningfulness. The meaningfulness score was calculated as the sum of four ritual dimensions: occurrence, affect, symbolic significance, and deliberateness. A couple score was calculated by averaging both partners' ritual meaningfulness scores. Higher scores indicated relatively higher ritual meaningfulness while lower scores indicated relatively lower ritual meaningfulness. Institutional support for the relationship. The Institutional Support Scale was designed for this study to assess the couple's perception of support for their relationship from sources such as family, friends, coworkers, neighbors, the lesbian community, and larger cultural institutions including legal, financial, religious, educational, and medical. The total ISS scores for both partners were averaged to form a couple score. Higher scores indicated relatively higher perceived levels of support while lower scores indicated relatively lower perceived levels of support. Demographic Variables A demographic information sheet (see Appendix B) was used to collect data regarding demographic variables. The following variables were assessed: age, duration of couple relationship, length of time cohabiting, location of residence, race-ethnicity, y education level, income, religion, and duration of lesbian identity. Data Analysis Multiple regression analyses were used to assess the contribution to relationship satisfaction of five predictor variables. The predictor variables for which data were

PAGE 125

11« analyzed were as follows: type of conflict resolution utilized by the partners, degree of agreement between partners regarding type of conflict resolution utilized, degree of relationship ritualization, level of ritual meaningfulness, and perceived level of institutional support for the relationship. Both inter-couple as well as intra-couple analyses were computed. Couple scores were generated by averaging individual partner scores. Description of the Population The population was composed of lesbian couples from various communities in the United States. It is difficult to determine the characteristics of lesbian populations as no census data exist for this group. A significant portion of the sample was drawn from the population of lesbian couples in Florida. Specifically, a majority of the couples for the Florida sample were drawn from the Alachua county area. The population of women in Alachua County, Florida, in 1998 was 77% White, 19% Black, and 4% Latina and "other" (U.S. Bureau of the Census, 1998a). The population of women in the State of Florida in 1997 was 75% White, 12% Black, and 13% Latina and "other" (State of Florida Bureau of Economic and Business Research, 1998). In the United States in 1998 the population of women was 75% White, 12% Black, 10% Latina, and 3% Asian (U.S. Bureau of the Census, 1998b). Sampling Procedures ' Network sampling techniques were used to solicit potential participants. The women in the sample were contacted through a variety of means. Advertisements for participants were posted in businesses frequented by lesbians such as local women's bookstores and gay-lesbian-bisexual community centers. Additionally, requests for

PAGE 126

119 participants were included in lesbian newsletters and via online groups. Volunteers were also sought at lesbian events such as sports competitions, concerts, workshops, and social gatherings. Beginning December 1, 1998, potential participants were invited to list their first names and telephone numbers or email addresses so that they could be contacted regarding the study. Beginning January 1, 1999, they were contacted by telephone, letter, or in person by the investigator in order to assess appropriateness for participation in the study. Potential participants were asked to provide information regarding their age (years), sexual orientation (lesbian, bisexual, transgendered, heterosexual), relationship status (single-coupled), commitment status (monogamous, nonmonogamous), duration of relationship (years), living status (together-separate), and household members (childrenno children). Couples in which both partners indicated an interest in participating in the study and stated that they were at least 1 8 years of age and lesbians in a monogamous relationship of at least 1 year's duration and who lived together without children were included in the study. Past research suggests that couples who have been in a relationship for at least 1 year have had time to sufficiently develop and participate in relationship rituals (Fiese, 1992, 1997; Fiese et al., 1993; Fiese & Kline, 1993). It was hypothesized that couples living with children would have developed family and relationship rituals that could be significantly different from couples who do not live with children. This hypothesis has been supported by past research (Fiese, 1992, 1997; Fiese, Hooker, Kotary, &Schwagler,1993; Fiese & Kline, 1993). Therefore, couples with children living in the home were excluded from the study. Furthermore, couples who have not been

PAGE 127

120 together at least 1 year may be in the hmerance period of relationship development and, therefore, may not have experienced much need for conflict resolution strategies. Similarly, couples who have not had to negotiate the daily demands of living together may not utilize conflict resolution styles in the same way that couple who cohabit do (Kurdek, 1988b, 1989, 1991, 1994a, 1994c, 1998; Kurdek & Schmitt, 1986a, 1986c). Therefore, only cohabiting couples who were in a committed and monogamous relationships of at least 1 year's duration were selected for the study. Participants selected for the study were given printed information explaining the purpose of the study and requesting their participation. They were informed of potential risks and benefits as a resuh of participation in the study and were given an opportunity to receive results of the study upon its completion. Participants were asked to sign and return an informed consent form. Additionally, participants were instructed to complete the survey independently and not to discuss their responses until after sealing the survey in the enclosed postage-paid envelope. Follow-up telephone contact was made approximately 2 weeks after mailing the survey packet to insure arrival and to offer to answer questions. One week after the telephone call, postcards were sent thanking participants for completing the survey and once again offering to answer questions and to share the results of the research upon completion of the investigation. Finally, upon completion of the study, postcards were mailed to participants offering to share the results of the study and again thanking participants for contributing to the research. Of the 175 packets distributed, a total of 90 (51%) completed and useable surveys were returned by the February 1. 1999, deadline. Six additional surveys were returned after the deadline but were not included in the analysis of the data. Three additional

PAGE 128

121 surveys were excluded due to missing data or because the couple did not meet the selection criteria (either single, nonmonogamous, living with children, not cohabhing, or relationship duration less than 1 year). Subjects ' The sample consisted of90 lesbian couples in the United States. All of the women self-identified as lesbians and considered themselves currently to be in a committed, monogamous relationship in which they lived together without children. Geographic location . A total of 90 couples (N=l 80) residing in California, Florida, Georgia, Illinois, Indiana, Maine, Massachusetts, Mississippi, New Hampshire, New York, North Carolina, Ohio, Pennsylvania, South Carolina, Tennessee, and Washington responded to the survey. A majority, 76% (69) of the couples in the sample resided in Florida. Five (6%) of the couples in the sample resided in Georgia. The remaining sixteen (18%) couples in the sample lived in either California, Illinois, Indiana, Maine, Massachusetts, Mississippi, New Hampshire, New York, North Carolina, Ohio, Pennsylvania, South Carolina, Tennessee, or Washington. Of the total participant couples, 41 (45%) resided in Alachua County, Florida. This portion of the total sample comprised 59% of the Florida sample. A majority of the 90 couples included in the sample, 49 (54%), resided in urban communities with a total population range of 2,500-100,000. A total of 25 (28%) couples in the sample resided in rural communities with a total population less than 2,500. The sample included 16 (18%) couples who resided in metropolitan communities with population greater than 100,000. Table 1 shows the distribution by community size for the sample.

PAGE 129

m Race . Of the 1 80 women in the sample, 90% (163) were of European descent (Caucasian). The remaining sample consisted to 6% (10) Latina women, 2% (3) women of AfricanAmerican descent, 1% (2) women of Asian descent, and 1% (2) women who classified themselves as biracial. Table 2 includes the frequency distribution by raceethnicity for the sample. . * Table 1 Frequency Distribution of Couples by Community Size Community Size Frequency (f) Percent (%) Cumulative f Cumulative % Rural 25 28 25 28 Urban 49 54 74 82 Metropolitan 16 18 90 100 Table 2 Race-Ethnic Distribution of the Sample Race-Ethnicity Frequency (f) Percent (%) Cumulative f Cumulative% White-European Descent 163 90.5 163 90.5 Black-African Descent 3 1.7 166 92.2 Latina-Hispanic IQ 5.6 176 97.8 Asian Descent 2 1.1 178 98.9 Bi-Racial, Other 2 1.1 180 100.0

PAGE 130

Spirituality . The spiritual and religious preference of the lesbians in this sample varied widely. Only 5% (8) of the women identified as Jewish, while 8% (15) identified as Catholic. Of the 1 80 women in the sample, 28% (50) identified as Protestant. A majority of the women in the sample, 59% (1 07), identified in nontraditional spiritual terms such as "Buddhist," "Pagan." "Nature," "free-flowing spirituality," or "none." Table 3 includes the frequency distribution by spirituality for the sample. Table 3 Spirituality Frequency Distribution Spirituality Frequency (f) Percent (%) Cumulative f Cumulative % Protestant 50 218 50 2X8 Catholic 15 8.3 65 36.1 Jewish 8 4.4 73 40.5 Others Pagan 26 14.4 99 54.9 Buddhist 8 4.4 107 593 Nature 14 7.8 121 67.1 None 59 32.9 180 100.0 Income . The annual income of the 180 women in the sample ranged from less than $5,000 annually to more than $70,000. A substantial number of the women, 42% (76), earned between $20,000 and $40,000 annually. Almost 74% (132) of the women in the sample earned more than $20,000 annually, and 52% (94) of the women earned more

PAGE 131

U -; 124 than $30,000 annually. Only 27% (48) of the women earned less than $20,000 annually, and 14% (25) earned less than $10,000 annually. Table 4 includes the frequency V distribution by income for the sample. Table 4 Annual Income Levels Income ($) Frequency (f) Percent (%) Cumulative f Cumulative % less than $5,000 8 4.4 8 4.4 $5,000-$9,999 17 9.4 25 13.8 $10,000-$ 14,999 12 6.7 37 20.5 $15,000-$ 19,999 11 6.1 48 26.6 $20,000-$29,999 38 21.1 86 47.7 $30,000-$39,999 38 21.1 124 68.8 $40,000-$49,999 25 13.9 149 82.7 $50,000-$59,999 14 7.8 163 90.5 ,v , $60,000-$69,999 4 2.2 167 92.7 ; more than $70,000 13 7.2 180 100.0 Education. The education level of the 180 women in the sample ranged from high school graduate to postdoctorate course work. Of the 180 women in the sample, 16% (29) had high school diplomas, 18% (33) had AA or AS degrees, and 27% (49) had BA • or BS degrees. Additionally, 26% (46) of the women had MA or MS degrees, and 13% (23) had PhD, JD, or MD degrees. None of the women in the sample endorsed middle

PAGE 132

125 school as the highest education level on the demographic data form. Table 5 includes the frequency distribution by education level for the sample. Table 5 V Education Levels FrpfiiiPtiov (f\ Percent (%) Cumulative f Cumulative % Middle School 0 0 High School 29 ; 16.1 ' 16.1 AAor AS 33 18.3 i 34.4 ' . BA or BS 49 27.2 HI 61.6 MA or MS 46 25.6 157 ,.,v.V 87.2 Ph.D., MD, JD 23 12.8 180 100 Age, length of relationship, cohabitation, lesbian identity . Table 6 includes the frequency distribution by age, relationship duration, cohabitation length, and duration of lesbian identity for the sample. The age of the women in the study spanned 50 years from a low of 1 9 years to a high of 69 years. The mean age was 41.9 years. The duration of the relationships for the couples in the sample ranged from 1 to 33 years. The mean length of time these couples had been in committed relationships was 7.18 years. The length of time cohabiting for the 90 couples in this sample ranged from 1 to 33 years. The mean length of cohabitation was 6.57 years. The amount of time the women in the sample had self-identified as lesbian ranged from 1 year to 51 years. The mean length of identification as lesbian was 17.9 years.

PAGE 133

126 This suggests that the women in the sample represented a variety of stages of lesbian identify development. Table 6 Age. Relat ionship Duration. Cohabitation Time, and Lesbian Identity Duration Factor Mean (years) Median (years) Standard Deviation Low (years) High (years) Age 41.97 43.5 10.6 19 69 Relationship Duration 7.18 10.5 5.9 1 33 Time Cohabiting 6.57 10.0 5.8 1 33 Time Lesbian-identified 17.88 20.5 9.8 1 51 Data Collection Each participant in the study was given a separate packet of instruments, coded in the upper right hand comer with couple number and partner number. Partners were given identical surveys. Partners were directed to respond privately and not to discuss their answers with each other until the surveys have been sealed in the return envelope. Each participant was asked to complete a demographic questionnaire, the Dyadic Adjustment Scale-Dyadic Satisfaction subscale, the Conflict Resolution Style Inventory (both self and partner forms), the Family Ritual Questionnaire, and an Institutional Support Scale devised by the author for a total of 130 items. These assessments took approximately 30 minutes to complete. The investigator was available to answer questions and process the assessment in person or by telephone.

PAGE 134

127 Instrumentation ' / In addition to a demographic questionnaire assessing age, length of relationship, race-ethnicity, education level, socioeconomic status, location of residence, household size, and religion, four instruments were used in this study. These instruments were (a) the Dyadic Adjustment Scale, (b) the Conflict Resolution Style Inventory (both self and partner forms), (c) the Family Ritual Questionnaire, and (d) an Institutional Support Scale designed for this investigation. , > Dyadic Adjustment Scale (DAS) ; j;--.:;• The Dyadic Adjustment Scale (DAS, Spanier, 1976) is a 32-item questionnaire designed to assess the quality of relationship dyads. Although Spanier (1976) made no specific mention of lesbian relationships, the DAS has been widely used in assessing relationship quality among gays and lesbians (Dailey, 1979; Donaldson, 1993; Easton, 1987; Eldridge & Gilbert, 1990; Jenkins, 1996; Kurdek & Schmitt, 1986b, 1986c) and is considered the most widely used measure of relationship satisfaction (Glenn, 1990). The DAS assesses four empirically verified components of dyadic adjustment using a 5-point Likert-type scale (Fitzpatrick & Best, 1 979; Spanier, 1 976; Spanier & Filsinger, 1983). Each of these components can be used as separate scales to measure dyadic satisfaction, dyadic cohesion, dyadic consensus, and affectional expression. The Dyadic Satisfaction score consists of 10 items designed to assess the amount of tension in the relationship and respondents' commitment to its continuance. This scale has been demonstrated to effectively assess global relationship satisfaction in samples of lesbian, gay, and heterosexual couples (e.g., Kurdek, 1998). The Dyadic Cohesion scale measures how often the couple engages in activities with one another and is comprised of

PAGE 135

128 five items. Thirteen items are used in the Dyadic Consensus scale to assess how often the couple is in agreement about significant matters such as money, religion, and division of household responsibilities. The Affectional Expression scale consists of four items that measure the degree to which the individual is satisfied with physical and sexual expression of affect in the relationship. Additionally, a global rating of relationship satisfaction, the Total Dyadic Adjustment score, may be obtained. Theoretically, this general relationship satisfaction score can range from 0 to 151 with higher scores indicative of greater "adjustment." Spanier's (1976) original research data indicated that the mean total score for the 218 married heterosexual couples was 1 14.8. Scores of 100 or greater are considered to be in the range of healthy fimctioning. * Many studies have assessed the reliability and validity of the DAS. Spanier ( 1 976) reported a range of internal consistency coefficient alphas: .73 for the Affectional Expression scale, .86 for the Dyadic Cohesion scale, .90 for the Dyadic Consensus scale, .94 for the Dyadic Satisfaction, and .96 for the Total Dyadic Adjustment scale. Other researchers have reported a range of Cronbach's coefficient alpha values for each of the scales: from .62 (DeTurck & Miller, 1986) to .73 (Spanier, 1976) for the Affectional Expression scale, from .72 (Johnson & Greenberg, 1985) to .92 (DeTurck & Miller, 1986) for the Dyadic Consensus scale, from .72 to .86 (DeTurk & Miller, 1986) for the Dyadic Cohesion scale, and from .77 (Antill & Cotton, 1982) to .94 (Spanier, 1976) for the Dyadic Satisfaction scale. Spanier (1976) initially reported a very high total reliability (r=. 96) for the DAS although a later study by Spanier and Thompson (1982) reported a Cronbach's alpha of .91, and more recently Johnson and Greenberg (1985) reported .84. Additionally, the DAS has demonstrated temporal

PAGE 136

129 stability. Using test-retest correlations, Stein, Girodo, and Dotzenroth (1982) found the reliability for the full scale to be .96 after 1 1 weeks while the reliabilities for each of the scales were .98 for Dyadic Consensus, .92 for Dyadic Satisfaction, .88 for Dyadic Cohesion, and .78 for Affectional Expression. The initial construction of the DAS consisted of a large pool of items drawn from all instruments on relationship quality that were available at the time. Items were then methodologically reduced through elimination of duplicate questions and discriminant analysis. In an assessment of the content validity of the DAS, three judges compared each item of the DAS with Spanier and Cole's original definition of dyadic adjustment. Irrelevant items were subsequently dropped from the scale (Spanier, 1976). Criterionrelated validity was established for the DAS in an evaluation in which divorced couples were differentiated from married couples and distressed couples were differentiated from nondistressed couples (Spanier, 1976; Spanier & Filsinger, 1983). Additionally, construct validity was demonstrated through a factor analytic comparison with the LockeWallace Marital Adjustment Scale. Correlation coefficients ranged from .86 to .88 for the two measures (Spanier, 1976). Kurdek (1992a) assessed the reliability and validity of the four scales of the DAS in a sample of 538 married heterosexual and 197 cohabiting lesbian and gay couples. Overall, he found the reliability data for each scale were "acceptable" regardless of couple type. However, he found that the Dyadic Cohesion, Affectional Expression, and Dyadic Consesus scales confounded the assessment of relationship satisfaction. Specifically, Kurdek (1992a) found that each of the four DAS scales did not contribute unique information to relations with other scores of relationship quality. Likewise, he

PAGE 137

130 identified that each of the four DAS scales did not contribute unique information to relations with individual-difference scores known to be linked with appraisals of relationship quality or to the prediction of dissolution of relationships. Kurdek (1992a) argued that most of the variance in these analyses was accounted for by the Dyadic Satisfaction score. In a later investigation of the construct validity of four different short forms of the DAS, Hunsley and colleagues (1995) analyzed data from 196 married or cohabiting Canadians. They found that the Dyadic Satisfaction subscale could be used as a short form substitute for the DAS without compromising construct validity. This study confirmed Kurdek's (1992a) prior research. In this investigation, the Dyadic Satisfaction subscale of the DAS was used to assess relationship satisfaction in lesbian couples. This 10-item scale had a potential score range from 0 to 50. Averaging the scores for both partners created a mean couple score. The mean score for the total sample (N = 180) was 40.32. This suggests that the sample of women were relatively happy with their relationships. Conflict Resolution Stvle Inventory rCRST) •> , < The Conflict Resolution Style Inventory (Kurdek, 1 994b) utilizes complementary self-report and partner-report versions of the same 1 6 items to assess four conflict resolution styles (positive problem solving, conflict engagement, withdrawal, and compliance) for each partaer in a couple. Previously developed self-report measures of conflict resolution style have either lacked a coherent conceptual framework or comprehensive documentation of psychometric properties (Bowman, 1 990; Boyd & Roach, 1977; Christensen, 1988; Snyder, 1981). On the other hand, the 16 items

PAGE 138

131 included on the CRSI are representative of four theoretically validated conflict resolution styles (Gottman & Krokoff, 1989). Furthermore, unlike any other measure of couple conflict resolution, the psychometric properties of the CRSI were empirically established using samples of lesbian couples, gay couples, and heterosexual couples (Kurdek, 1994b). The CRSI is based, in part, on Gottman' s (1994) conceptualization of couple stability as a function of each partner's individual style of conflict resolution (Bowman, 1990; Boyd & Roach, 1977; Gottman, 1994; Heavey, Layne, & Christensen, 1993). Accordingly, the CRSI relies on separate reports from each partner. This affords collection of information about each partner's conflict resolution style from two sources-the self and the partner. Each partner is asked to complete the 16-item CRSI-self scale • ,, and the parallel 16-item CRSI-partner scale. Both scales instruct respondents to rate the frequency (on a scale from l^never to 5=always) with which four different conflict resolution styles (positive problem solving, conflict engagement, withdrawal, and compliance) are used. Thus, each member of a couple dyad rates the frequency with which he-she uses each conflict resolution style as well as the frequency with which hisher partner uses each conflict resolution style (Kurdek, 1994b). Each of the conflict resolution styles-positive problem solving, conflict engagement, withdrawal, and compliance-is assessed by 4 items; thus, a total of 16 items representative of 1 6 conflict resolution strategies comprise the CRSI. Each of these 1 6 conflict resolution strategies was chosen for inclusion in the CRSI based on behavioral observations of couple interactions (Gottman & Krokoff, 1 989). Kurdek ( 1 994b) noted that these distinct conflict resolution styles were selected because of previously

PAGE 139

documented important links to changes in relationship satisfaction and relationship dissolution (Gottman, 1994). The positive problem-solving subscale measures the individual's use of negotiation and compromise. The conflict engagement subscale measures the individual's use of personal attacks and exploding. The withdrawal subscale measures the individual's use of tuning out one's partner and refusing to discuss issues. Finally, the compliance subscale measures the individual's use of giving in and not defending one's position. The initial validation studies of the CRSI were conducted over a 5 -year period with 75 gay couples, 51 lesbian couples, and 207 heterosexual couples. A majority of the sample was White, college educated, and employed. Face validity for both the CRSl-self and the CRSI-partaer was established as a result of item selection based on empirically documented observational data (Kurdek, 1994b). Reliability data for the CRSI-self and the CRSI-partner generalized across partner within couple and across type of couple (lesbian, gay, and heterosexual) over a 1-year period. Specifically, for each of the four CRSI-self scores the Pearson correlations ranged from .46 to .83 (Kurdek, 1994b). For the corresponding CRSI-partner scores Pearson correlations ranged from .54 to .83 (Kurdek, 1994b). All correlations were significant and moderate in size. Cronbach's alpha data demonstrated that the CRSI-self and CRSI-partaer scores for each type of couple as well as for each member within a couple were internally consistent. Particularly, the Cronbach's alpha scores for positive problem solving, conflict engagement, withdrawal, and compliance subscales of the CRSl-self ranged from .65 to .89 (Kurdek, 1 994b). The corresponding scores for each of the four conflict *

PAGE 140

133 resolution styles assessed by the CRSI-partner were also moderate in size and ranged from .80 to .91 (Kurdek, 1994b). The factor structure of the 1 6 items of the CRSI-self was assessed by a * confirmatory analysis. In the first step of this analysis, CRSI-self items for each member of a couple were restricted to load on only one of the four conflict resolution style factors. As expected, LISREL analyses resulted in an acceptable level of fit for the model for the CRSI-self (Kurdek, 1994b). A second model was tested based on the assumption that the four-factor model would be invariant across the types of couples (lesbian, gay, and heterosexual). This model also resulted in acceptable levels of fit for both partners in each couple type for the CRSI-self (Kurdek, 1994b). Similar analyses were conducted on the CRSI-partner to determine how well the items corresponded to the four-factor structure for each partner within the couple. As with the CRSI-self, the analyses for the CRSI-partner revealed an acceptable level of fit for both members of the couples in the sample. Additional analyses determined an acceptable level of fit for the model that hypothesized the four-factor structure to be invariant across couple type for the CRSIpartner (Kurdek, 1 994b). Four composite CRSI-self scores as well as four composite CRSI-partner scores were computed from the sum of ratings given for each of the four items related to positive problem solving, conflict engagement, withdrawal, and compliance strategies for each couple member. These scores were internally consistent and stable over a 1-year period (Kurdek, 1994b). Kurdek (1994b) evaluated four issues regarding validity of the CRSI-self and the CRSI-partner scores for each member of the couples in the sample. First, he investigated the degree of overlap between CRSI-self scores and CRSI-partner scores for each

PAGE 141

134 member of the same couple. He hypothesized that the convergent validity of both CRSI scores would be verified by high correlations between the CRSl-self scores and CRSIpartner scores for each partner of the same couple. He found moderate overlap between couple members' conflict engagement, withdrawal, and compliance scores with r values ranging from .29 to .63. The r values for positive problem solving ranged from .07 to .26. This is evidence of convergent validity, particularly for conflict engagement, withdrawal, and compliance scores. However, the relatively small overlap between CRSI-self and CRSI-partner scores for positive problem solving style across couple type may be due to a response bias. Positive problem solving--a socially desirable conflict resolution style-may be endorsed with greater frequency by individuals evaluating their own (or their partner's) conflict resolution style (Kurdek, 1998). Furthermore, between couple types there was very little difference regarding the frequency with which each of the four styles of conflict resolution was utilized. Concurrent and predictive validity of the CRSI-self and the CRSI-partner were assessed by examining the extent to which the CRSI-self and CRSI-partner scores were concurrently linked with relationship satisfaction, change in relationship satisfaction over time, and dissolution of the relationship. The results for lesbian and gay couples were similar to those for heterosexual married couples. Kurdek (1998) found that lesbian and gay couples as well as nonparent heterosexual married couples who used positive problem solving frequently and conflict engagement and withdrawal infrequently reported high relationship satisfaction. This finding is consistent with the pattern reported by heterosexual couples in other studies and evidence of the CRSI's concurrent

PAGE 142

135 validity (Kurdek, 1998). The compliance conflict resolution style was not significantly linked with relationship satisfaction. Some evidence of the predictive validity of the CRSI-self and CRSl-partner was obtained by examining links with change in relationship satisfaction over time as well as relationship dissolution. Specifically, among lesbian and gay couples and nonparent heterosexual married couples the frequent use of positive problem solving was linked s with increased relationship satisfaction over time. On the other hand, frequent use of conflict engagement and withdrawal was linked to decreased relationship satisfaction over time. Among the four conflict resolution styles, compliance was least likely to be ' related to change in relationship satisfaction over time. In terms of relationship dissolution, infrequent positive problem solving and frequent conflict engagement among lesbian, gay, and nonparent heterosexual married couples predicted dissolution. Kurdek (1998) noted that this pattern is consistent with prior research. Compliance was least likely to be related to dissolution. The reliability coefficients for each of the CRSI-Self and CRSl-Partner subscales for the sample of lesbian couples in this investigation were calculated using the Cronbach's alpha formula. The Cronbach's alpha formula was utilized to account for the Likert scale items on both instruments (Crocker & Algina, 1986). The coefficient alphas for the CRSI-Self subscales ranged from .766 for the withdrawal conflict resolution style to .853 for the compliance conflict resolution style. The coefficient alphas for the CRSIPartner subscales ranged from .848 for the compliance conflict resolution style to .901 for the positive problem solving conflict resolution style. These findings suggest that, depending on the subscale, 77% to 90% of the total score variance was from true score

PAGE 143

136 variance. These findings further indicated that respondents' performance was relatively consistent across items on the scale and may generalize to similar items within a comparable construct domain. The CRSI assessment of conflict resolution style calculated each partner's relative endorsement of strategies from each of four conflict resolution styles. Therefore, conflict resolution style was seen as a continuous rather than a categorical variable. This means that partners received scores on each conflict resolution style in accordance with the degree to which each style is used in resolving relationship disputes. The CRSI scores were used in three ways in this study: to evaluate the degree to which couples utilized each conflict resolution style, to determine the degree of agreement between partners' selfevaluations of their conflict resolution style, and to determine the degree of agreement between couple member's self-evaluations and partner-evaluations of conflict resolution style. Specifically, couple scores were calculated for each conflict resolution style by averaging both partners' CRSI-Self scores for each conflict resolution style. That is, the self-evaluations of both partners were averaged for each conflict resolution style to determine the degree to which the couple utilized each style. Scores from the CRSI-Self were used to determine the difference between partner's self-evaluations of conflict resolution style. That is, the CRSI-self scores for both partners were subtracted and the difference was used to determine the level of agreement between partners regarding each of the four conflict resolution styles. . Scores from the CRSI-Self and the CRSI-Partner were used to make comparisons between couple members regarding the both self and partner evaluations of one member of the couple. For instance, comparisons were made between couple member 1 's

PAGE 144

self-evaluation of her own conflict resolution style and couple member 2's evaluation of her partner's conflict resolution style. This score was computed from the difference between each couple member's self and partner evaluations for each of the four conflict resolution styles. ' ' ' tr ,r Family Ritual Questionnaire (FRO) • v , Rituals are patterned social interactions that include a prescription of roles (Mead, 1973) and an ascription of meaning (Moore & Myerhoff, 1977). Perceptions of ritual use by the couple were measured by the Family Ritual Questionnaire. The FRQ was used to assess the extent to which the couple uses ritual in their relationship. Specifically, the couple's use of ritual was assessed across seven settings and over eight dimensions. Summative scores for ritual meaningfuhiess and extent of relationship ritualization were calculated. The theory undergirding the FRQ is that families and couples use rituals to promote physical and mental health (Fiese & Kline, 1993; Imber-Black, 1988b; Rogers & HoUoway, 1991; Selvini-Palazzoli, Boscolo, Chechen, & Prata, 1977; Whiteside, 1988; Whiting, 1988). Particularly, rituals play a powerfiil role in assisting families with making transitions (Imber-Black, 1988a; van der Hart, 1983) as well as providing members with a sense of family identity and belonging (Bennett et al., 1 988). The FRQ is a 56-item self-report measure designed to assess the overall degree to which a family uses rituals in seven different settings including dinnertime, weekends, vacations, annual celebrations, special celebrations, religious holidays, and cultural and ethnic traditions. The forced-choice items also assess eight dimensions of rituals: occurrence, roles, routine, attendance, affect, symbolism, continuation, and

PAGE 145

138 deliberateness. Developed in 1991, the FRQ is designed to be administered to one or more members of a family unit. Respondents are asked to consider typical interactions in her or his family across seven settings. Eight pairs of statements that refer to each of the eight dimensions of rituals are presented for each setting. Respondents are instructed to choose one of the two statements in each pair that best reflects the family's typical interaction. The respondent is then asked to choose whether the statement is "really true" or "sort-of true" for the family. Each item is scored from 1 --which represents lower levels of ritualization-to 4which signifies higher levels of ritualization. The items of the FRQ were based on parts of the Wolin and Bennett Family Ritual Interview (Wolin & Bennett, 1984). Principal component analysis using a varimax rotation has revealed two distinct components of rituals identified by the FRQ: meaning and routines. The meaning factor describes the symbolic component of family rituals including the way that members make sense of family activities and interactions. The routine factor represents the degree to which roles are rigidly proscribed and the flexibility with which the family practices rituals. Scores from the dimensions of deliberateness, symbolic significance, affect, and occurrence scales loaded heavily on the meaning component. A summary score of these four dimensions comprises the FRQ Meaning score. The second component, routines, is comprised of items from the dimensions of roles and routines. A summary score of these two dimensions is used to compute the FRQ Routines score. The overall reliability (Cronbach alpha) of the FRQ subscales ranges from .58 for routines to .87 for dinnertime, religious holidays, and cultural traditions. Test-retest reliability is reported as .88 over a 4-week period.

PAGE 146

Face validity was established by administering, and subsequently modifying, the FRQ to individuals unfamiliar with the study or rituals. Professionals knowledgeable in the field of developmental psychology and family systems theory evaluated the content validity of the FRQ and recommended changes in item wording. Construct validity has been assessed by comparing the FRQ with the organization subscales of the Family Environment Scale (Moos & Moos, 1981). Further, the FRQ role subscale has been shown to have a positive correlation with anxiety and a negative correlation with lovability. The symbolic subscale of the FRQ was positively related to self-esteem and negatively related to anxiety (Fiese & Kline, 1993). The family ritual meaning factor has been significantly correlated with marital satisfaction and self-esteem (Fiese, 1992; Fiese et al., 1 993 ; Fiese & Kline, 1 993). Because the FRQ is a self-report measure, response bias reflective of social desirabiUty was a concern. The forced choice format was chosen to help eliminate this bias and statements were presented in a way that neither choice was seen as more desirable that the other (e.g., "some families regularly eat dinner together," but "other families rarely eat dinner together"). Likewise, the FRQ is reported to have no response bias on the basis of gender or socioeconomic status. The reliability coefficients for the FRQ settings and dimensions scales for the sample of lesbian couples in this investigation were calculated using the Cronbach's alpha formula. Additionally, reliability estimates for the FRQ Meaning factor. Routines factor, and Total Settings were calculated. The coefficient alphas for the seven FRQ settings subscales ranged from .558 for the vacation setting to .896 for the religious

PAGE 147

holidays setting. The vacation setting was the only FRQ setting subscale with a reliability estimate less than .700. The coefficient alphas for the eight FRQ dimensions subscales ranged from .489 for the occurrence dimension subscale to .742 for the deliberateness dimension subscale. The coefficient alphas for the continuity dimension and routines dimension subscales were .587 and .623, respectively. These are relatively low reliability estimates. Therefore, three of the eight ritual dimensions did not achieve a reliability coefficient of at least .70. ' ' . The coefficient alphas for the FRQ Meaning and Routines factors were .896 and .782, respectively. The coefficient alpha for the FRQ Total was .915. This finding suggest that about 91% of the total score variance was from true score variance. This finding further indicated that respondents' performance was relatively consistent across items on the scale and may generalize to similar items within a comparable construct domain. The FRQ total score was used to represent the degree of relationship ritualization. This score was the sum of scores across seven settings: dinnertime, weekend, vacation, annual celebrations, special celebrations, religious holidays, and cultural and ethnic traditions. A couple score was calculated by averaging both partners' relationship ritualization scores. The meaningfulness score was calculated as the sum of four ritual dimensions: occurrence, affect, symbolic significance, and deliberateness. A couple score was calculated by averaging both partners' ritual meaningfulness scores.

PAGE 148

141 Institutional Support for the Relationship The Institutional Support Scale created for this investigation measured partners' perceptions of institutional support for their relationship. The author created this scale to assess an array of possible sources of institutional support for couples' relationships. This instrument was based in part on Kurdek's (1998) barriers to leaving the relationship questions as well as social support literature. " * ; Kurdek (1 998) conceptualized institutional forces that affect relationship satisfaction in terms of barriers to leaving the relationship. He identified five types of barriers to leaving a relationship based on the following institutions: religion, family of origin, financial, social, and community. Although Kurdek (1998) did not differentiate between support from heterosexual and gay or lesbian sources, other researchers have found this distinction important when examining the social support networks of lesbians (Albro & Tully, 1979; Aura, 1985; Brown, 1989b; Mendola, 1981; Meyer, 1989). Consequently, the Institutional Support Scale includes items to assess support from lesbian and gay fi-iends and community as well as items to assess support from heterosexual fiiends and community. Although Kurdek (1998) identified five institutions that may affect lesbian and gay couples; other authors have suggested additional institutions of significance. Several writers have noted the importance of career, work place, and employment institutions in the lives of lesbian individuals (Bieschke & Matthews, 1 996; Boatwright, Gilbert, Forrest, & Ketzenberger, 1996; Chung, 1995; Fassinger, 1995; Pope, 1996; Sorensen & Roberts, 1997). The role of educational institutions has been studied by a number of researchers recently (Casper & Schultz, 1996; Fontaine, 1997; Jordan, Vaughan, &

PAGE 149

Woodworth, 1997; Rey & Gibson, 1997). Additionally, researchers have recently documented lesbians' interaction with healthcare institutions including medical (Allen, Glicken, Beach, & Nay lor, 1998; Rankow, 1995; Sorensen &. Roberts, 1997) and mental health systems (McDaniel, Cabaj, & Purcell, 1996; Morgan, 1992; Rankow, 1995; , .... Sorensen & Roberts, 1997). The interaction between lesbians and legal professionals has also been documented (McLeod & Crawford, 1998). Finally, although little has been written to date, lesbians may interact with social service and welfare agencies (Mallon, 1998; Mercier & Berger, 1989). Accordingly, items representative of each of these institutions were included in the item pool for the Institutional Support Scale. The item pool for the Institutional Support Scale was reviewed by a panel of three experts in the field of lesbian and gay studies. These independent reviewers evaluated the items for face validity. Only those approved by all three reviewers were retained. Reliability estimates were calculated for the ISS. The Cronbach' s alpha formula was utilized to account for the Likert scale items on the instrument (Crocker & Algina, 1986). The coefficient alpha for the ISS for the sample of lesbian couples in this investigation was .862. This finding suggests that about 86% of the total score variance was from true score variance. This was also indicative that respondents' performance was relatively consistent across items and could generalize to similar items from related construct domains. In this study, the total ISS score was used to represent perceived institutional support for the relationship. The total ISS scores for both partners were averaged to form a couple score.

PAGE 150

143 Demographic Questionnaire A demographic questionnaire was used to collect information regarding individual and couple characteristics. The following areas were assessed by the demographic questionnaire: age, duration of lesbian identity, duration of couple relationship, length of cohabitation, race-ethnicity, education level, socioeconomic status, location of residence, household size, and spirituality.

PAGE 151

CHAPTER 4 DATA ANALYSIS AND RESULTS ^ Analysis Procedures The purpose of this study was to examine the influence of five variables on the relationship satisfaction among lesbian couples. Four of these variables focused on aspects of partner interaction: (a) the extent of relationship ritualization in couple life, (b) the meaningfulness of relationship ritualization in couple life, (c) the type of conflict resolution style utilized by couples, and (d) the agreement between partners regarding the type of conflict resolution style used. The fifth variable — perceived institutional support — focused on the couples' participation in the larger social context. • The sample for this study of lesbian couples included only couples who lived together (without children) and had been in a committed monogamous relationship for at least 1 year. Specifically, the relationship satisfaction of lesbian couples was assessed by items from Spanier's (1976) Dyadic Adjustment Scale. The DAS scores for each partner were used in the analyses of individual data. Averaging the scores for each partner created a composite couple score; these scores were used in the analyses of couple data. Four conflict resolution styles were measured using Kurdek's ( 1 994b) Conflict Resolution Style Inventory: positive problem solving, conflict engagement, withdrawal, and compliance. The CRSI provides information regarding one's own evaluation of one's conflict resolution styles (CRSl-Self) as well as information about one's evaluation of one's partner's conflict resolution styles (CRSl-Partner). The CRSI assessment of 144

PAGE 152

145 conflict resolution style calculated each partner's relative endorsement of strategies from each of four conflict resolution styles. Therefore, conflict resolution style was seen as a continuous rather than a categorical variable. This means that partners received scores on each conflict resolution style in accordance with the degree to which each style is used in resolving relationship disputes. ' The CRSI scores were used in three ways in this study: to evaluate the degree to which couples utilize each conflict resolution style, to determine the degree of agreement between partners' self-evaluations of their conflict resolution style, and to determine the degree of agreement between couple member's self-evaluations and partner-evaluations of conflict resolution style. Specifically, couple scores were calculated for each conflict resolution style by averaging both partoers' CRSI-Self scores for each conflict resolution style. That is, the self-evaluations ofboth partners were averaged for each conflict resolution style to determine the degree to which the couple utilized each style. This couple score was used in analyses of couple data regarding the degree to which each couple used each of the four conflict resolution styles. Scores from the CRSI-Self were used to determine the extent of agreement between partner's self-evaluations of conflict resolution style. That is, the CRSI-self scores for both partners were subtracted and the difference was used to determine the level of agreement between partners regarding each of the four conflict resolution styles. This score was used in the analyses of couple data. Scores from the CRSI-Self and the CRSI-Partner were used to make comparisons between couple members regarding both self and partner evaluations of one member of the couple. This score was computed from the difference between each couple member's

PAGE 153

146 self and partner evaluations for each of the four conflict resolution styles. For instance, comparisons were made between couple member 1 's self-evaluation of her own conflict resolution style and couple member 2's evaluation of her partner's conflict resolution style. This score was computed from the difference between each couple member's self and partner evaluations for each of the four conflict resolution styles. This score was used in the analyses of individual data. f^i; Seven ritual settings and eight ritual dimensions were assessed by Fiese and Kline's (1994) Family Ritual Questionnaire. The relationship ritualization score was the sum of scores across seven settings: dinnertime, weekend, vacation, annual celebrations, special celebrations, religious holidays, and cultural and ethnic traditions. This relationship ritualization score was used for the evaluation of individual data associations involving total relationship ritualization. A couple score was calculated by averaging both partners' relationship ritualization scores and was used in the analysis of couple data. The meaningfulness score was calculated as the sum of four ritual dimensions: occurrence, affect, symbolic significance, and deliberateness. This ritual meaningfulness score was used for the evaluation of individual data associations involving ritual meaningfulness. A couple score was calculated by averaging both partners' rinial meaningfulness scores and was used in the analysis of couple data. An instrument created for this investigation— the Institutional Support Scale— assessed perceived institutional sources of support for the relationship. The total ISS score was used in analyses of individual data. The total ISS scores for both partners were averaged to form a couple score that was used in analyses of couple data.

PAGE 154

k;' 147 The analysis of data for this investigation was accomplished through the use of the SAS General Linear Model (GLM). Two regression models were developed to address the five research hypotheses: in one model couple data were analyzed (Model C) while in the other model individual data were analyzed (Model I). The goal of the regression analysis was to determine the relationship between an outcome measure, or • criterion variable, and relevant independent measures. This analysis can be done globally by evaluating the entire model, or it can be done for each individual variable when the i.f,. effects for all other variables are held constant. In this investigation, a series of regression models were developed to evaluate each of the five hypotheses for strength of associations and interactions on the predictor variables. These models designated the level of relationship satisfaction by DAS score as the criterion variable and conflict resolution style, agreement between partners regarding conflict resolution style, institutional support, relationship ritualization, and ritual meaning as independent variables. Model C was used to evaluate couple data. This model was used to evaluate hypotheses 1-5. The output variable for Model C was couple relationship satisfaction as measured by the DAS couple composite score. The input variables were couple conflict resolution style, agreement between partners' self-evaluations of conflict resolution style, couple institutional support, couple relationship ritualization, and couple ritual meaning. Unlike Model I, which evaluated individual partner data. Model C evaluated couple scores (N = 90). Model I evaluated individual data. This model was used in posthoc analyses to further evaluate hypothesis 1 . In this model the output variable was relationship

PAGE 155

satisfaction as measured by the DAS score. The input variables were conflict resolution style, agreement between couple members' self and partner evaluations of conflict resolution style, institutional support, relationship ritualization, and ritual meaning. Because hypothesis 1 was concerned with the possible influence of agreement between partners regarding conflict resolution style, individual data were used in this analysis (N = 180). The regression coefficients provide information regarding the direction of the relationship between the dependent variable and each independent variable. A positive coefficient indicates that an increase in the independent variable results in an increase in the dependent variable. A negative coefficient indicates that an increase in the independent variable results in a decrease in the dependent variable. The absolute value of the regression coefficient provides information regarding the degree to which a change in the independent variable affects a change in the dependent variable. For the purposes of determining levels of statistical significance, a Type I error rate of .05 was established. A decision to accept or reject the specific null research hypothesis was based on this predetermined attained significance level. Source data were roimded to the nearest hundredth. Posthoc analyses were conducted to determine possible mediating effects. In these analyses the independent variables were degree of relationship ritualization, ritual meanmg, and institutional support while the dependent variable was conflict resolution style. Figure 1 describes the specific variables for the regression models in this investigation.

PAGE 156

149 Institutional Relationship Supports Positive Problem Solving I Withdrawal Conflict Engagement! — L _ Conflict Resolution Style Agreement nL ^t Compliance I I — yTouple Relationship Satisfaction)^ Positive Problem Solvingf Conflict Resolution Style Conflict Engagement 1 Withdrawal I Conflict Resolution Style L ^ — [Compliance Ritual Ritual Meaningfulness Relationship Ritualization Figure 1. Operationalized Model with Resuhs of Hypothesis Testing (solid lines denote significant direct effects at p<.05; dashed lines denote nonsignificance; dotted lines denote significant mediating effects at p<.05)

PAGE 157

Analysis Results Model C . This model evaluated hypotheses 1, 2, 3, 4, and 5 using couple data rather than individual data (as was used in Model I) because differences between couples were of interest. Input variables were each of the four couple conflict resolution styles (conflict engagement, positive problem solving, withdrawal, and compliance) as well as differences in partner self-assessments regarding each of these conflict resolution styles.,. Couple averages for total relationship ritualization, ritual meaningfulness, and institutional supports were included among the independent variables in the regression equation. The output variable was the DAS couple composite score for relationship satisfaction. For Model C the main effects equation was significant (F = 8.486, 2>F = .0001) with this model accounting for 55% (R^ = .5536) of the variance in the level of DAS couple composite score for relationship satisfaction. Table 7 shows the sources of variance in Model C using couple data. The goal of the regression analysis was to determine what, if any, relationships existed between and independent variable and the outcome variable when the effects for all of the other variables were controlled. In this model two variables contributed significantly to the couple composite relationship satisfaction scores on the DAS. The variables found to have attained significance were conflict engagement conflict resolution style, t = -4.085, p<.05, posifive problem solving conflict resolution style, t = 2.325, E<.05. None of the other variables were found to contribute significantly to the outcome measure. An examination of the regression coefficients gives information regarding the magnitude and direction of the relationship between the independent variable and the

PAGE 158

151 Table 7 Source Table for Model C CCouple Data) to Test the Main Effects with DAS as Dependent Variable Source df Coefficient Estimate Standard Error of Estimate t-valuc Conflict Resolution Style (couple average) Problem Solving 1 0.7033 .3025 2.325 .0227* . Conflict Engagement 1 -0.9762 .2390 -4.085 .0001* Withdrawal 1 -0.3455 .2263 -1.526 .1310 Compliance , , -0.0097 .2094 -0.046 .9631 Partner Difference in Conflict Resolution Style Self Evaluation Problem Solving 1 0.1046 .2516 0.416 .6789 Conflict Engagement 1 0.2460 .2081 1 182 9407 Withdrawal 1 0.0951 .1717 0.554 .5813 Compliance 1 -0.0946 .1779 -0.532 .5963 1 0.0283 .0539 0.525 .6012 Relationship Ritualization (couple average) . ^ Ritual Meaningfulness (couple average) 1 -0.0524 .0884 -V.jyZ .5553 Institutional Support (couple average) 1 0.0124 .0605 0.206 .8376

PAGE 159

' 152 outcome variable. The results in Table 7 indicate that scores on the CRSI conflict engagement style were negatively associated with the outcome measure, DAS couple composite score for relationship satisfaction. That is, for every 1 -point increase on the CRSI conflict engagement scale there was a resultant decrease of .98 of a point on the DAS. The relationship between CRSI positive problem solving and DAS score, however, was positive in direction and slightly smaller. Therefore, for every 1 -point increase on the CRSI positive problem solving scale there was a resultant increase of .70 of a point on the DAS. None of the other variables in the regression model were statistically significant. ' Posthoc analyses using couple data were conducted to determine whether relationship supports, ritual meaningfulness, or total relationship ritualization might have a mediating effect rather than a direct effect on couple relationship satisfaction as had been originally hypothesized. For these analyses, the independent variables were couple averages for relationship ritualization, ritual meaningfulness, and institutional support; the output variable was couple average for conflict resolution style. Thus, four separate regression equations were created each with the independent variables relationship ritualization, ritual meaning, and relationship supports but with different output variables corresponding to each of the four conflict resolution styles (conflict engagement, positive problem solving, withdrawal, and compliance). For the conflict engagement conflict resolution style and compliance conflict resolution style the equations were not significant. The model in which conflict engagemem conflict resolution style was the dependent variable revealed an F = .460 and E>F = .71 13 while the model in which compliance conflict resolution style was the

PAGE 160

dependent variable revealed an F = 1.850 and 2>F = .1442. Thus neither of these models were statistically significant at the £<.05 level. This analysis of couple data suggests that the independent variables relationship ritualization, ritual meaning, and relationship supports do not have a direct effect on either the conflict engagement conflict resolution style or the compliance conflict resolution style. For the positive problem solving conflict resolution style the equation was significant (F = 4.393, e>F = .0063) with this model accounting for 13% (R^ = .1329) of the variance in the positive problem solving conflict resolution style score. Table 8 shows the sources of variance in the model. In this model two variables contributed substantially to the positive problem solving conflict resolution style score on the CRSI. The variables found to have attained significance were ritual meaningfulness, t = 3.29, p<.05, and total relationship ritualization, t = -2.76, p<.05. The other variable, perceived institutional supports for the relationship, was not found to contribute significantly to the outcome variable. i ^ . ^ , , Table 8 Source Table for the Mode l to Test the Indirect Effects with Positive Problem Solving Conflict R esolution Stvle as Dependent Variable Source df Coefficient Standard t-value p-value Estimate Error of Estimate Relationship Support 1 0.0341 .0299 1.140 .2576 Ritual Meaningfulness 1 0.1436 .0436 3.293 .0014* ' Relationship Ritualization 1 -0.0746 .0269 -2.764 .0070*

PAGE 161

An examination of the regression coefficients gives information regarding the magnitude and direction of the relationship between the independent variable and the outcome variable. The results in Table 8 indicate that scores on the FRQ meaning scale were positively associated with the outcome measure, positive problem solving conflict • resolution style. That is, for every 1 -point increase on the FRQ meaning scale there was a resultant increase of .14 of a point on the CRSI positive problem solving scale. On the other hand, the regression results suggest that the total relationship ritualization scores of the FRQ were negatively associated with the outcome measure, positive problem solving conflict resolution style. That is, for every 1 -point increase on the FRQ total ritualization score there was a resultant decrease of .08 of a point on the CRSI positive problem solving scale. Thus, ritual meaningfulness and relationship ritualization had an indirect effect on couple relationship satisfaction. That is, ritual meaningfulness and relationship ritualization effect positive problem solving conflict resolution style which, in turn, effects couple relationship satisfaction. For the withdrawal conflict resolution style the equation was significant (F = 4.547, E>F = .0052) with this model accounting for 14% (R^ = .1396) of the variance in the withdrawal conflict resolution style score. Table 9 shows the sources of variance in the model. In this model all of the variables contributed substantially to the withdrawal conflict resolution style score on the CRSI. The variables found to have attained significance were ritual meaningfulness, t = -2.41, e<.05, total relationship ritualization, t = 2.05, E<.05, and relationship supports, t = -2.64, p<.05. The results in Table 9 indicate that scores on the FRQ meaning scale were negatively associated with the outcome measure, withdrawal conflict resolution style.

PAGE 162

155 Table 9 Source Table for the Model to Test the Indirect Effects with Withdrawal Conflict Resolution Style as Dependent Variable Source df Coefficient Standard t-value p-value Estimate Error of Estimate Relationship Support 1 -0.0910 .0345 -2.635 .0100* Ritual Meaningfulness 1 -0.1212 J.0504 -2.406 .0183* Relationship Ritualization 1 -0.0638 .0312 2.048 .0437* That is, for every 1 -point increase on the FRQ meaning scale there was a resultant decrease of .12 of a point on the CRSI withdrawal scale. On the other hand, the regression results suggest that the total relationship ritualization scores of the FRQ were positively associated with the outcome measure, withdrawal conflict resolution style. That is, for every 1 -point increase on the FRQ total relationship ritualization score there was a resultant increase of .06 of a point on the CRSI withdrawal score. Furthermore, the results in Table 9 indicate that scores on the ISS measure of relationship supports were negatively associated with the outcome measure, withdrawal conflict resolution style. That is, for every 1 -point increase on the ISS there was a resultant decrease of .09 of a point on the CRSI withdrawal scale. However, this did not result in an indirect effect on couple relationship satisfaction, since withdrawal conflict resolution style did not effect ' couple relationship satisfaction. ModelJ. This model further evaluated hypothesis 1 using individual data rather than couple data because differences between couple members' self and partner assessments of conflict resolution style were being considered. Input variables were each

PAGE 163

156 of the four conflict resolution styles (conflict engagement, positive problem solving, withdrawal, and compliance) as well as partner agreement regarding each of these conflict resolution styles. Additionally, total relationship ritualization, ritual meanmgfiilness, and institutional support were included among the independent variables in the regression equation. The output variable was the DAS score for relationship satisfaction. For Model I the main effects equation was significant (F = 10.215, 2>F = .0001) with this model accounting for 40% (R^ = .4008) of the variance in the level of relationship satisfaction. Table 10 shows the sources of variance in Model I — the analysis of individual data. The goal of the regression analysis was to determine what, if any, relationships existed between and independent variable and the outcome variable when the effects for all of the other variables were controlled. In this model two variables contributed substantially to the relationship satisfaction scores on the DAS. The variables found to have attained significance were conflict engagement conflict resolution style, t = -5.078, E<.05, positive problem solving conflict resolution style, t = 2.853, p<.05. None of the other variables— including partner agreement regarding conflict resolution style— were found to contribute significantly to the outcome measure, relationship satisfaction. I ' An examination of the regression coefficients gives information regarding the magnitude and direction of the relationship between the independent variable and the outcome variable. The results in Table 10 indicate that scores on the CRSI conflict engagement style were negatively associated with the outcome measure, DAS relationship satisfaction score. That is, for every 1 -point increase on the CRSI conflict engagement scale there was a resultant decrease of .76 of a point on the DAS. The relationship between CRSI

PAGE 164

157 Table 10 Source Table for Model I (Individual Data) to Test the Main Effects with DAS as Dependent Variable Source df Coefficient Estimate Standard Error of Estimate t-value p-value Conflict Resolution Style Problem Solving 1 0.5598 .1962 2.853 .0049* Conflict Engagement 1 -0.7646 .1506 -5.078 .0001* ^v. Withdrawal 1 -0.0731 .1492 -0.490 .6249 Compliance 1 0.0604 .1196 0.504 .6146 Partner Self-Assessment Agreement about Conflict Resolution Style Problem Solving 1 -0.1497 .1257 -1.191 .2352 Conflict Engagement 1 0.2439 .1270 1.921 .0564 Withdrawal 1 -0.1343 .1236 -1.087 .( .2786 Compliance ^ 1 -0.1184 .1073 -1.104 .2712 Relationship Ritualization 1 0.0709 .0386 1.835 .0683 Ritual Meaningfulness 1 -0.0856 .0627 -1.367 .1736 Institutional Support 1 -0.0042 .0250 -0.167 .8677 E<.05

PAGE 165

positive problem solving and DAS score, however, was positive in direction and slightly smaller. Therefore, for every 1 -point increase on the CRSI positive problem solving scale there was a resultant increase of .56 of a point on the DAS. Hvpothesis Testing , . Five hypotheses were evaluated to test the theoretical assumptions of this research. Two linear regression models were developed. Model C tested hypotheses 1, 2, 3, 4, and 5 (using couple data) for statistical significance and Model 1 tested hypothesis 1 (using individual data) for statistical significance. The results for each hypothesis are described in the following paragraphs and summarized in Table 11. Hypothesis 1 stated there is no significant association between the level of relationship satisfaction reported by lesbian couples and the degree of partner agreement regarding the conflict resolution style utilized. This hypothesis was initially tested by Model C using couple data and later tested by Model I using individual data. The results of regression Model C did not demonstrate a statistically significant association between partner self-evaluation agreement regarding conflict resolution style and couple scores on the DAS measure of relationship satisfaction. Therefore, no statistical evidence existed to reject the null hypothesis. This model was further tested using Model I in order to evaluate differences between partner-evaluations and self-evaluations for each member of the couple. The results of regression Model I did not demonstrate a statistically significant association between partner agreement regarding conflict resolution style and scores on the DAS measure of relationship satisfaction. v ^ ;

PAGE 166

159 Table 11 Results of Hypothesis Testing Number Hypothesis Decision Hi There is no significant association between the level of relationship satisfaction reported by lesbian couples and the degree of partner agreement regarding the conflict resolution style utilized. Fail to Reject There is no significant association between the level of ' relationship satisfaction reported by lesbian couples and the type of conflict resolution utilized by each partner. Reject There is no significant association between the level of relationship satisfaction degree of relationship ritualization Fail to Reject H4 There is no significant association between the level of relationship satisfaction and the extent of ritual meaningfiilness reported by lesbian couples. Fail to Reject There is no significant association between the level of relationship satisfaction and the degree of perceived institutional support reported by lesbian couples. Fail to Reject Hypothesis 2 asserted there is no significant association between the level of relationship satisfaction and the type of conflict resolution style utilized by each partner. The results of the regression Model C demonstrated a statistically significant association between both the confiict engagement style of conflict resolution (t value = -4.09, 2<.05) :^ and the positive problem solving conflict resolution style (t value = 2.33, p<.05) and couple composite scores on the DAS measure of relationship satisfaction. For the conflict engagement style of conflict resolution the regression estimate (-0.9762) suggested an inverse relationship existed such that for every 1 -point increase on the CRSI a resultant decrease of .98 of a point on the DAS can be expected. Additionally, for the

PAGE 167

160 positive problem solving style of conflict resolution, the regression estimate (0.7033)suggested a positive relationship existed such that for every 1 -point increase on the CRSI a resultant increase of .70 of a point on the DAS can be expected. Data from the study supported the rejection of null hypothesis 2. ' Hypothesis 3 stated there is no significant association between the level of relationship satisfaction and the degree of relationship ritualization reported by lesbian couples. Based on the results of regression Model C, no significant association on the outcome variable was demonstrated in this sample. Therefore, no statistical evidence existed to reject the null hypothesis. Posthoc analyses demonstrated a statistically significant association between positive problem solving conflict resolution style and relationship ritualization (t value = -2.76, E<.05). For relationship ritualization the regression estimate (-0.0746) suggested a negative relationship existed such that for every 1 -point increase on the FRQ a resultant decrease of .08 of a point on the can be expected on the CRSI positive problem solving subscale. . " Hypothesis 4 asserted there is no significant association between the level of relationship satisfaction and the extent of ritual meaningfulness reported by lesbian couples. Based on the resuhs of regression Model C, no significant differences on the outcome variable were determined. Therefore, no statistical evidence existed to reject the null hypothesis. , ^. Posthoc analyses demonstrated a statistically significant association between the positive problem solving style of conflict resolution and ritual meaningfulness (t value = 3.29, p<.05). For ritual meaning the regression estimate (0. 1463) suggested a positive

PAGE 168

relationship existed such that for every 1 -point increase on the FRQ meaning factor a resultant increase of .14 of a point on the can be expected on the CRSI positive problem solving subscale. Similarly, posthoc analyses demonstrated a statistically significant association between withdrawal conflict resolution style and ritual meaningfulness (t value = -2.41, E<.05). For ritual meaning the regression estimate (-0.1212) suggested a negative relationship existed such that for every 1 -point increase on the FRQ meaning factor a resultant decrease of . 1 2 of a point on the can be expected on the CRSI withdrawal subscale. Thus, ritual meaningfulness appears to be a mediating variable. Hypothesis 5 stated there is no significant association between the level of relationship satisfaction and the degree of perceived institutional support reported by lesbian couples. Based on the results of regression Model C, no significant differences on the outcome variable were determined. Therefore, no statistical evidence existed to reject the null hypothesis. Posthoc analyses demonstrated a statistically significant association between withdrawal conflict resolution style and institutional support (t value = -2.64, p<.05). For institutional support the regression estimate (-0.0910) suggested a negative relationship existed such that for every 1 -point increase on the ISS a resultant decrease of .09 of a point on the can be expected on the CRSI withdrawal subscale. Thus, institutional support appears to be a mediating variable. Figure 2 summarizes the statistically significant finding from this study. Shown are the direct associations between couple relationship satisfaction and positive problem solving conflict resolution style as well as conflict engagement conflict resolution style.

PAGE 169

162 Additionally, the indirect associations between couple relationship satisfaction and ritual meaning and relationship ritualization are shown. r L Conflict En garment Style 1 1 Relationship Ritualization r 1 — — — — — — — > Relationship Satisfaction Ritua Positive Problem Solving Stylef J Meaning Figure 2. Diagram of Statistically Significant Effects. (Solid lines indicate direct effects; dashed lines indicate indirect effects). Summary This chapter presented discussion of the procedures for the analysis and the i, resuhs of this research. The outcome testing to accept or reject the study's five null research hypotheses was examined. Statistical evidence derived from the analysis of data supported the rejection of hypotheses 1 and 2. Although the null hypotheses 3, 4, and 5 were not rejected, further analyses of each suggested that the variables ritual meaningfuhiess, degree of relationship ritualization, and institutional support had mediating effects on the outcome measure rather than a direct effect as originally ' hypothesized.

PAGE 170

CHAPTERS DISCUSSION Overview of the Study The purpose of this study was to assess the influences of five variables on the relationship satisfaction reported by lesbian partners. The five variables examined were as follows: (a) the perceived conflict resolution style utilized by each partner, (b) the degree of agreement between partners in their assessments of their own and their partners' conflict resolution style, (c) the degree of relationship ritualization, (d) the level of ritual meaningflilness, and (e) the extent of perceived institutional support for the relationship. The theoretical models upon which this study was based addressed the contributions of internal couple dynamics and the external dynamics of the larger cultural context in which the couple is embedded. No existing theory integrates both perspectives. Consequently, this study was based upon two theoretical traditions: Gottman's (1994) balance theory and Kurdek's relationship appraisal model. Addhionally, this study is based upon Brown's (1989a) conceptual framework regarding the socio-political context in which lesbian couples live. Two theoretical premises underlie Gottman's (1994) balance theory. First, satisfied couples develop a ratio of positive to negative affective interactions as well as a ratio of positive to negative problem solving interactions. The second theoretical premise underlying balance theory is the notion that partners in satisfied couples develop a "fit" 163

PAGE 171

regarding preferred interactional style. Gottman (1994) postulated three types of "stable" couples: volatile, avoiding, and validating. Each of these couple types results in a "stable," satisfied relationship in which positive and negative interactions are balanced. Gottman' s model is based on observations of heterosexual couples—couples that benefit from the social sanctions given their relationships by the larger cultural context. Members of marginalized and stigmatized groups, such as lesbians, often do not receive the institutional support for their relationships afforded heterosexual couples. Brown (1989a) described three aspects of lesbian couples' cultural context that cut across all demographic characteristics: marginality, biculturalism, and normative creativity. Building on Brown's contextual model and Gottman' s balance conceptualization, Kurdek ( 1 998) hypothesized a model of relationship satisfaction applicable to both heterosexual married and homosexual cohabiting couples. Kurdek (1998) posited that relationship satisfaction was related to both internal forces within the relationship-such as shared agreements about problem solving, autonomy, equality, and intimacy-and external forces outside the relationship-such as institutional supports. Research Sample The research instrument was distributed to 175 lesbian couples. Of the 175 packets distributed, a total of 90 (51%) completed and useable surveys were returned by the February 1, 1999, deadline. Because lesbians do not enjoy social sanction for their relationships, identifying potential research participants can be difficult. Consequently, network sampling techniques were utilized. Therefore, the sample of lesbian couples in this study was not representative of the general population of lesbian couples in the United States.

PAGE 172

165 i ;" Efforts were made to solicit a diverse group of women representing various ages, classes, education levels, races, ethnicities, spiritual-religious affiliations, ages, geographic locations and community sizes by contacting lesbian organizations and individuals affiliated with lesbians from nondominant groups. In regard to age, geographic location, educational backgroimd, and spiritual-religious contexts, these efforts were relatively successful. Unfortunately, efforts to solicit the participation of women in nondominant class, racial, and ethnic groups were not as successful. The sample had a large age distribution (50 years) and included women from 16 states who lived in communities of various sizes. A significant portion (44%) of the sample lived in rural areas (28%) or metropolitan areas (16%). Over 34% of the women in the sample did not have a college degree, and 27% had annual incomes of less than $20,000. The spiritual affiliations of a majority of the women (55%) were nontraditional including Pagan, Buddhist, Goddess-based, and nature-oriented practices. The sample was predominantly White (90%); Latina women comprised 6% of the sample. The women in the sample represented a variety of lesbian identity development stages. The length of time the women in the sample had been self-defined as a lesbian spanned 50 years. Similarly, the relationship length of couples represented a variety of relationship development stages and spanned 33 years. ; ^ ' Association Between Co uple Relationship Satisfaction and Partner Agreement Regarding Conflict Resolution Stvle Utilized Hypothesis 1 stated there is no significant association between the level of relationship satisfaction reported by lesbian couples and the degree of partner agreement regarding the conflict resolution style utilized. This hypothesis was tested in two ways. First, using couple data, the association between couple relationship satisfaction and

PAGE 173

166 degree of agreement regarding partners' self-evaluations of conflict resolution style was tested using Model C. The results of this research did not support the rejection of this hypothesis. Second, using individual data, the association between partner relationship satisfaction and degree of agreement between each person 1 self-evaluation and person 2 partner-evaluation was tested using Model I. The results of this research did not prove to be statistically significant. Therefore, regardless of the model used (couple versus individual), there was no statistical evidence to support any significant association between relationship satisfaction and partner agreement regarding conflict resolution style-whether measured as parmer self-evaluation agreement or agreement between selfand parmer evaluations for each member of the couple. These results are inconsistent with Gottman's (1994) balance theory. According to this theory, partners in satisfied couples develop a "fif regarding preferred interaction style. He characterized three types of satisfied, "stable" couples: validating, volatile, and avoiding. Although these three couple types interact in very different ways, each form may lead to a satisfying and stable relationship if the partners find a mutually acceptable style of interaction. Gottman (1994) also noted that couples must be able to balance the negative interactions and affect in their relationships with significant amounts of positive interaction and affect. This point is discussed in a later section regarding the relationship between couple satisfaction and type of conflict resolution style. It would seem from this theory that agreement between partner's regarding each of the styles would significantly influence relationship satisfaction, yet in this sample this was not true for any of the conflict engagement styles.

PAGE 174

An analysis of participants' relative endorsement of the four conflict resolution styles revealed that a majority (90%) of the partners in the sample identified primarily with the positive problem solving style followed by the withdrawal style (5%), the conflict engagement style (1%), and the compliance style (1%). In addition, there were several partners (3%) with undifferentiated profiles; that is, they utilized two or more conflict styles equally rather than predominantly relying on one style. This trend of frequent endorsement of the positive problem-solving style is consistent with--but much larger than--what was predicted in evaluation studies of the CRSI. Kurdek ( 1 998) acknowledged the probability of a response bias regarding the positive problem-solving conflict resolution style. Positive problem solving~a socially desirable conflict resolution style-may be endorsed with greater frequency by individuals evaluating their ovra (or their partner's) conflict resolution style (Kurdek, 1 998). In this sample conflict resolution style was treated as a continuous rather than a categorical variable. Consequently, the design did not incorporate a strategy to equalize the number of participants by conflict resolution style. It is likely that the underrepresentation of partners in the sample who endorsed as their primary style of conflict resolution either the conflict engagement style, the withdrawal style, or the compliance style contributed to the finding that agreement regarding these styles did not significantly influence relationship satisfaction levels. However, it is surprising given the relative prevalence of partners in the sample endorsing the positive problem solving style of conflict resolution that no statistically significant association between conflict resolution style agreement and relationship satisfaction was found. One possible explanation for this finding might be that the spread of scores between partners regarding the degree to

PAGE 175

168 which positive problem-solving style was endorsed was not very large. In fact, analysis of frequency data revealed that a majority of partners highly endorsed the items on this scale. In fact, 64% of partner Is and 63% of partner 2s scored within 4 points of each other on the positive problem-solving style scale. This suggests that the data were . . skewed in the direction of higher endorsement of the positive problem-solving style items and were not well distributed within the scale. Additionally, because the score distribution for relationship satisfaction was skewed in the direction of increased satisfaction and the range was restricted, the variability of the outcome measure was restricted. Association Between Couple Relationship Satisfaction and Tvpe of Conflict Resolution Style Utilized Hypothesis 2 asserted there is no significant association between the level of relationship satisfaction and the type of conflict resolution style utilized by each partner. The results of this research supported the rejection of this hypothesis. That is, there was statistical evidence to support the notion that the couple's averaged self-evaluations of conflict resolution style significantly influenced couple relationship satisfaction scores. Specifically, both conflict engagement style and positive problem-solving style of conflict resolution significantly influenced couple relationship satisfaction. In the case of conflict engagement style of conflict resolution a statistically significant (p = .0001) inverse relationship between CRSI scores and DAS relationship satisfaction couple scores was substantiated by the regression analysis. On the other hand, in the case of positive problem-solving style of conflict resolution a statistically significant (p = .0227) positive relationship between CRSI scores and DAS relationship satisfaction couple scores was substantiated by the regression analysis. This means that the greater the

PAGE 176

degree of endorsement by both partners of conflict engagement style strategies of conflict resolution on the CRSI, the lower the DAS scores for couple satisfaction. The opposite was true for positive problem-solving style; the greater the degree of endorsement by both partners of positive problem-solving style strategies of conflict resolution on the CRSI, the higher the DAS score for couple satisfaction. The finding that relationship satisfaction reports were related to conflict resolution style was consistent with past research with samples of lesbian couples (e.g., Kurdek, 1998). These results offer support for Kurdek's (1998) relationship appraisal model as well as Gottman's (1994) balance theory of couple interaction. Couples who utilized positive problem-solving strategies such as "sitting down and discussing differences" and "finding ahematives that are acceptable to both partners" also reported higher levels of relationship satisfaction. Like Gottman's (1994) validating couple type, the satisfied partners who rely on the positive problem-solving style essentially find ways to balance negative interactions and affect with a large measure of positive interactions and affect. Gottman ( 1 994) also characterized a volatile couple type in which partners engaged in high levels of argument and conflict. These couples reported relatively high levels of relationship satisfaction only if they were successful at finding ways to balance the extreme negative interactions and affect with equally extreme and more frequent amounts of positive interactions and affect. It could be expected that the couples in this sample who relied on conflict engagement strategies of conflict resolution such as "launching personal attacks" and "throwing insults and digs" would report lower levels of couple relationship satisfaction-particularly if they did not find opportunities in other aspects of their relationships to balance out such negative interactions.

PAGE 177

170 Two of the conflict resolution styles assessed by the CRSI~withdrawal and compliance—did not influence the couple relationship satisfaction scores of this sample in a statistically significant manner. That is, the relationship satisfaction levels reported by couples who utilized such strategies as "remaining silent for long periods of time" or "tuning the other person out"--characteristic of the withdrawal conflict resolution stylewere not influenced by this style of settling disputes. Likewise, the relationship satisfaction levels reported by couples who utilized such strategies as "not defending my position" or giving in with little attempt to present my side of the issue"--characteristic of the compliance style of conflict resolution--also were not influenced by this manner of resolving conflicts. Gottman's (1994) avoiding couple type parallels the compliance and withdrawal conflict resolution styles. According to balance theory, avoiding type couples balance a low level of negative interaction and affect with a low level of positive interaction and affect. That is, they "pay" the price of calm with emotional distance. Yet, according to Gottman's (1994) theory, these couples may form stable and satisfying relationships. This theory predicts that satisfied partners who endorse such avoidancetype conflict resolution strategies must be able to develop a mutual "fit" with each other regarding this conflict resolution style and to develop methods to balance negative interactions and affect with positive interactions. As previously discussed, it is possible that the limited extent to which the women in this sample endorsed either withdrawal or compliance conflict resolution styles did not result in a very large distribution of scores across which associations could be assessed. An item analysis of participants' relative endorsement of the four conflict resolution styles revealed that a majority of the couples in the sample endorsed strategies consistent with the positive problem-solving style to a

PAGE 178

171 greater degree than they endorsed strategies of any other conflict resolution style. Furthermore, most of the couples in the sample were relatively satisfied with their relationships (mean DAS = 40 on a 50-point scale). Association Between Couple Relationship Satisfaction and Degree of Relationship Ritualization Hypothesis 3 stated there is no significant association between the level of relationship satisfaction and the degree of relationship ritualization reported by lesbian couples. The results of this research did not support the rejection of this hypothesis. That is, there was no statistical evidence to support the notion that relationship ritualization significantly influenced couple relationship satisfaction. However, later discussion shows the indirect effects of relationship ritualization as a predictor of couple relationship satisfaction after controlling for conflict resolution style. One explanation for this finding is that it supports prior research regarding the lack of marker events-such as weddings or the birth of a child--and the lack of rituals associated with these important occasions for lesbian couples (Carter & McGoldrick, 1989; Imber-Black & Roberts, 1992; Roth, 1985; Slater, 1995). Additionally, life cycle transitions such as retirement, serious illness, or death can be problematic as disqualification of the couple relationship by the larger society may lead to public ceremonies in which partners have to disguise their feelings and their relationship, thus intensifying their problems rather than promoting a sense of connectedness and continuity. Yet, in heterosexual contexts, rituals have been shown to foster a sense of identity and belonging for family members (Bennett et al., 1 992). An alternative explanation for the lack of association between relationship ritualization and couple relationship satisfaction may be found by examining the

PAGE 179

172 developmental state of the literature regarding ritual in couples and families. Much of the literature regarding family ritual has focused on the types and ftmctions of rituals in heterosexual family systems rather than on couple systems (Bennett et al., 1988; Bennett et al., 1987; Bossard & Boll, 1950; Boyce et al., 1983; Fiese, 1992; Fiese «& Kline, 1993; Imber-Black & Roberts, 1992; Kline, 1992; Mccubbin & McCubbin, 1987; Meredith, 1985; Meredith et al., 1989; Mize, 1988; Myerhoff, 1977; Reiss, 1981; Reiss & Oliveri, 1991; Roberts, 1988; Schvaneveldt & Lee, 1983; Wolin & Bennett, 1984; Wolin et al., 1980). Furthermore, a significant portion of this body of research has been done using qualitative methods. Only recently have instruments been available to assess ritual use and functions (Fiese & Kline, 1993). The latest research has begun to focus on the salience of rituals for family members (Fiese, 1992; Fiese & Kline, 1993; Fiese et al., 1993; January, 1996). Researchers have begun in recent years to include in their investigations the importance of rituals in the Ufe of marital partners or friendship dyads (Bruess, 1994; January, 1996). The empirical literature regarding family rituals is limited, and research regarding interpersonal rituals among couples is sparse; however, the empirical research regarding rituals in lesbian couples is nonexistent-although anecdotal accounts have supported the notion that ritual plays an important role in lesbian couple life (Laird, 1988; Slater, 1995). While it was important to use a standardized instrument for the evaluation of relationship ritualization and ritual meaning in this study of lesbian couples' lives, the selection of the FRQ presented some difficulties which were not initially apparent. As a result, it is possible that the FRQ did not validly assess rituals for this sample of lesbians. It is conceivable that the ritual lives of lesbian couples is significantly different from that

PAGE 180

of heterosexual couples~on which the FRQ was based. The FRQ, the instrument used to assess ritual settings and dimensions in this investigation, has never before been used with a sample of lesbians. Although the directions were modified to instruct couples to consider themselves a special case of "family" when encountering this word in each item, this abstraction may have presented difficulty for the women in the sample. Furthermore, it was expected that lesbian participants in this sample would be able to extrapolate from the examples of rituals given for each setting to include other rituals that might be more specific to their own experience. For instance, the directions for the Cultural and Ethnic Traditions setting of the FRQ read "Think of some cultural and ethnic traditions Some examples may be baptisms, naming ceremonies, bar mitzvahs, baking of particular ethnic foods, wakes, and funerals." Of course, the lesbian couples m this sample would not be attending baptisms because couples with children were not selected for the study. Additionally, some of the other Usted examples may have been perceived by certain couples as more available to them than other couples would have perceived the same example. For instance, lesbians that were relatively "out" to family and friends may have felt that any of the rituals listed would be open to them. However, relatively "closeted" couples may have felt that none of the rituals provided as examples were applicable to them. It was assumed that the lesbian couples in the sample might also feel free to think of additional rituals that were not listed such as Pride marches, croning ceremonies, or other cultural traditions specific to their lives as lesbians. Unsolicited feedback from participants in the study revealed that many of the participants found the FRQ heterosexist conceptually and linguistically. That is, they seemed to view parts of the FRQ as culturally biased much as a Chinese person might

PAGE 181

174 view a survey that asked about religious rites but included only items pertaining to Christianity and excluded items pertaining to Buddhism. Therefore, it is possible that the language of the FRQ interfered with the accurate assessment of rituals in this sample of lesbian couples. Lesbians may enjoy tremendous freedom to invent their relationships and roles in . a creative maimer (Brown, 1989b; Grahn, 1984). Lesbian couples-many of whom have not been afforded the traditional rituals afforded their heterosexual counterparts-may utilize what Brown (1989b) terms normative creativity to devise rituals of their own. These rituals certainly would not appear as examples on an instrument such as the FRQ, and it may not have occurred to the women in the sample to generalize from what was written to include these unique rituals in their thmking when responding to the FRQ. Conflict resolution stvle and relationship ritualization . As previously reported in this chapter, two of the conflict resolution styles were significant predictors of couple satisfaction score in this study of lesbian couples. The conflict engagement style of conflict resolution was negatively associated with couple relationship satisfaction while positive problem solvmg style of conflict resolution was positively associated with couple relationship satisfaction. Posthoc analyses were conducted to determine possible indirect effects of relationship ritualization on couple relationship satisfaction by examining the association of relationship ritualization and conflict resolution style. The results of these analyses revealed that couple relationship ritualization had a statistically significant negative effect on positive problem solving conflict resolution style (p = .0070). This means that as ' couple relationship ritualization scores increased the degree to which both partaers

PAGE 182

endorsed positive problem solving conflict resolution strategies decreased. In turn, as positive problem solving conflict resolution style scores decreased so did DAS couple scores for relationship satisfaction. Therefore, relationship ritualization had an indirect, negative effect on couple relationship satisfaction scores. These findings suggest that relationship ritualization as a whole indirectly impacted couple relationship satisfaction negatively for this sample of lesbians. Ritual meaningfulness may hold the key to understanding this outcome. The FRQ assessed eight dimensions of rituals: occurrence, affect, symbolic significance, deliberateness roles, routines, continuity, and attendance. Of these, the first four dimensions assess ritual meaningfiihiess. As is discussed in the following section, ritual meaningfiilness had a significant indkect effect on couple relationship satisfaction when conflict resolution style was controlled. It is possible that the four ritual dimensions that comprise relationship ritualization and that are not included in ritual meaningfulness account for the negative association found in this study. This would support research that suggests that relatively empty rituals contribute nothing positive--and possibly something negative--to couple and family relafionships (Fiese, 1992; Fiese & Kline, 1993; Imber-Black & Roberts, ' ' 1992; Myerhoff, 1977; Roberts, 1988; Wise. 1986). Association Between C ouple Relationship Satisfaction and Ritual Meaningfulness Hypothesis 4 asserted there is no significant association between the level of relationship satisfaction and the extent of ritual meaningfulness reported by lesbian couples. The results of this research did not support the rejection of this hypothesis. That is, there was no statistical evidence to support the notion that ritual meaningfiilness

PAGE 183

significantly influenced couple relationship satisfaction. However, posthoc analyses revealed an indirect positive association between ritual meaningfiilness and couple relationship satisfaction when conflict resolution style was controlled. The findings of this research are inconsistent with prior research in which the meaning of couple and family rituals has been shovra to be related to marital satisfaction in heterosexual couples (Fiese, 1992; Fiese et.al., 1993; Fiese & Kline, 1993). The concerns regarding the accuracy with which the FRQ assessed the ritual lives of lesbian couples that were detailed in the prior section may also explain the lack of a , statistically significant relationship between ritual meaningfiilness and couple relationship satisfaction. Conflict resolution style and ritual meaning . As previously reported in this chapter, two of the conflict resolution styles were significant predictors of couple satisfaction score in this study of lesbian couples. The conflict engagement style of conflict resolution was negatively associated with relationship satisfaction while positive problem-solving style of conflict resolution was positively associated with relationship satisfaction. Posthoc analyses were conducted to determine possible indirect effects of ritual meaningfiilness on relationship satisfaction by examining the association of ritual meaningfiilness and conflict resolution style. The results of these analyses revealed that ritual meaningfulness had a statistically significant positive effect on positive problemsolving conflict resolution style (p = .0014). This means that as ritual meaningfulness scores increased, so did the degree to which both partners endorsed positive problemsolving conflict resolution strategies. In turn, as positive problem-solving conflict

PAGE 184

resolution style scores increased, so did couple scores on the DAS measure of relationship satisfaction. Therefore, ritual meaningfulness appeared to have an indirect, positive effect on couple relationship satisfaction scores. This finding is consistent with prior research regarding the importance of meaningful-rather than hollow-rituals in couple and family relationships (Fiese, 1992; Fiese & Kline, 1993; Imber-Black & Roberts, 1992; Myerhoff, 1977; Roberts, 1988; Wise, 1986). Meaningful rituals offer couples and opportunity for positive interaction and shared positive affect. The indirect association between ritual meaningfulness and coupl relationship satisfaction offers support for Gottman's (1994) balance theory. Specifically, the findings of this study support the notion that increased meaningfulness of rituals~a positive interaction for couples--is associated with increased positive problem solving which, m turn, is associated with increased couple relationship satisfaction. ' Association Between Couple Relationship Satisfaction and Institutional Support for the Relationship Hypothesis 5 stated there is no significant association between the level of relationship satisfaction and the degree of perceived institutional support reported by lesbian couples. The results of this research did not support the rejection of this . hypothesis. That is, there was no statistical evidence to support the notion that couple's perceptions of institutional supports for the relationship significantly influenced couple relationship satisfaction. These findings are mconsistent with prior related research regarding support for the lesbian, gay, or bisexual individual (Albro & Tully, 1979; Aura, 1985; Chafetz et al., 1974; Grossman & Kemer, 1998; Jordan, 1995; Jordan & Deluty, 1998; Kurdek, 1988a;

PAGE 185

178 Kurdek & Schmitt, 1987a, 1987b; Levy, 1992; Meyer, 1989; Olson, 1988; Walsh, 1995; Wayment & Peplau, 1995; Wood, 1983). Much of this research supports the notion that support for lesbian individuals contributes to well-being. The focus of this investigation was on the role that support for the couple-rather than for the individual-played in the lives of lesbian women. Building on Brown's contextual model and Gottman's balance conceptualization, Kurdek (1 998) hypothesized a model of relationship satisfaction applicable to both heterosexual married and homosexual cohabiting couples. Kurdek (1998) posited that relationship satisfaction was related to both internal forces within the relationship-such as shared agreements about problem solving, autonomy, equality, and intimacy-and external forces outside the relationship-such as institutional supports. The degree of comfort which the women in this sample felt regarding their lesbian identities and the degree to which they were "out" may offer some explanation regarding the outcome of the study. The mean for the length of time that the women in this sample had considered themselves lesbian was nearly 1 8 years. It is likely, therefore, that these women were relatively comfortable with their lesbian identities and quite possible that they were relatively "out of the closet." Additionally, the very fact of their decision to participate in the survey lends credence to this notion. The mean score on the ISS measure of institutional support developed for this investigation was 93 -this was 12 units above the midpoint of the scale. This suggests that, as a group, the women in this sample felt that their relationships were more supported than not. It is possible that the relative homogeneity of the sample with regard to support for their relationships may have tempered the potential influence on relationship satisfaction.

PAGE 186

In addition, the instrument developed to assess institutional support for the relationship may have been impaired in its ability to assess reliably and validly the construct of institutional support. Although the Chronbach's alpha estimate of reliability for the total scale w^as adequate (.86), the reliability of items mtended to assess institutional supports rather than social and relational supports was poor (.66). Conflict resolution style and institutional support . Posthoc analyses were conducted to determine possible associations between institutional supports for the relationship and conflict resolution style. The results of these analyses revealed that institutional support had a statistically significant positive effect on withdrawal conflict resolution style (p = .0100). This means that as institutional support scores increased so did withdrawal conflict resolution style scores. However, withdrawal conflict resolution style was not significantly associated wdth couple relationship satisfaction; therefore, no indirect effect between institutional support for the relationship and couple relationship satisfaction was supported. Recommendations Implications for Therapy , Although the women in this mvestigation constituted what amounted to a nonclinical sample, the results of this study point out several factors to consider when working with lesbian couples in therapy. Conflict engagement style of conflict resolution was shown to be negatively associated with relationship satisfaction in lesbian couples when comparisons between couple composite scores were used. On the other hand, agreement between partners regarding the use of the conflict engagement style of conflict resolution was found to be positively associated with relationship satisfaction when

PAGE 187

180 individual scores were used. While these results regarding the associations between conflict engagement style of conflict resolution and relationship satisfaction may at first seem contradictory, when viewed in light of Gottman's (1994) balance theory, they make sense. It seems that couples' use of the conflict engagement style of conflict resolution in and of itself does not contribute to relationship satisfaction. In fact, the negative association between conflict engagement style and relationship satisfaction suggests that the opposite is true. However, for couples in which partners mutually agree to utilize this style of conflict resolution, a positive association with relationship satisfaction was y found. This finding suggests that rather than using therapy to attempt to alter the conflict engagement strategies of such couples, counselors working with couples in which conflictual interactions are prevalent may be able to assist partners in determining whether this style is compatible for both partners. If the conflict engagement style is mutually acceptable to both partners, therapists may then assist clients in finding ways to balance the potentially intense negative interactions and affect inherent in this style with opportunities for equally intense positive interactions and affect--such as meaningful rituals. Additionally, because positive problem-solving conflict resolution style was positively related to relationship satisfaction, it could be important for therapists to assist clients in developing constructive problem resolution skills. Additionally, both ritual meaning and relationship ritualization were shown to be positively associated with the positive problem-solving conflict resolution style. This means that ritual meaning and relationship ritualization had an indirect effect on relationship satisfaction in that they both influenced the positive problem-solving conflict style which, in turn, was directly

PAGE 188

: " 181 related to couple relationship satisfaction. For this reason, it could be important to assist lesbian clients in assessing, developing, and implementing meaningful rituals in their relationships. , ^ . ^ This study was designed to investigate how a nonclinical sample of lesbian couples describe their relationship functioning. The results of this investigation suggest possible preventative interventions. For instance, preventive interventions could be directed toward helping lesbian couples to enhance their relationships through the use of constructive problem-solving skill training. Furthermore, educational efforts could focus on helping lesbian couples identify the role of ritual in their relationships. To this end, lesbian couples could be assisted in determining the best ways to enhance the rituals in which they already participate and to create new and meaningful rituals together. Limitations of the Study There were a number of limitations inherent in this study. These limitations include concerns regarding data analytic techniques, sample selection, instrumentation, and the operationalization of variables. One of the primary concerns the data analytic techniques. This study relied heavily on correlational analyses. Correlational analyses do not allow one to make inferences about causal relationships. In addition, all intervening variables cannot be controlled. It is impossible to say that any of the variables assessed caused or created the level of relationship satisfaction reported by the couples in this study. Rather, it can be said that a relationship exists between that variable and the relationship satisfaction reported. Therefore, any inferences made regarding the direction of the relationships among the variables measured in this investigation were purely speculative.

PAGE 189

182 As is almost always the case with measurement, the bias of the use of self-report assessment tools by themselves is called into question (McCubbin & Thompson, 1991). It is conceivable that couples observations of their own behavior reflect, in part, a socially desirable response set. There is no way to know whether couples' observations of their • own behavior provide an accurate picture of their dynamics nor whether an outside observer could make a more accurate assessment. The use of the CRSI self and partner assessments may have minimized this limitation in regard to the conflict resolution style aspect of this study. However, the retrospective nature of the instruments required that participants reflect upon and generalize from events and interactions across time. This reflection calls into question the accuracy of the participant's memory, the influence of social desirability, and the potential effects of intervening variables and changing contexts. Additionally, the representativeness of the sample is limited both by the network sampling techniques utilized and the volunteer nature of participation in the study. It was difficult to obtain a truly representative sample of lesbian women for a number of reasons. First, some lesbians were unwilling to participate in this study because of fears that to do so would identify them as lesbians. The women who saw and responded to advertisements for participation in this investigation were likely be politically and/or socially active in a lesbian community. For these reasons the sample probably underrepresented lesbians at earlier stages of lesbian identity development. Likewise, women who were relatively closeted and, therefore, lacking in social and institutional supports were probably underrepresented as well. Finally, self-selection of couples may mean that those with more relationship difficulties may not have included themselves in the study.

PAGE 190

The restricted range of DAS scores for this sample supports this notion. Additionally, the instruments used in this study required a basic reading level and were somewhat lengthy so women with poor reading skills probably were not well represented in the sample. Additionally, because conflict resolution style was treated as a continuous rather than a categorical variable in this research, the design did not incorporate a strategy to equalize the number of participants by conflict resolution style. Furthermore, generalizability of the research was limited because of the network sampling techniques utilized. The research instruments used represent another limitation of this investigation. None of the instruments used has been normed on a population of lesbians although both the Dyadic Adjustment Scale and the Conflict Resolution Style Inventory have been used in studies with samples of lesbians and gay men. The Family Ritual Questionnaire had never before been used with a lesbian, gay, or bisexual sample. This instrument may have a heterosexual bias (Morin, 1977). Consequently, it is possible that this instrument neither adequately nor accurately assessed salient issues for lesbians. To use instruments normed on a population different from one's sample certainly creates the chance that the measures are not as applicable for the current sample as for members more similar to the normative group. ^ > { A further limitation concerns the culturally biased style of the Family Ritual Questionnaire and the complexity of the assessment task. Although the directions for the FRQ were modified for this study, the specific items and examples were not. The original instructions and each item of the FRQ refer to families" (e.g., "Some families observe cultural traditions . . . "). The instructions were modified to explain that a lesbian couple is a type of family, and participants were asked to consider themselves a "family"

PAGE 191

184 when responding to the items. Nonetheless, many women wrote unsolicited comments regarding the seemingly heterosexist language of the FRQ and expressed irritation at the apparent lack of inclusiveness. Additionally, the FRQ provided examples of rituals for each of the seven settings assessed which did not include any of the rituals common in lesbian lives such as croning ceremonies, annual Pride celebrations, or commitment ceremonies. It is possible that some women generalized to include rituals specific to the lesbian culture; however, much of the feedback written on the survey indicated that the examples of rituals seemed non-inclusive of alternate family forms. In retrospect, this seems to have presented a significant threat to the reliability and validity of the instrument. The Cronbach's alpha reliability estimates for 1 1 of the 15 FRQ subscales were low but adequate, ranging between .69 and .89. However, the alpha levels for the vacation setting (.56), occurrence dimension (.49), routines dimension (.62), and continuity dimension (.59) were very low. These four subscales constitute more than 25% of the total items on the FRQ. The ISS instrument developed for the investigation did not accurately assess the influence of institutional supports in this study. This instrument will need rigorous testing and modification if it is to contribute substantially to future research endeavors. Finally, the way in which variables were operationalized represents another limitation of the study. Conflict resolution styles were conceptualized as positive problem solving, conflict engagement, withdrawal, or compliance. Although operationalizing conflict resolution into only four styles helped organize the data and allowed the examination of differences, it posed limitations in that there may be other conflict resolution styles used by lesbians. Additionally, the rituals assessed in this

PAGE 192

185 investigation have been conceptualized in terms of eight dimensions occurring in seven settings. It is possible that operationalizing rituals in this manner may exclude potential rituals that occur in other settings. This seems to have been particularly true for this sample of women. Feedback from participants suggested that the definition of ritual afforded by the FRQ was too narrow and did not include rituals in which primarily lesbians participate. Finally, institutional support for the couple was conceptualized as a combination of civic and social sources of support from cultural institutions. Furthermore, civic sources of institutional support for the relationship included legal, health care, financial, religious, work place, and educational while social sources of institutional support for the relationship were represented by family of origin, intimate friends, and acquaintances. Conceptualizing institutional support for the relationship in this way may have limited the study by excluding other potential sources of institutional support. Suggestions for Future Study * --i; , ; Future investigations regarding lesbian couples' use of rituals may need to be done using qualitative rather than quantitative research methodology. Qualitative research may offer a better way to accurately understand the complexity and uniqueness of lesbian couples' experience of shared rituals. If the FRQ is used in future investigations with lesbian or gay or bisexual samples, it may need to be edited for inclusive language. Furthermore, the types of rituals assessed by the FRQ may need to be broadened in order to capture accurately the diversity and meaning of lesbian couples' ritual experience.

PAGE 193

186 A comparative, longitudinal study might be undertaken in which samples of lesbian couples, gay male couples, and heterosexual couples are surveyed regarding their use of rituals. This would afford an opportunity to determine whether rituals differ on the basis of couple type. Additionally, the influence of rituals over the various stages of relationship could be assessed. Furthermore, a longitudinal study would offer the opportunity to determine the possible influence of rituals on multiple measures of relationship satisfaction across time (e.g., point-in-time measures of satisfaction as well as the trajectory of change in satisfaction over time). , ; ; ' Because conflict resolution style was treated as a continuous rather than a categorical variable in this research, the design did not incorporate a strategy to equalize the number of participants by conflict resolution style. An analysis of participants' relative endorsement of the four conflict resolution styles assessed by the CRSI revealed that a majority of the partners in the sample identified with the positive problem-solving style (90%) followed by the withdrawal style (5%), the compliance style ( 1 %), and the conflict engagement style (1%). It seems that the latter three conflict resolution styles were grossly underrepresented in the sample. Future studies could include a sample in which each conflict resolution style was equally represented. If equal numbers of couples were assigned to each of the four conflict resolution styles, then comparisons could be made on the basis of predominant style utilized. This would allow researchers to determine whether couples with different types of conflict resolution styles pattern their ritual lives in similar or different manners. One might also choose a different method of measurement (e.g., observational rating of conflict).

PAGE 194

187 The ISS instrument developed for the investigation did not accurately assess the influence of institutional supports in this study. This instrument will need rigorous testing and modification if it is to contribute substantially to future research endeavors. Once again, it may be that future investigations regarding lesbian couples' sources of institutional support for their relationships could be better understood using qualitative rather than quantitative research methodology. The sample for this investigation was not representative of lesbians in the United States. Future studies could utilize random sampling techniques to address this limitation. Although this study was designed to investigate the relationship satisfaction levels of a nonclinical sample of lesbian couples, the spread of scores on the DAS was not very large. A larger, more diverse sample would address this concern in future research. Summary This chapter has provided a discussion of results and recommendations derived from a study of the influence of conflict resolution style, rituals use, and support for . relationship on the relationship satisfaction of lesbian couples. The variables that were statistically significant were discussed, and trends were examined for strengthening future investigations. The implications of these results for therapeutic intervention with couples were discussed as well.

PAGE 195

APPENDIX A LETTER TO PARTICIPANTS Dear Potential Research Participant: In an effort to better understand the nature of lesbian relationships, a research study is being conducted on the ways that lesbian partners resolve conflicts and use rituals in their relationships. This study will also investigate the sources of support that lesbian partners feel they have for their relationship. As a doctoral candidate in the Counselor Education Department at the University of Florida I am inviting you to participate in this study. Participation will require about 30 minutes. You will be asked to complete several questionnaires that assess your relationship satisfaction, your conflict resolution style, your rituals, and your sources of support. You can complete these questionnaires in the privacy of your home. It is important that you refrain from discussing your responses to the questions until you have completed the entire set of questionnaires and have sealed them in the envelope provided. The results will be strictly confidential and your name will not be made known to anyone or appear in any writing. The study requires that both members of a couple participate. Each partner must be at least 18 years of age and must consider herself a lesbian rather than bisexual, heterosexual, or transgendered. Further, partners must have lived together and considered themselves to be in a committed monogamous relationship for at least one year. Unfortunately, women who have children residing with them will not be able to participate in this study. Your participation in this study is completely voluntary. You may withdraw at any time. There are no known risks; however, if you feel that you need to speak with someone regarding issues stimulated by this survey you may call me for a referral. Participants will be offered a chance to "win" a bookstore gift certificate ($50). No immediate benefits are anticipated although you may request the results of the study and of your own responses. Please be assured that no other use, beyond this project, will be made of the information you provide. If you have questions about this research, you may contact me at my home (352)376-6171 or my faculty supervisor. Dr. Ellen Amatea, at (352)-392-073 1 . Either of us may be contacted m writing at 1212 Norman Hall, University of Florida, 3261 1. Questions or concerns about research participants' rights may be directed to the UFIRB 188

PAGE 196

189 office, University of Florida, Box 1 12250, Gainesville, FL, 3261 1; telephone (352)-3920433. If you would like to participate please sign the statement below. Enclosed please find a postage paid envelope in which to return the survey. Please mail it by January 1 5, 1999. ' ; • V :/^ : . Sincerely, 1 ' . Caroline Pace, Ed.S., LMHC Ellen Amatea, Ph.D. Principal Investigator ?,, Supervisor The Influence of Ritual Use, Institutional Support, and Conflict Resolution Style on Lesbian Couples' Relationship Satisfaction I have read the procedure described above. I voluntarily agree to participate in the procedure, and I have received a copy of this description. J ; . Signature of Participant Date

PAGE 197

APPENDIX B ' PARTICIPANT SELECTION PROTOCOL In order to participate in the study, both partners in a couple aged 1 8 or older living together in one household must agree to participate. Further, each partner must consider herself a lesbian and in a committed monogamous relationship of at least one year's duration. Women residing with children will not be included. In order to determine which potential participants will be assessed, the questions listed below will be asked. 1. Are you 18 or older? 2. Do you currently consider yourself a lesbian? Do you consider yourself bisexual, transgendered, or heterosexual? 3. Do you live with your partner? 4. Have you and your partner been in a committed monogamous relationship with each other for at least one year? 5. Do you and your partner live without children? Answers must be yes to the above five questions for both partners in order to participate in the study.

PAGE 198

APPENDIX C DEMOGRAPHIC DATA SHEET Questionnaire # Lesbian Relationship Study Thank you for participating in this research study regarding the ways that lesbian couples ' develop their relationships. Your responses to the survey questions that follow will help us better understand the ways that lesbian partners create a life together and settle differences. Your opinion is important; there are no right or wrong "answers" to these questions. Remember, your responses are confidential and anonymous. Please begin by providing the demographic information requested below. RaceEthnicity (circle one) " Caucasian Black , . Hispanic y . Native American Asian City of Residence State of Residence Education Level (circle one) Middle School ' High School AA-AS . BA-BS :;i ^.v. MA-MS Ph.D. Post-doc if Approximate Individual Armual Income (circle one) Under $5000 $30,000-$39,999 $5000-$9999 $40,000-$49,999 $10,000-$ 14,999 $50,000-$59,999 Age $15,000-$ 19,999 $20,000-$29,999 $60,000-$69,999 Over $70,000 191

PAGE 199

Length of Time in Present Relationship Length of Time Living with Present Partner Length of Time You Have Considered Yourself Lesbian-Gay Parental Status (circle one) children not living at home no children Religious Preference

PAGE 200

APPENDIX D INSTITUTIONAL SUPPORT SCALE PLEASE RATE THE STATEMENTS BELOW ON THE FOLLOWING SCALE 1 = STRONGLY DISAGREE, 2 = DISAGREE, 3 = AGREE, 4 = STRONGLY AGREE -Sr" SD D A SA 1 . My partner and I are a couple in the eyes of my close friends. ' . 12 3 4 2. My partner and I are included in decision making about social gatherings with acquaintances. 12 3 4 3. As a couple, my partner and I would NOT feel comfortable having a , , . commitment ceremony-wedding. 12 3 4 4. My partner and I are a couple in the eyes of healthcare institutions (e.g., hospital, dentist office). 12 3 4 5. My family would miss my partner if she did not attend family gatherings with me. 12 3 4 6. My partner and 1 are a couple in the eyes of the institution where I work (attend school). : ' 12 3 4 7. My partner and I own some things of value jointly. ' ' 12 3 4 8. Acquaintances would ask about my partner if she did not attend a gathering with me. 12 3 4 9. As a couple, my partner and I have a joint bank account. 12 3 4 10. My partner and I are NOT consulted in the planning of family gatherings 12 3 4 11. I would feel comfortable telling my legal counsel that my partner and I are a couple. 12 3 4 12. My partner and I are welcome as a couple at social functions hosted at my work (or school) site. 12 3 4 13. My partner and I are a couple in the eyes of my family. 12 3 4 14. My partner and I could NOT obtain a credit card as a couple. 12 3 4 15. I would feel comfortable telling people I work (attend school) with that my partner and I are a couple. 12 3 4 193

PAGE 201

194 16. My close friends recognize my partner's importance to me. 17. I would feel comfortable telling my healthcare provider that my partner and 1 are a couple. 18. I would NOT feel comfortable wearing a ring or other symbol signifying my relationship. 19. My healthcare provider would support my partner and I making healthcare decisions together. 20. My partner and I do NOT attend social gatherings with acquaintances as a couple. 21. As a couple, my partner and I could adopt a child if we so desired. 22. I feel that my relationship is supported in the community in which I live. 23. My partner and 1 could apply as a couple for a mortgage. 24. As a couple, my partner and I feel included in the planning of social gatherings with close friends. ' 25. My partner and I are recognized as a couple in the eyes of legal institutions. 26. My family includes my parmer and I in holiday and family celebrations (e.g.. Thanksgiving, birthdays). 27. The company I work for is NOT supportive of my partner and me as a couple. 28. Insurance companies in my community would cover my partner as a "spouse" on my policy. 29. 1 would feel comfortable asking a lawyer to draw up a Power of Attorney or will naming my partner as beneficiary. 30. Religious institutions in my community follow doctrine or teachings that afford lesbian couples the same status as heterosexual couples. 2 3 4 2 3 4 2 3 4 2 3 4 2 3 4 2 3 4 2 3 4 2 3 4 2 3 4 2 3 4 2 3 4 2 3 4 2 3 4 2 3 4 2 3 4

PAGE 202

REFERENCES Actelli, L. K. (1988). When spouses talk to each other about their relationships. Journal of Social and Personal Relationships. 5. 185-189. Adams, J. M., & Jones, W. H. (1997). The conceptualization of marital commitment: An integrative analysis. Journal of Personality and Social Psychology. 72. 1177-1196. Albro, J. C, & Tully, C. (1979). A study of lesbian life styles in the homosexual micro-culture and the heterosexual macro-culture. Journal of Homosexuality. 4 (4). 331334. Allen, L. B., Glicken, A. D., Beach, R. K., & Naylor, K. E. (1998). Adolescent health care experience of gay, lesbian, and bisexual young adults. Journal of Adolescent Healtlk23(4), 212-220. American Association of Marriage and Family Therapy. (1991). Code of ethics [brochure]. Washington, DC: Author. Anderson, K. K. (1990). Intimacy, fusion, and relationship quality of women in lesbian and heterosexual relationships (Doctoral dissertation, Oklahoma State University, 1990). Dissertation Abstracts International. 50. 4814R. Antill, J. K., & Cotton, S. (1982). Spanier's Dyadic Adjustment Scale: Some confirmatory analyses. Australian Psychologist. 17. 181-189. Armon, V. (1960). Some personality variables in overt female homosexuality. Journal of Projective Techniques. 24. 292-309. Aura, J. (1985). Women's social support: A comparison of lesbians and heterosexuals (Doctoral dissertation. University of California, 1985). Dissertation Abstracts International, 46-12B. 4389. • , . Bell, A. P., & Weinberg, M. S. (1978). Homosexualities: A study of diversity among me n and v^omen. New York: Simon and Schuster. Bell, A. P., Weinberg, M. S., & Hammersmith, S. K. (1981). Sexual preference: Its development in men and women. Bloomington, IN: Indiana University Press. 195

PAGE 203

196 Bern, S. L. (1974). The measurement of psychological androgyny. Journal of Consulting and Clinical Psychology. 42 (2). 155-162. Bennett, L. A., Wolin, S. J., & McAvity, K. J.(1988). Family identity, ritual, and myth: A cultural perspective on life cycle transitions. In C. J. Falicov (Ed.), Family transitions: Continuity and change over the life cycle (pp. 221-234). New York: Guilford Press. Beimett, L. A., Wolin, S. J., & Reiss, D. (1988). Deliberate family process: A strategy for protecting children of alcoholics. British Journal of Addiction. 83. 821-829. Bennett, L. A., Wolin, S. J., Reiss, D., & Teitelbaum, M. A. (1987). Couples at risk for transmission of alcoholism: Protective influences. Family Process. 26. 1 1 1-129. Berger, R. M. (1990). Passing: The impact on the quality of same-sex couple relationships. Social Work. 35. 328-332. Berger, R. M., & Mallon, D. (1993). Social support networks of gay men. Journal of Sociology and Social Welfare. 20 (11. 155-174. Betcher, W. (1987). Intimate play: Creating romance in everyday life. New York: Viking. „ Bieschke, K. J., & Matthews, C. (1996). Career counselor attitudes and behaviors toward gay, lesbian, and bisexual clients. Journal of Vocational Behavior. 48(2), 243-255. Blumstein, P., & Schwartz, P. (1983). American couples. New York: William Morrow. Boatwright, K. J., Gilbert, M. S., Forrest, L., & Kezenberger, K. (1996). Impact of identity development upon career trajectory: Listening to the voices of lesbian women. Journal of Vocational Behavior. 48 (2). 210-228. Bossard, J., & Boll, E. (1950). Ritual in family living. Philadelphia: University of Permsylvania Press. Bowman, M.L. (1990). Coping efforts and marital satisfaction: Measuring marital coping and its correlates. Journal of Marriage and the Family. 52. 463-474. Boyce, W. T., Jensen, E. W., James, S. A., & Peacock, J. L. (1983). The Family Routines Inventory: Theoretical origins. Social Science and Medicine. 1 7 (4). 1 93-700

PAGE 204

197 Boyd, L. A., & Roach, A. J. (1977). Interpersonal communication skills differentiating more satisfying from less satisfying marital relationships. Journal of Counseling Psychology. 24. 540-542. Bozett, F. W. (1987). Gav and lesbian parents. New York: Praeger. Braiker, H., & Kelley, H. H. (1979). Conflict in the development of close relationships. In R. L. Burgess & T. L. Huston (Eds.), Social exchange in developing relationships (pp. 135-168). New York: Academic Press. Brooks, L.S. (1981). Minority stress and lesbian women. Lexington, MA: D.C. Heath. Brown, L. S. (1989a). Lesbians, gay men, and their families: Coimnon clinical issues. Journal of Gay and Lesbian Psychotherapy. 1 (1). 65-77. Brown, L. S. (1989b). New voices, new visions: Toward a lesbian/gay paradigm for psychology. Psychology of Women Quarterly. 13. 445-458. Brown, L. S. (1991). Commentary on the special issue of The Counseling Psychologist: Counseling with lesbians and gay men. The Counseling Psychologist. 19 (2), 235-238. Brown, L. S. (1995). Lesbian identities: Concepts and issues. InA. R. D'Augelli & C. J. Patterson (Eds.), Lesbian, gay, and bisexual identities over the lifespan (pp. 3-23). New York: Oxford University Press. Bruess, C. J. S. (1994). "Bare-chested hugs" and "tough-guys' night": An inductive exammation of the form and function of interpersonal rituals in marriage and adult friendship (Doctoral dissertation, Ohio University, 1994). Dissertation Abstracts International. 55-07A. 1741. Bryant, A. S., & Demian. (1994). Relationship characteristics of American gay and lesbian couples: Findings from a a national survey. In L. A. Kurdek (Ed.), Social services for gav and lesbian couples (pp. 101-117). New York: Harrington Park Press/Haworth Press. Bucy, J. E. (1995). An exploratory study of family rituals, parenting stress, and developmental delay in early childhood (Doctoral dissertation, University of North Carolina, 1995). Dissertation Abstracts International. 57-Q2A. 575. Burch, B. (1986). Psychological merger in lesbian couples. Family Theranv. 9(3), 201-208. Butler, B. (1991). Ceremonies of the heart: Celebrating lesbian unions. Seattle, WA: Seal.

PAGE 205

198 Cabaj, R. P. (1988). Gay and lesbian couples: Lessons on human intimacy. Psychiatric Annals. 18 (1). 21-25. Caldwell, M.,&Peplau,L. A. (1984). The balance of power in lesbian relationships. Sex Roles. 10 (7-8). 587-599. Caldwell, M., & Peplau, L. A (1992). The balance of power in lesbian relationships. In W. R. Dynes & S. Donaldson (Eds.), Lesbianism (pp. 27-39). New York: Garland Publishing, Inc. Cardell, M., Finn, S., & Maracek, J. (1981). Sex-role identity, sex-role behavior, and satisfaction in heterosexual, lesbian, and gay male couples. Psychology of Women Quarterly. 5 (3). 488-494. . , Caron, S. L., & Ulin, M. (1997). Closeting and the quality of lesbian relationships. Families in Society: The Journal of Contemporary Human Services. 78. 413-419. . Carstensen, L. L., Gottman, J. M., & Levenson, R. W. (1995). Emotional behavior in long-term marriage. Psychology and Aging. 10 (1). 140-149. Carter, B., & McGoldrick, M. (1989). The changing family life cvcle: A framework for family therapy. Boston: Allyn and Bacon. Casper, V., & Schultz, S. (1996). Lesbian and gay parents encounter educators: Initiating conversations. In R. C. SavinWilliams, & K. M. Cohen (Eds.), The lives of lesbians, gays, and bisexuals: Children to adults (pp. 305-330). Ft. Worth, TX: Harcourt Brace College Publishers. Cass, V. A. (1979). Homosexual identity formation: A theoretical model. Journal of Homosexuality. 4 (3). 219-235. Cass, V. A. (1984). Homosexual identity formation: Testing a theoretical model. Journal of Sex Research. 20. 143-167. Chafetz, J. S., Sampson, P., Beck, P., & West, J. (1974). A study of homosexual women. Social Work. 19. 714-723 Chapman, B. E., & Brannock, J. C. (1987). Proposed model of lesbian identity development: An empirical examination. Journal of Homosexuality. 14, 69-80 Chesler, P. (1986). Mothers on trial: The battle for children and custody New York: McGraw-Hill.

PAGE 206

199 Christensen, A. (1988). Dysfunctional interactional patterns in couples. In P. NoUer & M. A. Fitzpatrick (Eds.), Perspectives on marital interaction (pp. 31-52). Clevedon, Avon, England: Multilingual Matters. Christensen, A, & Heavey, C. L. (1990). Gender and social structure in the demand/withdraw pattern of marital conflict. Journal of Personality and Social Psychology. 59. 73-81. Christensen, A., & Shenk, J. L. (1991). Communication, conflict, and psychological distance in nondistressed, clinical, and divorcing couples. Journal of Consulting and Clinical Psychology. 59. 45 8-463 . ;. Chung, Y. B. (1995). Career decision making of lesbian, gay, and bisexual individuals. Career Development Quarterly. 44 (2). 178-190. Clark, W. M., & Serovich, J. M. (1997). Twenty years and still in the dark? Content analysis of articles pertaining to gay, lesbian, and bisexual issues in marriage and family therapy journals. Journal of Marriage and Family Therapy. 23 (3). 239-253. Cobb, S. (1976). Social support as a moderator of life stress. Psychosomatic Medicine. 38. 300-314. Cohen, S., & Hoberman, H. (1983). Positive events and social supports as buffers of life change stress. Journal of Applied Social Psychology. 13 (2). 99-125. Cohen, S., Mermelstein, R., Kamarck, T., «fe Hoberman, H. (1985). Measuring the functional components of social support. In I. Sarason & B. Sarason (Eds.), Social support: Theory, research, and applications (pp. 73-94). Dordrecht, Netherlands: Martinez Nijhoff. Coleman, E. (1982). Developmental stages of the coming out process. Journal ^ of Homosexuality. 7 (2-3). 31-43. Cooke, B. D., Rossmann, M. M., McCubbin, H. I., & Patterson, J. M. (1988). Examining the definition and assessment of social support: A resource for individuals and families. Family Relations. 37. 211-216. Cotton, W. L. (1975). Social and sexual relationships of lesbians. Journal of Sex Research. 19 (1). 58-73. Coyne, J. C, & DeLongis, A. (1986). Gong beyond social support: The role of social relationships in adaptation. Journal of Consulting and Clinical Psychology. 54. 454-460. Cramer, D. (1986). Gay parents and their children: A review of the research and practical implications. Journal of Counseling and Development. 64. 504-507.

PAGE 207

200 Crocker, L., & Algina, J. (1986). Introduction to classical and modem test theory. New York: Holt, Rinehart, & Winston, Inc. Cummins, R. (1988). Perceptions of social support, receipt of supportive behaviors, and locus of control as moderators of the effects of chronic stress. American Journal of Community Psychology. 16 (5), 201-208. Curran, D. (1983). Traits of a healthy family. Minneapolis: Winston Press. Cutrona, C. (1984). Social support and stress in the transition to parenthood. Journal of Abnormal Psychology. 93. 378-390. Cutrona, C, & Russell, D. (1987). The provisions of social relationships and adaptation to stress. In W. H. Jones & D. Perlman (Eds.), Advances in personal relationships (Vol. 1, pp. 73-94). Greenwich, CT: JAI Press. Dahlheimer, D., & Feigal, J. (1991). Bridging the gap. Family Therapy Networker. 15 (11. 44-53. Dailey, D. M. (1979). Adjustment of heterosexual and homosexual couples in pairing relationships: An exploratory study. Journal of Sex Research. 15. 143-157. Dank,B. M. (1971). Coming out in the gay world. Psychiatry. 34. 180-197. D'Augelli, A. R., Collins, C, & Hart, M. M. (1987). Social support patterns of lesbian women in a rural helping network. Journal of Rural Community Psychology. m), 12-22. Davis, M. H., & Oathout, H. A. (1987). Maintenance of satisfaction in romantic relationships: Empathy and relational competence. Journal of Personality and Social Psychology. 53. 397-410. DeMonteflores,C.,&Schultz, S.J. (1978). Coming out: Similarities and differences for lesbians and gay men. Journal of Social Issues. 34. 59-72. DeTurck, M. A., & Miller, G. R. (1986). The effects of husbands' and wives' social cognition on their marital adjustment, conjugal power, and self-esteem. Journal of Marriage and the Family. 48. 715-724. Donaldson, C. (1993). Correlates of relationship satisfaction for lesbian couples (Doctoral dissertation, Hofstra University, 1993). Dissertation Abstracts International . 54-08B. 4369. Dom, B. A. (1990). An investigation into factors that contribute to successful long-term lesbian relationships (Doctoral dissertation. The Union Institute, 1990). Dissertation Abstracts International. 51-07B . 3560.

PAGE 208

Duffy, S. M., & Rusbult, C. E. (1986). Satisfaction and commitment in homosexual and heterosexual relationships. Journal of Homosexuality. 12 (2). 1-23. Durkheim, E. (1965). The elementary form of religious life. New York: The Free Press. Easton, M. A. (1987). The effect of unrealistic expectations on satisfaction and duration in male couples. Hemstead, NY: Hostra University. Eldridge, N. A., & Gilbert, L. A. (1990). Correlates of relationship satisfaction in lesbian couples. Psychology of Women Quarterly. 14. 43-62. Elliot, P. E. (1985). Theory and research on lesbian identity formation. International Journal of Women's Studies. 8 (11 64-71. Elliot, P. E. (1992). Theory and research on lesbian identity formation. InW. R. Dynes & S. Donaldson (Eds.), Lesbianism (pp. 94-101). New York: Garland Publishing, Inc. Epstein, N., DeGiovanni, I. S., & Jayne-Lazarus, C. (1978). Asserting training for couples. Journal of Behavior Therapy and Experimental Psychiatry. 9. 149-155. Evans, R. B. (1970). Sixteen personality factor questionnaire scores of homosexual men. Journal of ConsuUine and Clinical Psychology. 34 (21 212-215. Faderman, L. (1991). Odd girls and twilight lovers: A history of lesbian life in twentieth century America. New York: Columbia University Press. Falbo, T., & Peplau, L. A. (1980). Power strategies in intimate relationships. Journal of Personality and Social Psychology. 38. 618-628. Falco, K. L. (1991). Psychotherapy with lesbian clients . New York: Bruner/Mazel. Fassinger, R. E. (1995). From invisibility to mtegration: Lesbian identity in the workplace. Career Development Quarterly. 44 (21 148-167. Fiese, B. H. (1992). Dimensions of family rituals across two generations: Relation to adolescent identity. Family Process. 31. 1 5 1 1 62. Fiese, B.H. (1997). Family context in pediatric psychology from a transactional perspective: Family rituals and stories as examples. Journal of Pediatric Psvch olnpv 22 (2), 183-196.

PAGE 209

202 Fiese, B. H., Hooker, K. A., Kotary, L., & Schwagler, J. (1993). Family rituals in i the early stages of parenthood. Journal of Marriage and the Family. 55. 633-642. Fiese, B. H., & Kline, C. A. (1993). Development of the Family Ritual Questionnaire (FRQ): Initial reliability and validation studies. Journal of Family Psychology. 6. 290-299. Fischer, C. (1982). To dwell among friends. Chicago: University of Chicago Press. Fischer, 1. W. (1993). The relationship between attachment style and styles of conflict resolution in romantic relationships (Doctoral dissertation, Hofstra University, 1993). Dissertation Abstracts International. 55-0 IB . 6710. Fitzpatrick, M. A., & Best, P. (1979). Dyadic adjustment in relational types: Consensus, cohesion, affectional expression, and satisfaction in enduring relationships. Communication Monographs. 46. 167-178. >'' *' Fontaine, J. H. (1997). The sound of silence: Public school response to the needs of gay and lesbian youth. In M. B. Harris (Ed)., School experiences of gav and lesbian youth: The invisible minority fpp. 1 01 1 09V New York: Harrington Park Press/Haworth Press. Freedman, M. (1971). Homosexuality and psychological functioning. Belmont, CA: Brooks/Cole. Gargaetas, P. (1995). Altered states: Lesbian altermaking and the transformation of self. Women and Therapy. 16 r2/3'>. 95-105. The gay market. (1992, January). Overlooked Op inions, p. 7. >, Gibbs, E. D. (1989). Psychosocial development of children raised by lesbian mothers: A review of the research. Women and Therapy. 8. 65-75. Gilmore, T. C. (1996). Stress, coping, and adjustment among gay, lesbian, and bisexual youth (Doctoral dissertation, George Mason University, 1996). Dissertation Abstracts International, 57-04B. 2865. Glenn, N.D. (1990). Quantitative research on marital quality in the 1980s: A critical review. Journal of Marriage and the Family. 52. 8 1 8-83 1 . Gonsiorek, J. C, & Weinrich, J. D. (Eds.). (1991). Homosexuality: Rese^^rrh implication s for public p olicy Newbury Park, CA: Sage. Gottman, J. M. (1990). Children of gay lesbian parents. Journal of Homosexuality^ 1Q 177-196.

PAGE 210

203 Gottman, J. M. (1994). What predicts divorce: The relationship between marital processes and marital outcomes. Hillsdale, NJ: Lawrence Erlbaum Associates. Gottman, J. M. (1998). Psychology and the study of marriage. Annual Review ofPsvcholoev. 48. 169-197. Gottman, J. M., Coan, J., Carrere, S., & Swanson, C. (1998). Predicting marital happiness and stability from newlywed interactions. Journal of Marriage and the Family. 60, 5-22. Gottman, J. M., & Krokoff, L. J. (1989). Marital interaction and satisfaction: A longitudinal view. Journal of Consulting and Clinical Psychology. 57. 47-52. Gottman, J. M., & Levenson, R. W. (1992). Marital processes predictive of later dissolution: Behavior, physiology, and health. Journal of Personality and Social Psychology. 63 (21 221-233. Grahn, J. (1984). Another mother tongue: Gav words, gay worlds . Boston: Beacon Press. Green, R. (1972). Homosexuality as a mental illness. International Journal of Psychiatry. 10 (1X 77-98. Green, R., Bettinger, M., & Zacks, E. (1996). Are lesbian couples fused and gay male couples disengaged? Questioning gender straightjackets. In J. Laird & R. Green (Eds.). Lesbians and gavs in couples and families (pp. 185-230). San Francisco: JosseyBass. Green, R., Mandel, J. B., Hotvedt, M. E., Gray, J., & Smith, L. (1986). Lesbian mothers and their children: A comparison with solo parent heterosexual mothers and their children. Archives of Sexual Behavior. 15. 1 67-1 84. Grossman, A. H., & Kemer, M. S. (1998). Self-esteem and supportiveness as predictors of emotional distress in gay male and lesbian youth. Journal of Homosexuality. 35 (2V ?.5-39 , Hate crimes on the rise. (1998, October 24). Florida Times Union, p. Bl. Hatton, B. J. (1995). Personal power in lesbian relationships (Doctoral dissertation. University of Kentucky, 1995). Dissertation A bstracts Tntemationnl 5609A.3457. ' ' 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

PAGE 211

204 Heller, K. (1979). Lesbian families: Cultural and clinical issues. Social Work, 21380-385. Heller, K., & Swindle, R. (1983). Social networks, perceived social support, and coping with stress. In R. D. Felner, L. A. Jason, J. Moritsugu, & S. S. Farber (Eds.), Preventative psychology: Theory, research, and practice in community intervention (pp. 87-103). New York: Pergamon Press. Henderson, S. (1987). The social network, support, and neurosis: The function of attachment in adult life. British Journal of Psychiatrv. 131. 185-191. Hoeffer, B. (1981). Children's acquisition of sex-role behavior in lesbian-mother families. American Journal of Orthopsvchiatry. 5U 31). 536-543. Hojjat, M. (1997). Conflict-resolution styles and relationship satisfaction (Doctoral dissertation, Yale University, 1997). Dissertation Abstracts International. 58IIB. 6279. Hopkins, J. H. (1969). The lesbian personality. British Journal of Psvchiatrv. 115. 1433-1436. House, J. S. (1981). Work stress and social support . Reading, MA: AddisonWesley Publishing. Howard, J. A., Blumstein, P., & Schwartz, P. (1986). Sex, power, and influence tactics in intimate relationships. Journal of Personality and Social Psvcholoev. 51. 1-2109. Hunsley, J., Pinsent, C, Lefebvre, M., James-Tanner, S., & Vito, D. (1995). Assessment of couples, marriages, and families: Construct validity of the short forms of the Dyadic Adjustment Scale. Family Relations. 44. 231-237. Imber-Black, E. (1988a). Normative and therapeutic rituals in couples therapy. In E. Imber-Black, J. Roberts, & R. A. Whiting (Eds.), Rituals in families and familv therapy (pp. 1 13-134). New York: W. W. Norton. Imber-Black, E. (1988b). Ritual themes in families and family therapy. InE. Imber-Black, J. Roberts, & R. A. Whiting (Eds.), Rituals in families and family therapy (pp. 47-83). New York: W. W. Norton. Imber-Black, E. & Roberts, J. (1992 ). Rituals for our times: Celehratinp . he.ahn, and changing our lives an d our relationship s New York: HarperCollins.

PAGE 212

205 Jacob, T., Favorini, A., Meisel, S., & Anderson, C. (1978). The spouse, children, and family interactions of the alcoholic: Substantive findings and methodological issues. Journal of Studies on Alcohol. 39. 1231-1251. Jacobson, N. S., & Addis, M. E. (1993). Research on couples and couple therapy. What do we know? Where are we going. Journal of Consulting and Clinical Psychology. 61. 85-93. January, K.C. (1996). Family rituals and family functioning in Hispanic and Anglo families across the life cycle (Doctoral dissertation, Georgia State University, 1996). Dissertation Abstracts International. 57-08B. 5329. Jenkins, J. K. (1996). Family of origin differentiation and dyadic adjustment in lesbian relationships. Unpublished Master's thesis. Texas Woman's University, Denton, TX. Jenson, E., James, S., Boyce, W., & Hartnett, S. (1983). The family routines inventory: Development and validation. Social Science and Medicine. 17. 201-21 1. Johnson, S.M.,&Greenberg,L.S. (1985). Emotionally focused couples therapy: An outcome study. Journal of Consulting and Clinical Psychology. 53. 175184. Jordan, K. M. (1995). Coming out and relationship quality for lesbian women. (Doctoral dissertation. University of Maryland, 1995). Dissertation Abstracts International. 56-05B. 2941. Jordan, K.M.,&Deluty,R.H. (1998). Coming out for lesbian women: Its relation to anxiety, positive affectivity, self-esteem, and social support. Journal of Homosexuality. 35 r2'>. 4 1 -63 . Jordan, K. M., Vaughan, J. S., & Woodworth, K. J. (1997). 1 will survive: , Lesbian, gay, and bisexual youths' experience of high school. In M. B. Harris (Ed.), School experiences of g ay and lesbian youth: The invisible minority (pp. 17-33). New York: Harrmgton Park Press/Haworth Press. Kahn, R. L., & Antonucci, T. (1980). Convoys over the life course: Attachment, roles, and social support. In P. B. Baltes & O. Brom (Eds.), Lifespan development and behaYior(Vol. 3, pp. 221-234). Boston: Lexington Press. Kantrowitz, B. (1996, November 4). Parents come out. Newsweek , pp. 51-57. Kenny, D. A. (1996). Models of non-independence in dyadic research. Journal of Social a nd Personal Relationships. 13^ 279-294.

PAGE 213

206 Kinsey, A. C, Pomeroy, W. B., Martin, C. E., & Gebhard, P. H. (1953). Sexual behavior in the human female. Philadelphia: W. B. Saunders. Kirkpatrick, M., Smith, C, & Roy, R. (1981). Lesbian mothers and their children: A comparative study. Journal of Orthopsychiatry. 51 (3). 545-551. Kline, C. (1992). Family rituals, psychological separation, and college adjustment (Doctoral dissertation, Syracuse University, 1991). Dissertation Abstracts International. 54-02B. 1101. Krestan, J., & Bepko, C. (1980). The problem of fusion in the lesbian relationship. Family Process. 19. 277-290. Kurdek, L. A. (1988a). Perceived social support in gays and lesbians in cohabiting relationships. Journal of Personality and Social Psychology.54 (3). 504-509. Kurdek, L. A. (1988b). Relationship quality of gay and lesbian cohabiting couples. Journal of Homosexuality. 15 (3-4). 93-118. Kurdek, L. A. (1989). Relationship quality in gay and lesbian cohabiting couples: A 1 -year follow-up study. Journal of Social and Personal Relationships. 6 (1). 39-59. ? Kurdek, L. A. (1991). Correlates of relationship satisfaction in cohabiting gay and lesbian couples: Integration of contextual, investment, and problem-solving models. Journal of Personality and Social Psvchology. 61. 910-922. Kurdek, L. A. (1992a). Dimensionality of the Dyadic Adjustment Scale: Evidence from heterosexual and homosexual couples. Journal of Family Psychologv. 6(1), 22-35. Kurdek, L. A. (1992b). Relationship stability and relationship satisfaction in cohabitatmg gay and lesbian couples: A prospective longitudinal test of the contextual interdependence models. Journal of Social and Personal Relations. 9. 1 25-1 4? Kurdek, L. A. (1994a). Areas of conflict for gay, lesbian, and heterosexual couples: What couples argue about influences relationship satisfaction. Journal of Marriage and the Familv. 56. 923-934. Kurdek, L. A. (1994b). Conflict resolution styles in gay, lesbian, heterosexual nonparent, and heterosexual parent couples (The Ineffective Arguing Inventory and The Conflict Resolution Style Inventory). Journal of Marriage and the Familv 56 705722. — '

PAGE 214

207 Kurdek, L. A. (1994c). The nature and correlates of relationship quality in gay, lesbian, and heterosexual cohabiting couples: A test of the individual difference, interdependence, and discrepancy models. In B. Greene & G. M. Herek (Eds.), Psychological perspectives on lesbian and gav issues: Vol. 1. Lesbian and gay psychology: Theory, research, and clinical applications (pp. 133-155). Thousand Oaks, CA: Sage. Kurdek, L. A. (1995a). Developmental changes in relationship quality in gay and lesbian cohabiting couples. Developmental Psychology, 31. 86-94. Kurdek, L. A. (1995b). Predicting change in marital satisfaction from husbands' and wives' conflict resolution styles. Journal of Marriage and the Family. 57. 153-164. Kurdek, L. A. (1998). Relationship outcomes and their predictors: Longitudinal evidence from heterosexual married, gay cohabiting, and lesbian cohabiting couples. Journal of Marriage and the Family. 60. 553-568. Kurdek, L. A., & Schmitt, J. P. (1986a). Early development of relationship quality in heterosexual married, heterosexual cohabiting, gay, and lesbian couples. Developmental Psychology. 22. 305-309. Kurdek, L. A., & Schmitt, J. P. (1986b). Interaction of sex-role self-concept with relationship quality and relationship beliefs in married, heterosexual cohabitating, gay, and lesbian couples. Journal of Personality and Social Psychology. 51. 365-370. Kurdek, L. A., & Schmitt, J. P. (1986c). Relationship quality of partners in heterosexual married, heterosexual cohabiting, gay, and lesbian relationships. Journal of Personality and Social Psychology. 51. 71 1-720. Kurdek, L. A., & Schmitt, J. P. (1987a). Partner homogamy in married, heterosexual cohabitating, gay and lesbian couples. The Journal of Sex Research. 23. 212-232. Kurdek, L. A., & Schmitt, J. P. (1987b). Perceived emotional support from family and friends in members of homosexual, married, and heterosexual cohabitating couples. Journal of Homosexuality. 14. 57-68. Kus, R. J. (1985). Stages of coming out: An ethnographic approach. Western Journal of Nursing Research. 1 (2). 177-198. Laird, J. (1988). Women and ritual. In E. Imber-Black, J. Roberts, & R. Whiting (Eds.). Rituals in families and family therapy (pp. 331 -362V New York: Norton. Lakey, B., & Heller, K. (1988). Social support from a friend, perceived support, and social problem-solving. American Journal of Community Psychology. \ 6 (6). 811-

PAGE 215

208 Lange, S. J. (1989). Satisfaction and commitment in lesbian and heterosexual relationships: The roles of power, nurturance, and trust (Doctoral dissertation, University of Kentucky, 1989). Dissertation Abstracts International . 50-08B, 3676. Lazarus, R. S., & Launier, R. (1978). Stress-related transactions between person and environment. In L. A. Pervin & M. Lewis (Eds.), Perspectives in interactional psychology (pp. 287-327). New York: Plenum. Leavy, R. L., & Adams, E. M. (1986). Feminism as a correlate of self-esteem, self-acceptance, and social support among lesbians. Psychology of Women Quarterly. 10,321-326. Levine, H., & Evans, N. J. (1991). The development of gay, lesbian, and bisexual identities. In N. Evans & V. Wall (Eds.), Beyond tolerance: Gays, lesbians, and bisexuals on campus (pp. 1-23). Washington, DC: American Counseling Psychology Association. Levy, E. F. (1983). Social support and psychological disorder: A review. Journal of Community Psychology. 1 1 (1). 3-21. Levy, E. F. (1992). Strengthening the coping resources of lesbian families. Families in Societv: The Journal of Contemporary Human Services. 73. 23-3 1 . Lewis, R. A., & Spanier, G. B. (1979). Theorizing about the quality and stability of marriage. In W. R. Burr, R. Hill, R. I. Nye, & I. L. Reiss (Eds.), Contemporary theories about the family (Vol. 1, pp. 268-294). New York: Free Press. Lin, N., Ensel, W. M., Simeone, R. S., & Kuo, W. (1979). Social support, stressful life events, and illness: A model and empirical test. Journal of Health and Social Behavior. 40. 108-119. Lindenbaum, J. (1985). The shattering of an illusion: The problem of competition in lesbian relationships. Feminist Studies. 1. 85-103. Loney, J. (1971). An MMPl measurement of maladjustment in a sample of normal homosexual males. Journal of Clinical Psychology. 27. 486-488. , . Maarse, G. J. (1994). Gay men in sustained relationships: A study assessing attachment styles, sex-role orientation, and conflict resolution styles (Doctoral dissertation, California School of Professional Psychology, 1994). Dissertation Abstracts International. 55-06B. 2405. Macklin, E. D. (1980). Nontraditional family forms: A decade of research. Journal of Marriage and the Family. 42. 905-922. Maggiorre, D. J. (1988). Lesbianism: An annotated bibliography and g uide to the literature . 1976-1986. Metuchen, NJ: Scarecrow Press.

PAGE 216

209 Mallon, G. P. (1998). We don't exactly get the welcome wagon: The experiences of gay and lesbian adolescents in child welfare systems . New York: Columbia University Press. Markman, H. J., Floyd, F. J., Stanley, S. M., & Storaasli, R. D. (1988). Prevention of marital distress: A longitudinal investigation. Journal of Consulting and Clinical Psychology. 2. 478-482. Markman, H. J., Renick, M. J., Floyd, F., Stanley, S. M., & Clements, M. (1993). Preventing marital distress through communication and conflict management training: A 4and 5-year follow-up. Journal of Consuhing and Clinical Psychology. 61. 70-77. Markowitz, L. M. (1991). Homosexuality: Are we still in the dark? Family Therapy Networker. 15 (1). 26-35. Markowitz, L. M. (1993, September/October). We are family: Why gays and lesbians build strong bonds. Utne Reader. 59 . 82-83. Martin, A. (1982). Some issues in the treatment of gay and lesbian patients. Psychotherapy: Theory. Research, and Practice. 19 (3). 341-348. McCubbin, MA. & McCubbin, H. I. (1987). Family stress theory and assessment: The resiliency model of family stress, adjustment, and adaptation. In H. I. McCubbin (Ed.), Family assessment inventories for research and practice (pp. 2-32). Madison, WI: University of Wisconsin. McCubbin, M A, McCubbin, H. 1., Thompson, A. I., & Thompson, E. A. (1995). Resiliency in ethnic families: A conceptual model for predicting family adjustment and adaptation. In I. McCubbin (Series Ed.) & H. I. McCubbin, E. A. Thompson, A. I. Thompson, & J. E. Fromer (Volume Eds.), Resiliency in ethnic minority families: Native and immigrant American families. Volume 1 (pp. 3-48). Madison, WI: University of Wisconsin. McCubbin, H., «fe Thompson, A. I. (Eds.). (1991). Family assessment inventories for researc h and practice. Madison. WTUnivprdty nf Wkpr>nc;n-\/ff,rjicr.n McDaniel, J. S., Cabaj, R. P., & Purcell, D. W. (1996). Care across the spectrum of mental health settings: Working with gay, lesbian and bisexual patients in consultation-liaison services, inpatient treatment facilities, and community outpatient mental health centers. In R. P. Cabaj & T. S. Stein (Eds.), Textbook of homosexuality and mental health (pp. 687-704). Washington, DC: American Psychiatric Press, Inc. McGuire, D. K. (1995). Intimacy and internalized homophobia: Predictors of relationship satisfaction in lesbian couples (Doctoral dissertation. University of Akron, 1995). Dissertation Abstracts Tntemational. 56-n8R^ 4676

PAGE 217

210 McLeod, A., & Crawford, I. (1998). The postmodern family: An examination of the psychosocial and legal perspectives of gay and lesbian parents. In G. M. Herek (Ed.), Stigma and sexual orientation: Understanding prejudice against lesbians, gav men, and bisexuals (pp. 21 1-222). Thousand Oaks, CA: Sage Publications, Inc. Mead, M. (1973). Ritual in social crisis. In J. D. Shaughenessy (Ed.). The roots of ritual (pp. 87-101). Grand Rapids, MI: Erdmans Publishing. Melamed, D. K. (1992). Internalized homophobia and lesbian couple functioning (Doctoral dissertation. City University of New York, 1992). Dissertation Abstracts International. 53-07B. 3784. Mencher, J. (1984). Changing the lens on female personality development: A challenge to the notion of fusion in lesbian relationships. Unpublished masters thesis. Smith College School for Social Work, Northhampton, MA. Mencher, J. (1990). Intimacy in lesbian relationships: A critical re-examination of fusion. Work in Progress. 12. Stone Center Working Papers Series, Wellesley College. Mendola, M. (1981). The Mendola report: A new look at gay couples. New York: Crown Publishers. Mercier, L. R., & Berger, R. M. (1989). Social service needs of lesbian and gay adolescents: Telling it their way. Journal of Social Work and Human Sexuality. 8 (1). 75-95. . Meredith, W. (1985). The importance of family traditions. Wellness Perspectives. 2. 17-19. Meredith, W., Abbott, D., Lamanna, M., & Sanders, G. (1989). Rituals and family strengths. Familv Perspectives. 23. 75-83. Metz, M. F., Rosser, B. R. S., & Strapko, N. (1994). Differences in conflict resolution styles among heterosexual, gay, and lesbian couples. Journal of Sex Research. 31(4), 293-308. Meyer, C.J. (1989). Social support and couple closeness among lesbians (Doctoraldissertation, University of Minnesota, 1988). Dissertation Abstracts International. 50-ORA. 2667. Miller, J. B. (1984). The development of women's sense of self Work in Progress, 12, Stone Center working Papers Series, Wellesley College. Miller, J. B. (1986). What do we mean by relationships? Work in Progress. 22. Stone Center working Papers Series, Wellesley College.

PAGE 218

211 Miller, J. B. (1988). Connections, disconnections, and violations. Work in Progress. 33. Stone Center working Papers Series, Wellesley College. Mills, K. I. (1997, Summer). Stop the hate. Human Rights Campaign Quarterly . PP-8-9. . , , Minton, H. L., «& McDonald, G. J. (1984). Homosexual identity formation as a developmental process. Journal of Homosexuality. 9. 91-104. Mitchell, R. E., Billings, A. G., & Moos, R. H. (1982). Social support and wellbeing: Implications for prevention programs. Journal of Primary Prevention. 3(2). 7798. Mize, L. K. (1988). Family of origin ritual behavior and its impact on nevermarried women's current sense of meaning (Doctoral dissertation, Texas Women's University, 1988). Dissertation Abstracts International. 49-08B. 3141. Moore, S., & Myerhoff, B. (1977). Secular rituals: Form and meaning. Amsterdam: VanGorcum. ., . . Moos, R. H., & Moos, B. S. (1981). Familv Environment Scale. Palo Alto, CA: Consulting Psychologists Press. " Morgan, K. S. (1992). Caucasian lesbian's use of psychotherapy: A matter of attitude? Psychology of Women Quarterlv. 16. 127-130. Morin, S. F. (1977). Heterosexual bias in psychological research on lesbianism and male homosexuality. American Psychologist. 32. 629-637. Murphy, B.C. (1989). Lesbian couples and their parents: The effects of perceived parental attitudes on the couple. Journal of Counseling and Development. 68. 46-51. Murphy, B.C. (1992). Counseling lesbian couples: Sexism, heterosexism, and homophobia. In S. H. Dworkin & F. J. Gutierrez (Eds.), Counseling gay men and lesbians: Journey to the end of the rainbow (pp. 63-80). Alexandria, VA: American Association for Counseling and Development. Murstein, B. I. (1970). Stimulusvalue role: A theory of marital choice. The Journal of Marriage and the Family. 32. 465-481 . Myerhoff, B. G. (1977). We don't wrap herring in a printed page: Fusions, fictions, and continuity in secular ritual. In S. F. Moore & B. G. Myerhoff (Eds.), Secularritual (pp. 199-224). Aspen: VanGorcum.

PAGE 219

212 Neu, D. L. (1995). Women's empowerment through feminist rituals. Women and Therapy. 16 (2/31. 185-200. Ng, M. (1991). Family of origin differentiation and marital satisfaction in healthy-like, healthy-unlike, unhealthy-like, and unhealthy-unlike couples (Doctoral dissertation, Texas Woman's University, 1991). Dissertation Abstracts International. 5208A. 3090. Norton, H. S. (1995). Intimacy and relationship satisfaction in heterosexual, gay male, and lesbian couples (Doctoral dissertation, Massachusetts School of Professional Psychology, 1995). Dissertation Abstracts International. 56-06B. 3458. Oberstone, A. K., & Sukoneck, H. (1976). Psychological adjustment and life style of single lesbians and single heterosexual women. Psychology of Women Quarterly. 1 (2). 172-188. O'Leary, K. D., & Smith, D. A. (1991). Marital interaction. Aimual Review of Psychology. 42. 191-212. Olson, D.H. (1986). Circumplex model Vll: Validation studies and FACES III. Family Process. 25. 337-351. Olson, M.E. (1988). Factors that are related to the psychological well-being of lesbian women (Doctoral dissertation, California School of Professional Psychology, 1988). Dissertation Abstracts International. 49-lOB. 4553. O'Neill, M. R. (1998). Depression vulnerability in gay, lesbian, and bisexual young adults (Doctoral dissertation, State University of New York, 1998). Dissertation Abstracts International. 59-04B. 1864. Oring, E. (1984). Dyadic traditions. Journal of Folklore Research. 21. 19-28. Pagelow, M. (1980). Heterosexual and lesbian single mothers: A comparison of problems, coping, and solutions. Journal of Homosexuality. 5. 189-205. Park, J. (1994). The development of lesbian couple relationships (Doctoral dissertation, Boston University, 1994). Dissertation Abstracts international. 54-1 1 A . 3994. Paul, W., & Weinrich, J. D. (1982). Whom and what we study: Definition and scope of sexual orientation. In W. Paul, J. D. Weinrich, J. C. Gonsiorek, & M. E. Hotvedt (Eds.), Homosexua lity: Social, psychological, and biological issues (pp. 23-28). Beverly Hills: Sage. Pearlman, S. F. (1988). Distancing and connectedness: Impact on couple formation in lesbian relationships. Women and Therapy. 8. 77-88.

PAGE 220

213 Peplau, L. A. (1982). Research on homosexual couples: An overview. Journal of Homosexuality. 8 (2). 3-8. Peplau, L. A., & Cochran, S. D. (1990). A relational perspective on homosexuality. In D. P. Mc Whiter, S, A. Sanders, & J. M. Reinisch (Eds.), Homosexualitv/heterosexualitv: Concepts of sexual orientation (pp. 321-349). New York: Oxford University Press. Peplau, L. A., Padesky, C, & Hamiliton, M. (1982). Satisfaction in lesbian relationships. Journal of Homosexuality. 8 (2). 23-35. Pett, M. A., Lang, N., & Gander, A. (1992). Late-life divorce: Its impact on family rituals. Journal of Family Issues. 13 (4). 526-552. Pilisuk, M. (1982). Delivery of social support: The social inoculation. American Journal of Orthopsychiatry. 52 (1). 20-31. Pope, M. (1996). Gay and lesbian career counseling: Special career counseling issues. In A. L. Ellis «fe E. D. B. Higgle (Eds.), Sexuality on the job: Issues and services (pp. 91-105). New York: Harrington Park Press/Haworth Press. Procidano, M E., & Heller, K. (1983). Measures of perceived social support from friends and family: Three validation studies. American Journal of Community Psvchologv. li en. 1-24. Rankow, E. J. (1995). Lesbian health issues for the primary care provider. Journal of Family Practice. 40 (51. 486-493. Reiss, D. (1981). The family's construction of reality. Cambridge, MA: Harvard University. Reiss, D. (1982). The working family. A researcher's view of health in the household. American Journal of Psvchiatry. 139. 1412-1428. Reiss, D., & Oliveri, M. E. (1991). The family's conception of accountability and competence: A new approach to the conceptualization and assessment of family stress. Family Process. 30. 193-214. Rey, A. M., & Gibson, P. R. (1997). Beyond high school: Heterosexuals' selfreported anti-gay/lesbian behaviors and attitudes. In M. B. Harris (Ed.), School experiences of gav and lesbian youth: The invisible minority (pp. 65-84). New York: Harrington Park Press/Haworth Press.

PAGE 221

Richardson, D. (1981). Theoretical perspectives on homosexuality. In J. Hart & D. Richardson (Eds.), The theory and practice of homosexuality (pp. 5-37). London: Routledge & Kegan Paul. Riddle, D. I., & Sang, B. E. (1978). Psychotherapy with lesbians. Journal of Social Issues. 34. 84-100. Roberts, J. (1988). Setting the frame: Definition, functions, and typology of rituals. In E. Imber-Black, J. Roberts, & R. Whiting (Eds.), Rituals in families and family therapy (pp. 3-46). New York: W.W.Norton. Rogers, J. C, & HoUoway, R. L. (1991). Family rituals and the care of individual patients. Family Systems Medicine. 9. 249-259. Romano, K. (1990). The relation of perceived societal and familial homonegativism and homophobia and individual homophobia to relationship satisfaction and self-esteem in partnered lesbians (Doctoral dissertation, New York University, 1990). Dissertation Abstracts International. 5 1-04 A. 1171. Rosenthal, C. J., & Marshall, V. W. (1988). Generational transmission of family ritual. American Behavioral Scientist. 3 1 (6). 669-684. Roth, A. (1985). Psychotherapy with lesbian couples: Individual issues, female socialization, and the social context. Journal of Marital and Family Therapy. 1 1. 273286. Rusbult, C. E. (1980). Commitment and satisfaction in romantic associations: A test of the investment model. Journal of Experimental Social Psychology. 16. 1 72-1 86. Rusbult, C.E. (1983). A longitudinal test of the investment model: The development (and deterioration) of satisfaction and commitment in heterosexual involvements. Journal of Personality and Social Psychology. 45. 101-1 1 7. Saghir, M. T., & Robins, E. (1973). Male and female homosexuality: A comprehensive investigation. Baltimore: Williams & Wilkins. Sameroff, A. J., & Fiese, B. H. (1992). Family representations of development. In I. E. Sigel, A. V. McGillicuddy-DeLisi, & J. L. Goodnow (Eds.), Parental belief systems: The psycholog ical consequences for children (4* ed., pp. 347-369). Hillsdale, NJ: Erlbaum. Sang, B. E. (1978). Lesbian research: A critical evaluation. In G. Vida (Ed.), Our right to love (pp. 80-87). Englewood Cliffs, NJ: Prentice Hall. Schafer, C, Coyne, J. C, & Lazarus, R. S. (1981). The health-related functions of social support. Journal of Behavioral Medicine. 4 (4). 381-406

PAGE 222

Schneider, M. S. (1986). The relationships of cohabiting lesbian and heterosexual couples: A comparison. Psychology of Women Quarterly, 10, 234-239. Schvaneveldt, J., &. Lee, R. (1983). The emergence and practice of ritual in the American family. Family Perspective. 1 7. 137-143. Selvini-Palazzoli, S., Boscolo, L., Cechin, G., & Prata, G. (1977). Family rituals: A powerful tool in family therapy. Family Process, 1 6, 445-453. Siegelman, M. (1972). Adjustment of male homosexuals and heterosexuals. Archives of Sexual Behavior. 2 (1). 9-25. Silberkraus, S. L. (1995). A saphic sojourn: Evolution of a lesbian identity (Doctoral dissertation. University of Southern California, 1995). Dissertation Abstracts International. 5 1-01 A. 479. Slater, S. (1995). The lesbian family life cycle. New York: The Free Press. Snyder, D.K. (1981). Manual for the Marital Satisfaction Inventory . Los Angeles: Western Psychological Services. Snyder, N. P. (1991). The experience of intimacy: Interviews with lesbian women and heteroseximl women and men in long-term relationships (Doctoral dissertation, Harvard University, 1991). Dissertation Abstracts International. 52-06B. 3309. Solheim, C. A. (1990). Preference for change in family work roles, role conflict and conflict resolution style: A longitudinal study of wives, husbands, and couples (Doctoral dissertation. University of Minnesota, 1990). Dissertation Abstracts International. 51-09!. 3239. Sommers, M. (1982). The relationship between present social support networks and current levels of gay identity (Doctoral dissertation, California School of Professional Psychology, 1982). Dissertation Abstracts International. 43-06B. 1962. Sophie, J. (1986). A critical examination of stage theories of lesbian identity development. Journal of Homosexuality. \ 2 (2\ 39-51 , Sorensen, L., & Roberts, S. J. (1997). Lesbian uses of and satisfaction with mental health services: Results from Boston Lesbian Health Project. Journal of Homosexuality. 33 riV 35-49. Spanier, G. B. (1976). Measuring dyadic adjustment: New scales for assessing the quality of marriage and similar dyads. Journal of Marriage and the Family. 38. 1 5-78

PAGE 223

Spanier, G. B., & Filsinger, E. E. (1983). The dyadic adjustment scale. In E. E. Filsinger (EdX Marriage and family assessment (pp. 155-168). Beverly Hills, CA: Sage Publications. Spanier, G. B., & Thompson, L. (1982). A confirmatory analysis of the dyadic adjustment scale. Journal of Marriage and the Family. 44. 731-738. State of Florida Bureau of Economic and Business Research. 1998. Florida Statistical Abstract. 1998 . Gainesville. FL: University of Florida. Stein, S. J., Girodo, M., & Dotzenroth, S. (1982). The interrelationships and reliability of a multilevel behavior-based assessment package for distressed couples. Journal of Behavioral Assessment. 4. 343-360. > Stemp, P. S., Turner, R. J., & Noh, S. (1986). Psychological distress in postpartum period: the significance of social support. Journal of Marriage and the Family. 48 f2). 271-278. . . .. . . : Surrey, D. (1985). The "self-in-relation": A theory of women's development. Work in Progress. 13. Stone Center working Papers Series, Wellesley College. Tait, D. (1997). Stress and social support networks among lesbian and heterosexual women: A comparison study. Smith College Studies in Social Work. 67 (2). 213-224. Thompson, N. L., McCandless, B. R., & Strickland, B. R. (1971). Personal adjustment of male and female homosexuals. Journal of Abnormal Psychology. 78 (2). 237-240. Troiden, R. R. (1979). Becoming homosexual: A model of gay identity acquisition. Psychiatry. 42 (41. 362-373. Tuller,N. R. (1978). Couples: The hidden segment of the gay world. Journal of Homosexuality. 3 (4). 331-343. Turner, R. J., Frankel, B. G., & Levin, D. M. (1983). Social support: Conceptualization, measurement, and implications for mental health. Research in Community Mental Health. 3. 67-11 1 . Turner, V. (1967). The forest of symbols: Aspects of Ndembu ritual. Ithaca, NY: Cornell University Press. Turner, V. (1969). The ritual process: Structure and anti-structure. Ithaca, NY: Cornell University Press. U.S. Bureau of the Census, 1998a. County and citv extra: annual metro, city, and county data book 1998. Washmgton, DC: U.S. Department of Commerce.

PAGE 224

U.S. Biireau of the Census, 1998b. Statistical abstract of the United States. 1998 . Washington, DC: U.S. Department of Commerce. Valenti, R. M. (1997). The effects of utlizing positive ritual behavior in marriage on marital satisfaction and intimacy (Doctoral dissertation, Seton Hall University, 1997). Dissertation Abstracts International. 5 8-09 A. 3432. van der Hart, O. (1983). Rituals in psychotherapy: Transition and continuity. New York: Irvington. Vargo, S. (1987). The effects of women's socialization on lesbian couples. In Boston Lesbian Psychologies Collective (Eds.), Lesbian psychologies: Explorations and challenges (pp. 161-173). Urbana: University of Illinois Press. Walsh, W. M. (1995). Social support factors among lesbians with a positive sexual identity (Doctoral dissertation, California School of Professional Psychology, 1995). Dissertation Abstracts International. 56-05B. 2891. Wayment, H. A., & Peplau, L. A. (1995). Social support and well-being among lesbian and heterosexual women: A structural modeling approach. Personality and Social Psychology Bulletin. 21 (11). 1 189-1 199. . Weiss, R. (1974). The provision of social relationships. In Z. Rubin (Ed.), Doing unto others (pp. 17-26). Englewood Cliffs, N J: Prentice-Hall. Wendell, J. S. (1986). An exploratory examination of the relationship between dyadic differences in lesbian sexual orientation disclosure and dyadic adjustment of lesbian couples in committed relationships (Doctoral dissertation. New York University, 1985). Dissertation Abstracts International. 47-04A. 1200. Whiteside, M. F. (1988). Creation of family identity through ritual performance in early remarriage. In E. Imber-Black, J. Roberts, & R. A. Whiting (Eds.), Rituals in families and family therapy (pp. 276-304). New York: W. W. Norton. Whiting, R. A. (1988). Guidelines to designing therapeutic rituals. In E. ImberBlack, J. Roberts, & R. A. Whiting (Eds.), Rituals in families and family therapy (pp. 84109). New York: W. W. Norton. Wilcox, B. L., & Vemberg, E. (1985). Conceptual and theoretical dilemmas facing social support research. In I. Sarason & B. Sarason (Eds.), Social support: Theory, res earch, and applications (pp. 3-20). Netherlands: Martinez Nijhoff. Wilson, M. L., & Green, R. L. (1971). Personality characteristics of female homosexuals. Psychological Reports. 28. 407-412.

PAGE 225

218 Wise, G. (1986). Family routines, rituals, and traditions: Grist for the family mill and buffers against stress. In S. Van Zandt (Ed.), Family strengths 7: Vital connections (pp. 32-41). Lincoln, NE: Center for Family Strengths. Wolin, S. J., & Bennett, L. A. (1984). Family rituals. Family Process. 23. 401420. Wolin, S. J., Bennett, L., Noonan, D., & Teitelbaum, M. (1980). Disrupted family rituals: A factor in the intergenerational transmission of alcoholism. Journal of Studies in Alcohol. 41. 199-214. Wood, M.J. (1983). Lesbian and heterosexual couples: A comparison of available social support and three indicators of effective couple functioning (Doctoral dissertation. University of Maryland, 1983). Dissertation Abstracts International. 4502B. 692. Zacks, E., Green, R., & Marrow, J. (1988). Comparing lesbian and heterosexual couples on the circumplex model: An initial investigation. Family Process. 27. 471-484.

PAGE 226

BIOGRAPHICAL SKETCH In 1981 Caroline Pace graduated from the University of Missouri with a Bachelor of Science degree in secondary education with a major in science and minors in journalism and mathematics. In 1986 she received a Master of Education degree from the University of Missouri in school guidance and counseling. In 1991 she received a Specialist in Education degree in counseling from the University of Florida. In 1992 she began the doctoral program in couples and family therapy at the University of Florida. 219

PAGE 227

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. '* Ellen Amatea, Chairperson Professor of Counselor Education I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Mary Howard-Hamilton Associate Professor of Counselor Education I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality as a dissertation for the degree of Doctor of Philosophy. M. David Miller Professor of Foundations of Educatioii I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality as a dissertation for the degree of Doctor of Philosophy. Phyllis Meek Associate Professor of Counselor Education This dissertation was submitted to the Graduate Faculty of the College of Education and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. May, 1999 ;an, Colleg Dean, Graduate School