Assessing nontraditional couples: Validity of the marital satisfaction inventory–revised with gay, lesbian, and cohabiting heterosexual couples
Means-Christensen, Adrienne J
Thirty-one gay male couples and 28 lesbian couples were compared with 36 cohabiting heterosexual couples using the Marital Satisfaction Inventory-Revised (MSI-R), a multidimensional measure of relationship functioning intended for use with both traditional and nontraditional couples. Analyses of scales’ internal consistency and factor structure supported the construct validity of this measure with nontraditional couples. Analyses of mean profiles on the MSI-R indicated that cohabiting opposite-gender and same-gender couples were more alike than different, and were more similar to nondistressed samples of married heterosexual couples from the general community than to couples in therapy. Implications of current findings for clinical assessment and intervention are considered, and directions for future research are proposed.
To what extent does the empirical literature inform clinical assessment and interventions with nontraditional couples? Although the conceptualization of couples has expanded over the last 25 years to include gay male and lesbian couples as well as cohabiting heterosexual dyads, both research and clinical interventions with nontraditional couples have lagged behind. Studies of couple interventions have emphasized married couples as the research participants and intended consumers. Similarly, assessment techniques developed with married heterosexual dyads rarely have been examined for their potential application to nontraditional couples. In the absence of studies examining the generalizability of assessment tools’ psychometric properties to cohabiting or gay and lesbian couples, little confidence can be placed in their use for describing relationship functioning or directing clinical interventions in these populations.
Findings concerning the prevalence of nontraditional couples in the United States highlight the importance of examining clinical interventions and assessment techniques with these groups. Research indicates that approximately 4% of men identify themselves as exclusively or mostly gay, and 3% of women identify themselves as exclusively or primarily lesbian (Barringer, 1993), although some studies report somewhat lower percentages (e.g., Laumann, Gagnon, Michael, & Michaels, 1994). In the United States Census for the year 2000, approximately 11 million Americans (3.8%) identified themselves as cohabiting with an unmarried partner (U.S. Census Bureau, 2001). Same-gender couples constitute approximately 2% of all households, translating into roughly one million same-gender couples throughout the United States (Clark & Fields, 1999). Moreover, state prohibitions against recognizing same-gender couples as married are likely to result in gay male and lesbian couples comprising a disproportionately high percentage of cohabiting dyads but not being identified as such.
Although reliable data regarding participation in couple therapy by nontraditional couples are sparse, the few available findings regarding gay and lesbian couples document the importance of addressing both common and unique concerns and examining clinical techniques specific to this population. Previous research indicates that gay males and lesbians participate in psychotherapy at higher rates than their heterosexual counterparts (Bell & Weinberg, 1978; Morgan, 1992). Modrcin and Wyers (1990) found that 54% of the lesbian couples and 32% of the gay couples in their community sample had previously sought the services of a professional to address problems in their current relationship; moreover, 86% of lesbians and 60% of gay males in their sample stated that they would seek professional help in the future if problems arose in their relationship.
Various factors may contribute to the lag in research and clinical techniques developed specifically for nontraditional couples. One factor may include institutional biases against same-gender couples, including lack of legal equality with married heterosexual couples in most states, historical views of homosexuality as a mental disorder in previous versions of the American Psychiatric Association’s (APA) Diagnostic and Statistical Manual (DSM; APA, 1952, 1968), and continuing biases against unmarried heterosexual and same-gender couples by many religious denominations and political groups. Another factor may be ungrounded assumptions that gay male and lesbian relationships are different from heterosexual relationships in fundamental ways, such as being generally short-term and nonmonogamous (Bell & Weinberg, 1978) and therefore more difficult to study or less deserving of it. The modest research in this area, however, indicates that these relationships are quite similar to heterosexual relationships on a number of dimensions, including relationship-maintenance behaviors (Haas & Stafford, 1998), the effects of gender role on communication problems (Julien, Arellano, & Turgeon, 1997), aspects of relationship quality, such as liking of partner (Kurdek & Schmitt, 1986b) and the relation between stage of relationship and relationship quality (Kurdek & Schmitt, 1986a). Findings also suggest that gay and lesbian couples may be more similar to married heterosexual couples than to cohabiting heterosexual couples in terms of their relationship satisfaction and love for partner (Kurdek & Schmitt, 1986b). A third factor underlying the paucity of research with nontraditional couples may involve researchers’ desire to preserve the relative homogeneity of their samples at the potential cost of their findings’ generalizability.
Although several studies of gay and lesbian couples have adopted clinical measures developed primarily with heterosexual married couples (e.g., Kurdek, 1994; Kurdek & Schmitt, 1986b; Metz, Rosser, & Strapko, 1994; Testa, Kinder, & Ironson, 1987), very few have specifically examined the psychometric properties of these measures when used with nontraditional couples as compared to findings for the heterosexual married couples on which they were developed. Exceptions include two studies examining measures of fear of intimacy (Greenfield & Thelen, 1997) and couple functioning (Burgoyn, 2001) with gay and lesbian couples. In both studies, findings generally supported use of the measure with nontraditional couples, but limited data for the measure with diverse populations of distressed and nondistressed couples render comparisons of nontraditional with traditional couples more tenuous.
STUDY OVERVIEW
This study sought to address two questions. First, to what extent does the Marital Satisfaction Inventory-Revised (MSI-R; Snyder, 1997), a multidimensional measure of relationship satisfaction developed with married heterosexual couples and widely used in research and clinical applications, retain its psychometric properties when used with cohabiting heterosexual or gay and lesbian couples? Second, given evidence supporting use of this measure with nontraditional couples, how do cohabiting heterosexual couples and gay and lesbian couples compare with each other and with previous samples of heterosexual married couples from community and clinic populations?
From one perspective, gay and lesbian couples logically share similar challenges in their relationships to those confronting heterosexual couples. For example, independent of sexual orientation, all couples need to define boundaries separating their own relationship from others, resolve conflicts and reach decisions together, deal with issues of emotional and sexual intimacy, and manage resources of time and money. However, gay and lesbian couples also face unique concerns from having grown up in, and continuing to contend with, a predominantly heterosexual society promulgating both formal and informal policies discriminatory against individuals with a same-gender sexual orientation. Indeed, because of the stigmatization they experience beginning early in life, the psychosocial development of gays and lesbians has been regarded by several researchers as similar to that of ethnic and racial minorities (Cass, 1984; Coleman, 1982; Fassinger, 1991; Ponse, 1978; Troiden, 1989).
Because of their potential differences from heterosexual couples in various aspects of both individual and relationship functioning, assessment techniques used for clinical or research purposes with gay and lesbian couples need to be evaluated for both their reliability and construct validity. Whenever instruments developed for one group are applied to a new group, their psychometric integrity in the new application needs to be evaluated separately (Cuellar, 1998; Okazaki & Sue, 1995). For example, scales’ internal consistency and factor structure need to be examined to determine whether the linkage of overall relationship satisfaction to emotional expressiveness, problem resolution, and conflict in such areas as finances, use of discretionary time, and the sexual relationship is similar for both nontraditional couples and test-development samples.
Moreover, a measure could retain its reliability and validity across diverse populations but still reflect group mean differences influencing the interpretation of test findings for either community or clinical samples. For example, if mean profiles for community samples of gay and lesbian couples were to differ substantially from those for community samples of heterosexual married or cohabiting unmarried couples, this could potentially alter the normative framework and interpretive guidelines for assessing relationship dysfunction when the assessment measure is applied to same-gender couples in a clinical setting.
This study examined use of the MSI-R with nontraditional couples. The MSI-R was developed to address both psychometric and clinical concerns in evaluating couples’ relationships. The MSI-R is a multidimensional, self-report measure of relationship satisfaction with more than 20 years of empirical and clinical study supporting its reliability, validity, and utility in assessing and treating distressed couples (Snyder & Aikman, 1999). The inventory includes 150 true-false items and permits comparisons of partners’ independent evaluations of their relationship across 13 profile scales, and comparison of these evaluations to profiles typical of couples in therapy or from the general population. In the revision of the original instrument (MSI; Snyder, 1979), the MSI-R items were modified to refer to relationship and partner, rather than marriage and spouse, to facilitate its use with nontraditional couples. Current analyses explored the internal consistency of MSI-R scales and the instrument’s factor structure for both cohabiting heterosexual and same-gender couples and compared results for these samples with those reported previously for heterosexual married couples. Mean MSI-R profiles were derived for nontraditional couples sampled in this study, and these profiles were compared to previous findings for married, heterosexual community, and clinic samples.
METHOD
Participants and Procedures
Gay male and lesbian couples. A sample of 31 cohabiting gay male couples and 28 cohabiting lesbian couples were recruited from three diverse communities in Texas, including a major metropolitan area, a midsized community that is home to a major teaching and research university, and a semirural community in the southernmost region of Texas. This sociodemographically diverse sample was intended to preclude concerns regarding generalizability hampering many previous studies of gay and lesbian couples dominated by primarily Caucasian samples (Brown, 1995; Julien et al., 1997; Modrcin & Wyers, 1990). Gay and lesbian couples initially were made aware of the study through community organizations, flyers and advertisements posted in offices serving gay and lesbian residents, and by word of mouth by previous participants. For approximately two-thirds of the sample, participants contacted the investigators for study materials, and an appointment was scheduled during which research assistants went to the homes of participants, reviewed the informed consent procedure with them, and left study materials for participants to complete. Research assistants returned to the couple’s home later that day to pick up the study materials and to answer any questions. The remaining one-third of the sample received research materials directly from a gay and lesbian organization in their community and returned materials to the investigators by mail. There were no significant differences in couples’ responses to measures across the two sampling procedures.
Gay male participants averaged 36.6 years in age (SD = 8.9) and had been in a relationship with their current partner an average of 9.1 years. Ethnic representation included 53% non-Hispanic White American and 42% Hispanic. Gay participants averaged 15.9 years of education; 87% were employed outside the home and, among these, approximately 54% held professional or managerial positions and 42% held administrative, clerical, or technical positions. Lesbian participants averaged 32.9 years in age (SD = 8.8) and had been in a relationship with their current partner an average of 4.9 years. Ethnic representation included 56% non-Hispanic White American and 40% Hispanic. Lesbian participants averaged 16.5 years of education; 95% were employed outside the home and, among these, approximately 58% held professional or managerial positions, and 33% held administrative, clerical, or technical positions.
Cohabiting heterosexual couples. A sample of 36 unmarried, cohabiting heterosexual couples was recruited from the same university community from which some of the gay male and lesbian couples were sampled. Recruitment procedures were similar and included flyers and advertisements posted throughout the community and word of mouth by previous participants. These participants contacted the investigators, an appointment was scheduled, and research assistants delivered study materials to their home and retrieved them later that same day.
For cohabiting heterosexual couples, men averaged 23.1 years in age (SD = 3.8) and women averaged 21.1 years (SD = 2.1). Couples’ current relationship averaged 1.2 years in length. Ethnic representation included 77% non-Hispanic White American, 14% Hispanic, 4% African American, and 4% Asian American. Participants averaged 15.3 years of education; 69% were employed outside the home and, among these, approximately 20% held professional or managerial positions and 44% held administrative, clerical, or technical positions.
Measures
Marital Satisfaction Inventory-Revised. All respondents completed the MSI-R-a 150-item, true-false, self-report measure of relationship functioning designed to identify both the nature and intensity of distress in distinct areas of partners’ interaction. The MSI-R includes two validity scales, one global distress scale, and 10 additional scales assessing specific dimensions of the relationship. Each of the scales, excluding the validity scales (INC and CNV) and role orientation (ROR), are scored in a direction whereby higher scores reflect higher levels of relationship distress. Scale names, abbreviations, and brief descriptions are as follows:
Inconsistency (INC). A validity scale assessing the individual’s consistency in responding to item content. (High scores reflect greater inconsistency.) Conventionalization (CNV). A validity scale assessing individuals’ tendencies to distort the appraisal of their relationship in a socially desirable direction. (High scores reflect denial of common relationship shortcomings.)
Global Distress (GDS). Measures individuals’ overall dissatisfaction with the relationship.
Affective Communication (AFC). Evaluates individuals’ dissatisfaction with the amount of affection and understanding expressed by their partner.
Problem-Solving Communication (PSC). Assesses the couple’s general ineffectiveness in resolving differences.
Aggression (AGG). Measures the level of intimidation and physical aggression experienced by respondents from their partner.
Time Together (TTO). Evaluates the couple’s companionship as expressed in time shared in leisure activity.
Disagreement About Finances (FIN). Measures relationship discord regarding the management of finances.
Sexual Dissatisfaction (SEX). Assesses dissatisfaction with the frequency and quality of intercourse and other sexual activity.
Role Orientation (ROR). Evaluates the respondent’s advocacy for a traditional versus nontraditional orientation toward marital and parental gender roles. (High scores reflect a nontraditional orientation.)
Family History of Distress (FAM). Reflects the disruption of relationships within the respondent’s family of origin.
Dissatisfaction With Children (DSC). Assesses the relationship quality between respondents and their children as well as parental concern regarding one or more children’s emotional and behavioral well-being.
Conflict Over Child Rearing (CCR). Evaluates the extent of conflict between partners regarding childrearing practices.
The MSI-R is administered to each partner separately and requires approximately 25 minutes to complete. Individuals’ responses are scored along the 13 profile scales using normalized T-scores based on gender-specific norms. Thus, for each scale the mean score for nondistressed couples from the community is 50, with a standard deviation of 10. In this study, the ROR scale was not administered to gay or lesbian participants because items refer specifically to distribution of gender-linked roles in heterosexual dyads, and the scale has diminished relevance to same-gender couples. (Sample items include, “The man should be the head of the family,” and “The most important thing for a woman is to be a good wife and mother.”) In addition, because fewer than 10% of either heterosexual cohabiting couples or gay and lesbian couples reported having children, the two child-related scales (DSC and CCR) were omitted from this study.
Previous studies have supported the internal consistency and temporal stability of MSI-R scales, as well as their convergent and discriminant validity (see Snyder & Aikman, 1999, for a review). Specifically, coefficients of internal consistency for the MSI-R scales derived from a combined sample of 2,040 individuals from the community and 100 individuals in couple therapy ranged from .70 to .93 (M = .82); temporal stability coefficients derived from 210 individuals from the community retested after 6 weeks ranged from .74 to .88 (M = .79). In a comparison of 50 clinic couples and 77 community couples matched on demographic indices, each of the MSI-R scales discriminated between the community and clinic couples at p
RESULTS
Internal Consistency of MSI-R Scales
To what extent do the MSI-R scales retain their internal consistency with nontraditional couples? Scales’ internal consistency was examined separately for cohabiting heterosexual couples and for gay male and lesbian couples, and results for each were compared to findings for heterosexual married couples reported in the test manual (Snyder, 1997). Coefficients of internal consistency (Cronbach’s alpha) and mean inter-item correlations are presented in Table 1. For a combined sample of 1,020 heterosexual married couples from the community and 50 married couples in therapy (Snyder, 1997), Cronbach’s alpha coefficients for the 10 scales also included in this study ranged from .78 to .93, M = .84. By comparison, Cronbach’s alpha coefficients for these same scales for 36 heterosexual cohabiting couples sampled in this study ranged from .75 to . 93 (M = .81), and for 59 gay and lesbian couples the coefficients ranged from .78 to .91 (M = .83, excluding ROR). Mean inter-item correlations for these scales for cohabiting couples ranged from .23 to .40 (M = .28) and for gay and lesbian couples ranged from .22 to .36 (M = .31). Overall, these findings confirm that the MSI-R scales retain their high levels of internal consistency across traditional married and nontraditional samples and that, within the latter group, internal consistency coefficients do not differ significantly across unmarried heterosexual versus gay and lesbian couples.
Intercorrelations and Factor Structure of MSI-R Scales
How do the scales’ factor structures with nontraditional couples support the construct validity of the MSI-R with these groups? Whereas analyses of internal consistency examine items’ interrelatedness within scales, analyses of scale correlations and factor structures examine the interrelatedness of constructs across scales. In this study, correlation and factor analyses of individuals’ scores were used to examine the MSI-R’s structural equivalence across cohabiting heterosexual versus gay and lesbian couples and to compare factor structures derived from these groups to that reported for married heterosexual couples. Results of these analyses are presented in Tables 2 and 3.
For a sample of 1,070 community and clinic married couples, Snyder (1997) reported a three-factor solution accounting for 64% of the common variance shared among scales. The first factor was composed primarily of broadband measures of relationship affect (CNV, GDS, and TTO) and communication difficulties (AFC and PSC), and secondarily by specific domains of relationship conflict (AGG, FIN, SEX, and CCR). A second factor was defined primarily by scales reflecting child-related concerns (DSC and CCR), whereas the third and smallest factor was defined almost entirely by individuals’ views toward gender-linked marital and parental roles (ROR). Because the two child-related scales (DSC and CCR) were not scored in the present study of nontraditional couples, and because ROR was not administered to gay and lesbian couples, comparisons of factor structure across cohabiting heterosexual versus gay and lesbian couples were restricted to analyses of the remaining nine scales. Because emergence of a child-related factor was precluded in the present nontraditional samples, a two-factor solution was anticipated. However, given the additional deletion of ROR to facilitate comparison across heterosexual and same-gender dyads, the scale composition of the anticipated second factor was not predicted beforehand.
Analyses used principal-components factor analysis with orthogonal rotation, with retention of components having initial eigenvalues of 1.0 or greater. For heterosexual cohabiting couples, two factors emerged accounting for 42.6% and 13.3% of the common variance, respectively. As anticipated, similar to findings for heterosexual married couples, the first factor was composed primarily of broadband measures of relationship affect (CNV, GDS, and TTO) and communication difficulties (AFC and PSC), and secondarily by specific domains of relationship conflict (AGG and FIN). The second, smaller factor for cohabiting couples was defined primarily by difficulties in the sexual relationship (SEX), history of conflict in the family of origin (FAM), and the absence of physical and verbal aggression (AGG).
For gay and lesbian couples, again two factors emerged accounting for 52.7% and 12.8% of the common variance, respectively. Similar to results for cohabiting heterosexual couples, the first factor was composed primarily of broad-band measures of relationship affect (CNV, GDS, and TITO) and communication difficulties (AFC and PSC), and secondarily by specific domains of relationship conflict (AGG and FIN), but also by difficulties in the sexual relationship (SEX). The second, smaller factor for gay and lesbian couples was defined primarily by history of conflict in the family of origin (FAM) and the absence of physical and verbal aggression (AGG), but also by modest loadings reflecting difficulties in the sexual relationship (SEX) and the absence of disagreements about finances (FIN).
Because differences in the relation of sexual satisfaction (SEX) to overall relationship satisfaction (GDS) might vary more as a function of gender than sexual orientation, the correlation between these two scales was examined for each of the four gender-by-sexual-orientation subgroups. Correlations between SEX and GDS for heterosexual men and women were .15 and .30, respectively-considerably less than those for gay and lesbian participants (.44 and .54, respectively)and less than the correlation between these two scales for the standardization sample (.53 pooled across gender). The lower correlations for heterosexual men in this study may relate to their relatively greater satisfaction and reduced variability of scores in the sexual domain (M = 44.9 and SD = 7.88, compared to M = 50.0 and SD = 9.63 for heterosexual men in the standardization sample).
Thus, overall the factor structures of MSI-R scales for nontraditional couples have significant similarities to the structure obtained for heterosexual married couples-particularly in terms of a large, dominant component reflecting overall relationship affect and communication efficacy. The most striking difference between heterosexual cohabiting and same-gender dyads is the stronger linkage of satisfaction with the sexual relationship to overall relationship affect and communication among gay male and lesbian couples.
Mean MSI-R Profiles for Nontraditional Couples
To what extent does the MSI-R indicate significant differences in relationship functioning across heterosexual cohabiting versus gay and lesbian couples? And to what extent do scale scores for these nontraditional groups differ from those reported for distressed and nondistressed heterosexual married couples? Mean MSI-R profiles for heterosexual and gay male participants are presented in Figure 1, and for lesbian and heterosexual female participants in Figure 2. Multivariate analyses of variance (MANOVAs) comparing the four groups showed significant effects for gender (male vs. female, F [10, 177] = 2.26, p = .02); for sexual orientation (same-gender vs. opposite-gender preference, F [10, 177] = 2.01, p = .04); and for the gender-by-sexual-orientation interaction (F [10, 1771 = 2.21, p = .02). However, after controlling for age, ethnicity, education, and occupational level of respondents, only the effect for gender remained significant (F [10, 135] = 2.55,p
Subsequent univariate analyses of variance (ANOVAs) and pairwise mean comparisons indicated that, compared to cohabiting heterosexual males, gay males were somewhat less likely to describe their sexual relationship (SEX) as highly satisfying. Lesbian participants were more likely than gay males to report better-than-average satisfaction with affective communication (AFC) in their relationship. Combined across sexual orientation, women were slightly more likely than men to describe their relationship in a socially desirable manner (CNV), as more satisfying overall (GDS), and as having better quality of leisure time together (TTO). No significant differences were found across the four groups on scales reflecting consistency of responding (INC), effectiveness of problem-solving communication (PSC), aggression (AGG), disagreement about finances (FIN), or history of family distress (FAM). The statistical significance of these univariate comparisons did not change after controlling for sociodemographic differences.
However, in all cases, the actual magnitude of group differences for the above comparisons was modest. Of 41 pairwise scale comparisons (21 across gender within sexual orientation, and 20 across sexual orientation within gender), in only four instances did the absolute difference reach even half a standard deviation (5 T-score points): Heterosexual males, compared to their gay male counterparts, reported greater satisfaction with their sexual relationship (SEX), lesbian participants, compared to gay males, responded in a more socially desirable direction (CNV), reported greater satisfaction with affective communication in their relationship (AFC), and reported greater satisfaction with the quality of leisure time together (TTO).
The meaning of group mean profiles obtained in the present study for these nontraditional couples is enhanced when compared to findings for heterosexual nondistressed married couples from the general community and distressed couples in therapy. In every instance, each of the four profiles presented in Figures 1 and 2 are more similar to profiles for heterosexual married couples from the community (obtaining mean scores of 507) than to distressed couples assessed in clinical settings (typically obtaining mean scores exceeding 607). Using a criterion of 5 T-score points (one-half a standard deviation) to denote a meaningful difference (a criterion based on scales’ internal consistency and standard error of difference), the only differences between profiles derived in this study for cohabiting heterosexual and same-gender couples were that, compared to traditional community samples: (a) lesbian participants responded in a more socially desirable direction (CNV), reported greater satisfaction with affective communication in their relationship (AFC), and reported greater satisfaction with the quality of leisure time together (TIO); (b) both male and female cohabiting heterosexual participants reported more egalitarian role orientation attitudes (ROR), and (c) cohabiting heterosexual males reported greater satisfaction with their sexual relationship (SEX).
DISCUSSION
How do current findings bear on the usefulness of the MSI-R with nontraditional couples? First, results confirm that the MSI-R scales retain their high levels of internal consistency across traditional married and nontraditional samples and that, within the latter group, internal consistency coefficients do not differ significantly across unmarried heterosexual versus gay and lesbian couples. These findings indicate that MSI-R items intended to assess specific constructs of relationship functioning such as communication processes or specific domains of conflict demonstrate the same level of construct coherence for unmarried heterosexual or same-gender couples as for married opposite-gender couples. These results also indicate that the magnitude of score differences within individuals across scales or between partners on the same scale (i.e., the standard error of difference) required to be considered significant remains generally the same for nontraditional as for traditional couples-that is, approximately 56 T-score points.
Second, overall the factor structures of MSI-R scales for nontraditional couples have significant similarities to the structure obtained for heterosexual married couples, particularly in terms of a large, dominant component reflecting overall relationship affect and communication efficacy. This finding supports the construct validity of MSI-R scales in that the linkage of overall relationship satisfaction to emotional expressiveness, problem resolution, and conflict in such areas as finances, use of discretionary time, and the sexual relationship is similar for cohabiting same-gender and opposite-gender couples when compared to each other and to heterosexual married samples. The most striking difference between heterosexual cohabiting and same-gender dyads is the stronger linkage of satisfaction with the sexual relationship to overall relationship affect and communication among gay and lesbian couples, although this may result in part from reduced variability in sexual satisfaction scores for the heterosexual group.
These findings supporting MSI-R scales’ reliability and construct validity complement subsequent mean score comparisons of cohabiting heterosexual men and women with gay men and lesbians and with previous groups of married heterosexual couples from community or clinic samples. What do these comparisons indicate? First, similar to findings by others (cf., Haas & Stafford, 1998; Kurdek, 1991, 1992, 1994; Kurdek & Schmitt, 1986a; Metz et al., 1994), cohabiting same-gender and opposite-gender couples were more alike than different, and more similar to nondistressed samples of married heterosexual couples from the general community than to couples in therapy. Statistically significant differences were few and modest in magnitude. There were no differences among groups in this study on measures of problemsolving communication (PSC), verbal or physical aggression (AGG), conflict over finances (FIN), or history of distress in the family of origin (FAM). Moreover, multivariate comparisons among groups did not reveal significant effects for sexual orientation or for gender-by-sexual-orientation interaction after controlling for sociodemographic differences.
Second, group differences observed in this study, although rarely exceeding even one-half a standard deviation (5 T-score points) on these scales, were generally consistent with findings from the few previous studies comparing same-gender and opposite-gender couples. For example, findings in this study that lesbians reported greater satisfaction with the quality of emotional expressiveness and shared leisure time in their relationship than either gay men or heterosexual women are consistent with findings of high levels of relationship satisfaction and emotional and recreational intimacy among lesbians reported by others (cf., Kurdek, 1988, 1989; Metz et al., 1994; Peplau, Padesky, & Hamilton, 1982; Zacks, Green, & Marrow, 1988). In contrast, the modest trend for gay men to appraise their relationship more positively than heterosexual men as reported by Metz et al. (1994) was not replicated in this study. Although gay men in this study did not report any less relationship satisfaction when compared to a previous sample of heterosexual married men, the cohabiting heterosexual men sampled in this study reported greater than usual satisfaction in their sexual relationship, with a trend toward greater satisfaction with leisure time spent with their partner.
Clinical Implications
What implications do these findings have for clinical assessment and intervention? First, current findings support the usefulness of the MSI-R as a measure of relationship functioning with both cohabiting heterosexual couples and with gay male and lesbian couples. The generalizability of previous findings regarding internal consistency and factor structure for the MSI-R to these nontraditional couples provides some confidence when extending previous findings regarding scales’ criterion-related and predictive validity to these groups. Independent reviews of the MSI-R have emphasized its systematic empirical development, linkage of scale scores to interpretive guidelines on an actuarial basis, and usefulness for planning and evaluating therapeutic interventions (Boen, 1988; Burnett, 1987; Dixon, 1985; Fowers, 1990; Waring, 1985). Although additional studies linking MSI-R scores to independent measures of relationship functioning with unmarried same-gender and opposite-gender couples will be required to replicate both external scale correlates and scale elevations corresponding to specific relationship characteristics, current findings support the judicious application of previously derived interpretive guidelines with nontraditional couples.
Second, the similarity of community samples of cohabiting heterosexual and gay and lesbian couples in this study to nondistressed community samples of married heterosexual couples derived previously suggests that, when nontraditional couples do obtain elevated MSI-R scale scores reflecting relationship distress, these can be attributed at least in part to actual relationship difficulties, rather than entirely to marital status or sexual orientation. That is, in clinical assessment with nontraditional couples, interpretive ranges on the MSI-R scales suggesting “possible or moderate problem” and “definite or extensive problem” likely serve equally well as interpretive ranges when applied to these couples as when applied to married heterosexual couples. As with heterosexual married couples, clinical findings from the MSI-R should be carefully integrated with other interview and observational findings in developing a case formulation and comparing partners’ descriptions of their relationship with each other or with other clinical and nondistressed groups. What caveats should supplement these general conclusions? First, although psychometric findings
indicate that gay male and lesbian couples are more similar to than different from heterosexual couples, it is critical to retain a keen appreciation for the unique dynamics and sociocultural stressors encountered by same-gender couples. Brown (1995) provides a particularly useful discussion of the impact of homophobia and heterosexism on gay and lesbian couples. She notes that “many of the difficulties faced by such couples have as their primary etiology the oppressive social and interpersonal environments in which lesbians and gay men are required to operate” (p. 275). Separate from external legal and social sanctions, gay men and lesbians may themselves hold unfounded negative views regarding their own potential for enduring and fulfilling intimate relationships because of their own socialization experiences.
Second, although stereotypic gender-role attitudes and cognitive, emotional, and interpersonal styles reflecting “femininity” or “masculinity” warrant examination in any couple regardless of sexual orientation, it is particularly important to examine issues of gender role with gay male and lesbian couples. To work effectively with same-gender couples, therapists must have a clear understanding of both the strengths and deficits of each gender role as they affect relationship functioning (Brown, 1995). In the United States and other Western cultures, men’s socialization emphasizes competitiveness, assertiveness, autonomy, selfconfidence, instrumentality, and the tendency not to express intimate feelings; in comparison, women’s socialization emphasizes nurturance, emotional expressiveness, verbal exploration of emotions, and warmth (Maccoby, 1990). Because they are more likely to share similar socialization histories, same-gender partners may initially have greater familiarity with their partner’s gender-linked emotional and interpersonal styles.
Items comprising the ROR scale on the MSI-R are worded in such a way as to render them less meaningful to same-gender couples. Hence, future research may benefit from careful development of gender-role measures more applicable to gay and lesbian couples. Nevertheless, understanding the potential for gender-role conflicts in same-gender couples can be vital to intervening successfully in their relationships; careful clinical inquiry in this domain remains critical. For example, Blumenstein and Schwartz (1983) noted that because same-gender partners may be rejecting or reevaluating both male and female stereotypic gender roles, they may be more likely to experience relational conflict as they attempt to develop new patterns of emotional expressiveness, conflict resolution, balance of power, or instrumental responsibilities within their own relationship. Similarly, Brown (1995) noted that because competitiveness is such a salient aspect of male socialization, with gay male couples it is important to explore how either partner is communicating excess value or devaluation to his partner or himself based on income or other status criteria. Julien et al. (1997) suggest making partners aware of any rigidity in their gender-role attitudes, teaching them new skills associated with gender traits they do not exhibit, and instructing them in how to use these skills to modify their relationship and address areas of conflict more constructively.
In this study, opposite-gender and same-gender couples did not differ in history of conflict in their respective families of origin (FAM scale), although gay males showed a nonsignificant trend toward greater distress in this domain. Given both anecdotal and empirical evidence concerning parents’ adverse reactions to their adult children’s disclosure of a gay or lesbian sexual orientation (Brown, 1995), minimal group differences on the FAM scale in the present study are somewhat surprising; these results may be attributable in part to the length of relationship for these gay men and lesbian participants (averaging 9.1 and 4.9 years, respectively) possibly contributing to partial acceptance by members of their extended families. In their study of lesbian couples, Eldridge and Gilbert (1990) found that more than one-half of the women had not told their fathers about their sexual orientation, and one-third had not told their mothers. In a study of gay male couples (Berger, 1990), 17% of the participants had not told their parents of their sexual orientation; moreover, conflicts with family members were among the most persistent sources of difficulty reported by these participants.
Although several studies suggest that gay male and lesbian couples are more likely than heterosexual couples to avail themselves of professional services to address problems in their current relationship, other findings indicate that the outcome of counseling may frequently be compromised by the therapist’s lack of knowledge regarding the clinical and empirical literature relevant to same-gender couples or by the therapist’s own explicit or implicit heterosexist bias. For example, Modrcin and Wyers (1990) found that 38%-79% of the gay and lesbian couples who had sought relationship counseling found the experience unhelpful. A survey by the American Psychological Association’s Committee on Lesbian and Gay Concerns (1991) found that many experienced practitioners held attitudes toward same-gender couples that ranged from ignorant, but well-meaning to viciously attacking. Brown (1995) noted the importance of recognizing gay and lesbian couples as a minority group confronting unique intrapersonal as well as sociocultural stress, while rejecting social biases to regard their sexual orientation as deviant or pathological. Equally important to working effectively with gay and lesbian couples is therapists’ commitment to remaining current with emerging clinical and empirical literature in this domain.
Future Research
Research with nontraditional couples, particularly research emphasizing clinical assessment and intervention techniques, remains in its infancy. Relatively few studies of gay male and lesbian couples have used married or unmarried heterosexual couples as a comparison group; failure to do so precludes the ability to disentangle significant effects resulting from marital status, gender, sexual orientation, and the interaction of these factors (Metz et al., 1994). In addition, much of the research with couples generally, and with samegender couples specifically, has been conducted with Caucasians age 25-40 (Brown, 1995). Although this investigation included an intentional effort to increase the representation of minority participants among gay and lesbian couples, the samples’ lack of correspondence to U.S. Census characteristics and sociodemographic differences between heterosexual and same-gender couples, even when controlled statistically, potentially limit the generalizability of current findings. Hence, future studies would benefit from more systematic sampling of nontraditional couples across ethnicity, age, socioeconomic status, and related domains.
Little is known about gay and lesbian couples from a developmental lifespan perspective; even less is known about cultural differences in gender-role socialization and heterosexist biases and their impact on same-gender couples’ relationship functioning. Despite a modest trend toward increased parenthood among gay male and lesbian couples through adoption, birth, or custody from previous relationships (Brown, 1995), almost nothing is known about the impact of child rearing on same-gender couples compared to the literature examining the impact of children on traditional married couples across the family life cycle. As more same-gender couples pursue the opportunity to rear children, efforts to expand research on the family life cycle with this population will become increasingly important.
Although the wording of MSI-R items was revised to increase this measure’s usefulness with nontraditional couples, the original instrument was nevertheless conceived primarily from an emphasis on heterosexual couples. Consequently, additional domains or specific issues within domains particularly germane to gay and lesbian couples may not be adequately assessed by this measure. Although gay and lesbian couples confront many of the same challenges as do heterosexual couples, they also contend with issues and concerns unique to same-gender couples. Hence, although current findings suggest that the MSI-R is relevant and appropriate for assessing gay and lesbian couples in both clinical and research applications, by itself this measure, like any other, is incomplete.
Finally, compared to controlled outcome studies of clinical interventions with heterosexual couples, empirical findings regarding couple therapy for same-gender couples are virtually nonexistent. Until relationship therapy with gay and lesbian couples undergoes the same scrutiny as therapy with heterosexual couples, clinicians working with same-gender couples are left to draw on empirically supported interventions for relationship issues and processes common to couples regardless of sexual orientation, while making every effort to recognize and contend with the unique developmental and sociocultural stressors their gay and lesbian clients confront in both their individual and relationship lives.
REFERENCES
Abbott, B. V, Snyder, D. K., Gleaves, D. H., Hahlweg, K., & Klann, N. (2000, August). Cross-cultural reliability and validity of the German Marital Satisfaction Inventory-Revised. Paper presented at the meeting of the American Psychological Association, Washington, DC.
American Psychiatric Association. (1952). Diagnostic and statistical manual of mental disorders (DSM). Washington, DC: Author.
American Psychiatric Association. (1968). Diagnostic and statistical manual of mental disorders (DSM-II). Washington, DC: Author.
Barringer, F. (1993, April 25). Polling on sexual issues has its drawbacks. New York Times, Section 1, p. 23.
Bell, A. P, & Weinberg, M. S. (1978). Homosexualities: A study of the diversity among men and women. New York: Simon & Schuster.
Berger, R. M. (1990). Men together: Understanding the gay couple. Journal ofHomosexuality, 19(3), 31-49. Blumstein, P., & Schwartz, P. (1983). American couples. New York: Morrow.
Boen, D. L. (1988). A practitioner looks at assessment in marital counseling. Journal of Counseling and Development, 66, 484-486.
Brown, L. S. (1995). Therapy with same-gender couples: An introduction. In N. S. Jacobson and A. S. Gurman, (Eds.), Clinical handbook of couple therapy (pp. 274-291). New York: Guilford.
Burgoyn, R. W. (2001). The Relationship Assessment Measure for same-sex couples (RAM-SSC): A standardized instrument for evaluating gay couple functioning. Journal of Sex and Marital Therapy, 27, 279-282.
Burnett, P. (1987). Assessing marital adjustment and satisfaction: A review. Measurement and Evaluation in Counseling and Development, 20,113-121.
Cass, V C. (1984). Homosexual identity formation: Testing a theoretical model. Journal of Sex Research, 20, 143-167.
Clark, C., & Fields, J. M. (August, 1999). First glance: Preliminary analysis of relationship, marital status, and grandparents items on the Census 2000 dress rehearsal. Paper presented at the annual meeting of the American Sociological Association, Chicago.
Coleman, E. (1982). Developmental stages of the coming out process. Journal of Homosexuality, 7, 31-43.
Committee on Lesbian and Gay Concerns. (1991). Final report of the task force on bias in psychotherapy with lesbians and gay men. Washington, DC: American Psychological Association.
Cuellar, I. (1998). Cross-cultural clinical psychological assessment of Hispanic Americans. Journal of Personality Assessment, 70, 71-86.
Dixon, D. N. (1985). Review of the Marital Satisfaction Inventory. In J. V. Mitchell (Ed.), The ninth mental measurements yearbook: (Vol. 1, pp. 894-895). Lincoln, NE: University of Nebraska Press.
Eldridge, N. S., & Gilbert, L. A. (1990). Correlates of relationship satisfaction in lesbian couples. Psychology of Women Quarterly, 14, 4342.
Fassinger. R. E. (1991). The hidden minority: Issues and challenges in working with lesbian women and gay men. Counseling Psychologist, 19, 151-176.
Fowers, B. J. (1990). An interactional approach to standardized marital assessment: A literature review. Family Relations, 39, 368-377.
Greenfield, S., & Thelen, M. (1997). Validation of the fear of intimacy scale with a lesbian and gay male population. Journal of Social and Personal Relationships, 14, 707-716.
Haas, S. M., & Stafford, L. (1998). An initial examination of maintenance behaviors in gay and lesbian relationships. Journal of Social and Personal Relationships, 15, 846-855.
Julien, D., Arellano, C., & Turgeon, L. (1997). Gender issues in heterosexual, gay and lesbian couples. In W. K. Halford & H. J. Markman (Eds.), Clinical handbook of marriage and couples interventions (pp. 107-127). New York: Wiley.
Kurdek, L. A. (1988). 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 I-year follow-up study. Journal of Social and Personal Relationships, 6, 39-59.
Kurdek, L. A. (1991). Correlates of relationship satisfaction in cohabiting gay and lesbian couples: Integration of contextual, investment, and problem-solving methods. Journal of Personality and Social Psychology, 61, 910-922.
Kurdek, L. A. (1992). Relationship stability and relationship satisfaction in cohabiting gay and lesbian couples: A prospective longitudinal test of the contextual and interdependence models. Journal of Social and Personal Relationships, 9, 125-142. Kurdek, L. A. (1994). Areas of conflict for gay, lesbian, and heterosexual couples: What couples argue about influences relationship satisfaction. Journal of Marriage and the Family, 56, 923-934.
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). Relationship quality of partners in heterosexual married, heterosexual cohabiting, and gay and lesbian relationships. Journal of Personality and Social Psychology, 51, 711-720.
Laumann, E. O., Gagnon, J. H., Michael, R. T., & Michaels, S. (1994). The social organization of sexuality: Sexual practices in the United States. Chicago: University of Chicago Press.
Maccoby, E. E. (1990). Gender and relationships. American Psychologist, 45, 513-520.
Metz, M. E., Rosser, B. R. S., & Strapko, N. (1994). Differences in conflict-resolution styles among heterosexual, gay, and lesbian couples. Journal of Sex Research, 31, 293-308.
Modrcin, M. J., & Wyers, N. L. (1990). Lesbian and gay couples: Where they turn when help is needed. Journal of Gay and Lesbian Psychotherapy, 1(3), 89-104.
Morgan, K. S. (1992). Caucasian lesbians’ use of psychotherapy: A matter of attitude? Psychology of Women Quarterly, 16, 127-130.
Negy, C., & Snyder, D. K. (1997). Ethnicity and acculturation: Assessing Mexican American couples’ relationships using the Marital Satisfaction Inventory-Revised. Psychological Assessment, 9, 414-421.
Negy, C., & Snyder, D. K. (2000). Relationship satisfaction of Mexican American and non-Hispanic White American interethnic couples: Issues of acculturation and clinical intervention. Journal of Marital and Family Therapy, 26, 293-304.
Negy, C., Snyder, D. K., & Diaz-Loving, R. (2002, August). A cross-national comparison of Mexican and Mexican American couples using the Marital Satisfaction Inventory-Revised. Paper presented at the meeting of the American Psychological Association, Chicago.
Okazaki, S., & Sue, S. (1995). Methodological issues in assessment research with ethnic minorities. Psychological Assessment, 7, 367-375.
Peplau, L. A., Padesky, C., & Hamilton, M. (1982). Satisfaction in lesbian relationships. Journal of Homosexuality, 8(2), 23-35. Posse, B. (1978). Identities in the lesbian world: The social construction of self. Westport, CT: Greenwood.
Snyder, D. K. (1979). Marital Satisfaction Inventory. Los Angeles, CA: Western Psychological Services.
Snyder, D. K. (1997). Manual for the Marital Satisfaction Inventory-Revised. Los Angeles, CA: Western Psychological Services.
Snyder, D. K., & Aikman, G. G. (1999). The Marital Satisfaction Inventory-Revised. In M. E. Maruish (Ed.), Use of psychological testing for treatment planning and outcomes assessment (2nd ed., pp. 1173-1210). Mahwah, NJ: Erlbaum.
Spasojevic, J., Heifer, R. W., & Snyder, D. K. (2000). Effects of posttraumatic stress and acculturation on marital functioning in Bosnian refugee couples. Journal of Traumatic Stress, 13, 205-217.
Testa, R. J., Kinder, B. N., & Ironson, G. (1987). Heterosexual bias in the perception of loving relationships of gay males and lesbians. Journal of Sex Research, 23(2), 163-172.
Troiden, R. R. (1989). The formation of homosexual identities. Journal of Homosexuality, 17, 43-73.
U.S. Census Bureau. (2001). Census 2000 profiles of general demographic characteristics: United States. Washington, D.C.: Author.
Waring, E. M. (1985). Review of the Marital Satisfaction Inventory. In J. V. Mitchell (Ed.), The ninth mental measurements yearbook: (Vol. 1, pp. 895-96). Lincoln, NE: University of Nebraska Press.
Zacks, E., Green, R., & Marrow, J. (1988). Comparing lesbian and heterosexual couples on the circumplex model: An initial investigation. Family Process, 27, 471-484.
Adrienne J. Means-Christensen
University of California, San Diego
Douglas K. Snyder
Texas A&M University
Charles Negy
University of Central Florida
Adrienne Means-Christensen, PhD, Anxiety and Traumatic Stress Disorders Research Program, University of California at San Diego; Douglas K. Snyder, PhD, Department of Psychology, Texas A&M University; Charles Negy, PhD, Department of Psychology, University of Central Florida.
An earlier version of this paper was presented at the XXVIII Interamerican Congress of Psychology, Santiago, Chile, July 31, 2001.
Correspondence concerning this article should be addressed to Adrienne Means-Christensen, Anxiety and Traumatic Stress Disorders Research Program, University of California at San Diego, 8950 Villa La Jolla Drive, Suite 2243, La Jolla, California 92037. E-mail: adrienne@ucsd.edu; or to Douglas K. Snyder, Department of Psychology, Texas A&M University, College Station, Texas 77843-4235. E-mail: dks@psyc.tamu.edu
Copyright American Association for Marriage and Family Therapy Jan 2003
Provided by ProQuest Information and Learning Company. All rights Reserved
