Factor structure and validity of the Lesbian, Gay, and Bisexual Knowledge and Attitude Scale for Heterosexuals (LGB-KASH)
TABLE OF CONTENTS Chapter Page I. INTRODUCTION 1 II. LITERATURE REVIEW 5 Sexual Identity and Socialization 6 Heterosexual Attitudes Towards LGB Individuals 8 LGB-KASH 10 Validity of the LGB-KASH 12 Traditional-homonegativity 12 Modern-homonegativity 14 Religiosity 15 Religious fundamentalism 16 Summary 17 Research Questions and Hypotheses 19 III. METHOD 21 Participants 21 Instruments 24 Modern Homonegativity Scale 26 Religious Fundamentalism Scale 27 Functional Assessment of Chronic Illness Therapy-Spirituality Scale 28 Procedure 29 IV. RESULTS 30 Descriptive Statistics 30 Confirmatory Factor Analysis 31 Post Hoc Re-Specification of Five-Factor Oblique Model 36 Confirmatory factor analysis with revised model 40 LGB-KASH Subscales Internal Consistency and Intercorrelations 41 Convergent Validity 42 Additional Analyses: Statistical Differences Among Demographic Groups....45 Gender group differences 45 Racial group differences 46 Christian faith group differences 49 Political orientation group differences 51 V. DICUSSION 53 Factor Structure of the LGB-KASH 53 Construct Validity of the LGB-KASH 55 Religious and political affiliation 56 Race 56 Gender differences 59 Vl l l
Religiosity and spirituality 59 Limitations 60 Implications and Future Research 62 REFERENCES 64 APPENDIX A: Figures 78 APPENDIX B: Lesbian, Gay, and Bisexual Knowledge and Attitude Scale for Heterosexuals (LGB-KASH) 85 APPENDIX C: Modern Homonegativity Scale (MHS) 89 APPENDIX D: Religious Fundamentalism Scale (RFS) 92 APPENDIX E: Functional Assessment of Chronic Illness Therapy-Spirituality Scale 95 APPENDIX F: Tables 1 and 2 for Maximum Likelihood Estimates for Indicator Variables for the 28-Item LGB-KASH Five-Factor Oblique and Second-Order Model 97 APPENDIX G: Tables 1 and 2 for Maximum Likelihood Estimates for Indicator Variables for the Revised 21-Item LGB-KASH Five-Factor Oblique Model and Second-Order Model 101 IX
TABLES Table Page 1 Frequencies of Demographic Variables Categorized by Number of Males, Females, and Total Percent 22 2 LGB-KASH Item Means and Standard Deviations for Present Study and Worthington and Colleagues' Study 30 3 Goodness-of-Fit Indicators for the Competing Hypothetical Models and Worthington and Colleagues' Model for the 28-Item LGB-KASH 34 4 Goodness-of-Fit Indicators for the Revised 21-Item LGB-KASH and Worthington and Colleagues' Model 42 5 Means, Standard Deviations, and Internal Consistency Estimates for LGB-KASH Subscale Scores 42 6 Correlations Among LGB-KASH Subscales and Selected Predictor Variables 44 7 Follow-up ANOVAs to One-Way MANOVA for Gender on Predictor Variables and Means (Standard Deviations) 46 8 Follow-up ANOVAs to One-Way MANOVA for Race on Predictor Variables 48 9 Means (Standard Deviations) for Race on Predictor Variables 49 10 Follow-up ANOVAs to One-Way MANOVA for Christian Group Type on Predictor Variables and Means (Standard Deviations) 51 11 Follow-up ANOVAs to One-Way MANOVA for Political Orientation Type on Predictor Variables and Means (Standard Deviations) 52 x
LIST OF FIGURES gures Page Diagram of model depicting the individual and social identity development processes 79 Diagram of model depicting six biopsychosocial factors that influence sexual identity development 81 Diagram of the five-factor model depicting the five factors Hate, LGB Knowledge, LGB Civil Rights, Religious Conflict, and Internalized Affirmativeness 83 XI
LGB-KASH Factor Structure 1 CHAPTER I: INTRODUCTION Lesbian, gay, and bisexual (LGB) issues have been increasingly politicized in the United States with regards to same-sex marriage and adoption rights, health benefits for partners, discrimination rights within one's employment, and anti-violence laws (Rimmerman, 2001, 2008; Rimmerman, Wald, & Wilcox, 2000). For instance, during the 2004 presidential election, the issue of same-sex marriage deluged the media, subsequently leading to a congressional hearing to reexamine the Defense of Marriage Act and to consider a constitutional amendment banning same-sex marriage (National Conference of State Legislatures, 2004). Media coverage of LGB issues, such as the heated and politicized same-sex marriage debate, presents an opportunity for heterosexuals to consider their beliefs and attitudes towards LGB civil rights issues and homosexual lifestyles (Worthington, Savoy, Dillon, & Vernaglia, 2002; Worthington, Dillon, & Becker-Schutte, 2005). Over the last three decades there has been an increasingly positive trend in attitude and acceptance towards LGB people (Herek, 2000; Sherrill & Yang, 2000; Wilcox & Wolpert, 2000; Yang, 2000). At the same time research findings indicate that there has been an increase in reported violent offenses against LGB individuals (Lacayo, 1998; Skolnik et al., 2008) who are likely to be exposed to harassment, violence, or discrimination in high schools (D'Augelli, Pilkington, & Hershberger, 2002; Kosciw, Diaz, & Greytak, 2007; Herek, 2009; Human Rights Watch, 2001), university campuses (Cotton-Huston & Waite, 2000; D'Augelli, 1992; Herek, 1993, 2002; Hinrichs & Rosenberg, 2002; Rankin, 2003; Rhoades, 1994; Schwartz & Lindley, 2005; Sullivan, 1998; Waldo, 1998), and employment settings (Herek, 2009; Waldo, 1999). Researchers contend that these contradictory findings are indicative of modern-day ambivalent
LGB-KASH Factor Structure 2 heterosexual attitudes towards homosexuality (Morrison & Bearden, 2007; Morrison & Morrison, 2002; Morrison, Morrison, & Franklin, 2009; Worthington et al., 2002, 2005). That is, in the past, homosexuality was clearly not accepted because of traditional homonegativity characterized by moral objections (e.g., the belief that homosexuality is a sin). Even though in current times some heterosexuals still openly reject homosexuality based on moral reasons, others are ambivalent about their beliefs, feelings, and acceptance towards homosexuals' lifestyles and civil rights (Morrison & Bearden, 2007; Morrison & Morrison, 2002; Morrison et al., 2009; Worthington et al., 2002, 2005). A diverse range of life experiences and contextual factors contribute to the development of rejecting and ambivalent attitudes towards LGB individuals, including religious beliefs (Altemeyer & Hunsberger, 2005; Monson & Oliphant, 2007; Worthington et al., 2005), political conservatism (Herek, 2002; Hunsberger, 1996; Laythe, Finkel, & Kirkpatrick, 2001; Olson, Cadge, & Harrison, 2006; Rowatt, LaBouff, Johnson, Froese, & Tsang, 2009), extent of sexual identity exploration (Eliason, 1995; Worthington et al., 2002; Worthington, Savoy, Navarro, & Hampton, 2008), gender socialization (Kilanski, 2003; Kimmel, 1994; Kite & Wiley, 1996; Worthington et al., 2002), and prevalence of homonegativity within one's immediate environment (Bieschke, Perez, & Debord, 2007; Herek, 1995, 2000; Worthington et al., 2002). Attitudes that vary from prejudicial to LGB-affirmative are expected to develop as a result of these influences. Historically, however, research has not considered heterosexuals' positive attitudes towards LGB individuals, but has emphasized homonegative attitudes (Herek, 2000, 2004; Morrison & Bearden, 2007; Parrott, Adams, & Zeichner, 2002; Schwanberg, 1993).
LGB-KASH Factor Structure 3 The literature examining attitudes towards LGB individuals has focused on the uni-dimensional constructs of homonegativity and homophobia, which have relatively synonymous meanings that describe negative feelings and attitudes towards LGB individuals (Herek, 2000, 2004; O'Donohue & Casselles, 1993; Rothblum & Bond, 1996; Weinberg, 1972). However, recent research has suggested that heterosexual attitudes towards LGB individuals are multi-dimensional (Worthington et al., 2002; 2005). Worthington et al. (2005) developed the Lesbian, Gay, Bisexual Knowledge and Attitude Scale for Heterosexuals (LGB-KASH), which is comprised of multiple factors. These factors are represented by items that reflect both negative and positive attitudes towards LGB individuals. Worthington and colleagues concluded from confirmatory factor analysis with college students and university staff that their scale consisted of five factors including (a) avoidance, feelings of discomfort, and violence towards LGB people; (b) knowledge of LGB history and symbols; (c) attitudes toward LGB civil rights; (d) conflicted attitudes caused by religious beliefs; and (e) extent of feeling comfortable with attraction to a same-sex individual, having friends who identify as LGB, and willingness to participate in LGB social activism. The authors also assessed the validity of the instrument by examining the relationship of the LGB-KASH subscales scores to several constructs, two of which will be examined in this study: religiosity and traditional homonegativity. The purposes of this study were to examine the construct validity of the LGB- KASH with a diverse college student population. First, confirmatory factor analysis was conducted to attempt to replicate the factor structure of the instrument found by Worthington and colleagues. Second, the concurrent validity of the instrument was examined by assessing the relationship of students' scores on the LGB-KASH subscales
LGB-KASH Factor Structure 4 to their scores on scales assessing modern-homonegativity and religious fundamentalism. These two constructs are expected to be related to several scales of the LGB-KASH because they capture factors that are hypothesized to influence heterosexual attitudes toward LGB populations and issues. Limited research has been conducted with the construct of spiritual well-being and attitudes towards homosexuality. Therefore, this construct was included in the study to explore whether it differed from religious fundamentalism. Although this study focuses on the assessment of heterosexual attitudes towards LGB individuals, it is first necessary to consider the development of heterosexual attitudes. Therefore, the next chapter, which includes the literature review, provides a description of Worthington and colleagues' model of sexual identity that captures the factors that influence heterosexual attitudes. Understanding the sexual identity model will help in understanding how Worthington and colleagues developed items reflecting heterosexual attitudes and knowledge that comprised the five factors of the LGB-KASH. Subsequently, the description of the LGB-KASH scale will follow, including definitions of the various dimensions and results of factor analysis of scores in the LGB-KASH. Next, two of the constructs that Worthington and colleagues used to validate the scale (traditional-homonegativity and religiosity) and the constructs used in the present study (modern-homonegativity and religious fundamentalism) are presented.
LGB-KASH Factor Structure 5 CHAPTER II: LITERATURE REVIEW This chapter describes constructs that have historically been used to refer to attitudes towards homosexuality. Subsequently, a description of the Worthington et al. (2002, 2005) sexual identity model provides an explanation of individual sexual identity dimensions and social sexual identity dimensions. The sexual identity model provides a framework that helps to explain how Worthington and colleagues conceptualize the development of heterosexual attitudes towards LGB individuals, which is described next. After the discussion of the development of heterosexual attitudes, a description of Worthington and colleagues scale development of the LGB-KASH is provided that includes a description of confirmatory analysis results. Additionally, a discussion of validity findings with the LGB-KASH is provided that includes correlations with homonegative scales and religiosity scales. Finally, the constructs of traditional homonegativity and religiosity used in Worthington and colleagues' study are compared to the constructs of modern homonegativity and religious fundamentalism used in this study. Historically, constructs that have been used to describe heterosexual attitudes toward LGB populations and issues have included heterosexism, homophobia, and homonegativity. Heterosexism refers to a societal ideology that rests on the belief that heterosexuality is the norm and that privileges or opportunities associated with this lifestyle are entitlements (Herek, 1995, 2000; Worthington, et al., 2002, 2005). Weinberg (1972) coined the term homophobia referring to the intense irrational fear and intolerance of being around a gay or lesbian person. Empirical research indicates that heterosexuals' prejudices towards LGB people are not characterized as a phobia in the clinical sense (O'Donohue & Caselles, 1993; Shields & Harriman, 1984). That is, heterosexual
LGB-KASH Factor Structure 6 negative attitudes towards sexual minorities are not manifested through physiological reactions such as with other phobias (e.g., claustrophobia, arachnophobia). Instead, homonegativity has been conceptualized as prejudicial attitudes toward and devaluation of homosexuals (Hudson & Ricketts, 1980; Morrison & Morrison, 2002). Worthington et al. (2005) have proposed that heterosexual attitudes toward sexual minorities reflect other dimensions besides sexual prejudices and intolerance. Furthermore, they contend that heterosexual attitudes are one aspect of the person's overall sexual identity. Sexual Identity and Socialization In the past, most models of sexual identity development focused on sexual minorities (Cass, 1979; Fassinger & Miller, 1996; Hoffman, 2004; McCarn & Fassinger, 1996; Worthington et al., 2002). Worthington et al. (2002) proposed a model to describe the processes of sexual identity development among heterosexual individuals that could be expanded to include the sexual identity development of LGB individuals. Worthington and colleagues conceptualized sexual identity in terms of two reciprocal components: an individual and a social sexual identity. The individual aspect of sexual identity includes six dimensions (see Figure 1, Appendix A): (a) sexual needs (desire, impulses), (b) sexual values (judgments, acceptance), (c) preferences for sexual activities (kissing, sexual intercourse), (d) partner characteristics (physical and emotional attributes), (e) sexual orientation identity (personal definition such as heterosexual or lesbian), and (f) modes of sexual preferences (verbal or nonverbal, indirect or direct). The social component of sexual identity is comprised of two dimensions: (a) social membership identity (e.g., recognizing oneself as a member of a heterosexual group) and (b) attitudes towards LGB individuals. Attitudes towards LGB individuals, one of the two dimensions of the social component of sexual
LGB-KASH Factor Structure 7 identity, was the focus of this study. People's progression through the individual and social identity processes is believed to be influenced by six biopsychosocial factors: (a) biology (biological predispositions); (b) microsocial context (influence of family, peers, etc.); (c) gender socialization (fitting into expected gender roles); (d) culture (specific to time and place); (e) religious orientation (role of religion); and (f) systemic homonegativity, prejudice, and heterosexual privilege (see Figure 2, Appendix A). Basing their model on Marcia's (1987) ego identity statuses, Worthington et al. (2002) identified five statuses of individual sexual identity that reflect the extent of a person's exploration and commitment related to the six dimensions of one's individual components of sexual identity (i.e., sexual needs, sexual values, preferences for sexual activities, partner characteristics, sexual orientation identity, and modes of sexual preferences). Three of these statuses—unexplored commitment, active exploration, and diffusion—are equivalent to Marcia's foreclosure, moratorium, and diffusion statuses, respectively. Worthington and colleagues identified two additional individual statuses named deepening commitment and synthesis. Deepening commitment, which resembles Marcia's (1987) achieved identity status, is characterized by movement towards commitment to one's identified sexual needs and values, preferences for sexual activities, partner characteristics, and modes of sexual expression, following exploration. However, Worthington and colleagues noted that heterosexuals may transition from unexplored commitment to deepening commitment to heterosexuality without engaging in active exploration. That is, the authors hypothesized that a heterocentrist environment by default fosters the crystallization of a sexual identity that conforms to the dominant-heterosexual culture. Individuals can move from deepening commitment toward three alternative
LGB-KASH Factor Structure 8 trajectories: active exploration, diffusion, or synthesis. Achievement of the synthesis status involves integration of the individual and social sexual identity dimensions into one's sense of self. Because the individual and social components of sexual identity are reciprocal, levels of exploration and commitment are expected to influence the dimensions of both components including attitudes toward LGB individuals, which is one of the two dimensions of the social component of sexual identity. Heterosexual Attitudes Towards LGB Individuals Worthington et al. (2002, 2005) have identified several contextual factors, such as gender socialization, systemic homonegativity, and religious doctrine that are believed to influence heterosexuals' attitudes towards LGB individuals. A central assumption of the heterosexual identity model is that the dominant discourse defines heterosexual gender role behaviors as normative and homosexual behaviors as deviant. Consequently, messages from the media, family, and community typically portray homosexual lifestyles and behaviors negatively, and individuals who express gender role behaviors that are inconsistent with gender norms often experience discrimination. In other words, the process of traditional gender socialization fosters homonegative attitudes, which serve to preserve the privileged status of heterosexuality. In a heterocentrist context, self-definition as a heterosexual is primarily based on the rejection of what one is not—a homosexual—rather than in the affirmation of what one is (Worthington et al., 2002, 2005). In order to develop an affirming stance towards LGB individuals, heterosexuals need to explore their own sexuality to form a secure and positive heterosexual identity. Additionally, positive attitudes towards LGB individuals are formed to the extent heterosexuals become knowledgeable about LGB issues and aware of both the prevalence of homonegative messages and the denial of civil rights to
LGB-KASH Factor Structure 9 LGB individuals (e.g., history, LGB community). Finally, religious teachings that condemn homosexuality may engender ambivalent attitudes among individuals who recognize the prevalence of homonegativity in the community, but who experience conflict between their acceptance of homosexuals and their adherence to religious beliefs. According to Worthington et al.'s (2002) model, sexual identity exploration that is outside one's expected norms also influences a heterosexual person's attitudes towards homosexuality partly by helping the person develop comfort in dealing with same-sex attraction feelings, interest and motivation to participate in LGB social activism, and openness to having friends identified as LGB. In the process of sexual identity development, active exploration involves evaluation and experimentation at a cognitive or behavioral level that is related to the dimensions of individual sexual identity (perceived sexual needs, sexual activities, characteristics of partners, sexual values, sexual orientation identity, and preferred modes of sexual expression). In order for sexual active exploration to influence positive attitudes toward homosexuality, the sexual exploration must be meaningful and beyond what is expected within the person's social context such as (a) experimenting with persons outside of one's sexual identity, racial, and socioeconomic group or (b) exploring with different types of sexual activities (Worthington et al., 2002, 2005). Therefore, sexual exploration that exclusively conforms to the norms of one's culture is unlikely to generate positive attitudes towards homosexual identified people or lifestyles. Compared to those who have not engaged in active sexual exploration, heterosexuals who have engaged in cognitive and behavioral sexual exploration beyond expected norms will likely have a stronger sense of self- awareness and security regarding their own sexual identities and a greater understanding of diversity of sexual expression among others (Worthington et al., 2005). Therefore, it is
LGB-KASH Factor Structure 10 expected that increased active exploration of the individual's sexual identity will be associated with heterosexuals' attitudes of acceptance of LGB persons and issues. In sum, Worthington et al. (2002, 2005) proposed that heterosexual attitudes towards LGB individuals encompass the following dimensions: (a) systemic homonegativity; (b) conflicted attitudes towards LGB individuals due to religious beliefs and values; (c) knowledge of LGB history, symbols, and community; (d) attitudes toward LGB civil rights issues; and (e) sexual self-awareness and affirming attitudes of being associated with an LGB individual. They developed a scale, titled the Lesbian, Gay, Bisexual Knowledge and Attitude Scale for Heterosexuals (LGB-KASH), to assess heterosexuals' knowledge of and attitudes toward gay, lesbian, and bisexual individuals and issues in terms of these five dimensions. The next section provides a brief overview of the development of the LGB-KASH. LGB-KASH The LGB-KASH is a multi-factorial scale that assesses knowledge and attitudes toward LGB individuals in domains that have not been captured in previous measures, including religious conflicted attitudes, knowledge of LGB symbols and history, sense of comfort with one's sexuality, willingness to participate in LGB social activism, and violent homonegativity. Typically, scales have assessed heterosexual attitudes toward sexual minorities uni-dimensionally, ranging from condemnation to tolerance of LGB individuals (Herek, 1994; Worthington et al., 2002). In several studies, Worthington and colleagues examined the factor structure (via exploratory and confirmatory factor analysis) and validity of the LGB-KASH with primarily White college students from several Midwestern universities. Their analysis indicated that the scale is comprised of five factors that are consistent with their theoretical model.
LGB-KASH Factor Structure 11 Presently, the only published research examining the LGB-KASH's factor structure and validity are the studies Worthington et al. (2005) conducted as part of the scale's development. Exploratory factor analysis of the LGB-KASH yielded five factors that reflect to an extent components of Worthington and colleagues sexual identity development model. These five factors were labeled Hate, LGB Knowledge, LGB Civil Rights, Religious Conflict, and Internalized Affirmativeness (see Figure 3, Appendix A). Hate refers to attitudes about avoidance, self-consciousness, hatred, and violence toward LGB individuals shaped by systemic homonegative experiences. LGB knowledge reflects an individual's knowledge base regarding the history, symbols, and organizations related to the LGB community. LGB civil rights capture beliefs about the rights of LGB individuals with respect to issues such as same-sex marriage, child rearing, health care, and insurance benefits. Religious conflict refers to conflicted beliefs and ambivalent attitudes towards LGB individuals due to religious beliefs. Internalized affirmativeness captures self-awareness, comfort with having an LGB friend or feeling attracted to someone of the same sex, and a willingness to engage in LGB proactive social activism. Results of a confirmatory factor analysis with the LGB-KASH in a subsequent study indicated that the five-factor oblique model and the second-order model evidenced a mediocre fit with the data. The second-order model, however, did not improve the fit over the five-factor oblique model. Worthington and colleagues (2005) examined the factor structure and validity of the LGB-KASH, which included primarily (85%) European White college students and adults recruited through the web. A unique aspect of the present study is that an ethnically diverse group of college students were sampled to examine factor structure and validity of the LGB-KASH, replicating Worthington et al.'s findings. Therefore, in this study, a confirmatory factor analysis was conducted to
LGB-KASH Factor Structure 12 examine whether the factor structure of the LGB-KASH would replicate with an ethnically diverse group of college students. Validity of the LGB-KASH To assess the convergent and concurrent validity of the scale, Worthington et al. (2005) examined the relationship of the LGB-KASH subscales to several, two of which will be examined in this study: religiosity and traditional-homonegativity. These two constructs both have limitations that provide a rationale for using new constructs in this study's validity assessment of the LGB-KASH scale. Even though findings provided evidence for the validity of the LGB-KASH subscales, there are some limitations with the measurement of traditional homonegativity and religiosity used by Worthington et al. that will be discussed in the following subsection. In addition, an overview of the corresponding constructs of modern-homonegativity and religious fundamentalism will be discussed as appropriate variables to assess the validity of the LGB-KASH subscales. Traditional homonegativity. Worthington et al. (2005) examined the relation of the LGB-KASH subscales to a measure that assessed attitudes towards LGB individuals called the Attitudes Towards Lesbians and Gays (ATLG; Herek, 1984). They found that the ATLG was moderately associated with the LGB-KASH subscales. However, the ATLG assesses traditional homonegative attitudes that may have changed over the last couple of decades in the United States because of the politicalization of homosexuality (Tucker & Potocky-Tripodi, 2006; Herek, 1994; Morrison & Bearden, 2007; Morrison & Morrison, 2002; Morrison et al., 2009). The construct of traditional homonegativity is constrained because it characterizes attitudes and feelings that are founded primarily on biblically-based moral objections and stereotypes of homosexual people and their lifestyles (Herek, 1984; 1994, 2000; Morrison & Morrison, 2002; Morrison et al., in
LGB-KASH Factor Structure 13 press; Morrison, Parriag, & Morrison, 1999; Worthington et al., 2005). Although traditional homonegativity demonstrates moderate association with factors of the LGB- KASH, its' definition does not capture the increasingly prevalent modern day heterosexual attitudes towards LGB individuals (Tucker & Potocky-Tripodi, 2006; Herek, 1994; Morrison & Bearden, 2007; Morrison & Morrison, 2002; Morrison et al., 2009; Worthington et al., 2005). Current-day homonegativity is characterized by people believing LGB individuals are inappropriately asking for benefits and rights (Herek, 1994; Morrison & Morrison, 2002; Morrison, Morrison, & Franklin, in press). Additionally, modern-day prejudices are comprised of ambivalent attitudes, rather than strict biblically-based moral objections about homosexual people and lifestyles (Morrison & Morrison, 2002; Morrison et al., 2009; Raja & Stokes, 1998). Herek (1984) has noted that because the ATLG was validated in 1984, it requires either modification or replacement by a more modern instrument that reflects the complexity of current heterosexual attitudes. Over the last couple of decades the ATLG has been widely used, with results indicating that college students have increasingly responded with higher tolerance, which could reflect an increased acceptance of LGB individuals (Altemeyer, 2001; Mohipp & Morry, 2004; Schellenberg, Hirt, & Sears, 1999; Simoni, 1996; Waldo & Kemp, 1997). Scholars contend that individuals may respond with lower levels of homonegativity on the ATLG measure because they are responding to items that reflect outdated prejudices against homosexuality (Herek, 1994; Morrison & Morrison, 2002; Worthington et al., 2005). Items on the ATLG reflect an old-fashioned prejudice against gay men and lesbians based on traditional religious moral beliefs and misconceptions about homosexuality (e.g., "Male homosexuality is a perversion," "The growing number of lesbians indicates a