563834 research-article2014
JIVXXX10.1177/0886260514563834Journal of Interpersonal ViolenceGemberling et al.
Article
Lesbian, Gay, and Bisexual Identity as a Moderator of Relationship Functioning After Sexual Assault
Journal of Interpersonal Violence 1–22 © The Author(s) 2014 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/0886260514563834 jiv.sagepub.com
Tess M. Gemberling,1 Robert J. Cramer,1 Rowland S. Miller,1 Caroline H. Stroud,1 Ramona M. Noland,1 and James Graham2
Abstract Sexual assault is unfortunately common, especially among lesbian, gay, and bisexual (LGB) individuals. Yet, the associations of such victimization have not yet been extensively established in the areas of sexual identity and romantic relationship functioning. Accordingly, the present study examined the associations between lifetime sexual assault, LGB identity, and romantic relationship functioning in a sample of 336 LGB individuals. A history of sexual assault was associated with attachment anxiety and several sexual identity components (i.e., higher levels of acceptance concerns, identity uncertainty, internalized homonegativity, and identity superiority). Furthermore, an association of sexual assault and attachment avoidance was moderated by internalized homonegativity. Finally, a more secure LGB identity was associated with healthier romantic relationship functioning. Collectively, these findings are applicable to services for LGB sexual assault victims, suggesting the incorporation of treatment that bolsters LGB identity and couple functioning. Limitations and future directions are discussed.
1Sam
Houston State University, Huntsville, TX, USA Care Center, Santa Fe, NM, USA
2Southwest
Corresponding Author: Tess M. Gemberling, Department of Psychology, Sam Houston State University, 1806 Avenue J, Huntsville, TX 77340, USA. Email: [email protected]
Downloaded from jiv.sagepub.com at CARLETON UNIV on October 20, 2015
2
Journal of Interpersonal Violence
Keywords lesbian, gay, and bisexual identity; lifetime sexual assault; romantic relationships Sexual assault, usually defined as forced or coerced sexual activity, is unfortunately common (Douglas, Burgess, Burgess, & Ressler, 2006). In the general population, approximately 11% to 18% of women and 1% to 3% of men report experiencing sexual assault sometime during their lives (Basile, Chen, Black, & Saltzman, 2007; Black et al., 2011; Tjaden & Thoennes, 1998). However, rates of victimization are often higher for lesbian, gay, or bisexual (LGB) individuals. According to one review of the literature, 16% to 85% of lesbian or bisexual (LB) women and 12% to 54% gay or bisexual (GB) men report lifetime sexual assault (LSA; Rothman, Exner, & Baughman, 2011). These events are consequential. Both general and LGB samples of sexual assault victims report higher rates of personal difficulties, such as depression, anxiety, post-traumatic stress disorder (PTSD) symptoms, sexual dysfunction, and social withdrawal, than nonvictimized individuals do (e.g., Choudhary, Smith, & Bossarte, 2012; Cramer, McNiel, Holley, Shumway, & Boccellari, 2012; Oltmanns & Emery, 2010). Despite interpersonal and psychological difficulties associated with sexual assault victimization, its associations with romantic relationship functioning are complex and understudied, especially among LGB individuals. Because LGB individuals report experiencing higher rates of sexual victimization than heterosexuals do, and because they may encounter additional romantic relationship complications that emerge from minority status, examination of these connections is necessary to identify potential problematic impacts of sexual victimization (Balsam, Rothblum, & Beauchaine, 2005; Mohr & Fassinger, 2006; Rothman et al., 2011). The present study sought to remedy this gap in the literature by examining the associations between LSA, LGB identity factors, and romantic relationship functioning. As such, these three constructs and their interactions are reviewed in the following sections.
Romantic Relationship Functioning Correlates of Sexual Victimization There have been several studies investigating the associations of sexual victimization with subsequent outcomes in romantic relationships, but we were unable to locate any that examined sexual assault victimization across the
Downloaded from jiv.sagepub.com at CARLETON UNIV on October 20, 2015
Gemberling et al.
3
lifespan—that is, forced or coerced sexual activity that occurred at any time during an individual’s lifetime (Douglas et al., 2006). Instead, studies have examined the association between romantic relationships and sexual victimization separately by age at victimization, choosing between adulthood and childhood victimization, among heterosexual females. We also did not find studies that specifically addressed the romantic relationship functioning of LGB or male sexual assault victims. When adults are sexually assaulted, they report various impairments in romantic relationship functioning that include relationship disruption, sexual dysfunction, mismatched communication, and victim blame (e.g., Burgess & Holmstrom, 1979; Smith, 2005). In particular, two investigations (Miller, Williams, & Bernstein, 1982; Rodkin, Hunt, & Cowan, 1982) have identified changes in romantic attachment that subsequently affect relationship dynamics. Specifically, following an assault, a female victim often becomes more dependent on her partner for emotional support. Initially, no relationship disturbance is apparent, with her partner meeting her increased needs. However, over time, the victim’s requirements and her partner’s provided support may become mismatched, with her needs exceeding his understanding. This limited literature would imply that sexual assault victims are likely to face attachment anxiety, defined as nervous concern about the quality of their relationships. However, the literature has yet to comment on the other aspects of romantic relationship functioning such as relationship satisfaction, commitment, or other forms of attachment. Children who are victimized also encounter romantic difficulties as adults (DiLillo, 2001). Victims are often less satisfied with their relationships, have difficulty trusting and confiding in their partners, experience more couple discord and conflict, and have more negative perspectives of their partners, such as perceiving them to be more controlling and less caring. For example, childhood sexual abuse has been associated with lower relationship satisfaction, poorer communication, and less trust of romantic partners than is observed among nonvictimized females (DiLillo & Long, 1999; Hunter, 1991). Again applying these results to the present analysis, this literature suggests that childhood sexual abuse victims experience lower levels of relationship satisfaction, but it again does not comment on relationship commitment or attachment styles.
LGB Identity: Definition and Associations While LGB individuals experience personal complications (e.g., depression or anxiety) and interpersonal complications (e.g., sexual or romantic difficulties) resulting from sexual assault, as identified by investigations using
Downloaded from jiv.sagepub.com at CARLETON UNIV on October 20, 2015
4
Journal of Interpersonal Violence
sexual assault victims from the general population, LGB individuals may also be at risk of additional issues relating to sexual orientation identity (Cramer, McNiel, et al., 2012). Contemporary conceptualizations of LGB identity incorporate components such as reaction to stigma and victimization, social acceptance concerns, and general well-being (e.g., Herek, Gillis, & Cogan, 2009; Mohr & Kendra, 2011). The present paper adopts a multidimensional model of LGB identity inclusive of eight social and cognitive facets. Mohr and colleagues have formulated one conceptualization of sexual orientation identity. Based on their analyses of the Lesbian, Gay, and Bisexual Identity Scale (LGBIS), they (Mohr & Fassinger, 2000; Mohr & Kendra, 2011) delineated eight different components of LGB identity: (a) Concealment Motivation (CM, desiring to keep one’s LGB identity private), (b) Identity Uncertainty (IU, being confused about one’s sexual orientation identity), (c) Internalized Homonegativity (IH, adopting heterosexist views and dismissing one’s LGB identity), (d) Difficult Process (DP, considering one’s LGB identity development as subjectively challenging), (e) Acceptance Concerns (AC, worry pertaining to social acceptance and fear of anti-LGB identity stigma), (f) Identity Superiority (IS, preferring LGB rather than heterosexual people), (g) Identity Centrality (IC, embracing one’s LGB identity as central to personal identity), and (h) Identity Affirmation (IA, privately and publicly affirming one’s LGB identity). This conceptualization includes a variety of LGB identity aspects, including both negatively (e.g., internalized homonegativity) and positively valenced (e.g., identity centrality) facets. Moreover, the LGBIS is informed by social (e.g., acceptance concerns), cognitive (e.g., internalized homonegativity), and behavioral (e.g., identity affirmation) aspects. Of the many concerns that may arise for LGB sexual assault victims, a few studies have addressed the role of sexual identity components as sexual victimization correlates: For both gay men and lesbians, internalized homophobia (equivalent to Mohr’s concept of internalized homonegativity) was considered an important variable in the symptomatology of sexual victimization, being reported at a slightly higher rate by sexually assaulted gay males than usually found in the community and was a significant predictor of depression, experiential avoidance, and PTSD symptoms (Gold, Dickstein, Marx, & Lexington, 2009; Gold, Marx, & Lexington, 2007). Furthermore, sexual identity confusion (likely an equivalent to Mohr’s identity uncertainty) has been identified as a common association of sexual assault among male victims, either heterosexual or gay (Coxell & King, 2010; Peterson, Voller, Polusny, & Murdoch, 2011; Tewksbury, 2007), with one study finding that 70% of their sample had “long-term crises with their sexual orientation” (Walker, Archer, & Davies, 2005, p. 76). While indicating that LGB sexual
Downloaded from jiv.sagepub.com at CARLETON UNIV on October 20, 2015
Gemberling et al.
5
assault victims often experience difficulties in their sexual identity conceptualization, the literature has yet to examine the associations between the full range of sexual identity components and sexual victimization. In one of the few studies to address such associations, Mohr and Fassinger (2006) investigated the connections between four different aspects of LGB identity (i.e., internalized homonegativity, stigma sensitivity, identity confusion, and identity superiority) and same-sex relationship quality. Each of the LGB identity variables was negatively associated with relationship quality and accounted for unique variability of relationship quality. Various studies have found correlations between internalized homophobia and problematic relationship functioning (e.g., Balsam & Szymanski, 2005; Frost & Meyer, 2009), indicating the importance of individual LGB identity for LGB romantic relationships.
The Present Study The present study integrated sexual assault victimization, sexual identity conceptualization, and romantic relationships among LGB persons under the following hypotheses: Hypothesis 1 (H1): LSA would be associated with lower relationship satisfaction and commitment, and more attachment anxiety or avoidance, compared with non-victimized counterparts. Hypothesis 2 (H2): LSA would be correlated with a more negative (i.e., higher concealment motivation, internalized homonegativity, identity uncertainty, acceptance concerns, difficult process, identity superiority) and less positive (i.e., lower identity affirmation, identity centrality) LGB identity. Hypothesis 3 (H3): A more negative (i.e., higher concealment motivation, internalized homonegativity, identity uncertainty, acceptance concerns, difficult process, identity superiority) and less positive (i.e., lower identity affirmation, identity centrality) LGB identity would be correlated with lower relationship satisfaction and commitment, and more attachment avoidance and anxiety. Hypothesis 4 (H4): As an exploratory analysis, examination of any moderating influences of LGB identity on the LSA–romantic relationship link.
Method This study was a secondary analysis of Cramer, Stroud, Miller, and Graham’s (2012) Project Legacy database. The present study only included sexual
Downloaded from jiv.sagepub.com at CARLETON UNIV on October 20, 2015
6
Journal of Interpersonal Violence
assault, LGB identity, and romantic relationship–specific aspects of the initial project in its analysis.
Procedure Participants were recruited from Legacy Community Health Services, a comprehensive care, federally funded medical clinic that provides health care to underserved populations in Houston, Texas. The study was described to potential participants as a voluntary health survey, and was advertised by research assistants and Legacy staff in person and through fliers in two departments. During recruitment, assistants and staff confirmed that potential participants met the study’s three criteria: (a) a minimum of 18 years of age, (b) at least a 10th grade reading level, and (c) LGB self-identification. If participants met the criteria and agreed to participate, they signed the consent form (after being given the chance to ask questions) and completed the questionnaire packet in the primary care department waiting room.
Participants Participants were 3361 self-identified lesbian, gay, or bisexual individuals. Collectively, the sample was 42.1 years of age on average with a standard deviation of 11.1 years. The sample included 238 (73.7%) males and 78 (24.1%) females. By ethnicity, the sample was comprised of 135 (41.8%) Caucasian, 132 (40.9%) African American, 37 (11.5%) Hispanic, and 19 (5.9%) “Other” individuals (Asian, Native American, Biracial, etc.). With regard to romantic relationship status, 200 (61.9%) individuals were single, 18 (5.6%) were dating, 73 (22.6%) were in a monogamous relationship, and 31 (9.6%) were in an open relationship. Concerning the number of sexual partners, participants reported a median of two lifetime sexual partners with a standard deviation of 11.1 partners. Of the several health factors assessed, 46 (14.2%) individuals reported having been diagnosed with AIDS, and 207 (64.1%) individuals reporting being HIV-positive. Finally, the sample varied in their sexual assault status. Most (n = 218; 67.5%) reported no sexual assault victimization, but 105 (32.5%) individuals reported sexual assault victimization.
Measures LSA. LSA was assessed through a broad victimization question included in the demographics section of the questionnaire. Previous studies—Basile et al. (2007) and Choudhary et al. (2012), for example—have also used a
Downloaded from jiv.sagepub.com at CARLETON UNIV on October 20, 2015
Gemberling et al.
7
singular yes/no question to establish sexual assault status. Participants were asked “Have you ever been a victim of any of the following crimes?” Those checking the sexual assault box were coded as having experienced such victimization during their lifetime. LGB identity. Mohr and Kendra’s (2011) LGBIS was designed to assess LGB individuals’ current thoughts and feelings relating to their LGB identity. The LGBIS includes 27 items that are averaged into eight subscales, namely, concealment motivation (3 items, α = .79, present analysis α = .72), identity uncertainty (4 items, α = .89, present analysis α = .78), internalized homonegativity (3 items, α = .86, present analysis α = .72), difficult process (3 items, 1 reverse coded, α = .82, present analysis α = .52), acceptance concerns (3 items, α = .76, present analysis α = .59), identity superiority (3 items, α = .81, present analysis α = .63), identity centrality (5 items, 1 reverse coded, α = .85, present analysis α = .62), and identity affirmation (3 items, α = .89, present analysis α = .83). Relationship satisfaction. Hendrick’s (1988) Relationship Assessment Scale (RAS) was developed to assess general satisfaction in romantic relationships. The RAS incorporates seven items (α = .86, present analysis α = .89) that form a unifactorial measure. Individuals who were not in a romantic relationship at the time of questionnaire completion were prompted to think of their most recent romantic relationship while responding. Relationship commitment. Arriaga and Agnew’s (2001) Commitment Scale (CS) was created to assess romantic relationship commitment. The CS contains 12 items that can be interpreted in two manners, either through its three subscales or its total score (X. B. Arriaga, personal communication, October 10, 2013). For these analyses, we chose to utilize the total score (present analysis α = .90), preferring to examine general commitment instead of its three components (i.e., psychological attachment, long-term orientation, and intent to persist). As with the RAS, individuals who were not in a romantic relationship at the time of questionnaire completion were prompted to think of their most recent romantic relationship while responding. Attachment styles. Wei, Russell, Mallinckrodt, and Vogel’s (2007) 12-item Experiences in Close Relationships Scale–Short Form (ECR-SF) was constructed to efficiently assess two dimensions of adult attachment to one’s intimate partners. The first of these is anxiety about abandonment or attachment anxiety, which involves “a fear of interpersonal rejection or abandonment, an excessive need for approval from others, and distress when one’s
Downloaded from jiv.sagepub.com at CARLETON UNIV on October 20, 2015
8
Journal of Interpersonal Violence
partner is unavailable or unresponsive” (p. 188). This subscale contains six items (1 reverse coded) and typically α range = .78 to .86; however, in the present administration, α = .60. The second dimension is avoidance of intimacy or attachment avoidance, which involves “fear of dependent and interpersonal intimacy, an excessive need for self-reliance, and reluctance to self disclose” (p. 188). This subscale also contains six items (three reverse coded) and usually has α range = .78 to .88; in the present data, however, α = .65.
Results Preliminary Analyses Preliminary correlations and between-group tests were performed to determine if demographic variables (i.e., age, gender, ethnicity, relationship status, HIV status, AIDS status, and number of sexual partners) were related to LGB identity components (LGBIS subscales) and/or romantic relationship variables (i.e., RAS, CS, and ECR-SF subscales). To control for Type I error, an alpha level of .01 was used to test for covariance. The significant results for the LGBIS subscales were as follows: (a) Gender, with females reporting less acceptance concerns and identity centrality than males; (b) Ethnicity, with African American individuals reporting more concealment motivation and more of a difficult process than “Other” individuals, Caucasian individuals reporting less identity uncertainty and more identity affirmation and identity centrality than African American individuals, and African American individuals reporting more internalizing homonegativity than Caucasian and “Other” individuals; (c) Relationship status, with individuals in an open relationship reporting less concealment motivation than single or dating individuals, and single individuals reporting more identity uncertainty and internalizing homonegativity than those in a monogamous or open relationship; (d) HIV status, with HIV+ individuals reporting more concealment motivation and internalized homonegativity than HIV− individuals. Test statistics, means, and standard deviations for each analysis are displayed in Table 1. The significant results for the relationship variables (i.e., RAS, CS, and ECR-SF subscales) were the demographics: (a) relationship status, with single and dating individuals reporting less relationship satisfaction and commitment than those in a monogamous or open relationship, although dating individuals and those in an open relationship did not significantly differ from each other with regard to relationship satisfaction, and single individuals reporting more relationship avoidance than those in a monogamous or open relationship, and (b) HIV status, with HIV+ individuals reporting
Downloaded from jiv.sagepub.com at CARLETON UNIV on October 20, 2015
Gemberling et al.
9
less relationship commitment than HIV− individuals. Test statistics, means, and standard deviations for each analysis are displayed in Table 1.
Hypothesis Testing2 Hypothesis 1. A MANCOVA was used to test the first hypothesis, including the LSA question (predictor variable), the RAS total, CS total and ECR-SF subscales (criterion variables), and the dummy-coded variables for the covariates relationship status and HIV status (control variables). LSA displayed a significant multivariate effect, Wilks’s λ = .95, F(4, 290) = 3.72, p = .006, η2p = .05. LSA was associated with attachment anxiety, with those reporting a history of sexual assault having more anxiety about abandonment than those without such victimization (see Table 2 for test statistics, means, and standard deviations). Hypothesis 2. A MANCOVA was used to test the second hypothesis, including the LSA question (predictor variable), the LGBIS subscales (criterion variables), and the dummy-coded variables for the covariates gender, ethnicity, relationship status, and HIV status (control variables). LSA again displayed a significant multivariate effect, Wilks’s λ = .94, F(8, 292) = 2.21, p = .027, η2p = .06. LSA was associated with four LGB identity components (LGBIS subscales), specifically higher levels of acceptance concerns, identity uncertainty, internalized homonegativity, and identity superiority (see Table 2 for test statistics, means, and standard deviations). Hypothesis 3. A multivariate regression was used to test the third hypothesis, including the LGBIS subscales (predictor variables); the RAS total, CS total, and ECR-SF subscales (criterion variables); and the dummy-coded variables for the covariates relationship status and HIV status.3 Four LGB identity components displayed a significant multivariate effect: acceptance concerns, Wilks’s λ = .92, F(4, 278) = 5.81, p < .001, η2p = 08; concealment motivation, Wilks’s λ = .94, F(4, 278) = 4.61, p = 001, η2p = .06; identity uncertainty, Wilks’s λ = .94, F(4, 278) = 4.38, p = 002, η2p = .06; and identity affirmation, Wilks’s λ = .95, F(4, 278) = 3.39, p = .010, η2p = .05. In particular, acceptance concerns was positively associated with attachment anxiety. Concealment motivation was positively associated with relationship commitment, and negatively associated with attachment avoidance. Identity uncertainty was negatively associated with relationship satisfaction and commitment, and was positively associated with avoidance of intimacy. Identity affirmation was negatively associated with attachment anxiety and avoidance. Test statistics for each aspect are displayed in Table 3.
Downloaded from jiv.sagepub.com at CARLETON UNIV on October 20, 2015
10
Journal of Interpersonal Violence
Table 1. Test Statistics, Means, and Standard Deviations for Significant Preliminary Analyses. Test Statistics
M (SD)
LGBIS Subscale or Demographic RRF Scale
F
Gender
Acceptance Concerns
8.86
.003 −0.40 (d)
Identity Centrality
8.53
.004 −0.38 (d)
p
d or
η2
Concealment Motivation
4.04
.008 .04 (η2)
Identity Uncertainty
3.57
.014 .03 (η2)
Internalized Homonegativity
4.21
.006 .04 (η2)
Difficult Process
3.71
.012 .03 (η2)
Identity Affirmation
5.99
.001 .05 (η2)
Identity Centrality
3.66
.013 .03 (η2)
Concealment Motivation
5.02
.002 .05 (η2)
Identity Uncertainty
5.97
.001 .05 (η2)
Internalized Homonegativity
5.92
.001 .05 (η2)
Relationship status
RAS (Satisfaction)
14.70
JIVXXX10.1177/0886260514563834Journal of Interpersonal ViolenceGemberling et al.
Article
Lesbian, Gay, and Bisexual Identity as a Moderator of Relationship Functioning After Sexual Assault
Journal of Interpersonal Violence 1–22 © The Author(s) 2014 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/0886260514563834 jiv.sagepub.com
Tess M. Gemberling,1 Robert J. Cramer,1 Rowland S. Miller,1 Caroline H. Stroud,1 Ramona M. Noland,1 and James Graham2
Abstract Sexual assault is unfortunately common, especially among lesbian, gay, and bisexual (LGB) individuals. Yet, the associations of such victimization have not yet been extensively established in the areas of sexual identity and romantic relationship functioning. Accordingly, the present study examined the associations between lifetime sexual assault, LGB identity, and romantic relationship functioning in a sample of 336 LGB individuals. A history of sexual assault was associated with attachment anxiety and several sexual identity components (i.e., higher levels of acceptance concerns, identity uncertainty, internalized homonegativity, and identity superiority). Furthermore, an association of sexual assault and attachment avoidance was moderated by internalized homonegativity. Finally, a more secure LGB identity was associated with healthier romantic relationship functioning. Collectively, these findings are applicable to services for LGB sexual assault victims, suggesting the incorporation of treatment that bolsters LGB identity and couple functioning. Limitations and future directions are discussed.
1Sam
Houston State University, Huntsville, TX, USA Care Center, Santa Fe, NM, USA
2Southwest
Corresponding Author: Tess M. Gemberling, Department of Psychology, Sam Houston State University, 1806 Avenue J, Huntsville, TX 77340, USA. Email: [email protected]
Downloaded from jiv.sagepub.com at CARLETON UNIV on October 20, 2015
2
Journal of Interpersonal Violence
Keywords lesbian, gay, and bisexual identity; lifetime sexual assault; romantic relationships Sexual assault, usually defined as forced or coerced sexual activity, is unfortunately common (Douglas, Burgess, Burgess, & Ressler, 2006). In the general population, approximately 11% to 18% of women and 1% to 3% of men report experiencing sexual assault sometime during their lives (Basile, Chen, Black, & Saltzman, 2007; Black et al., 2011; Tjaden & Thoennes, 1998). However, rates of victimization are often higher for lesbian, gay, or bisexual (LGB) individuals. According to one review of the literature, 16% to 85% of lesbian or bisexual (LB) women and 12% to 54% gay or bisexual (GB) men report lifetime sexual assault (LSA; Rothman, Exner, & Baughman, 2011). These events are consequential. Both general and LGB samples of sexual assault victims report higher rates of personal difficulties, such as depression, anxiety, post-traumatic stress disorder (PTSD) symptoms, sexual dysfunction, and social withdrawal, than nonvictimized individuals do (e.g., Choudhary, Smith, & Bossarte, 2012; Cramer, McNiel, Holley, Shumway, & Boccellari, 2012; Oltmanns & Emery, 2010). Despite interpersonal and psychological difficulties associated with sexual assault victimization, its associations with romantic relationship functioning are complex and understudied, especially among LGB individuals. Because LGB individuals report experiencing higher rates of sexual victimization than heterosexuals do, and because they may encounter additional romantic relationship complications that emerge from minority status, examination of these connections is necessary to identify potential problematic impacts of sexual victimization (Balsam, Rothblum, & Beauchaine, 2005; Mohr & Fassinger, 2006; Rothman et al., 2011). The present study sought to remedy this gap in the literature by examining the associations between LSA, LGB identity factors, and romantic relationship functioning. As such, these three constructs and their interactions are reviewed in the following sections.
Romantic Relationship Functioning Correlates of Sexual Victimization There have been several studies investigating the associations of sexual victimization with subsequent outcomes in romantic relationships, but we were unable to locate any that examined sexual assault victimization across the
Downloaded from jiv.sagepub.com at CARLETON UNIV on October 20, 2015
Gemberling et al.
3
lifespan—that is, forced or coerced sexual activity that occurred at any time during an individual’s lifetime (Douglas et al., 2006). Instead, studies have examined the association between romantic relationships and sexual victimization separately by age at victimization, choosing between adulthood and childhood victimization, among heterosexual females. We also did not find studies that specifically addressed the romantic relationship functioning of LGB or male sexual assault victims. When adults are sexually assaulted, they report various impairments in romantic relationship functioning that include relationship disruption, sexual dysfunction, mismatched communication, and victim blame (e.g., Burgess & Holmstrom, 1979; Smith, 2005). In particular, two investigations (Miller, Williams, & Bernstein, 1982; Rodkin, Hunt, & Cowan, 1982) have identified changes in romantic attachment that subsequently affect relationship dynamics. Specifically, following an assault, a female victim often becomes more dependent on her partner for emotional support. Initially, no relationship disturbance is apparent, with her partner meeting her increased needs. However, over time, the victim’s requirements and her partner’s provided support may become mismatched, with her needs exceeding his understanding. This limited literature would imply that sexual assault victims are likely to face attachment anxiety, defined as nervous concern about the quality of their relationships. However, the literature has yet to comment on the other aspects of romantic relationship functioning such as relationship satisfaction, commitment, or other forms of attachment. Children who are victimized also encounter romantic difficulties as adults (DiLillo, 2001). Victims are often less satisfied with their relationships, have difficulty trusting and confiding in their partners, experience more couple discord and conflict, and have more negative perspectives of their partners, such as perceiving them to be more controlling and less caring. For example, childhood sexual abuse has been associated with lower relationship satisfaction, poorer communication, and less trust of romantic partners than is observed among nonvictimized females (DiLillo & Long, 1999; Hunter, 1991). Again applying these results to the present analysis, this literature suggests that childhood sexual abuse victims experience lower levels of relationship satisfaction, but it again does not comment on relationship commitment or attachment styles.
LGB Identity: Definition and Associations While LGB individuals experience personal complications (e.g., depression or anxiety) and interpersonal complications (e.g., sexual or romantic difficulties) resulting from sexual assault, as identified by investigations using
Downloaded from jiv.sagepub.com at CARLETON UNIV on October 20, 2015
4
Journal of Interpersonal Violence
sexual assault victims from the general population, LGB individuals may also be at risk of additional issues relating to sexual orientation identity (Cramer, McNiel, et al., 2012). Contemporary conceptualizations of LGB identity incorporate components such as reaction to stigma and victimization, social acceptance concerns, and general well-being (e.g., Herek, Gillis, & Cogan, 2009; Mohr & Kendra, 2011). The present paper adopts a multidimensional model of LGB identity inclusive of eight social and cognitive facets. Mohr and colleagues have formulated one conceptualization of sexual orientation identity. Based on their analyses of the Lesbian, Gay, and Bisexual Identity Scale (LGBIS), they (Mohr & Fassinger, 2000; Mohr & Kendra, 2011) delineated eight different components of LGB identity: (a) Concealment Motivation (CM, desiring to keep one’s LGB identity private), (b) Identity Uncertainty (IU, being confused about one’s sexual orientation identity), (c) Internalized Homonegativity (IH, adopting heterosexist views and dismissing one’s LGB identity), (d) Difficult Process (DP, considering one’s LGB identity development as subjectively challenging), (e) Acceptance Concerns (AC, worry pertaining to social acceptance and fear of anti-LGB identity stigma), (f) Identity Superiority (IS, preferring LGB rather than heterosexual people), (g) Identity Centrality (IC, embracing one’s LGB identity as central to personal identity), and (h) Identity Affirmation (IA, privately and publicly affirming one’s LGB identity). This conceptualization includes a variety of LGB identity aspects, including both negatively (e.g., internalized homonegativity) and positively valenced (e.g., identity centrality) facets. Moreover, the LGBIS is informed by social (e.g., acceptance concerns), cognitive (e.g., internalized homonegativity), and behavioral (e.g., identity affirmation) aspects. Of the many concerns that may arise for LGB sexual assault victims, a few studies have addressed the role of sexual identity components as sexual victimization correlates: For both gay men and lesbians, internalized homophobia (equivalent to Mohr’s concept of internalized homonegativity) was considered an important variable in the symptomatology of sexual victimization, being reported at a slightly higher rate by sexually assaulted gay males than usually found in the community and was a significant predictor of depression, experiential avoidance, and PTSD symptoms (Gold, Dickstein, Marx, & Lexington, 2009; Gold, Marx, & Lexington, 2007). Furthermore, sexual identity confusion (likely an equivalent to Mohr’s identity uncertainty) has been identified as a common association of sexual assault among male victims, either heterosexual or gay (Coxell & King, 2010; Peterson, Voller, Polusny, & Murdoch, 2011; Tewksbury, 2007), with one study finding that 70% of their sample had “long-term crises with their sexual orientation” (Walker, Archer, & Davies, 2005, p. 76). While indicating that LGB sexual
Downloaded from jiv.sagepub.com at CARLETON UNIV on October 20, 2015
Gemberling et al.
5
assault victims often experience difficulties in their sexual identity conceptualization, the literature has yet to examine the associations between the full range of sexual identity components and sexual victimization. In one of the few studies to address such associations, Mohr and Fassinger (2006) investigated the connections between four different aspects of LGB identity (i.e., internalized homonegativity, stigma sensitivity, identity confusion, and identity superiority) and same-sex relationship quality. Each of the LGB identity variables was negatively associated with relationship quality and accounted for unique variability of relationship quality. Various studies have found correlations between internalized homophobia and problematic relationship functioning (e.g., Balsam & Szymanski, 2005; Frost & Meyer, 2009), indicating the importance of individual LGB identity for LGB romantic relationships.
The Present Study The present study integrated sexual assault victimization, sexual identity conceptualization, and romantic relationships among LGB persons under the following hypotheses: Hypothesis 1 (H1): LSA would be associated with lower relationship satisfaction and commitment, and more attachment anxiety or avoidance, compared with non-victimized counterparts. Hypothesis 2 (H2): LSA would be correlated with a more negative (i.e., higher concealment motivation, internalized homonegativity, identity uncertainty, acceptance concerns, difficult process, identity superiority) and less positive (i.e., lower identity affirmation, identity centrality) LGB identity. Hypothesis 3 (H3): A more negative (i.e., higher concealment motivation, internalized homonegativity, identity uncertainty, acceptance concerns, difficult process, identity superiority) and less positive (i.e., lower identity affirmation, identity centrality) LGB identity would be correlated with lower relationship satisfaction and commitment, and more attachment avoidance and anxiety. Hypothesis 4 (H4): As an exploratory analysis, examination of any moderating influences of LGB identity on the LSA–romantic relationship link.
Method This study was a secondary analysis of Cramer, Stroud, Miller, and Graham’s (2012) Project Legacy database. The present study only included sexual
Downloaded from jiv.sagepub.com at CARLETON UNIV on October 20, 2015
6
Journal of Interpersonal Violence
assault, LGB identity, and romantic relationship–specific aspects of the initial project in its analysis.
Procedure Participants were recruited from Legacy Community Health Services, a comprehensive care, federally funded medical clinic that provides health care to underserved populations in Houston, Texas. The study was described to potential participants as a voluntary health survey, and was advertised by research assistants and Legacy staff in person and through fliers in two departments. During recruitment, assistants and staff confirmed that potential participants met the study’s three criteria: (a) a minimum of 18 years of age, (b) at least a 10th grade reading level, and (c) LGB self-identification. If participants met the criteria and agreed to participate, they signed the consent form (after being given the chance to ask questions) and completed the questionnaire packet in the primary care department waiting room.
Participants Participants were 3361 self-identified lesbian, gay, or bisexual individuals. Collectively, the sample was 42.1 years of age on average with a standard deviation of 11.1 years. The sample included 238 (73.7%) males and 78 (24.1%) females. By ethnicity, the sample was comprised of 135 (41.8%) Caucasian, 132 (40.9%) African American, 37 (11.5%) Hispanic, and 19 (5.9%) “Other” individuals (Asian, Native American, Biracial, etc.). With regard to romantic relationship status, 200 (61.9%) individuals were single, 18 (5.6%) were dating, 73 (22.6%) were in a monogamous relationship, and 31 (9.6%) were in an open relationship. Concerning the number of sexual partners, participants reported a median of two lifetime sexual partners with a standard deviation of 11.1 partners. Of the several health factors assessed, 46 (14.2%) individuals reported having been diagnosed with AIDS, and 207 (64.1%) individuals reporting being HIV-positive. Finally, the sample varied in their sexual assault status. Most (n = 218; 67.5%) reported no sexual assault victimization, but 105 (32.5%) individuals reported sexual assault victimization.
Measures LSA. LSA was assessed through a broad victimization question included in the demographics section of the questionnaire. Previous studies—Basile et al. (2007) and Choudhary et al. (2012), for example—have also used a
Downloaded from jiv.sagepub.com at CARLETON UNIV on October 20, 2015
Gemberling et al.
7
singular yes/no question to establish sexual assault status. Participants were asked “Have you ever been a victim of any of the following crimes?” Those checking the sexual assault box were coded as having experienced such victimization during their lifetime. LGB identity. Mohr and Kendra’s (2011) LGBIS was designed to assess LGB individuals’ current thoughts and feelings relating to their LGB identity. The LGBIS includes 27 items that are averaged into eight subscales, namely, concealment motivation (3 items, α = .79, present analysis α = .72), identity uncertainty (4 items, α = .89, present analysis α = .78), internalized homonegativity (3 items, α = .86, present analysis α = .72), difficult process (3 items, 1 reverse coded, α = .82, present analysis α = .52), acceptance concerns (3 items, α = .76, present analysis α = .59), identity superiority (3 items, α = .81, present analysis α = .63), identity centrality (5 items, 1 reverse coded, α = .85, present analysis α = .62), and identity affirmation (3 items, α = .89, present analysis α = .83). Relationship satisfaction. Hendrick’s (1988) Relationship Assessment Scale (RAS) was developed to assess general satisfaction in romantic relationships. The RAS incorporates seven items (α = .86, present analysis α = .89) that form a unifactorial measure. Individuals who were not in a romantic relationship at the time of questionnaire completion were prompted to think of their most recent romantic relationship while responding. Relationship commitment. Arriaga and Agnew’s (2001) Commitment Scale (CS) was created to assess romantic relationship commitment. The CS contains 12 items that can be interpreted in two manners, either through its three subscales or its total score (X. B. Arriaga, personal communication, October 10, 2013). For these analyses, we chose to utilize the total score (present analysis α = .90), preferring to examine general commitment instead of its three components (i.e., psychological attachment, long-term orientation, and intent to persist). As with the RAS, individuals who were not in a romantic relationship at the time of questionnaire completion were prompted to think of their most recent romantic relationship while responding. Attachment styles. Wei, Russell, Mallinckrodt, and Vogel’s (2007) 12-item Experiences in Close Relationships Scale–Short Form (ECR-SF) was constructed to efficiently assess two dimensions of adult attachment to one’s intimate partners. The first of these is anxiety about abandonment or attachment anxiety, which involves “a fear of interpersonal rejection or abandonment, an excessive need for approval from others, and distress when one’s
Downloaded from jiv.sagepub.com at CARLETON UNIV on October 20, 2015
8
Journal of Interpersonal Violence
partner is unavailable or unresponsive” (p. 188). This subscale contains six items (1 reverse coded) and typically α range = .78 to .86; however, in the present administration, α = .60. The second dimension is avoidance of intimacy or attachment avoidance, which involves “fear of dependent and interpersonal intimacy, an excessive need for self-reliance, and reluctance to self disclose” (p. 188). This subscale also contains six items (three reverse coded) and usually has α range = .78 to .88; in the present data, however, α = .65.
Results Preliminary Analyses Preliminary correlations and between-group tests were performed to determine if demographic variables (i.e., age, gender, ethnicity, relationship status, HIV status, AIDS status, and number of sexual partners) were related to LGB identity components (LGBIS subscales) and/or romantic relationship variables (i.e., RAS, CS, and ECR-SF subscales). To control for Type I error, an alpha level of .01 was used to test for covariance. The significant results for the LGBIS subscales were as follows: (a) Gender, with females reporting less acceptance concerns and identity centrality than males; (b) Ethnicity, with African American individuals reporting more concealment motivation and more of a difficult process than “Other” individuals, Caucasian individuals reporting less identity uncertainty and more identity affirmation and identity centrality than African American individuals, and African American individuals reporting more internalizing homonegativity than Caucasian and “Other” individuals; (c) Relationship status, with individuals in an open relationship reporting less concealment motivation than single or dating individuals, and single individuals reporting more identity uncertainty and internalizing homonegativity than those in a monogamous or open relationship; (d) HIV status, with HIV+ individuals reporting more concealment motivation and internalized homonegativity than HIV− individuals. Test statistics, means, and standard deviations for each analysis are displayed in Table 1. The significant results for the relationship variables (i.e., RAS, CS, and ECR-SF subscales) were the demographics: (a) relationship status, with single and dating individuals reporting less relationship satisfaction and commitment than those in a monogamous or open relationship, although dating individuals and those in an open relationship did not significantly differ from each other with regard to relationship satisfaction, and single individuals reporting more relationship avoidance than those in a monogamous or open relationship, and (b) HIV status, with HIV+ individuals reporting
Downloaded from jiv.sagepub.com at CARLETON UNIV on October 20, 2015
Gemberling et al.
9
less relationship commitment than HIV− individuals. Test statistics, means, and standard deviations for each analysis are displayed in Table 1.
Hypothesis Testing2 Hypothesis 1. A MANCOVA was used to test the first hypothesis, including the LSA question (predictor variable), the RAS total, CS total and ECR-SF subscales (criterion variables), and the dummy-coded variables for the covariates relationship status and HIV status (control variables). LSA displayed a significant multivariate effect, Wilks’s λ = .95, F(4, 290) = 3.72, p = .006, η2p = .05. LSA was associated with attachment anxiety, with those reporting a history of sexual assault having more anxiety about abandonment than those without such victimization (see Table 2 for test statistics, means, and standard deviations). Hypothesis 2. A MANCOVA was used to test the second hypothesis, including the LSA question (predictor variable), the LGBIS subscales (criterion variables), and the dummy-coded variables for the covariates gender, ethnicity, relationship status, and HIV status (control variables). LSA again displayed a significant multivariate effect, Wilks’s λ = .94, F(8, 292) = 2.21, p = .027, η2p = .06. LSA was associated with four LGB identity components (LGBIS subscales), specifically higher levels of acceptance concerns, identity uncertainty, internalized homonegativity, and identity superiority (see Table 2 for test statistics, means, and standard deviations). Hypothesis 3. A multivariate regression was used to test the third hypothesis, including the LGBIS subscales (predictor variables); the RAS total, CS total, and ECR-SF subscales (criterion variables); and the dummy-coded variables for the covariates relationship status and HIV status.3 Four LGB identity components displayed a significant multivariate effect: acceptance concerns, Wilks’s λ = .92, F(4, 278) = 5.81, p < .001, η2p = 08; concealment motivation, Wilks’s λ = .94, F(4, 278) = 4.61, p = 001, η2p = .06; identity uncertainty, Wilks’s λ = .94, F(4, 278) = 4.38, p = 002, η2p = .06; and identity affirmation, Wilks’s λ = .95, F(4, 278) = 3.39, p = .010, η2p = .05. In particular, acceptance concerns was positively associated with attachment anxiety. Concealment motivation was positively associated with relationship commitment, and negatively associated with attachment avoidance. Identity uncertainty was negatively associated with relationship satisfaction and commitment, and was positively associated with avoidance of intimacy. Identity affirmation was negatively associated with attachment anxiety and avoidance. Test statistics for each aspect are displayed in Table 3.
Downloaded from jiv.sagepub.com at CARLETON UNIV on October 20, 2015
10
Journal of Interpersonal Violence
Table 1. Test Statistics, Means, and Standard Deviations for Significant Preliminary Analyses. Test Statistics
M (SD)
LGBIS Subscale or Demographic RRF Scale
F
Gender
Acceptance Concerns
8.86
.003 −0.40 (d)
Identity Centrality
8.53
.004 −0.38 (d)
p
d or
η2
Concealment Motivation
4.04
.008 .04 (η2)
Identity Uncertainty
3.57
.014 .03 (η2)
Internalized Homonegativity
4.21
.006 .04 (η2)
Difficult Process
3.71
.012 .03 (η2)
Identity Affirmation
5.99
.001 .05 (η2)
Identity Centrality
3.66
.013 .03 (η2)
Concealment Motivation
5.02
.002 .05 (η2)
Identity Uncertainty
5.97
.001 .05 (η2)
Internalized Homonegativity
5.92
.001 .05 (η2)
Relationship status
RAS (Satisfaction)
14.70
