<%BANNER%>

Understanding homonegative attitudes through sex, race, and gender role ideologies

University of Florida Institutional Repository

PAGE 1

UNDERSTANDING HOMONEGATIVE ATTI TUDES THROUGH SEX, RACE, AND GENDER ROLE IDEOLOGIES: AN AN ALYSIS OF THE 1972-1998 GENERAL SOCIAL SURVEY By CHRISTIAN GROV A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS UNIVERSITY OF FLORIDA 2003

PAGE 2

Copyright 2003 by Christian Grov

PAGE 3

This thesis is dedicated to Scott-Lee Cash.

PAGE 4

ACKNOWLEDGMENTS This paper would have not been possible without the consistent support of several faculty members at the University of Florida. First and foremost I would like to thank Kendal Broad for helping spawn my interest in gender and sexuality research, in addition to always making herself available for both personal and professional advice. Secondly, I wish to thank my committee members, Constance Shehan and Charles Peek. Charles has always been there to shed light where there was empirical darkness, aiding me in my growth as a graduate student and quantitative researcher. Connie has always been there to offer both moral and emotional support. She has had the ability to give my work substance and ground when I was at a loss for direction. Finally, I wish to thank John Henretta. As the former Graduate Coordinator of the Department of Sociology, John has offered me both logical and hard-to-come-by advice, guiding me into the transition of becoming a graduate student. As the departmental chair, John has expressed a superior sense of responsibility and moral obligation to all graduate students. He has always pointed me in the right direction, whether in life or academically. In total, I would not be where I am today without the combined support of Chuck, John, Kendal and Connie. Distinctively and profoundly, they have influenced my past and shed light upon my future. iv

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES............................................................................................................vii LIST OF FIGURES.........................................................................................................viii ABSTRACT.......................................................................................................................ix CHAPTER 1 INTRODUCTION........................................................................................................1 2 BACKGROUND AND LITERATURE.......................................................................3 Sex and Gender Role Ideologies...................................................................................4 Racial Differences in Homonegativity.........................................................................8 Additional Influential Variables.................................................................................10 Education.............................................................................................................11 Religious Fundamentalism..................................................................................12 Age......................................................................................................................12 Political Ideology.................................................................................................13 Hypotheses..................................................................................................................13 3 RESEARCH DESIGN AND METHODS..................................................................15 Sample........................................................................................................................15 Measures.....................................................................................................................16 4 RESULTS...................................................................................................................21 Univariate Analysis....................................................................................................21 Bivariate Analysis.......................................................................................................22 Multivariate Analysis..................................................................................................25 Intervening and Suppressor Effects.....................................................................27 Nonadditivity.......................................................................................................31 Race interactions..........................................................................................33 Gender interactions......................................................................................36 Interactions with sex, race, and the gender role scale..................................36 v

PAGE 6

5 DISCUSSION AND CONCLUSION........................................................................39 6 LIMITATIONS AND RECOMMENDATIONS.......................................................44 LIST OF REFERENCES...................................................................................................47 BIOGRAPHICAL SKETCH.............................................................................................52 vi

PAGE 7

LIST OF TABLES Table page 3-1 Factorial analysis and Chronbachs alpha for HomoScale and RoleScale..................18 4-1 Univariate statistics......................................................................................................22 4-2 Sample size and means by year...................................................................................22 4-3 Correlation matrix for all quantitative variables..........................................................24 4-4 Pearson correlation between HomoScale and RoleScale............................................24 4-5 Racial and gender differences using T-tests................................................................25 4-6 Model summaries........................................................................................................28 4-7 Model summaries with nonadditivity..........................................................................32 vii

PAGE 8

LIST OF FIGURES Figure page 4-1 Interactive effects of education and race on homonegativity......................................34 4-2 Interactive effects of religious fundamentalism and race on homonegativity.............35 4-3 Interactive effects of race and political ideology on homonegativity.........................36 4-4 Interactive effects of education and gender on homonegativity..................................37 4-5 Interactive effects of gender and the RoleScale on homonegativity...........................38 4-6 Interactive effects of race and the RoleScale on homonegativity...............................38 viii

PAGE 9

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master or Arts UNDERSTANDING HOMONEGATIVE ATTITUDES THROUGH SEX, RACE, AND GENDER ROLE IDEOLOGIES: AN ANALYSIS OF THE 1972-1998 GENERAL SOCIAL SURVEY By Christian Grov May 2003 Chair: Constance Shehan Cochair: Charles W. Peek Major Department: Sociology This study attempts to analyze homonegativity as an expression of contempt for violations in perceived gender role ideologies. Society assigns and constricts individuals into strict behavioral patterns based on ideas about gender roles. Individuals who violate these roles are often perceived as deviant and subsequently treated as such. Additionally, previous research has demonstrated both racial and gender differences in homonegative attitudes with African Americans and males displaying more homonegativity. This study gives considerable attention to the interaction between race and gender. Data used are from the 1977-1998 General Social Survey. The General Social Survey is a face-to-face interview administered to a representative sample of the Englishspeaking American public over the age of 18. Gender role ideologies were measured using a scale of questions about attitudes toward women (e.g., women in politics, women in the workforce, women and the family). Homonegativity was measured using a scale of ix

PAGE 10

questions about attitudes toward homosexuals (e.g., allowing a homosexual to teach in a college, allowing a book written by a homosexual to remain in a library). The data show a strong relationship between supporting traditional gender role ideologies and being homonegative. Additionally, Caucasian homonegativity is more sensitive to changes in educational attainment, religious fundamentalism, political ideology, and gender role attitudes. Similarly, male homonegative attitudes are more sensitive to changes in education and gender role ideologies. x

PAGE 11

CHAPTER 1 INTRODUCTION There are a large number of characteristics an individual may possess that guide the individuals belief and attitudinal schemas. Individuals possess both biological drives and social characteristics that can sometimes be used to better explain social phenomenon. Features of an individuals religious background, race, educational attainment, upbringing, immediate environment, socioeconomic status, etcetera all interact with one another and impact how individuals interact with others. This analysis attempts to better predict the characteristics in individuals that allow them to hold negative attitudes toward gays and lesbians. Specifically, this analysis addresses how gender role ideologies, race and gender impact attitudes toward lesbians and gays. Demographic (control) characteristics are also taken into consideration. There is never one reason why an individual displays homonegative affect or behaviors. However social science has been able to identify several key predictors (e.g., religious fundamentalism, age, and educational attainment.) This study explores the facet of how attitudes about sex roles impact homophobia. Moreover, it explores the probable connection between a traditional sex role ideology and a liberal or egalitarian sex role ideology. The idea of sex roles was proposed after World War II and originally applied to the area of family studies (Ferree 1990). Individuals in the traditional family were seen as fulfilling certain roles (e.g., women nurture and men provide). Feminist thought of the 1960s and later challenged these ideologies, as they were seen as contributing to and justifying the subjugation of women (e.g., women should raise children rather than seek 1

PAGE 12

2 employment or education). Although sex role theory was originally applied to the family, and later rejected for its apparent flaws, members of society today still hold belief patterns that men are to fit into a certain set of behavior patterns just as women are (Ferree 1990). Homosexual individuals violate the sex roles assigned to them by society, as they are not procreative and oftentimes do not adhere to a standard set of masculine or feminine behaviors. This is further complicated by the stereotypical images society upholds when envisioning homosexuals as individuals and a group (e.g., flamboyant-sex-crazed gay men and masculine-ized butch-lesbian women) (Faderman 1991). These flawed images are perpetuated by a number of factors (e.g., media misrepresentation and sensationalism, individuals who are closeted about their sexuality, the persistence of homophobia). The purpose of this analysis is not to unveil all factors leading to the continuances of homophobia, but rather explicate interactions among gender, race, and gender role ideologies and how they impact social attitudes. Conclusions can expose the intricate linkage between the subjugation of women and the subjugation of other sexual minorities, and can add to the understandings of societal views of sex-role ideologies. Finally, our understanding of why potential universal gender and racial differences exist is quite limited. Analyzing confounding effects will better help unmask why differences are perpetuated. Implications could lead to protocol that may help individuals to think outside of the gender box to understand the fluidity of gender rather than to polarize roles.

PAGE 13

CHAPTER 2 BACKGROUND AND LITERATURE The scientific study of gays did not emerge until the later half of the last century, and a most of the research pathologized the behavior. It was not until 1973 that the American Psychiatric Association, which had classified it as a mental disorder, removed homosexuality from the Diagnostic and Statistical Manual. Understandably, the body of academic research on gays is limited, hence so is the specific study of homonegativity. In 1973 Weinberg documented a term in academic literature describing negative attitudes toward lesbians and gays. That term was homophobia and directly meant the dread of being in close quarters with homosexuals and in the case of homosexuals themselves, self-loathing (p. 4). Over the years the term has been redefined as attitudinal structures toward gays and lesbians have been better studied. In 1983, Gramick created the term homosexphobia to describe a fear of homosexuality. Other terms since created include sexual prejudice, homonegativity (Herek 2000) homosexism, (Hansen 1982) and homoprejudice (Logan 1996). Although homophobia is the most common term (because it was the first), many social scientists have identified an inherent flaw with it. Using a medical model, homophobia as a term has pathological implications that may not be inherent. For the purposes of this paper the term homonegativity will be used to define negative attitudes toward lesbians, gays, bisexuals, and transgenders 1 1 For the purposes of simplicity the terms LGBT and/or gay will be used synonymously to describe lesbians, gays, bisexuals, and transgenders alike. 3

PAGE 14

4 Substantial negative impacts of homonegativity on gays have been noted. Gay youth are seven times more likely to attempt suicide when compared to other teens (Baumrind 1995). Additionally, gay teens have higher levels of self-reported stress and substance abuse. Finally, these teens are more likely to experience rejection by friends and also by family. The United States Gay and Lesbian Task Force found that more than 75% of lesbians and 90% of gay men had been verbally harassed because of their sexuality (Herek 1988). In 1988, Herek found that 50% of gay males and more than a third of lesbian females reported being threatened with physical violence. The Kaiser Family Foundation (2001) conducted an opinion survey among the general public with a specific oversample of gays. Among the gay sample, Kaiser found a large majority felt there was more acceptance of homosexuality today compared to a few years ago. On the contrary, almost three-quarters of this gay sub sample had experienced a significant amount of both prejudice and discrimination because of their sexual orientation. Finally, about one-third of the sample indicated that their family (or a family member) has refused to accept them because of their sexuality. The Kaiser Foundation found that most of the general public reports knowing someone who is gay. Secondly the general public also agrees that there is greater acceptance of gays today compared to a few years ago. Nonetheless, about half of the sample felt that homosexual behavior is morally wrong. Although public support may be increasing toward gays, homonegativity in some forms is still widespread and needs attention as to its causal and contributing factors. Sex and Gender role Ideologies Research by and large has found that heterosexual males, in general, display more homonegativity than females (DAugelli and Rose 1990, Fishbein 1996, Herek 1986,

PAGE 15

5 1988, 1994, 2002, Hudson and Ricketts 1980, Johnson et al. 1997, Kite 1984, Klamen et al. 1999, Morin and Garfinklel 1978, Oliver and Hyde 1993, Pratte 1993, Price 1982, Seltzer 1992, Thompson et al. 1985, Whitley and Kite 1995). Males are generally socialized to more stringent hetero-normative gender roles, whereas females are given more freedom in the gender roles they may fulfill. An individual who violates her or his gender role is more likely to be treated as deviant (Laner and Laner 1979, 1980). Moreover, Laner and Laner (1979) found that lesbians and gays experience homonegative affect, in part, because of their perceived sex-role deviance. Specifically, a male who violates his gender role is likely to be treated more as a deviant than a female who violates her socially defined role (Herek 1994, Kite and Whitley 1996). Herek (1984) found homonegative individuals more often were highly authoritarian, had traditional attitude toward gender roles, were more negative toward other minority groups, had less education, and were male. Deaux and Kite (1987) found that heterosexual attitudes toward gays and lesbians are influenced by a generalized gender belief system, defined as a set of beliefs and opinions about males and females and about the purported qualities of masculinity and femininity (p. 97). They further discuss that this gender-belief-system drives stereotypes including attitudes toward appropriate gendered behavior, toward men and women, and toward those who violate these stereotypes. Research supports that womens homonegative attitudes are more inclined to change compared to mens (Lance 1992). Kite and Whitley (1996) argue that . rigid gender roles arguably require [heterosexual] men to eschew homosexuality. Because men are expected to avoid feminine traits or activities, and because gay persons . are often thought to be deviant from appropriate gender roles, men may feel pressured to display

PAGE 16

6 antigay prejudice (p. 338). In summary, there is a strong link between traditional gender role ideologies and homonegativity. More important are the potential differences in the links between gender role ideologies and homonegavity that may exist based on gender. Earlier research supports that greater levels of homonegativity are derived from more stringent gender role ideologies, that men succumb to more stringent ideologies than women, and that men are more homonegative. Research has continually identified the existence of gender differences in attitudes toward gays with emphasis that men tend to be more homonegative. Empirical reasoning explaining this distinction lacks development. Heterosexuality is equated ideologically with normal masculinity and normal femininity, whereas homosexuality is equated with violating the norms of gender (Herek 1988, p. 97). Society has polarized understandings of gender (Bem 1993); forcing individuals to either conform or face consequences. Polar ideologies are not new to social understanding but rather indicative of human schematic understandings (e.g., black or white, fat or thin, male or female, good or evil, heaven or hell). Heterosexuals tend to expresses more negative attitudes toward homosexual individuals of the same sex; this pattern is more pronounced among men than women (Kite and Whitley 1996). Researchers typically do not distinguish between attitudes toward gays and lesbians. Most often the terms gay or homosexual are used to describe both. Gender biases could lead respondents to consider attitudes toward gay men before they consider attitudes toward lesbians. Herek (2002) proposes that individual attitudes toward gays and lesbians may reflect that individuals attitudes toward her or his own sexuality.

PAGE 17

7 Because of homosexualitys stigmatized status, many heterosexuals wish to avoid being labeled gay or lesbian, and this concern is probably stronger among men in the U.S. society. Some individuals may feel a particular need to distance themselves from gay people because they have experienced homosexual desired or engaged in same-sex behaviors, which they regard as extremely unacceptable and inconsistent with their self-concept (p 43). In essence, a mans sense of masculine identity may encounter more insecurity, causing him to display more homonegative affect. This is in concurrence with earlier researched indicating that men have less liberty in the gender/sex role behaviors they may display. In contrast, women have more liberty in the gender/sex roles they may undertake. This may give a female individual more security in her attitudes toward individuals who violate their gender roles. Additionally, this helps explain why homonegative affect is sometimes less severe for lesbians than gay men. In accord with the previous discussion, there is one research study that refutes all previous findings indicating males are more homophobic than females. Proulx (1997) sampled 553 Brazilian college university students and found a complete inversion of gender and homophobia (i.e., women were more homonegative than men and homonegativity was more pronounced between heterosexual women toward lesbians than between heterosexual men and gay men). Proulx hypothesized that these differences are probably due to Brazilan culture, in that the women of this area undergo rigid gender roles. This data clearly supports Hereks (2002) postulations about the correlations between rigid gender role structures and subsequent attitudes toward gays and lesbians. The Proulx study is a counter example that supports the hypothesis and offers reasoning as to why these characteristics may exist. In sum, research has clearly defined three things. First there has been discussion on the impact of gender role ideologies on homonegativity finding that traditional gender

PAGE 18

8 role ideologies contribute to greater homonegativity. Second has been the discussion of gender differences in homonegativity suggesting that, in most cases, females are less homonegative than males. Finally, there is the discussion that gender role ideologies are dependent upon gender, finding that males are subjected to more stringent gender role structures giving them more traditional belief patterns. Unfortunately, there is little to no empirical research that moves far beyond this to develop other influential factors that may cause or mask these relationships. This study will attempt to examine the impact of gender and gender role ideologies on homonegativity, and how interaction between the two may better explain why differences continue to exist. The items selected for the scale measuring gender role ideologies assess several aspects of attitudes about gender roles but they center only on attitudes toward women. Although simply assessing attitudes toward women may not ascertain all aspects of gender role ideologies, they do help assess a significant portion of them. Ideally this assessment would include a variety of gender role attitudes (e.g., attitudes about mens behaviors), however the General Social Survey does not elaborate into this aspect. Racial Differences in Homonegativity Racial or ethnic makeup is a second predictor of homonegavity. The General Social Survey conglomerates race and ethnicity into three categories: Caucasians, African Americans, and other. In this study, the other group has been omitted due to lack of sample size and complex variability in race and ethnicity that is sure to exist within such a category. When comparing African Americans to Whites it is plausible to think that, because of the civil rights and equality struggles that African Americans have endured over the course of the last millennium, African Americans could be more compassionate to other minority groups being discriminated against in similar fashions. In this type of

PAGE 19

9 discussion, gays would be categorized as a sexual minority. In contrast, it is plausible to believe that because African Americans may value a more traditional family model, are on average more religious, and hold more stringent constructs of gender role identifications, they would thereby be more homonegative. Research seems to support the latter model more. Chng and Moore (1991) found that African American males were more homophobic than Caucasian males based on the information reported from nine African American respondents. It was found that African American females displayed less tolerance toward homosexuals than Caucasian females (Ernst et al. 1991). Black et al. (1998) studied the relationship of homophobia and sexism among social work students. Their findings support research that correlates gender roles and homonegativity. They additionally found that changes in sexist and corresponding homophobic attitudes might occur with greater ease among Caucasians than among African Americans. Unfortunately, Black et al. offer no assessment explaining this interaction other than citing the 1995 study by Blee and Tickamyer. Blee and Tickamyer noted that gender roles specifically were more multidimensional for African American men than Caucasian men. In all, African American homonegativity could be better explained through the differences between Caucasians and African Americans in their attitudes about gender roles. In these circumstances, research has both reified the relationship between gender roles and homonegativty, while considering racial and gender differences. Klamen et al. (1999) study of medical students found that the African American respondents were more likely to agree with statements such as Homosexuality is immoral (p 58). The same study also found that, when compared to other ethnic groups,

PAGE 20

10 African Americans were much more likely to oppose statements such as Homosexuals should have equal employment opportunity (p 59). Wills and Crawfords 2000 study found that African Americans were more likely to attribute the cause of homosexuality to a moral choice while other ethnic groups tended to say a combination of biology and environment (p 101). In general, it is found that condemnation against homosexuals is greatest in African American communities when compared to Caucasians (Freudenberg 1989, Staples 1982, Waldner et al. 1999). In contrast, African American homosexuals are not rejected by their families to the extent that Caucasian homosexuals are (Greaves 1987). This could be due to stronger family ties within the African American community. It has clearly been identified that homonegativity in African Americans is greater than that in Caucasians. Little research addresses the reasons why this may be the case. Harper (1991) suggested that this might be due to African Americans greater sense of nationalism that emerged from the civil rights movements of the 1960s. This suggests that homosexuality is viewed or can be viewed as un-American. This study attempts to discover both the influence race has over homonegativity in addition to assessing its confounding effects with other variables such as education, gender role ideologies, religious fundamentalism, and political ideology. As in the case of gender, much research has discussed the existence of this relationship, however little has been done to empirically examine why it exists. Additional Influential Variables In analyzing potential racial or gender differences in homonegativity or the impact of gender role ideologies on homonegativity, it is necessary to consider other common variables identified as having influence over homonegative attitudes. This analysis is

PAGE 21

11 necessary because these other influential variables can vary across both race and gender. Only when we control for these differences, might we be able to obtain a clearer picture of the effects of race, gender, and gender role ideologies. Education It is a general fact that a college education tends to increase ones sense of liberalness. Individuals with lower levels of education are less likely to have exposure to the vast aspects of many social issues. Many activists have hailed high schools as being some of the most homophobic institutions in existence. In general, neither texts nor teachers in secondary institutions cover the environmental, social, and biological aspects of homosexuality. Understandably these aspects are not touched upon in any lower levels of schooling either (e.g., middle school or elementary school). Colleges are known for fostering social change; nonetheless anti-gay sentiments are [still] prevalent (and even fostered) at the university level (Walters and Hayes 1998, p. 3). Whitleys 1987 study of college students found that older students were less negative toward homosexuals than were the younger ones. First year college students, in general, displayed negative attitudes toward homosexuality (Young and Whertvine 1982). Although both Whitleys and Young and Whertvines research from the 1980s may seem dated, many researchers have consistently found that the less education a person has, the more likely she or he is to display homonegative attitudes (Beran, et al. 1992, Bowman 1979, Glenn and Weaver 1979, Irwin and Thompson 1977, Nyberg and Alston 1976, Price and Hsu 1992). Wills and Crawford (2000) found that people with a masters or doctorate degree were least likely to agree with questions concerning gays going to hell as a result of sexual orientation. They also found that respondents with a high school diploma or below were more likely to attribute homosexuality to a moral

PAGE 22

12 choice, rather than environment or biology. Finally, they found respondents with a high school degree or below were less in favor of homosexuals having equal civil rights. Religious Fundamentalism Certain religious segments have been most noted for expressing homonegavity (e.g., the Christian coalition). In this study, level of religious fundamentalism, in contrast to liberalism, will be controlled for. Admittedly, there are many factors of religion that can impact homonegativity such as extent of intrinsic beliefs, denomination, frequency of church attendance, etc. On average, Catholics are usually more tolerant of homosexuality than are Protestants (Bierly 1985, Wills and Crawford 2000). People whose religion has a more fundamentalist orientation are more likely to be homonegative compared to individuals with religions that have less fundamentalist orientations (Herek and Glunt 1993). Herek (1994) found that simply being more religious would cause a person to exhibit a more anti-homosexual attitude. In contrast, research supports that individuals high in several aspects of religiosity will internalize their beliefs; thereby they display less homonegativty than individuals who use religion as an explanation for current social climates (Batson et al. 1986). Age The children of both Generation D and Generation X have been growing up in an era where gays are being portrayed by the media in a more positive light. Some of the only negative media attention that a member of Generation X would have heard about gays may have centered on the HIV/AIDS epidemic. For the most part, it is now socially accepted that HIV/AIDS is not a gay disease, and hence much of that attribution has been dispelled. Members of previous generation cohorts, such as the baby boomers and those

PAGE 23

13 before them, grew up in eras where they saw homosexuality was still classified under a medical model as a mental disorder. Similarly, they lived in times where it was acceptable for the police to harass lesbians and gays, and in general to display homonegative affects and behaviors. Individuals growing up and being raised in a society where gays are more accepted would be more socially conditioned to accept homosexuality. Similarly, individuals who grew up and were raised to believe that some/all aspects of homosexuality are in some part wrong would be more secure in those beliefs. The older an individual is (or the longer they spent in a time period where homonegativity was more acceptable) the more homonegative attitudes they will continue to display. Political Ideology Research has supported that homonegativity is positively correlated to the level of political conservativism a person displays (Whitcomb 2001). Heaven and Oxman (1999) found that males were more likely display conservative attitudes and that these attitudes correlated strongly with homonegativity. Among members of the military Marine Corps, researchers found that homonegativity and conservative attitudes correlated strongly (Estrada 1999). There is a clear positive relationship between politically conservative attitudes and homonegativity. Hypotheses Based on the literature review and corresponding theoretical framework built for this study, the following relationships are believe to be demonstrated by the data: Individuals with more traditional gender role ideologies will have more homonegative attitudes. Males will display more traditional gender role ideologies than females.

PAGE 24

14 Males will display more homonegativity than females. African Americans will display more homonegativity than Caucasians.

PAGE 25

CHAPTER 3 RESEARCH DESIGN AND METHODS Sample The General Social Survey, a time series study based on a probability sample, is a survey of opinions and beliefs administered face to face every few years to a representative sample of the United States population. Its content briefly addresses a plethora of topics including attitudes about the government, abortion, welfare, suicide, pornography, work, family etc. Sample sizes for each survey year are roughly under 3,000 individuals with an over sample of African Americans. Data span from 1972 through 2000. This study uses data from the 1977, 1985, 1988, 1989, 1990, 1991, 1993, 1994, 1996, and 1998 datasets (N = 6133) 2 In order to be qualified for sample consideration respondents must have been between the ages of 18 and 89 years, English-speaking, and not institutionalized at the time of the survey (i.e., whose who reside in nursing homes, incarcerated individuals, or those committed to mental institutions are not surveyed). Significant limitations of this survey include the exclusion of non-English speaking households and individuals living on U.S. Military bases. The response rate is approximately 75%. 2 Data were originally to be taken from the 1998 dataset alone. Unfortunately, not a great enough number of respondents answered all questions relevant to the two scales. To counteract, all years were included into the sample selection. Again, not all questions were asked every year. The final set of years utilized reflects those in which all items on the selected scales were asked to respondents (1977, 1985, 1988, 1989, 1990, 1991, 1993, 1994, 1996, and 1998 See Table 3) Additionally the sample size of 6133 does not reflect all respondents from those years, but rather those remaining who answered all questions used in this study. 15

PAGE 26

16 Measures Two scales will be used for this analysis. The dependent variable is constructed from an array of questions assessing attitudes toward homosexuals. As mentioned, gender role ideologies constitute one of the major explanatory variables in this study. Attitudes about gender roles were assessed using an array of questions specifically about the roles of women. The Homonegativity scale (HomoScale) was be derived from the following series of questions 3 : HOMOSEX: What about sexual relations between two adults of the same sex do you think it is . SPKHOMO: And what about a man who admits that he is a homosexual. Suppose this admitted homosexual wanted to make a speech in your community. Should he be allowed to speak, or not? COLHOMO: Should a homosexual be allowed to teach in a college or university, or not? LIBHOMO: If someone in your community suggested that a book, written by a homosexual, in favor of homosexuality should be taken out of your public library, would you favor removing this book or not? The question HOMOSEX 4 is coded on a scale from 1 to 4 (Strongly Disagree, Disagree, Agree, and Strongly Agree), while the other three mnemonics are dichotomized (e.g., yes or no, allow or not allow). The responses of the HOMOSEX variable were collapsed into dichotomous categories in order to match the other variables in the scale. This additionally prevents overweighing the responses for other items in the scale. Items 3 The homonegativity scale originally included the variable ASKSEXOR: Before giving an individual secret or top-secret clearance, the government should have the right to ask a person their sexual orientation? The inclusion of this variable into the scale presented a significant problem, as there were a limited number of individuals this question was asked of. The sample size was cut virtually in half with the inclusion of this variable. 4 The mnemonic for ASKSEXOR also was on a 4-point scale similar to HOMOSEX, however omitted (See comment 2).

PAGE 27

17 on the homonegativity scale were reordered and added together so that a score of zero represents complete acceptance of homosexuality while a score of 4 represents the highest level of homonegativity. The second scale on attitudes about gender roles (RoleScale) includes the following questions: FEFAM: Now I'm going to read several more statements. As I read each one, please tell me whether you strongly agree, agree, disagree, or strongly disagree with it. For example, here is the statement: D. It is much better for everyone involved if the man is the achiever outside the home and the woman takes care of the home and family. FECHLD: Now I'm going to read several more statements. As I read each one, please tell me whether you strongly agree, agree, disagree, or strongly disagree with it. For example, here is the statement: A. A working mother can establish just as warm and secure a relationship with her children as a mother who does not work. FEHOME: Do you agree or disagree with this statement? Women should take care of running their homes and leave running the country up to men. FEPOL: Tell me if you agree or disagree with this statement: Most men are better suited emotionally for politics than are most women. FEWORK: Do you approve or disapprove of a married woman earning money in business or industry if she has a husband capable of supporting her? The questions FEFAM and FECHLD are on a four-point scale (similar to HOMOSEX). The remaining variables are on a two-point scale of Agree or Disagree. Both FEFAM and FECHLD were recoded into dichotomous categories much like HOMOSEX. Like the homonegativity scale, the gender role scale will range from 0 to 5 with higher scores indicating more traditional attitudes about gender roles (in contrast to lower scores indicating more egalitarian attitudes). Scales are being used rather than individual items because the items selected represent variations in similar characteristics.

PAGE 28

18 Both the homonegativity scale and gender role scale were assessed for reliability using factorial analysis and Varimax rotation (Table 3-1). Varimax rotation identified one factor for each scale. Follow-up analysis computed Chronbachs alpha for both scales (a = 0.783 for the homonegativity scale and a = 0.697 for the gender roles scale) (Tables 4-1 and 4-2). Overall mean value for the homonegativity scale was 1.627 with insignificant skew and slightly platykurtotic (kurtosis = -1.026). The gender role scale had an overall mean of 1.510 and was neither significantly skewed nor kurtotic (Table 4-1). Table 3-1 Factorial analysis and Chronbachs alpha for HomoScale and RoleScale H VariableFactorAlpha if item deletedomoScaleHomosex0.58270.8228Chronbach Alpha = 0.7827Colhomo0.86040.6734Spkhomo0.84660.6922Libhomo0.81240.7104RoleScaleFefam0.76390.5992Chronbach Alpha = 0.6966Fechld0.60580.6736Fepol0.69960.6459Fehome0.78840.5999Fework0.48790.7048N=6133Source, General Social Survey 1977-1998 As discussed, gender and race were also included in this analysis. Gender was dichotomized into male and female (F = 1) while race was dichotomized into African American and Caucasian (Af. Amer. = 1). Finally, the additional control factors were included. These factors include the age at the time of the interview (in years), highest educational attainment of the respondent (in years), religious fundamentalism, and the political ideology of the respondents. Religious fundamentalism was measured by asking respondents how they identified their

PAGE 29

19 religious beliefs (fundamental, moderate or liberal) 5 Political ideology was measured using a seven-point scale in which respondents were asked to indicate where they fall in their political beliefs (1 being extremely liberal and 7 being extremely conservative). Finally, educational attainment was measured by asking respondents what was the highest grade completed. Responses could vary from 0 (no formal schooling) to 20 (a doctoral degree or greater than four years of schooling beyond a bachelors degree). As discussed, research has identified that older individuals, and individuals with less education display more homonegativity. Furthermore, individuals who identify as politically conservative or religiously fundamental also have a tendency to display more homonegative affect. Univariate, Bivariate, and Ordinary Least Squares multivariate analysis were used to see how gender role attitudes, gender and race confound to explain homonegative affect. All models will take into consideration the year the data were collected, which were coded into a series of dummy variables. For modeling purposes, data from the year 1998 were excluded as this year represented the reference category. As discussed, homonegative attitudes have been in flux over the course of the last quarter century. Consistent consideration for the year of data collection will help better show the effects of independent variables net of year. Model building will consider the effects of all variables discussed simultaneously (Model 2, Table 4-6), the impact of gender role ideologies alone (Model 1), the impact of 5 In the assessment of both scales and all independent variables, non-response to any one of the questions constituted exclusion from the sample. In the end, 6133 individuals remained. See also, comment 2. Fundamentalism was coded 1 through 3 with higher values corresponding to liberal orientation.

PAGE 30

20 gender alone (Model 3) and race alone (Model 9). Gender role ideologies, gender, and race are believed to provide substantial impacts on homonegativity. To assess the impacts of other confounding variables on the significance of gender and race, additional models were run (Models 4-8 and 10). Finally, assessment of nonadditivity between both gender and race with other variables was taken into consideration (Models 11-16, Table 4-7).

PAGE 31

CHAPTER 4 RESULTS Univariate Analysis Approximately 11.3% of the sample was comprised of African American individuals. The remaining 88.7% were identified as Caucasian. The sample was comprised of approximately 56% female and 44% male 6 The mean age was roughly 44.5 years with a median of 41 years (Table 4-1). As stated, ages varied from 18 to 89 years. The mean education of the sample was 12.88 years, or basically just short of one year of college. The median education reported was the 12 th grade. Religious fundamentalism was coded by asking respondents whether their religious denomination was fundamentalist, moderate, or liberal (coded 1 through 3 with higher values indicating more liberalism). Mean value for fundamentalism was 1.94 with both the median and mode at 2 (moderate). Finally, for political ideology, respondents were asked how they identified on a scale from 1 through 7 with a response of 1 indicating extremely liberal while a response of 7 corresponded to extremely conservative. Mean response was 4.07 (moderate) with a median and mode of 4 as well. Distributions for all variables were approximately normal, with the exception of a slightly platykurtotic distribution for religious fundamentalism (Kurtosis = -1.26). The final sample size was 6133 (Table 4-2). 6 Race was dummy coded with values of 0 corresponding to Caucasian and 1 corresponding to African Americans. Gender was dummy coded with the 1 group representing females. 21

PAGE 32

22 Table 4-1 Univariate statistics H R A F A E FP * N S ValueFrequencyPercentMeanSt. DevSkewKurtosisMinMaxomoScale*0149624.391.6261.4120.546-1.026041208734.03287314.2335679.254111018.1oleScale**0213234.761.5101.5140.756-0.507051143023.322101916.62372011.7445268.5853064.99fam0.11301emale0.56201ge44.47316.9320.515-0.6451889ducation12.8783.020-0.1710.767020undamental1.9430.7590.095-1.25813olitical ID4.0711.383-0.071-0.45617 Increasing values correspond to increasing HomoNegativity* Increasing values correspond to increasing traditional gender role ideologies=6133ource, General Social Survey 1977-1998 Table 4-2 Sample size and means by year * N S YearSample sizeHomoScale mean*RoleScale mean**197711632.0122.321198511931.8811.59919883441.7211.43919893591.5851.33719903291.6231.35719913621.6191.46119933971.4841.23719946931.3451.08219966811.2481.14119986121.2061.223 Higher values correspond to more negative attitudes* Higher values correspond to more traditional gender role attitudes= 6133ource, General Social Survey 1977-1998 Bivariate Analysis Pearsons Correlations were assessed for all quantitative variables (e.g., both scales, respondent age, education, political ideology, and religious fundamentalism) (Table 4-3). Correlations were significant in all relationships except for those between religious

PAGE 33

23 fundamentalism and age. The homonegative scale had a positive moderate correlation with the gender role scale supporting the first hypothesis that individuals expressing traditional gender role ideologies were also more homonegative (r = 0.48, p < 0.01). The homonegative scale also had positive weak correlations with age and political ideology (r = 0.26, p < 0.01, and r = 0.22, p < 0.01 respectively), supporting expected relationships between both older individuals displaying more homonegativity and politically conservative individuals expressing more homonegativity. Additionally, there was a moderate negative correlation between respondents education and reported homonegativity (r = -0.41, p < 0.01) supporting that individuals with a higher education will report less homonegativity. Further, there is weak negative correlation between the homonegative scale and religious fundamentalism (r = -0.295, p < 0.01), supporting that religiously liberal individuals report less homonegativity. Noteworthy is the substantial correlation between the homonegative scale and the gender role scale (r = 0.48, p < 0.01). Finally, correlations among independent variables, although significant, were relatively weak ruling out possible multicolinearity (Table 4-3). Because the focus of this project ascertains the relationship between homonegative attitudes and gender role ideologies, correlations by year were assessed between the two scales (Table 4-4). Overall, the strength and direction of the relationship is relatively consistent. There is a slight dip in the strength of the correlation in 1989 (r 1989 = 0.34), however not a substantial drop. Nonetheless, the strength of the relationships between the scales was at its highest point a year prior (r 1988 = 0.55).

PAGE 34

24 Table 4-3 Correlation matrix for all quantitative variables H A FPol V* HomoScaleRoleScaleAgeEduc-ationFund-amentalPol. viewsomoscale10.480.26-0.41-0.2950.22Rolescale10.33-0.36-0.180.19ge1-0.240.01 0.12Education10.23-0.05undamental1-0.14. Views1alues correspond to Pearson's correlation NOT Significant a > 0.05N=6133Source, General Social Survey, 1977-1998 Table 4-4 Pearson correlation between HomoScale and RoleScale S YearCorrelation*19770.4919850.4719880.5519890.3419900.4119910.4319930.4719940.3819960.4519980.47*All significant a p < 0.01ource, General Social Survey 1977-1998 Assessment of potential racial or gender differences were calculated using independent sample t-tests (Table 4-5). Focusing on racial differences, we find that African Americans report a mean higher score on the homonegative scale. This supports the fourth hypothesis that African Americans will display more homonegativity than Caucasians. There are no significant racial differences on answers toward the gender role scale. Nonetheless, there are significant racial differences in age, education, religious fundamentalism, and political ideology. In sum, we find that African Americans in the sample are approximately three years younger, have about a year less education, are more fundamentalist, and politically liberal than Caucasians in the sample.

PAGE 35

25 Table 4-5 Racial and gender differences using T-tests Variable T valuep-valueMean for WhitesMean for African AmericansRaceHomoscale-4.73<.011.5961.86RoleScale-0.050.61Age4.32<.0144.841.9Education7.72<.0112.9812.05Fundamental18.12<.0121.46Pol. Views3.74<.014.093.89Mean for MenMean for WomenGenderHomoScale1.780.08RoleScale4.72<.051.611.43Age-2.96<.0143.745Education3.85<.011312.7Fundamental4.56<.011.991.90Pol. Views1.570.11N=6133Source, General Social Survey 1977-1998 In analysis of gender differences, we find no differences in mean scores on the homonegative scale. This does not support the third hypothesis that men will display more homonegative attitudes than women. As expected, we do find gender differences in that women are more egalitarian in their gender role ideologies, supporting the second hypothesis. Additionally, women are about a year and a half older on average, have less education, and report more fundamentalist religious ideologies than men. Finally, we find no significant gender differences in political ideology. Hypothetical stipulations and previous research have both supported that women tend to display less homonegativity than males. Controlling for differences between males and females in age, education, gender role ideologies, and/or religious fundamentalism may help to better expose the relationship between gender and homonegativity Multivariate Analysis In an effort to understand the effects of the independent variables on the homonegative scale while controlling for the effects of other variables, multivariate

PAGE 36

26 analysis was conducted using Ordinary Least Squares methods. Model 1 (Table 4-6) presents the effects of the gender role scale on the homonegative scale while controlling for the variability in each year the data were collected. As stated, dummy coding was allowed for each year data were collected from and 1998 is excluded from all models as it acts as the reference group. In Model 1, we see the effect of traditional gender views on homonegativity is positive and significant. Increases in traditional gender role ideologies correspond to increases in homonegativity (Std. B = 0.466, p < 0.05). Additionally, significant to all models is the apparent spike attributed to the data from the year 1985. It is believed that this spike in overall predicted homonegativity could be better attributed to the historical events surrounding the discovery of HIV/AIDS in the early 1980s and its greater incidences among gay men. In 1984, nationwide recognition was given to the existence of this epidemic. At that time, the disease was mainly affecting gay men. Societal initial responses were to both blame the disease on gay men and also allow justification as punishment for immoral behavior. This spike observed consistently in 1985 could be evidence of a period effect being experienced in response anti-gay sentiments associated with the HIV/AIDS epidemic. With time, there was an increase in recognition that this disease was not necessarily a gay disease. This could explain the observed drops in homonegativity attributed to subsequent years of data. Finally, while controlling for the years data were collected, we see that more recent years (in particular 1994 and 1996) do not contribute significantly compared to 1998 in explaining homonegativity. This can be described as a plateau effect. In essence, we see an overall pattern of decrease in homonegativty across the years with diminishing effects in more recent years.

PAGE 37

27 Model 2 considers all variables introduced in this study (Full Model). Controlling for the effects of gender, race, education, religious fundamentalism, political ideology, and the years the data were collected, we still observe a significant contribution of gender role ideologies on attitudes toward homosexuals. As stated, all models that include the gender role scale as a predictor, find that the gender role scale contributes the most predictive ability (demonstrated by its Standardized Estimate). In analysis of other control factors, we find all expected relationships confirmed with the exception of predicted racial differences. In Model 2, we find that race is not a significant predictor of attitudes toward homosexuals while controlling for other factors. Otherwise, net of the effects of other variables, we find that men, older individuals, those with lower levels of education, those expressing fundamentalist religious ideologies, those with traditional gender role ideologies and those expressing conservative political ideologies are all predicted to display more homonegative attitudes than their counterparts. Using a subsequent F-test, model 2 provides overall explanatory power of the homonegativity scale (F = 209 p < 0.05) 7 In analysis of this power, we find that approximately 35% of the variability in the homonegativity scale can be better explained by the gender role scale, race, gender, education, political ideology, age and religious fundamentalism compared to using the mean of the homonegativity scale alone (R 2 = 0.353). Intervening and Suppressor Effects The data at hand present both intervening and suppressor effects related to expected relationships among race and gender and their effects on homonegativity. As remarked in 7 Significant F-statistics were obtained from all models discussed.

PAGE 38

28 Table 4-6 Model summaries** Model 1Model 2Model 3Model 4Model 5Model 6Model 7Model 8Model 9Model 10Intercept.7192.1201.2343.7272.357.2392.1422.59.6731.55RoleScale.466 (.434)*.300 (.280)*.351 (.327)*.320 (.299)*.465 (.433)*.427 (.398)*Afam.016 (.071).063 (.281)*.016 (.069)Female-.025 (-.072)*-.017 (-.047)-038 (.108)*-.033 (-.097)*-.027 (-.076)*-.015 (-.042)-.027 (-.075)*Age.101 (.008)*.267 (.022)*.087 (.007)*.108 (.009)*Education-.218 (-.102)*-.385 (-.180)*-.252 (-.118)*-.214 (-.100)*Fundamental-.172 (-.319)*-.288 (-.536)*-.189 (-.352)*-.212 (-.394)*Pol. View.113 (.115)*1977.079 (.286)*.067 (.242)*.223 (.804)*.129 (.463)*.198 (.712)*.233 (.804)*.057 (.204)*.320 (.214)*.082 (.295)*.074 (.266)*1985.131 (.469)*.114 (.408)*.188 (.672)*.139 (.496)*.174 (.619)*.196 (.699)*.116 (.413)*.059 (.411)*.135 (.483)*.127 (.452)*1988.062 (.378)*.051 (.311)*.083 (.514)*.057 (.353)*.071 (.433)*.091 (.561)*.052 (.321)*.115 (.296)*.061 (.375)*.054 (.239)*1989.048 (.286)*.038 (.226)*.063 (.378)*.043 (.261)*.051 (.307)*.067 (.402)*.050 (.241)*.048 (.214)*.059 (.294)*.041 (.245)*1990.050 (.315)*.046 (.290)*.067 (.417)*.053 (.330)*.063 (.397)*.068 (.429)*.046 (.287)*.036 (.291)*.052 (.327)*.050 (.311)*1991.044 (.266)*.035 (.208)*.068 (.410)*.049 (.292)*.056 (.337)*.069 (.405)*.037 (.223)*.046 (.196)*.046 (.275)*.038 (.229)*1993.040 (.228)*.024 (.137).048 (.273)*.025 (.146).039 (.225)*.049 (.279)*.028 (.160)*.033 (.143).040 (.232)*.035 (.201)*1994.035 (.156)*.024 (.106).031 (.138).021 (.093).020 (.090).035 (.154)*.029 (.128)*.025 (.101).036 (.160)*.027 (.121)1996.001 (.034)-.0005 (-.002).009 (.041).003 (.012).002 (.011).014 (.062).005 (.204).023 (.009).007 (.031).003 (.012)R-Squared.242.353.043.184.126.114.308.341.246.287F-Value19520927.512579.871.7209227166205Dependent Variable: HomoScaleN=6133* p < 0.05**Values correspond to standardized parameter estimates. Those in parentheses correspond to the parameter estimate.Source General Social Survey 1977-1998

PAGE 39

29 bivariate analysis of gender and homonegativity, we found that there was weak evidence of a bivariate relationship (men and women did not differ significantly in their homonegative attitudes). However, in Model 2 (the full model), we find that gender does provide a significant effect while controlling for other variables. This evidence now supports the second hypothesis that males are more homonegative. Further analysis was conducted to unveil what variables contributed to this suppressor effect. Models 3 through 8 are those being used to discuss the relationship between gender and homonegativity. Model 3 demonstrates that there is no evidence of a relationship between gender and homonegativity (even while controlling for the year the data were collected from). However Models 4, 5 and 6 all show that the relationship becomes significant when we control for age, education, and religious fundamentalism. As noted in bivariate analysis, women in the sample tended to be older, more religiously fundamental, and have less educated, which all contribute to higher levels of homonegativity. When controlling for any one of these factors, the previously masked relationship becomes prevalent. To add to the complexity of this relationship, the gender role scale was reintroduced into the model. In bivariate analysis we found that women in the sample tended to have more egalitarian attitudes about gender roles. When the gender role scale was reintroduced in the presence of gender, education, age, or any combination of the three, we found that gender was again no longer significant. The lack of relationship remained intact until religious fundamentalism was controlled for 8 In essence, the gender role scale better explains gender differences in homonegativity until the level of religious 8 For simplicity, not all model combinations were shown

PAGE 40

30 fundamentalism is taken into consideration. Respondents age and education can also be used to unmask the relationship between gender and homonegativity however not while controlling for gender role ideologies. In either situation, the data suggest the implications religious doctrines can have on both gender role ideologies, on homonegativity and gender. As briefly mentioned, the data also present an intervening relationship between race and homonegativity. In bivariate analysis, we found extremely strong evidence of a relationship between race and homonegativity, with African Americans displaying more homonegative attitudes than Caucasians. In Model 2 (the full model), we see that race no longer acts as a significant predictor of homonegativity while controlling for other effects. As with unveiling the relationship between gender and homonegativity, several models were assessed to unveil the intervening relationship between race and homonegativity. In all models assessed (using combinations of other independent variables) we find that race remains intact as a significant predictor of homonegativity until religious fundamentalism is controlled for 9 (Models 9 and 10). Bivariate analysis has indicated that African Americans are more fundamental in their religious ideology than Caucasians. When this factor is taken into consideration, race alone no longer acts as a significant predictor of homonegativity. Previous research has implicated African Americans as having higher levels of homonegativity. Although this may be the case, these findings demonstrate that this relationship may be the result of religious doctrine, and not necessarily race. Admittedly, there is a historical relationship in the religious differences among Caucasians and African Americans; however in the analysis of 9 Not all model combinations shown

PAGE 41

31 homonegativity, inadvertent focus on racial differences could probably be better discussed as a product of religious affiliation and dogma, even while controlling for the effects of gender role ideologies. Nonetheless, a discussion of interaction/nonadditivity may help to better understand this relationship. Nonadditivity Finally, multivariate analysis took into consideration nonadditive relationships among the independent variables. Analysis of nonadditive relationships helps to better understand interactive effects among variables (Table 4-7). Rather, non-consideration of interactive models assumes identical change for one variable over the variability of another. It would predict, for example, that males and females experience the same predicted change in homonegativity as level of education increased. In theory, either males or females may be more sensitive to increases or decreases in education and this can be reflected in predicted homonegativity. Nonadditivity in this study will give consideration that males may differ from females just as blacks may differ from whites across the variation of other independent variables. Models 11 through 14 take into consideration the interaction between both race and gender on other independent variables. Models 15 and 16 take into consideration the interaction between race and gender on the gender role scale.

PAGE 42

32 Table 4-7 Model summaries with nonadditivity** Model 11Model 12Mode I l 13Model 14Model 15Model 16ntercept2.1852.1622.0832.2312.1812.107RoleScale.300 (.279)*.300 (.280)*.299 (.279)*.301 (.280)*.268 (.249)*.314 (.292)*Afam-.112 (-.500)*-.052 (-.233)*.083 (.368)*.016 (.072).016 (.072).055 (.244)*Female-.026 (-.074)*-.024 (-.069)*-.025 (-.072)*-.111 (-.316)*-.054 (-.154)*-.026 (-.074)*Age.102 (.009)*.101 (.008)*.100 (.008)*.102 (.008)*.100 (.008)*.101 (.008)*Education-.231 (-.108)*-.217 (-.101)*-.219 (-.102)*-.237 (-.111)*-.218 (-.102)*-.218 (-.102)*Fundamental-.170 (-.317)*-.183 (-.340)*-.170 (-.317)*-.172 (-.320)*-.172 (-.320)*-.171 (-.318)*Pol. View.114 (.116)*.112 (.114)*.214 (.127)*.113 (.116)*.112 (.115)*.112 (.114)*RaceXeduc.132 (.047)*RaceXfund.074 (.200)*RaceXpol. View-.072 (-.076)*SexXeduc.090 (.019)*SexXroleScale.051 (.054)*RaceXroleScale-.056 (-.113)*1977.067 (.242)*.066 (.240)*.067 (.240)*.068 (.243)*.067 (.241)*.067 (.242)*1985.114 (.408)*.114 (.405)*.113 (.404)*.114 (.409)*.114 (.407)*.115 (.410)*1988.050 (.308)*.050 (.307)*.051 (.310)*.051 (.311)*.051 (.311)*.051 (.312)*1989.038 (.226)*.037 (.224)*.037 (.224)*.038 (.227)*.038 (.228)*.038 (.227)*1990.046 (.289)*.046 (.288)*.046 (.287)*.046 (.291)*.046 (.287)*.047 (.292)*1991.035 (.208)*.034 (.205)*.034 (.206)*.035 (.208)*.034 (.204)*.035 (.209)*1993.024 (.138).023 (.132).024 (.137).024 (.139).024 (.135).0252 (.145)*1994.023 (.103).023 (.104).023 (.104).024 (.104).024 (.106).0245 (.109)1996-.0006 (-.003)-.001 (-.006)-.0003 (-.001)-.0002 (-.001)-.0008 (-.003)-.0006 (-.003)R-Squared.355.354.354.354.354.355F-Value198197197197197197Dependent variable: HomoScaleN=6133* p<.05**Values correspond to standardized parameter estimates. Those in parentheses correspond to the parameter estimate.Source General Social Survey 1977-1998

PAGE 43

33 Race interactions In analysis of the interaction of race and education 10 (Model 11), we find that the effect of having higher levels of education does not produce the same rate of predicted decrease in homonegativity between whites and blacks. Additionally, we find that African Americans with no education are expected to have lower levels of homonegativity than whites. However, as the level of education increases, African Americans will not experience the same rate of predicted decrease in homonegativity. At approximately 10 years of education, controlling for other effects, we observe that the two lines cross. In other words, beyond 10 years of education we find that whites benefit from higher levels of education predicts lower levels of homonegativity than African Americans at the same level of education; whereas, before ten years of education, African Americans are less homonegative than whites at the same level (Figure 4-1) 11 An interesting aspect of this interaction is the almost reverse effect education has on African Americans (i.e., African Americans with lower levels of education are expected to have lower homonegativity than whites with the same education, compared to reverse that occurs when the two groups have higher levels). Finally, important in this relationship, is to take into consideration that increases in education do indeed predict lower levels of homonegativity for both groups, however the impact is greater for Caucasians. This implication suggests whites homonegativity is more sensitive to changes in education whereas African Americans homonegativity is more immune. 10 Interactions were also assessed for Education and Religious Fundamentalism, and Gender and Religious Fundamentalism, Gender and Race, however these interactions were not significant and will not be discussed. 11 All figures shown are to scale however the scales shown differ in effort to visually demonstrate the interactive relationships discussed interactive relationships discussed

PAGE 44

34 cale mo 0.511.522.53048121620Education in YearsHoS Whites Blacks Figure 4-1 Interactive effects of education and race on homonegativity In Model 12, we take into consideration the interaction between race and religious fundamentalism. This interaction is highly similar to the race and education interaction. African Americans expressing fundamental religious ideologies are expected to be less homonegative than Caucasians at the same level of fundamentalism. However, as with the interaction between education and race, the predicted lines cross as religious fundamentalism moves toward more liberal religious ideologies (Figure 4-2). African Americans expressing liberal religious ideologies are expected to be more homonegative than Caucasians in the same category, controlling for other effects. In this situation, we see that having religiously liberal ideologies benefits Caucasians more than African Americans in their attitudes toward gays and lesbians. As with education, having religiously liberal ideologies does decrease predicted values in homonegativity, but not at the same rate across races. Finally, the evidence surrounding Model 10 (religious fundamentalism better explaining race in predicted homonegativity) is quite interesting. It was previously argued that racial differences in negative attitudes toward lesbians and

PAGE 45

35 gays could be better attributed to differences in religious fundamentalism. Nonetheless, Model 12 proposes that religious fundamentalism acts independent of races. e moS 11.11.21.31.41.51.61.71.81.9123Fundamental Moderate LiberalHocal Whites Blacks Figure 4-2 Interactive effects of religious fundamentalism and race on homonegativity Model 13 assesses the interaction between race and political ideologies. Contrary to the interactions between race-education and race-fundamentalism, the interaction between race and political ideologies predicts liberal African Americans to be more homonegative than Caucasians with the same political ideology (controlling for other factors) (Figure 4-3). However, having a more conservative political ideology has a greater impact on homonegativity for Caucasians than African Americans, in that Caucasians responding as conservative or extremely conservative are expected to display more homonegativity than African Americans with the same ideology (controlling for other factors). In sum of interactions with race, we find an overall pattern of effect that is greater for Caucasians than for African Americans. The slope of the parameter estimate for the African American group has, in all circumstances, been closer to zero. In other words, we find that African American homonegativity is more consistent across the variations of

PAGE 46

36 political ideologies, religious fundamentalism, and education, whereas increases or decreases in these variables presents greater impacts for Caucasians levels of homonegativity. 1. 0.50.70.91.11.31.571234567Political IdeologyHomoScale Whites Blacks Figure 4-3 Interactive effects of race and political ideology on homonegativity. Gender interactions Model 14 takes into consideration the interactive effects of gender and education (Figure 4-4). In this circumstance, we find that women overall display less homonegativity. Additionally, we find that the benefit to higher levels of education on decreasing predicted homonegativity has a more significant impact for men than women. Womens lack of decrease in homonegativity for higher levels of education could probably be a result of lower overall homonegativity to begin with. Interactions with sex, race, and the gender role scale Finally, analysis of interaction with the gender role scale was assessed. Model 15 analyzes interaction between gender and the gender role scale while Model 16 considers racial interactions with the gender role scale. Interaction finds that women expressing more egalitarian attitudes toward gender roles will subsequently display less

PAGE 47

37 homonegative affect than males in the same category. However women who display more traditional gender role ideologies are actually predicted to display higher levels of homonegativity than males in the same category (controlling for other variables) (Figure 4-5). e Scal omo H 0.511.522.53048121620Education In Years Males Females Figure 4-4 Interactive effects of education and gender on homonegativity The final interactive effect to be discussed accounted for the interaction of race and the gender role scale, and its impact on Homonegativity (Model 16). In this analysis we find that African Americans holding egalitarian attitudes about gender roles are expected to be more homonegative than Caucasians with similar gender role attitudes. However the impact of having more traditional gender roles has a greater substantial impact for Caucasians, to where predicted values in homonegativity actually surpass those of African Americans ( Figure 4-6). Findings indicate that increases in the gender role scale do correspond to increases in homonegativity for both Caucasians and African Americans, however the impacts of having more traditional gender role ideologies is much greater for Caucasians than for African Americans (at least in explaining negative attitudes toward lesbians and gays).

PAGE 48

38 0.50.70.91.11.31.51.71.92.12.32.5012345RoleScaleHomoScale Males Females Figure 4-5 Interactive effects of gender and the RoleScale on homonegativity. 0.511.522.5012345RoleScale Whites Blacks e Scal mo Ho Figure 4-6 Interactive effects of race and the RoleScale on homonegativity. All nonadditive models discussed underwent subsequent F-tests 12 to determine if the interactive models provided a significantly better model fit than Model 2 (the full model). In all instances, interactive models did indeed provide significant F-values (p < 0.05). In other words, allowing for interaction actually provided more explanatory ability than models ignoring this feature. 12 Not Shown

PAGE 49

CHAPTER 5 DISCUSSION AND CONCLUSION The purpose of this study was to discuss why gender and racial differences in attitudes toward homosexuals might differ. Additionally, this study aimed to discuss these differences as a result of the impact of gender role ideologies. Findings indicate that gender differences in homonegativity can be hidden by normative differences in age, education, gender role ideologies or religious fundamentalism. Furthermore, male homonegativity is easier influenced by educational changes than females. In the case of race, multivariate analysis finds that prominent racial differences between blacks and whites homonegativity can be better explained by religious fundamentalism. In considering interaction, the data indicate whites attitudes toward homosexuals have a tendency to be more sensitive to changes in education, religious fundamentalism and gender role ideologies. Finally, in the consideration of the impacts of gender role ideologies, we find confirmed expected relationships that traditional gender role ideologies correlate with increases in homonegativity; even while controlling for the effects of race, gender, education, religious fundamentalism, political ideology, and year data were collected from. Substantial to this correlation was its strength these ideologies have, over other variables, in uniquely explaining homonegative attitudes, in addition to the sensitivity in homonegativity experienced by females as a result of changes in gender role ideologies. Previous research on homonegativity has identified correlates among individuals that stimulate homonegativity, in addition to providing some reasoning why these 39

PAGE 50

40 relationships may exist. This study has built a bridge in that it has done both. The data have demonstrated linkages between gender role ideologies, race, gender, religious fundamentalism, age and education with subsequent attitudes toward lesbians and gays. Just like race and gender, it is assumed that gender role ideologies precede homonegativity and not the reverse. In this circumstance we can now better understand the effects that gender role schemas play in our every day lives. Barbara Risman in her book Gender Vertigo (1998) has proposed that gender should be as insignificant as the length of ones toes or the color of ones eyes. Nonetheless, gender is at the forefront of everything we do. The elimination of gender inequality can only be accomplished through the elimination of gender. Although there may be many factors influencing homonegativity, the data have demonstrated that gender role schemas are still the best predictor, even when controlling for these other influential factors. In essence, this study helps expose the link between gender inequality and homonegativity. Although an elimination of gender inequality could have a significant impact on reducing homonegativity, the data also demonstrate that there are many other factors beyond gender role ideologies that influence homonegativity. Future research could analyze potential linkages between gender role ideologies and any number of the other independent variables considered in this study. An altering of the structures of gender and gender role ideologies will subsequently have effects on any number of other variables (i.e., challenging gender role ideologies will not only effect homonegativity but also institutions such as the government, religion, families). Previous research have almost universally identified that men tend to express more homonegative attitudes than women. This study has argued that this effect could largely

PAGE 51

41 be attributed to the impacts of gender role ideologies. Importantly noted was the lack of an apparent gender difference in bivariate analysis of homonegativity (i.e., men and women did not significantly differ in homonegative attitudes). Multivariate analysis helped to unveil this relationship and it was suggested that the initial relationship might have been masked by gender differences in the sample (i.e., men were younger on average and tended to have more education on average). Many of the previous discussions on gender differences in homonegativity were able to identify this difference through bivariate analysis and little attention was given toward unmaking this relationship. Additionally, many previous researchers had a tendency to use more convenient samples such as college students, medical students, etc. The General Social Survey uses a probability-based sample that is far more accurate at portraying the American population than a college sample. Although this research has suggested bivariate differences were not discovered because of other factors, further research is needed to analyze why or why not differences may exist. This study also suggests the need for understanding African American homonegativity. Research has continually demonstrated that racial discrepancies exist, but has rarely offered as explanation for why this may be the case. Bivariate analysis supported previous research in conclusions about racial differences. Nonetheless, this relationship was better explained through religious fundamentalism. This acted as a preliminary indication and explanation as to why these differences may exist, however the incorporation of nonadditive effects in the multivariate models further complicated any assumptions. In all circumstances of interaction discussed, we find that African Americans are less sensitive to the changes in other variables; and although political

PAGE 52

42 ideology, religious fundamentalism, and education do provide explanatory power, changes across these variables for African Americans are minute compared to those of Caucasians. Understandably, there may be influences not addressed by this study that could better explain African American homonegativity. The lack of variability/sensitivity could be attributed to the smaller sample size of this group in the sample or poor choice of variables that may confound homonegativity for African Americans. Additionally the African American sample is under representative of black men. Again, this provides further reason to devote considerable attention to African American homonegativity. In analyzing racial differences it is equally important to assess class differences. It could be possible that class differences are what cause this plateau effect among African Americans. In analyzing gender interactions with education, a unique relationship was discovered. As predicted, women display less homonegativity than men in similar educational categories. However the benefit of having a higher level of education to reduce homonegativity is more substantial in males. This could be due to several reasons. First, it is possible that because female homonegativity is already low, there is little that an increase in education can due to reduce this further. Second, is the idea that education may not be as good a predictor for reducing homonegativity among females. This would be similar to the discussion of interaction and African American homonegativity. When considering interactions between gender role ideologies (the RoleScale) and gender, a unique unexpected relationship was discovered. Egalitarian women were expected to be less homonegative than men in the same category; however female adherence to gender role ideologies greatly increases predicted homonegativity (to where

PAGE 53

43 is surpasses the predicted homonegativity of males in the same category). Since traditional gender role ideologies typically have more substantial negative impacts on women, I argue that women who support these values have embedded themselves with a much stricter gender role reality. In this study, women overall were expected to display more egalitarian attitudes toward gender roles (especially those specifically concerning the roles and behaviors of women). However, in the event that women display traditional roles (contrary to normative expectations), we find individuals who embrace structural norms more stringent than society imposes. These individuals, I propose, are classic illustrations of the impacts that stringent gender role ideologies can have on attitudinal structures (specifically toward gays and lesbians, or any individuals/groups who violate supposed gender norms). These women have adopted these stringent traditional gender role ideologies and allowed them to influence how they see the world. I call this phenomenon a belief effect experienced by these women. The effect of these females believing in strict traditional gender role ideologies, contrary what may be expected of females, creates a strict gender role schema observed indirectly through homonegativity. Future research on the belief effect observed with this data may help to confirm the deep-rooted linkages between gender role ideologies and attitudes toward sexual minorities.

PAGE 54

CHAPTER 6 LIMITATIONS AND RECOMMENDATIONS There are several features of this study that could benefit from refinement in the future. The first is unidimensionality of the gender role scale. Gender role ideologies were operationalized by considering attitudes about the equality of women. Ideals of gender and sex roles span much further than attitudes about women. Ideally, a scale of gender role attitudes would encompass many dimensions that gender role ideologies are demonstrated through (e.g., childrearing responsibilities, housework, attitudes about men, attitudes about how to raise children, attitudes toward gender neutrality). Additionally, the questions forming the homonegativity scale do not specify the gender of the homosexual being described in addition to being rather one-dimensional. As discussed, attitudes toward lesbians versus gays are not universal. It secondly omits bisexual individuals along with transgender and inter-sexed individuals. There is substantial research indicating that attitudes and behaviors may not necessarily match up. Additionally, there is the stipulation that just because individuals may not agree with homosexuality does not mean they would deny homosexuals equal rights. And although both scales used do represent an array of questions, variability in the scales boarder what a truly quantitative variable should be. Hence there may be some question as to whether Ordinary Least Squares methodology was best used for the Mulivariate analysis. Nonetheless, univariate analysis did provide relatively strong support that the variables and scales used do approximate normal distributions. 44

PAGE 55

45 Additionally, this study gave little focus to the effect of changes over time. A substantial body of research in the area of studying homonegativity has been devoted to observed changes over time. In an effort to increase sample size and statistical power, twenty years of data were conglomerated. Although the year the data were collected from was controlled for, the data taken from 1985 does implicate a possible period effect that was only briefly addressed by this study. Either the use of a single large cross-sectional dataset, or further analysis of potential age, period, or cohort effects should be considered in future research. The General Social Survey, although a very strong research tool, in itself presents issues of validity and accuracy. This study is primarily assessing attitudes and omits behaviors. It is generally accepted that attitudes compliment behaviors (and vice versa) however it is firmly established that this may not be true all the time (i.e., front-stage versus back-stage behavior). Additionally, it is taken for granted that survey respondents are completely honest. The General Social Survey is given face to face; hence it is possible that respondents may respond dishonestly on a variety of questions in effort to avoid disapproval from the interviewer. The sexual orientation of the respondent could have significant influence over their responses. Gay respondents would be far more likely to respond in favor of attitudes toward homosexuals independent of many other control variables. There are many ways to assess homosexual behavior including self-identification, feelings and behaviors. There is a behavioral measure used in the General Social Survey that assess whether respondents have had sexual relations with someone of the same or opposite sex. Future research could consider this variable and assess its potential influence.

PAGE 56

46 As discussed, considerable attention has been given to analyzing gender differences in homonegativity. This study has suggested that these differences would largely be as a result of differing attitudes about gender roles. Nonetheless, this study finds that gender differences still prevail even while controlling for gender role ideologies. In essence, what variables could better explain the gender differences between male and female homonegative attitudes? Finally, considerable attention was devoted to racial differences among African Americans and Caucasians, however no attention was allotted to any other racial or ethnic categories. This is a huge limitation in that a significant portion of the United States is comprised of a variety of ethnic minorities. Future research must consider the plethora of ethnic diversity that should exist in a representative sample of the United States population.

PAGE 57

LIST OF REFERENCES American Psychiatric Association (2000) Diagnostic and Statistical Manual of Mental Disorders : DSM-IV-TR. Washington DC.: American Psychiatric Association. Batson, C. D., Flink, C. H., Schoenrade, P. A., Fultz, J., & Pych, V. (1986). Religious orientation and overt versus covert racial prejudice. Journal of Personality and Social Psychology, 40, 290-302. Batson, C. D., Schoenrade, P. A., & Pych, V. (1985). Brotherly love or self-concern? Behavioural consequences of religion. In L. B. Bron (Ed.), Advances in the psychology of religion. Oxford: Pergamon Press. Baumrind, D. (1995). Commentary on sexual orientation: Research and social Implications. Developmental Psychology, 31, 130-136. Bem, S. L. (1993). The lenses of gender: Transforming the debate on sexual inequality. New Haven, CT: Yale University Press. Beran, N. J., Claybaker, C., Dillon, C., & Haverkamp, R. J. (1992). Attitudes toward minorities: A comparison of homosexuals and the general population. Journal of Homosexuality, 23, 65-83. Bierly, M. M. (1985). Prejudice toward contemporary outgroups as a generalized attitude. Journal of Applied Social Psychology, 15, 189-199. Black, B., Oles, T. P., and Moore, L., (1998). The relationship between attitudes; homophobia and sexism among social work students. Affila Journal of Women and Social work; 13, 2, 166-190. Blee, K., & Tickamyer, A. (1995). Racial differences in mens attitudes about womens gender roles. Journal of Marriage and the Family, 57, 21-30. Bowman, R. (1979). Public attitudes toward homosexuality in New Zealand. International Review of Modern Sociology, 9, 229-238. Chng, C. & Moore, A. (1991). College students beliefs and behaviors about AIDS: Implication for family life educators. Family Relations, 40, 258-263. 47

PAGE 58

48 DAugelli, A. R., & Rose, M. L. (1990). Homophobia in a university community: Attitudes and experiences of heterosexual freshman. Journal of College Student Development, 31, 484-491. Davis, J.A. and T. W. Smith. 1998. General Social Surveys, 1972-2000. Cumulative online codebook and data file ( www.icpsr.umich.edu/gss ). Chicago, accessed November 16, 2002. Deaux, K., & Kite, M. E. (1987). Thinking about gender. In B. B. Hess &M. M Ferree (Eds.), Analyzing Gender: A handbook of social science research (p. 920117). Newbury Park, CA: Sage. Ernst, F., Francis, R., Nevels, H. & Lemen, C. (1991). Condemnation of homosexuality in the black community: A gender-specific phenomenon? Archives of Behavior, 20, 579-585. Estrada, A. X. & Weiss, D. J. (1999). Attitudes of military personnel towards homosexuals. Journal of Homosexuality. 37, 4, 83-97. Faderman, L. (1991). Odd girls and twilight lovers: A history of lesbian life in Twentieth-Century America. New York, Penguin. Ferree, M. M. (1990). Beyond Separate Spheres: Feminism and Family Research. Journal of Marriage and the Family, 52, 866-884. Fishbein, H.D. (1996). Peer prejudice and discrimination. Boulder, CO. Westview Press. Freudenberg, N. (1989). Preventing AIDS: A guide to effective education for the prevention of HIV infection. American Public Health Association, Washington DC. Glenn, N., & Weaver, C. N. (1979). Attitudes towards premarital, extramarital, and homosexual relations in the U.S. in the 1970s. Journal of Sex Research, 15, 108-118. Gramick, J. (1983). Homophobia. A new challenge. Social Work, 28, 137-141. Greaves, W. L. (1987). The black community. In H. Dalton & S. Burris (Eds.). AIDS and the law (pp. 281-289). New Haven: Yale University Press. Hansen, G. L. (1982). Androgyny, sex-role orientation, and homosexism. Journal of Psychology, 112, 39-45. Harper, P. B. (1991). Eloquence and epitaph: Black Nationalism and the homophobic impulse in response to the death of Max Robinson. Social Text, 28, 68-86.

PAGE 59

49 Heaven, P. & Oxman, L. N. (1999). Human values, conservatism and stereotypes of homosexuals. Personality and Individual Differences. 27,1, 109-118. Herek, G. M. (1984). Beyond homophobia: A social psychological perceptive on attitudes toward lesbians and gay men. Journal of Homosexuality, 10, 1-21. Herek, G. M. (1986). On heterosexual masculinity: some psychical consequences of the social construction of gender and sexuality. American Behavioral Scientist, 29, 563-577. Herek, G. M. (1988). Heterosexuals attitudes toward lesbians and gay men: Correlates and gender differences. Journal of Sex Research, 25, 451-477. Herek, G. M., & Glunt, E. K. (1993). Interpersonal contact and heterosexuals attitudes toward gay men: Results from a national survey. The Journal of Sex Research, 30, 239-244. Herek, G. M. (1994). Assessing heterosexuals attitudes towards lesbian and gay men: a review of empirical research with the ATLG scale. In b. Green and G. M. Herek (Eds.), Contemporary Perspectives on lesbian and Gay Issues in Psychology (p. 206-228) Newbury Park, CA: Sage. Herek, G. M. (2000). The Psychology of sexual prejudice. Current Directories in Psychological Science, 9, 19-22. Herek, G. M. (2002) Gender gaps in public opinion about lesbians and gay men. Public Opinion Quarterly. 66, 40-66. Hudson, W. W. & Ricketts, W. A. (1980). A strategy for measurement of homophobia. Journal of Homosexuality, 5, 357-372 Irwin, P., & Thompson, N. L. (1977). Acceptance of the rights of homosexuals: A social profile. Journal of Homosexuality, 3, 107-121. Johnson, M. E., Brems, C., and Alford-Keating, P. (1997) Personality correlates of homophobia. Journal of Homosexuality. 34, 57-69. Kaiser Family Foundation (2001). Inside-Out: a report on the experiences of lesbians, gays, and bisexuals in America an the publics views on issues and policies related to sexual orientations. #3193 www.kff.org Kaiser Family Foundation, accessed April 8, 2003. Kite, M. E. (1984). Sex differences in attitudes toward homosexuals: A meta-analytic review. Journal of Homosexuality, 10, 69-81.

PAGE 60

50 Kite, M. E. and Whitley, B. E. (1996) Sex differences in attitudes toward homosexual persons, behaviors and civil rights: a meta-analysis. Personality and Social Psychology Bulletin, 22, 336-353. Klamen, D. L., Grossman, L. S., Kopacz, D. R. (1999) Medical student homophobia Journal of Homosexuality, 37, 53-63. Lance, L. M. (1992). Changes in homophobic views as related to interaction with gay persons: a study in the reduction of tensions. International Journal of Group Tension, 22, 291-299. Laner, M. R., & Laner, R.H., (1979). Personal style or sexual preference: Why gay men are disliked. International Review of Modern Sociology, 9, 215-228. Laner, M. R., & Laner, R.H., (1980). Sexual preference or personal style? Why lesbians are disliked. Journal of Homosexuality, 5, 339-356. Logan, C. R. (1996) Homphobia? No, Homoprejudice. Journal of Homosexuality. 31, 31-53. Morin, S., & Garfinkle, E. (1978). Male Homophobia. Journal of Social Issues, 34, 29-47. Nyberg, K. L., & Alston, J. P. (1976). Analysis of public attitudes toward homosexual behavior. Journal of Homosexuality, 2, 99-107. Oliver, M. B., & Hyde, J. S. (1993). Gender differences in sexuality: A meta-analysis. Psychological Bulletin, 114, 29-51. Pratte, T. (1993). A comparative study of attitudes toward homosexuality: 1986 and 1991. Journal of Homosexuality, 26, 77-83. Price, J. H. (1982). High school students attitudes toward homosexuality. Journal of School Health, 52, 469-474. Price, V. & Hsu, M. (1992). Public opinions about AIDS policies: The role of misinformation and attitudes towards homosexuals. Public Opinion Quarterly, 56, 29-52. Proulx, R. (1997) Homophobia in Northeastern Brazilian University Students. Journal of Homosexuality. 34, 47-57. Risman, B. (1998) Gender Vertigo: American Families in Transistion. New Haven, CT. Yale University Press.

PAGE 61

51 Seltzer, R. (1992) The social location of those holding antihomosexual attitudes. Sex Roles, 26, 391-398. Staples, R. (1982). Black masculinity: The Black males role in American society. San Francisco: Black Scholar Press. Thompson, E.H., Grisanti, C., & Pleck, J.H. (1985). Attitudes toward the male role and their correlates. Sex Roles, 13, 413-427. Waldner, L. K., Sikka, A., Baig, S. (1999). Ethnicity and sex differences in university students knowledge of AIDS, fear of AIDS, and homophobia. Journal of Homosexuality, 37, 117-133. Walters, A. S., & Hayes, D. M. (1998). Homophobia within schools: Challenging the culturally sanctioned dismissal of gay students and colleagues. Journal of Homosexuality, 35, 1-23. Weinberg, G. (1973). Society and healthy homosexual. New York: Anchor books. Whitcomb, D. H. (2001). Development, implementation, and applications of an instrument to measure relationships among sexual orientation, political orientation, and socioeconomic status in an undergraduate population. Dissertation Abstracts international. 61, 9-B, 5062. Whitley, B. E., Jr. (1987). The relationship of sex-role orientation to heterosexuals attitudes toward homosexuals. Sex Roles, 17, 103-113. Whitley, B. E., Jr., & Kite, M. E. (1995). Sex differences in attitudes toward homosexuality: A comment on Oliver and Hyde (1993). Psychological Bulletin, 117, 146-154. Wills, G., & Crawford, R. (2000). Attitudes toward homosexuality in Shreveport-Bossier City, Louisiana. Journal of Homosexuality, 38, 97-115. Young, M., & Whertvine, J. (1982). Attitudes of heterosexual students toward homosexual behavior. Psychological Reports, 51, 673-674.

PAGE 62

BIOGRAPHICAL SKETCH Christian Grov, born in 1980, was the second child of Sharon and Kjell Grov. Both Sharon and Kjell were foreign immigrants to the United States in search of a better life for themselves and their children. Raised in south Florida, Christian graduated salutatorian of Stranahan High School, Ft. Lauderdale in 1998 and immediately moved into the honors program at Broward Community College. After only two semesters, Christian finished his associates degree and moved to Gainesville, Florida in pursuit of a bachelors degree in psychology. Upon enrolling in one course in sociology, Christian decided to double major in both psychology and sociology. In May of 2001, Christian graduated with honors in both areas. That August, he began the masters program in sociology at the University of Florida, where he focused on the study of gender, families, and human sexuality. Two years later, Christian graduated with a Master of Arts degree. In the fall of 2003, Christian will be beginning the doctoral program in sociology at the City University of New York, Graduate Center (CUNY) in New York City. 52


Permanent Link: http://ufdc.ufl.edu/UFE0000692/00001

Material Information

Title: Understanding homonegative attitudes through sex, race, and gender role ideologies : an analysis of the 1972-1998 general social survey
Physical Description: x, 52 p.
Language: English
Creator: Grov, Christian ( Dissertant )
Shehan, Constance L. ( Thesis advisor )
Peek, Charles ( Reviewer )
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2003
Copyright Date: 2003

Subjects

Subjects / Keywords: Sociology thesis,M.A   ( local )
Dissertations, Academic -- UF -- Sociology   ( local )

Notes

Abstract: This study attempts to analyze homonegativity as an expression of contempt for violations in perceived gender role ideologies. Society assigns and constricts individuals into strict behavioral patterns based on ideas about gender roles. Individuals who violate these roles are often perceived as deviant and subsequently treated as such. Additionally, previous research has demonstrated both racial and gender differences in homonegative attitudes with African Americans and males displaying more homonegativity. This study gives considerable attention to the interaction between race and gender. Data used are from the 1977-1998 General Social Survey. The General Social Survey is a face-to-face interview administered to a representative sample of the English- speaking American public over the age of 18. Gender role ideologies were measured using a scale of questions about attitudes toward women (e.g., women in politics, women in the workforce, women and the family). Homonegativity was measured using a scale of questions about attitudes toward homosexuals (e.g., allowing a homosexual to teach in a college, allowing a book written by a homosexual to remain in a library). The data show a strong relationship between supporting traditional gender role ideologies and being homonegative. Additionally, Caucasian homonegativity is more sensitive to changes in educational attainment, religious fundamentalism, political ideology, and gender role attitudes. Similarly, male homonegative attitudes are more sensitive to changes in education and gender role ideologies. x
Subject: attitudes, gender, general, homonegativity, homophobia, ideologies, interaction, norc, race, sex, sexroles, social, survey
General Note: Title from title page of source document.
General Note: Includes vita.
Thesis: Thesis (M.A.)--University of Florida, 2003.
Bibliography: Includes bibliographical references.
General Note: Text (Electronic thesis) in PDF format.

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0000692:00001

Permanent Link: http://ufdc.ufl.edu/UFE0000692/00001

Material Information

Title: Understanding homonegative attitudes through sex, race, and gender role ideologies : an analysis of the 1972-1998 general social survey
Physical Description: x, 52 p.
Language: English
Creator: Grov, Christian ( Dissertant )
Shehan, Constance L. ( Thesis advisor )
Peek, Charles ( Reviewer )
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2003
Copyright Date: 2003

Subjects

Subjects / Keywords: Sociology thesis,M.A   ( local )
Dissertations, Academic -- UF -- Sociology   ( local )

Notes

Abstract: This study attempts to analyze homonegativity as an expression of contempt for violations in perceived gender role ideologies. Society assigns and constricts individuals into strict behavioral patterns based on ideas about gender roles. Individuals who violate these roles are often perceived as deviant and subsequently treated as such. Additionally, previous research has demonstrated both racial and gender differences in homonegative attitudes with African Americans and males displaying more homonegativity. This study gives considerable attention to the interaction between race and gender. Data used are from the 1977-1998 General Social Survey. The General Social Survey is a face-to-face interview administered to a representative sample of the English- speaking American public over the age of 18. Gender role ideologies were measured using a scale of questions about attitudes toward women (e.g., women in politics, women in the workforce, women and the family). Homonegativity was measured using a scale of questions about attitudes toward homosexuals (e.g., allowing a homosexual to teach in a college, allowing a book written by a homosexual to remain in a library). The data show a strong relationship between supporting traditional gender role ideologies and being homonegative. Additionally, Caucasian homonegativity is more sensitive to changes in educational attainment, religious fundamentalism, political ideology, and gender role attitudes. Similarly, male homonegative attitudes are more sensitive to changes in education and gender role ideologies. x
Subject: attitudes, gender, general, homonegativity, homophobia, ideologies, interaction, norc, race, sex, sexroles, social, survey
General Note: Title from title page of source document.
General Note: Includes vita.
Thesis: Thesis (M.A.)--University of Florida, 2003.
Bibliography: Includes bibliographical references.
General Note: Text (Electronic thesis) in PDF format.

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0000692:00001


This item has the following downloads:


Full Text












UNDERSTANDING HOMONEGATIVE ATTITUDES THROUGH SEX, RACE, AND
GENDER ROLE IDEOLOGIES: AN ANALYSIS OF THE 1972-1998 GENERAL
SOCIAL SURVEY
















By

CHRISTIAN GROV


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF ARTS

UNIVERSITY OF FLORIDA


2003

































Copyright 2003

by

Christian Grov

































This thesis is dedicated to Scott-Lee Cash.















ACKNOWLEDGMENTS

This paper would have not been possible without the consistent support of several

faculty members at the University of Florida. First and foremost I would like to thank

Kendal Broad for helping spawn my interest in gender and sexuality research, in addition

to always making herself available for both personal and professional advice. Secondly, I

wish to thank my committee members, Constance Shehan and Charles Peek. Charles has

always been there to shed light where there was empirical darkness, aiding me in my

growth as a graduate student and quantitative researcher. Connie has always been there to

offer both moral and emotional support. She has had the ability to give my work

substance and ground when I was at a loss for direction. Finally, I wish to thank John

Henretta. As the former Graduate Coordinator of the Department of Sociology, John has

offered me both logical and hard-to-come-by advice, guiding me into the transition of

becoming a graduate student. As the departmental chair, John has expressed a superior

sense of responsibility and moral obligation to all graduate students. He has always

pointed me in the right direction, whether in life or academically. In total, I would not be

where I am today without the combined support of Chuck, John, Kendal and Connie.

Distinctively and profoundly, they have influenced my past and shed light upon my

future.
















TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ................................................................................................. iv

LIST OF TA BLE S .................. ............ .................. ....... ............ .. vii

LIST OF FIGURES ............. .. ..... ...... ........ ....... ............. ............. viii

ABSTRACT .............. ......................................... ix

CHAPTER

1 IN TRODU CTION ................................................. ...... .................

2 BACKGROUND AND LITERATURE .............................................. ...............3

Sex and G ender R ole Ideologies............................................ ........................... 4
Racial D differences in H om onegativity ........................................ ...... ............... 8
A additional Influential V ariables ........................................ ........................... 10
E d u c atio n ............................................... ....................... 1 1
R religious Fundam entalism ............................................ ............................... 12
A g e .............................................................................. 12
P political Ideology ....................................... ......................... 13
H y p o th e se s ........................................................ ................ 13

3 RESEARCH DESIGN AND METHODS ................ ..................................15

S a m p le .................................................................................................................. 1 5
M e a su re s ............................................................................................................... 1 6

4 R E S U L T S .............................................................................2 1

U nivariate A naly sis .............................................................. 2 1
B iv ariate A n aly sis....................................................................................................... 2 2
Multivariate Analysis...................................... ........ 25
Intervening and Suppressor Effects ........................................27
N onadditivity ..................................... ........ ................... 31
Race interactions. .............................................................. 33
G ender interactions ....................... ...... ... ................36
Interactions with sex, race, and the gender role scale ................................36


v











5 DISCUSSION AND CONCLUSION ........................................ ...................... 39

6 LIMITATIONS AND RECOMMENDATIONS ....................... ..........................44

L IST O F R E FE R E N C E S ............................................................................. .............. 47

B IO G R A PH IC A L SK E T C H ...................................................................... ..................52
















LIST OF TABLES

Table pge

3-1 Factorial analysis and Chronbach's alpha for HomoScale and RoleScale ................18

4 -1 U niv ariate statistics........................................................................... .....................22

4-2 Sam ple size and m eans by year ................................. ............... ............... 22

4-3 Correlation matrix for all quantitative variables............................... ............... 24

4-4 Pearson correlation between HomoScale and RoleScale. .........................................24

4-5 Racial and gender differences using T-tests. .......................................................25

4-6 M odel sum m aries ........................... .......................... .... .... ........ ........28

4-7 M odel summaries with nonadditivity ...................................................................... 32
















LIST OF FIGURES


Figure pge

4-1 Interactive effects of education and race on homonegativity .............. ...................34

4-2 Interactive effects of religious fundamentalism and race on homonegativity ...........35

4-3 Interactive effects of race and political ideology on homonegativity. ........................36

4-4 Interactive effects of education and gender on homonegativity..............................37

4-5 Interactive effects of gender and the RoleScale on homonegativity. ..........................38

4-6 Interactive effects of race and the RoleScale on homonegativity. .............................38















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master or Arts

UNDERSTANDING HOMONEGATIVE ATTITUDES THROUGH SEX, RACE, AND
GENDER ROLE IDEOLOGIES: AN ANALYSIS OF THE 1972-1998 GENERAL
SOCIAL SURVEY

By

Christian Grov

May 2003

Chair: Constance Shehan
Cochair: Charles W. Peek
Major Department: Sociology

This study attempts to analyze homonegativity as an expression of contempt for

violations in perceived gender role ideologies. Society assigns and constricts individuals

into strict behavioral patterns based on ideas about gender roles. Individuals who violate

these roles are often perceived as deviant and subsequently treated as such. Additionally,

previous research has demonstrated both racial and gender differences in homonegative

attitudes with African Americans and males displaying more homonegativity. This study

gives considerable attention to the interaction between race and gender.

Data used are from the 1977-1998 General Social Survey. The General Social

Survey is a face-to-face interview administered to a representative sample of the English-

speaking American public over the age of 18. Gender role ideologies were measured

using a scale of questions about attitudes toward women (e.g., women in politics, women

in the workforce, women and the family). Homonegativity was measured using a scale of









questions about attitudes toward homosexuals (e.g., allowing a homosexual to teach in a

college, allowing a book written by a homosexual to remain in a library).

The data show a strong relationship between supporting traditional gender role

ideologies and being homonegative. Additionally, Caucasian homonegativity is more

sensitive to changes in educational attainment, religious fundamentalism, political

ideology, and gender role attitudes. Similarly, male homonegative attitudes are more

sensitive to changes in education and gender role ideologies.














CHAPTER 1
INTRODUCTION

There are a large number of characteristics an individual may possess that guide the

individual's belief and attitudinal schemas. Individuals possess both biological drives and

social characteristics that can sometimes be used to better explain social phenomenon.

Features of an individual's religious background, race, educational attainment,

upbringing, immediate environment, socioeconomic status, etcetera all interact with one

another and impact how individuals interact with others. This analysis attempts to better

predict the characteristics in individuals that allow them to hold negative attitudes toward

gays and lesbians. Specifically, this analysis addresses how gender role ideologies, race

and gender impact attitudes toward lesbians and gays. Demographic (control)

characteristics are also taken into consideration.

There is never one reason why an individual displays homonegative affect or

behaviors. However social science has been able to identify several key predictors (e.g.,

religious fundamentalism, age, and educational attainment.) This study explores the facet

of how attitudes about sex roles impact homophobia. Moreover, it explores the probable

connection between a traditional sex role ideology and a liberal or egalitarian sex role

ideology. The idea of sex roles was proposed after World War II and originally applied to

the area of family studies (Ferree 1990). Individuals in the traditional family were seen as

fulfilling certain roles (e.g., women nurture and men provide). Feminist thought of the

1960s and later challenged these ideologies, as they were seen as contributing to and

justifying the subjugation of women (e.g., women should raise children rather than seek









employment or education). Although sex role theory was originally applied to the family,

and later rejected for its apparent flaws, members of society today still hold belief

patterns that men are to fit into a certain set of behavior patterns just as women are

(Ferree 1990). Homosexual individuals violate the sex roles assigned to them by society,

as they are not procreative and oftentimes do not adhere to a standard set of masculine or

feminine behaviors. This is further complicated by the stereotypical images society

upholds when envisioning homosexuals as individuals and a group (e.g., flamboyant-sex-

crazed gay men and "masculine-ized" butch-lesbian women) (Faderman 1991). These

flawed images are perpetuated by a number of factors (e.g., media misrepresentation and

sensationalism, individuals who are closeted about their sexuality, the persistence of

homophobia). The purpose of this analysis is not to unveil all factors leading to the

continuances of homophobia, but rather explicate interactions among gender, race, and

gender role ideologies and how they impact social attitudes.

Conclusions can expose the intricate linkage between the subjugation of women

and the subjugation of other sexual minorities, and can add to the understandings of

societal views of sex-role ideologies. Finally, our understanding of why potential

universal gender and racial differences exist is quite limited. Analyzing confounding

effects will better help unmask why differences are perpetuated. Implications could lead

to protocol that may help individuals to think outside of the gender box to understand the

fluidity of gender rather than to polarize roles.














CHAPTER 2
BACKGROUND AND LITERATURE

The scientific study of gays did not emerge until the later half of the last century,

and a most of the research pathologized the behavior. It was not until 1973 that the

American Psychiatric Association, which had classified it as a mental disorder, removed

homosexuality from the Diagnostic and Statistical Manual. Understandably, the body of

academic research on gays is limited, hence so is the specific study of homonegativity.

In 1973 Weinberg documented a term in academic literature describing negative

attitudes toward lesbians and gays. That term was "homophobia" and directly meant "the

dread of being in close quarters with homosexuals and in the case of homosexuals

themselves, self-loathing" (p. 4). Over the years the term has been redefined as

attitudinal structures toward gays and lesbians have been better studied. In 1983, Gramick

created the term "homosexphobia" to describe a fear of homosexuality. Other terms since

created include "sexual prejudice," "homonegativity" (Herek 2000) "homosexism,"

(Hansen 1982) and "homoprejudice" (Logan 1996). Although "homophobia" is the most

common term (because it was the first), many social scientists have identified an inherent

flaw with it. Using a medical model, homophobia as a term has pathological implications

that may not be inherent. For the purposes of this paper the term homonegativity will be

used to define negative attitudes toward lesbians, gays, bisexuals, and transgenders1.




1 For the purposes of simplicity the terms LGBT and/or gay will be used synonymously to describe
lesbians, gays, bisexuals, and transgenders alike.









Substantial negative impacts of homonegativity on gays have been noted. Gay

youth are seven times more likely to attempt suicide when compared to other teens

(Baumrind 1995). Additionally, gay teens have higher levels of self-reported stress and

substance abuse. Finally, these teens are more likely to experience rejection by friends

and also by family. The United States Gay and Lesbian Task Force found that more than

75% of lesbians and 90% of gay men had been verbally harassed because of their

sexuality (Herek 1988). In 1988, Herek found that 50% of gay males and more than a

third of lesbian females reported being threatened with physical violence.

The Kaiser Family Foundation (2001) conducted an opinion survey among the

general public with a specific oversample of gays. Among the gay sample, Kaiser found a

large majority felt there was more acceptance of homosexuality today compared to a few

years ago. On the contrary, almost three-quarters of this gay sub sample had experienced

a significant amount of both prejudice and discrimination because of their sexual

orientation. Finally, about one-third of the sample indicated that their family (or a family

member) has refused to accept them because of their sexuality.

The Kaiser Foundation found that most of the general public reports knowing

someone who is gay. Secondly the general public also agrees that there is greater

acceptance of gays today compared to a few years ago. Nonetheless, about half of the

sample felt that homosexual behavior is morally wrong. Although public support may be

increasing toward gays, homonegativity in some forms is still widespread and needs

attention as to its causal and contributing factors.

Sex and Gender role Ideologies

Research by and large has found that heterosexual males, in general, display more

homonegativity than females (D'Augelli and Rose 1990, Fishbein 1996, Herek 1986,









1988, 1994, 2002, Hudson and Ricketts 1980, Johnson et al. 1997, Kite 1984, Klamen et

al. 1999, Morin and Garfinklel 1978, Oliver and Hyde 1993, Pratte 1993, Price 1982,

Seltzer 1992, Thompson et al. 1985, Whitley and Kite 1995). Males are generally

socialized to more stringent hetero-normative gender roles, whereas females are given

more freedom in the gender roles they may fulfill. An individual who violates her or his

gender role is more likely to be treated as deviant (Laner and Laner 1979, 1980).

Moreover, Laner and Laner (1979) found that lesbians and gays experience

homonegative affect, in part, because of their perceived sex-role deviance. Specifically, a

male who violates his gender role is likely to be treated more as a deviant than a female

who violates her socially defined role (Herek 1994, Kite and Whitley 1996). Herek

(1984) found homonegative individuals more often were highly authoritarian, had

traditional attitude toward gender roles, were more negative toward other minority

groups, had less education, and were male. Deaux and Kite (1987) found that

heterosexual attitudes toward gays and lesbians are influenced by a "generalized gender

belief system," defined as "a set of beliefs and opinions about males and females and

about the purported qualities of masculinity and femininity" (p. 97). They further discuss

that this gender-belief-system drives stereotypes including attitudes toward appropriate

gendered behavior, toward men and women, and toward those who violate these

stereotypes. Research supports that women's homonegative attitudes are more inclined to

change compared to men's (Lance 1992). Kite and Whitley (1996) argue that rigid

gender roles arguably require [heterosexual] men to eschew homosexuality. Because men

are expected to avoid feminine traits or activities, and because gay persons are often

thought to be deviant from appropriate gender roles, men may feel pressured to display









antigay prejudice" (p. 338). In summary, there is a strong link between traditional gender

role ideologies and homonegativity. More important are the potential differences in the

links between gender role ideologies and homonegavity that may exist based on gender.

Earlier research supports that greater levels of homonegativity are derived from more

stringent gender role ideologies, that men succumb to more stringent ideologies than

women, and that men are more homonegative.

Research has continually identified the existence of gender differences in attitudes

toward gays with emphasis that men tend to be more homonegative. Empirical reasoning

explaining this distinction lacks development. "Heterosexuality is equated ideologically

with 'normal' masculinity and 'normal' femininity, whereas homosexuality is equated

with violating the norms of gender" (Herek 1988, p. 97). Society has polarized

understandings of gender (Bem 1993); forcing individuals to either conform or face

consequences. Polar ideologies are not new to social understanding but rather indicative

of human schematic understandings (e.g., black or white, fat or thin, male or female,

good or evil, heaven or hell).

Heterosexuals tend to expresses more negative attitudes toward homosexual

individuals of the same sex; this pattern is more pronounced among men than women

(Kite and Whitley 1996). Researchers typically do not distinguish between attitudes

toward gays and lesbians. Most often the terms gay or homosexual are used to describe

both. Gender biases could lead respondents to consider attitudes toward gay men before

they consider attitudes toward lesbians. Herek (2002) proposes that individual attitudes

toward gays and lesbians may reflect that individuals attitudes toward her or his own

sexuality.









"Because of homosexuality's stigmatized status, many heterosexuals wish to avoid
being labeled gay or lesbian, and this concern is probably stronger among men in
the U.S. society. Some individuals may feel a particular need to distance
themselves from gay people because they have experienced homosexual desired or
engaged in same-sex behaviors, which they regard as extremely unacceptable and
inconsistent with their self-concept" (p 43).

In essence, a man's sense of masculine identity may encounter more insecurity,

causing him to display more homonegative affect. This is in concurrence with earlier

researched indicating that men have less liberty in the gender/sex role behaviors they may

display. In contrast, women have more liberty in the gender/sex roles they may

undertake. This may give a female individual more security in her attitudes toward

individuals who violate their gender roles. Additionally, this helps explain why

homonegative affect is sometimes less severe for lesbians than gay men.

In accord with the previous discussion, there is one research study that refutes all

previous findings indicating males are more homophobic than females. Proulx (1997)

sampled 553 Brazilian college university students and found a complete inversion of

gender and homophobia (i.e., women were more homonegative than men and

homonegativity was more pronounced between heterosexual women toward lesbians than

between heterosexual men and gay men). Proulx hypothesized that these differences are

probably due to Brazilan culture, in that the women of this area undergo rigid gender

roles. This data clearly supports Herek's (2002) postulations about the correlations

between rigid gender role structures and subsequent attitudes toward gays and lesbians.

The Proulx study is a counter example that supports the hypothesis and offers reasoning

as to why these characteristics may exist.

In sum, research has clearly defined three things. First there has been discussion on

the impact of gender role ideologies on homonegativity finding that traditional gender









role ideologies contribute to greater homonegativity. Second has been the discussion of

gender differences in homonegativity suggesting that, in most cases, females are less

homonegative than males. Finally, there is the discussion that gender role ideologies are

dependent upon gender, finding that males are subjected to more stringent gender role

structures giving them more traditional belief patterns. Unfortunately, there is little to no

empirical research that moves far beyond this to develop other influential factors that

may cause or mask these relationships. This study will attempt to examine the impact of

gender and gender role ideologies on homonegativity, and how interaction between the

two may better explain why differences continue to exist.

The items selected for the scale measuring gender role ideologies assess several

aspects of attitudes about gender roles but they center only on attitudes toward women.

Although simply assessing attitudes toward women may not ascertain all aspects of

gender role ideologies, they do help assess a significant portion of them. Ideally this

assessment would include a variety of gender role attitudes (e.g., attitudes about men's

behaviors), however the General Social Survey does not elaborate into this aspect.

Racial Differences in Homonegativity

Racial or ethnic makeup is a second predictor of homonegavity. The General Social

Survey conglomerates race and ethnicity into three categories: Caucasians, African

Americans, and other. In this study, the "other" group has been omitted due to lack of

sample size and complex variability in race and ethnicity that is sure to exist within such

a category. When comparing African Americans to Whites it is plausible to think that,

because of the civil rights and equality struggles that African Americans have endured

over the course of the last millennium, African Americans could be more compassionate

to other minority groups being discriminated against in similar fashions. In this type of









discussion, gays would be categorized as a sexual minority. In contrast, it is plausible to

believe that because African Americans may value a more traditional family model, are

on average more religious, and hold more stringent constructs of gender role

identifications, they would thereby be more homonegative. Research seems to support the

latter model more.

Chng and Moore (1991) found that African American males were more

homophobic than Caucasian males based on the information reported from nine African

American respondents. It was found that African American females displayed less

tolerance toward homosexuals than Caucasian females (Ernst et al. 1991). Black et al.

(1998) studied the relationship of homophobia and sexism among social work students.

Their findings support research that correlates gender roles and homonegativity. They

additionally found that changes in sexist and corresponding homophobic attitudes might

occur with greater ease among Caucasians than among African Americans.

Unfortunately, Black et al. offer no assessment explaining this interaction other than

citing the 1995 study by Blee and Tickamyer. Blee and Tickamyer noted that gender

roles specifically were more multidimensional for African American men than Caucasian

men. In all, African American homonegativity could be better explained through the

differences between Caucasians and African Americans in their attitudes about gender

roles. In these circumstances, research has both reified the relationship between gender

roles and homonegativty, while considering racial and gender differences.

Klamen et al. (1999) study of medical students found that the African American

respondents were more likely to agree with statements such as "Homosexuality is

immoral" (p 58). The same study also found that, when compared to other ethnic groups,









African Americans were much more likely to oppose statements such as "Homosexuals

should have equal employment opportunity" (p 59). Wills and Crawford's 2000 study

found that "African Americans were more likely to attribute the cause of homosexuality

to a moral choice while other ethnic groups tended to say a combination of biology and

environment" (p 101).

In general, it is found that condemnation against homosexuals is greatest in African

American communities when compared to Caucasians (Freudenberg 1989, Staples 1982,

Waldner et al. 1999). In contrast, African American homosexuals are not rejected by their

families to the extent that Caucasian homosexuals are (Greaves 1987). This could be due

to stronger family ties within the African American community. It has clearly been

identified that homonegativity in African Americans is greater than that in Caucasians.

Little research addresses the reasons why this may be the case. Harper (1991) suggested

that this might be due to African Americans greater sense of nationalism that emerged

from the civil rights movements of the 1960s. This suggests that homosexuality is viewed

or can be viewed as un-American.

This study attempts to discover both the influence race has over homonegativity in

addition to assessing its confounding effects with other variables such as education,

gender role ideologies, religious fundamentalism, and political ideology. As in the case of

gender, much research has discussed the existence of this relationship, however little has

been done to empirically examine why it exists.

Additional Influential Variables

In analyzing potential racial or gender differences in homonegativity or the impact

of gender role ideologies on homonegativity, it is necessary to consider other common

variables identified as having influence over homonegative attitudes. This analysis is









necessary because these other influential variables can vary across both race and gender.

Only when we control for these differences, might we be able to obtain a clearer picture

of the effects of race, gender, and gender role ideologies.

Education

It is a general fact that a college education tends to increase ones sense of

liberalness. Individuals with lower levels of education are less likely to have exposure to

the vast aspects of many social issues. Many activists have hailed high schools as being

some of the most homophobic institutions in existence. In general, neither texts nor

teachers in secondary institutions cover the environmental, social, and biological aspects

of homosexuality. Understandably these aspects are not touched upon in any lower levels

of schooling either (e.g., middle school or elementary school). Colleges are known for

fostering social change; nonetheless "anti-gay sentiments are [still] prevalent (and even

fostered) at the university level" (Walters and Hayes 1998, p. 3).

Whitley's 1987 study of college students found that older students were less

negative toward homosexuals than were the younger ones. First year college students, in

general, displayed negative attitudes toward homosexuality (Young and Whertvine

1982). Although both Whitley's and Young and Whertvine's research from the 1980s

may seem dated, many researchers have consistently found that the less education a

person has, the more likely she or he is to display homonegative attitudes (Beran, et al.

1992, Bowman 1979, Glenn and Weaver 1979, Irwin and Thompson 1977, Nyberg and

Alston 1976, Price and Hsu 1992). Wills and Crawford (2000) found that people with a

masters or doctorate degree were least likely to agree with questions concerning gays

going to hell as a result of sexual orientation. They also found that respondents with a

high school diploma or below were more likely to attribute homosexuality to a moral









choice, rather than environment or biology. Finally, they found respondents with a high

school degree or below were less in favor of homosexuals having equal civil rights.

Religious Fundamentalism

Certain religious segments have been most noted for expressing homonegavity

(e.g., the Christian coalition). In this study, level of religious fundamentalism, in contrast

to liberalism, will be controlled for. Admittedly, there are many factors of religion that

can impact homonegativity such as extent of intrinsic beliefs, denomination, frequency of

church attendance, etc.

On average, Catholics are usually more tolerant of homosexuality than are

Protestants (Bierly 1985, Wills and Crawford 2000). People whose religion has a more

fundamentalist orientation are more likely to be homonegative compared to individuals

with religions that have less fundamentalist orientations (Herek and Glunt 1993). Herek

(1994) found that simply being more religious would cause a person to exhibit a more

anti-homosexual attitude. In contrast, research supports that individuals high in several

aspects of religiosity will internalize their beliefs; thereby they display less

homonegativty than individuals who use religion as an explanation for current social

climates (Batson et al. 1986).

Age

The children of both Generation D and Generation X have been growing up in an

era where gays are being portrayed by the media in a more positive light. Some of the

only negative media attention that a member of Generation X would have heard about

gays may have centered on the HIV/AIDS epidemic. For the most part, it is now socially

accepted that HIV/AIDS is not a gay disease, and hence much of that attribution has been

dispelled. Members of previous generation cohorts, such as the baby boomers and those









before them, grew up in eras where they saw homosexuality was still classified under a

medical model as a mental disorder. Similarly, they lived in times where it was

acceptable for the police to harass lesbians and gays, and in general to display

homonegative affects and behaviors. Individuals growing up and being raised in a society

where gays are more accepted would be more socially conditioned to accept

homosexuality. Similarly, individuals who grew up and were raised to believe that

some/all aspects of homosexuality are in some part wrong would be more secure in those

beliefs. The older an individual is (or the longer they spent in a time period where

homonegativity was more acceptable) the more homonegative attitudes they will continue

to display.

Political Ideology

Research has supported that homonegativity is positively correlated to the level of

political conservativism a person displays (Whitcomb 2001). Heaven and Oxman (1999)

found that males were more likely display conservative attitudes and that these attitudes

correlated strongly with homonegativity. Among members of the military Marine Corps,

researchers found that homonegativity and conservative attitudes correlated strongly

(Estrada 1999). There is a clear positive relationship between politically conservative

attitudes and homonegativity.

Hypotheses

Based on the literature review and corresponding theoretical framework built for

this study, the following relationships are believe to be demonstrated by the data:

* Individuals with more traditional gender role ideologies will have more
homonegative attitudes.
* Males will display more traditional gender role ideologies than females.






14


* Males will display more homonegativity than females.
* African Americans will display more homonegativity than Caucasians.















CHAPTER 3
RESEARCH DESIGN AND METHODS

Sample

The General Social Survey, a time series study based on a probability sample, is a

survey of opinions and beliefs administered face to face every few years to a

representative sample of the United States population. Its content briefly addresses a

plethora of topics including attitudes about the government, abortion, welfare, suicide,

pornography, work, family etc. Sample sizes for each survey year are roughly under

3,000 individuals with an over sample of African Americans. Data span from 1972

through 2000. This study uses data from the 1977, 1985, 1988, 1989, 1990, 1991, 1993,

1994, 1996, and 1998 datasets (N = 6133)2. In order to be qualified for sample

consideration respondents must have been between the ages of 18 and 89 years, English-

speaking, and not institutionalized at the time of the survey (i.e., whose who reside in

nursing homes, incarcerated individuals, or those committed to mental institutions are not

surveyed). Significant limitations of this survey include the exclusion of non-English

speaking households and individuals living on U.S. Military bases. The response rate is

approximately 75%.





2 Data were originally to be taken from the 1998 dataset alone. Unfortunately, not a great enough number of
respondents answered all questions relevant to the two scales. To counteract, all years were included into
the sample selection. Again, not all questions were asked every year. The final set of years utilized reflects
those in which all items on the selected scales were asked to respondents (1977, 1985, 1988, 1989, 1990,
1991, 1993, 1994, 1996, and 1998 See Table 3) Additionally the sample size of 6133 does not reflect all
respondents from those years, but rather those remaining who answered all questions used in this study.









Measures

Two scales will be used for this analysis. The dependent variable is constructed

from an array of questions assessing attitudes toward homosexuals. As mentioned,

gender role ideologies constitute one of the major explanatory variables in this study.

Attitudes about gender roles were assessed using an array of questions specifically about

the roles of women.

The Homonegativity scale (HomoScale) was be derived from the following series
of questions
* HOMOSEX: "What about sexual relations between two adults of the same sex -
do you think it is ...

* SPKHOMO: "And what about a man who admits that he is a homosexual. Suppose
this admitted homosexual wanted to make a speech in your community. Should he
be allowed to speak, or not?"

* COLHOMO: "Should a homosexual be allowed to teach in a college or university,
or not?"

* LIBHOMO: "If someone in your community suggested that a book, written by a
homosexual, in favor of homosexuality should be taken out of your public library,
would you favor removing this book or not?"

The question HOMOSEX4 is coded on a scale from 1 to 4 (Strongly Disagree,

Disagree, Agree, and Strongly Agree), while the other three mnemonics are dichotomized

(e.g., yes or no, allow or not allow). The responses of the HOMOSEX variable were

collapsed into dichotomous categories in order to match the other variables in the scale.

This additionally prevents overweighing the responses for other items in the scale. Items


3 The homonegativity scale originally included the variable ASKSEXOR: "Before giving an individual
secret or top-secret clearance, the government should have the right to ask a person their sexual
orientation?" The inclusion of this variable into the scale presented a significant problem, as there were a
limited number of individuals this question was asked of. The sample size was cut virtually in half with the
inclusion of this variable.

4 The mnemonic for ASKSEXOR also was on a 4-point scale similar to HOMOSEX, however omitted (See
comment 2).









on the homonegativity scale were reordered and added together so that a score of zero

represents complete acceptance of homosexuality while a score of 4 represents the

highest level of homonegativity.

The second scale on attitudes about gender roles (RoleScale) includes the following

questions:

* FEFAM: "Now I'm going to read several more statements. As I read each one,
please tell me whether you strongly agree, agree, disagree, or strongly disagree
with it. For example, here is the statement: D. It is much better for everyone
involved if the man is the achiever outside the home and the woman takes care of
the home and family."

* FECHLD: "Now I'm going to read several more statements. As I read each one,
please tell me whether you strongly agree, agree, disagree, or strongly disagree
with it. For example, here is the statement: A. A working mother can establish just
as warm and secure a relationship with her children as a mother who does not
work. "

* FEHOME: Do you agree or disagree with this statement? Women should take care
of running their homes and leave running the country up to men.

* FEPOL: Tell me if you agree or disagree with this statement: Most men are better
suited emotionally for politics than are most women.

* FEWORK: Do you approve or disapprove of a married woman earning money in
business or industry if she has a husband capable of supporting her?

The questions FEFAM and FECHLD are on a four-point scale (similar to

HOMOSEX). The remaining variables are on a two-point scale of "Agree or Disagree."

Both FEFAM and FECHLD were recorded into dichotomous categories much like

HOMOSEX. Like the homonegativity scale, the gender role scale will range from 0 to 5

with higher scores indicating more traditional attitudes about gender roles (in contrast to

lower scores indicating more egalitarian attitudes). Scales are being used rather than

individual items because the items selected represent variations in similar characteristics.










Both the homonegativity scale and gender role scale were assessed for reliability

using factorial analysis and Varimax rotation (Table 3-1). Varimax rotation identified one

factor for each scale. Follow-up analysis computed Chronbach's alpha for both scales (a

= 0.783 for the homonegativity scale and a = 0.697 for the gender roles scale) (Tables 4-

1 and 4-2). Overall mean value for the homonegativity scale was 1.627 with insignificant

skew and slightly platykurtotic kurtosiss = -1.026). The gender role scale had an overall

mean of 1.510 and was neither significantly skewed nor kurtotic (Table 4-1).

Table 3-1 Factorial analysis and Chronbach's alpha for HomoScale and RoleScale

Alpha if item
Variable Factor deleted
HomoScale Homosex 0.5827 0.8228 Chronbach Alpha = 0.7827
Colhomo 0.8604 0.6734
Spkhomo 0.8466 0.6922
Libhomo 0.8124 0.7104

RoleScale Fefam 0.7639 0.5992 Chronbach Alpha 0.6966
Fechld 0.6058 0.6736
Fepol 0.6996 0.6459
Fehome 0.7884 0.5999
_Fework 0.4879 0.7048
N=6133
Source, General Social Survey 1977-1998


As discussed, gender and race were also included in this analysis. Gender was

dichotomized into male and female (F = 1) while race was dichotomized into African

American and Caucasian (Af. Amer. = 1).

Finally, the additional control factors were included. These factors include the age

at the time of the interview (in years), highest educational attainment of the respondent

(in years), religious fundamentalism, and the political ideology of the respondents.

Religious fundamentalism was measured by asking respondents how they identified their









religious beliefs (fundamental, moderate or liberal)5. Political ideology was measured

using a seven-point scale in which respondents were asked to indicate where they "fall"

in their political beliefs (1 being extremely liberal and 7 being extremely conservative).

Finally, educational attainment was measured by asking respondents what was the

highest grade completed. Responses could vary from 0 (no formal schooling) to 20 (a

doctoral degree or greater than four years of schooling beyond a bachelors degree). As

discussed, research has identified that older individuals, and individuals with less

education display more homonegativity. Furthermore, individuals who identify as

politically conservative or religiously fundamental also have a tendency to display more

homonegative affect.

Univariate, Bivariate, and Ordinary Least Squares multivariate analysis were used

to see how gender role attitudes, gender and race confound to explain homonegative

affect. All models will take into consideration the year the data were collected, which

were coded into a series of dummy variables. For modeling purposes, data from the year

1998 were excluded as this year represented the reference category. As discussed,

homonegative attitudes have been in flux over the course of the last quarter century.

Consistent consideration for the year of data collection will help better show the effects

of independent variables net of year.

Model building will consider the effects of all variables discussed simultaneously

(Model 2, Table 4-6), the impact of gender role ideologies alone (Model 1), the impact of



5 In the assessment of both scales and all independent variables, non-response to any one of the questions
constituted exclusion from the sample. In the end, 6133 individuals remained. See also, comment 2.
Fundamentalism was coded 1 through 3 with higher values corresponding to liberal orientation.






20


gender alone (Model 3) and race alone (Model 9). Gender role ideologies, gender, and

race are believed to provide substantial impacts on homonegativity. To assess the impacts

of other confounding variables on the significance of gender and race, additional models

were run (Models 4-8 and 10). Finally, assessment of nonadditivity between both gender

and race with other variables was taken into consideration (Models 11-16, Table 4-7).














CHAPTER 4
RESULTS

Univariate Analysis

Approximately 11.3% of the sample was comprised of African American

individuals. The remaining 88.7% were identified as Caucasian. The sample was

comprised of approximately 56% female and 44% male6. The mean age was roughly 44.5

years with a median of 41 years (Table 4-1). As stated, ages varied from 18 to 89 years.

The mean education of the sample was 12.88 years, or basically just short of one year of

college. The median education reported was the 12th grade. Religious fundamentalism

was coded by asking respondents whether their religious denomination was

fundamentalist, moderate, or liberal (coded 1 through 3 with higher values indicating

more liberalism). Mean value for fundamentalism was 1.94 with both the median and

mode at 2 (moderate). Finally, for political ideology, respondents were asked how they

identified on a scale from 1 through 7 with a response of 1 indicating "extremely liberal"

while a response of 7 corresponded to "extremely conservative." Mean response was 4.07

(moderate) with a median and mode of 4 as well. Distributions for all variables were

approximately normal, with the exception of a slightly platykurtotic distribution for

religious fundamentalism (Kurtosis = -1.26). The final sample size was 6133 (Table 4-2).






6 Race was dummy coded with values of 0 corresponding to Caucasian and 1 corresponding to African
Americans. Gender was dummy coded with the 1 group representing females.












Table 4-1 Univariate statistics


HomoScale* 0 1496 24.39 1.626 1.412 0.546 -1.026 0 4
1 2087 34.03
2 873 14.23
3 567 9.25
4 1110 18.1


RoleScale** 0 2132 34.76 1.510 1.514 0.756 -0.507 0 5
1 1430 23.32
2 1019 16.62
3 720 11.74
4 526 8.58
5 306 4.99
Afam 0.113 0 1
Female 0.562 0 1
Age 44.473 16.932 0.515 -0.645 18 89
Education 12.878 3.020 -0.171 0.767 0 20
Fundamental __1.943 0.759 0.095 -1.258 1 3
Political ID 4.071 1.383 -0.071 -0.456 1 7
Increasing values correspond to increasing HomoNegativity
** Increasing values correspond to increasing traditional gender role ideologies
N=6133
Source, General Social Survey 1977-1998


Table 4-2 Sample size and means by year

Sample HomoScale RoleScale
Year size mean* mean**
1977 1163 2.012 2.321
1985 1193 1.881 1.599
1988 344 1.721 1.439
1989 359 1.585 1.337
1990 329 1.623 1.357
1991 362 1.619 1.461
1993 397 1.484 1.237
1994 693 1.345 1.082
1996 681 1.248 1.141
1998 612 1.206 1.223
Higher values correspond to more negative attitudes
** Higher values correspond to more traditional gender role attitudes
N 6133
Source, General Social Survey 1977-1998

Bivariate Analysis


Pearson's Correlations were assessed for all quantitative variables (e.g., both scales,


respondent age, education, political ideology, and religious fundamentalism) (Table 4-3).


Correlations were significant in all relationships except for those between religious


Frequency


Percent


St. Dev


Kurtosis Min Max









fundamentalism and age. The homonegative scale had a positive moderate correlation

with the gender role scale supporting the first hypothesis that individuals expressing

traditional gender role ideologies were also more homonegative (r = 0.48, p < 0.01). The

homonegative scale also had positive weak correlations with age and political ideology (r

= 0.26, p < 0.01, and r = 0.22, p < 0.01 respectively), supporting expected relationships

between both older individuals displaying more homonegativity and politically

conservative individuals expressing more homonegativity. Additionally, there was a

moderate negative correlation between respondent's education and reported

homonegativity (r = -0.41, p < 0.01) supporting that individuals with a higher education

will report less homonegativity. Further, there is weak negative correlation between the

homonegative scale and religious fundamentalism (r = -0.295, p < 0.01), supporting that

religiously liberal individuals report less homonegativity. Noteworthy is the substantial

correlation between the homonegative scale and the gender role scale (r = 0.48, p < 0.01).

Finally, correlations among independent variables, although significant, were relatively

weak ruling out possible multicolinearity (Table 4-3).

Because the focus of this project ascertains the relationship between homonegative

attitudes and gender role ideologies, correlations by year were assessed between the two

scales (Table 4-4). Overall, the strength and direction of the relationship is relatively

consistent. There is a slight dip in the strength of the correlation in 1989 (r1989 = 0.34),

however not a substantial drop. Nonetheless, the strength of the relationships between the

scales was at its highest point a year prior (r1988 = 0.55).











Table 4-3 Correlation matrix for all quantitative variables

HomoS RoleS Educ- Fund- Pol.
cale cale Age ation mental views
Homoscale 1 0.48 0.26 -0.41 -0.295 0.22
Rolescale 1 0.33 -0.36 -0.18 0.19
Age 1 -0.24 0.01 0.12
Education 1 0.23 -0.05
Fundamental __1 -0.14
Pol. Views 1
Values correspond to Pearson's correlation
NOT Significant a > 0.05
N=6133
Source, General Social Survey, 1977-1998

Table 4-4 Pearson correlation between HomoScale and RoleScale

Year Correlation*
1977 0.49
1985 0.47
1988 0.55
1989 0.34
1990 0.41
1991 0.43
1993 0.47
1994 0.38
1996 0.45
1998 0.47
*All significant a p < 0.01
Source, General Social Survey 1977-1998

Assessment of potential racial or gender differences were calculated using


independent sample t-tests (Table 4-5). Focusing on racial differences, we find that


African Americans report a mean higher score on the homonegative scale. This supports


the fourth hypothesis that African Americans will display more homonegativity than


Caucasians. There are no significant racial differences on answers toward the gender role


scale. Nonetheless, there are significant racial differences in age, education, religious


fundamentalism, and political ideology. In sum, we find that African Americans in the


sample are approximately three years younger, have about a year less education, are more


fundamentalist, and politically liberal than Caucasians in the sample.











Table 4-5 Racial and gender differences using T-tests

Mean for
Mean for African
Variable T value p-value Whites Americans
Race Homoscale -4.73 .01 1.596 1.86
RoleScale -0.05 0.61
Age 4.32 .01 44.8 41.9
Education 7.72 .01 12.98 12.05
Fundamental 18.12 .01 2 1.46
Pol. Views 3.74 <.01 4.09 3.89

Mean for Mean for
Men Women
Gender HomoScale 1.78 0.08
RoleScale 4.72 <.05 1.61 1.43
Age -2.96 <.01 43.7 45
Education 3.85 <.01 13 12.7
Fundamental 4.56 <.01 1.99 1.90
Pol. Views 1.57 0.11
N=6133
Source, General Social Survey 1977-1998

In analysis of gender differences, we find no differences in mean scores on the


homonegative scale. This does not support the third hypothesis that men will display


more homonegative attitudes than women. As expected, we do find gender differences in


that women are more egalitarian in their gender role ideologies, supporting the second


hypothesis. Additionally, women are about a year and a half older on average, have less


education, and report more fundamentalist religious ideologies than men. Finally, we find


no significant gender differences in political ideology. Hypothetical stipulations and


previous research have both supported that women tend to display less homonegativity


than males. Controlling for differences between males and females in age, education,


gender role ideologies, and/or religious fundamentalism may help to better expose the


relationship between gender and homonegativity


Multivariate Analysis

In an effort to understand the effects of the independent variables on the


homonegative scale while controlling for the effects of other variables, multivariate









analysis was conducted using Ordinary Least Squares methods. Model 1 (Table 4-6)

presents the effects of the gender role scale on the homonegative scale while controlling

for the variability in each year the data were collected. As stated, dummy coding was

allowed for each year data were collected from and 1998 is excluded from all models as it

acts as the reference group. In Model 1, we see the effect of traditional gender views on

homonegativity is positive and significant. Increases in traditional gender role ideologies

correspond to increases in homonegativity (Std. B = 0.466, p < 0.05). Additionally,

significant to all models is the apparent "spike" attributed to the data from the year 1985.

It is believed that this spike in overall predicted homonegativity could be better attributed

to the historical events surrounding the discovery of HIV/AIDS in the early 1980's and

its greater incidences among gay men. In 1984, nationwide recognition was given to the

existence of this epidemic. At that time, the disease was mainly affecting gay men.

Societal initial responses were to both blame the disease on gay men and also allow

justification as punishment for immoral behavior. This spike observed consistently in

1985 could be evidence of a period effect being experienced in response anti-gay

sentiments associated with the HIV/AIDS epidemic. With time, there was an increase in

recognition that this disease was not necessarily a gay disease. This could explain the

observed drops in homonegativity attributed to subsequent years of data. Finally, while

controlling for the years data were collected, we see that more recent years (in particular

1994 and 1996) do not contribute significantly compared to 1998 in explaining

homonegativity. This can be described as a plateau effect. In essence, we see an overall

pattern of decrease in homonegativty across the years with diminishing effects in more

recent years.









Model 2 considers all variables introduced in this study (Full Model). Controlling

for the effects of gender, race, education, religious fundamentalism, political ideology,

and the years the data were collected, we still observe a significant contribution of gender

role ideologies on attitudes toward homosexuals. As stated, all models that include the

gender role scale as a predictor, find that the gender role scale contributes the most

predictive ability (demonstrated by its Standardized Estimate). In analysis of other

control factors, we find all expected relationships confirmed with the exception of

predicted racial differences. In Model 2, we find that race is not a significant predictor of

attitudes toward homosexuals while controlling for other factors. Otherwise, net of the

effects of other variables, we find that men, older individuals, those with lower levels of

education, those expressing fundamentalist religious ideologies, those with traditional

gender role ideologies and those expressing conservative political ideologies are all

predicted to display more homonegative attitudes than their counterparts. Using a

subsequent F-test, model 2 provides overall explanatory power of the homonegativity

scale (F = 209 p < 0.05)7. In analysis of this power, we find that approximately 35% of

the variability in the homonegativity scale can be better explained by the gender role

scale, race, gender, education, political ideology, age and religious fundamentalism

compared to using the mean of the homonegativity scale alone (R2 = 0.353).

Intervening and Suppressor Effects

The data at hand present both intervening and suppressor effects related to expected

relationships among race and gender and their effects on homonegativity. As remarked in


Significant F-statistics were obtained from all models discussed.















Table 4-6 Model summaries**
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Model 9 Model 10
Intercept .719 2.120 1.234 3.727 2.357 .239 2.142 2.59 .673 1.55
RoleScale .466 (.434) .300 (.280) .351 (.327) .320 (.299) .465 (.433) .427 (.398) *
Afam .016 (.071) .063 (.281) .016 (.069)
Female -.025 (-.072) -.017 (-.047) -038 (.108) -.033 (-.097) -.027 (-.076) -.015 (-.042) -.027 (-.075) *
Age .101 (.008) .267 (.022) .087 (.007) .108 (.009) *
Education -.218 (-.102) -.385 (-.180) -.252 (-.118) -.214 (-.100) *
Fundamental -.172 (-.319) -.288 (-.536) -.189 (-.352) -.212 (-.394) *
Pol. View .113 (.115) *


1977 .079 (.286) .067 (.242) .223 (.804) .129 (.463) .198 (.712) .233 (.804) .057 (.204) .320 (.214) .082 (.295) .074 (.266) *
1985 .131 (.469) .114(.408) .188 (.672) .139(.496) .174(.619) .196 (.699) .116(.413) .059 (.411) .135 (.483) .127(.452) *
1988 .062(.378) .051 (.311) .083 (.514) .057(.353) .071 (.433) .091 (.561) .052 (.321) .115(.296) .061 (.375) .054(.239) *
1989 .048 (.286) .038 (.226) .063 (.378) .043 (.261) .051 (.307) .067 (.402) .050 (.241) .048 (.214) .059 (.294) .041 (.245) *
1990 .050(.315) .046 (.290) .067 (.417) .053 (.330) .063 (.397) .068 (.429) .046 (.287) .036(.291) .052(.327) .050 (.311) *
1991 .044 (.266) .035 (.208) .068 (.410) .049 (.292) .056 (.337) .069 (.405) .037 (.223) .046 (.196) .046 (.275) .038 (.229) *
1993 .040 (.228) .024 (.137) .048 (.273) .025 (.146) .039 (.225) .049 (.279) .028 (.160) .033 (.143) .040 (.232) .035 (.201) *
1994 .035 (.156) .024 (.106) .031 (.138) .021 (.093) .020 (.090) .035 (.154) .029 (.128) .025 (.101) .036 (.160) .027 (.121)
1996 .001 (.034) -.0005 (-.002) .009 (.041) .003 (.012) .002 (.011) .014(.062) .005 (.204) .023 (.009) .007 (.031) .003 (.012)


R-Squared .242 .353 .043 .184 .126 .114 .308 .341 .246 .287
F-Value 195 209 27.5 125 79.8 71.7 209 227 166 205
Dependent Variable: HomoScale
N=6133
*p<0.05
**Values correspond to standardized parameter estimates. Those in parentheses correspond to the parameter estimate.
Source General Social Survey 1977-1998









bivariate analysis of gender and homonegativity, we found that there was weak evidence

of a bivariate relationship (men and women did not differ significantly in their

homonegative attitudes). However, in Model 2 (the full model), we find that gender does

provide a significant effect while controlling for other variables. This evidence now

supports the second hypothesis that males are more homonegative. Further analysis was

conducted to unveil what variables contributed to this suppressor effect. Models 3

through 8 are those being used to discuss the relationship between gender and

homonegativity. Model 3 demonstrates that there is no evidence of a relationship between

gender and homonegativity (even while controlling for the year the data were collected

from). However Models 4, 5 and 6 all show that the relationship becomes significant

when we control for age, education, and religious fundamentalism. As noted in bivariate

analysis, women in the sample tended to be older, more religiously fundamental, and

have less educated, which all contribute to higher levels of homonegativity. When

controlling for any one of these factors, the previously masked relationship becomes

prevalent.

To add to the complexity of this relationship, the gender role scale was reintroduced

into the model. In bivariate analysis we found that women in the sample tended to have

more egalitarian attitudes about gender roles. When the gender role scale was

reintroduced in the presence of gender, education, age, or any combination of the three,

we found that gender was again no longer significant. The lack of relationship remained

intact until religious fundamentalism was controlled for8. In essence, the gender role scale

better explains gender differences in homonegativity until the level of religious


8 For simplicity, not all model combinations were shown









fundamentalism is taken into consideration. Respondent's age and education can also be

used to unmask the relationship between gender and homonegativity however not while

controlling for gender role ideologies. In either situation, the data suggest the

implications religious doctrines can have on both gender role ideologies, on

homonegativity and gender.

As briefly mentioned, the data also present an intervening relationship between race

and homonegativity. In bivariate analysis, we found extremely strong evidence of a

relationship between race and homonegativity, with African Americans displaying more

homonegative attitudes than Caucasians. In Model 2 (the full model), we see that race no

longer acts as a significant predictor of homonegativity while controlling for other

effects. As with unveiling the relationship between gender and homonegativity, several

models were assessed to unveil the intervening relationship between race and

homonegativity. In all models assessed (using combinations of other independent

variables) we find that race remains intact as a significant predictor of homonegativity

until religious fundamentalism is controlled for9 (Models 9 and 10). Bivariate analysis

has indicated that African Americans are more fundamental in their religious ideology

than Caucasians. When this factor is taken into consideration, race alone no longer acts as

a significant predictor of homonegativity. Previous research has implicated African

Americans as having higher levels of homonegativity. Although this may be the case,

these findings demonstrate that this relationship may be the result of religious doctrine,

and not necessarily race. Admittedly, there is a historical relationship in the religious

differences among Caucasians and African American's; however in the analysis of


9 Not all model combinations shown









homonegativity, inadvertent focus on racial differences could probably be better

discussed as a product of religious affiliation and dogma, even while controlling for the

effects of gender role ideologies. Nonetheless, a discussion of interaction/nonadditivity

may help to better understand this relationship.

Nonadditivity

Finally, multivariate analysis took into consideration nonadditive relationships

among the independent variables. Analysis of nonadditive relationships helps to better

understand interactive effects among variables (Table 4-7). Rather, non-consideration of

interactive models assumes identical change for one variable over the variability of

another. It would predict, for example, that males and females experience the same

predicted change in homonegativity as level of education increased. In theory, either

males or females may be more sensitive to increases or decreases in education and this

can be reflected in predicted homonegativity. Nonadditivity in this study will give

consideration that males may differ from females just as blacks may differ from whites

across the variation of other independent variables. Models 11 through 14 take into

consideration the interaction between both race and gender on other independent

variables. Models 15 and 16 take into consideration the interaction between race and

gender on the gender role scale.

















Table 4-7 Model summaries with nonadditivity**
Model 11 Model 12 Model 13 Model 14 Model 15 Model 16
Intercept 2.185 2.162 2.083 2.231 2.181 2.107
RoleScale .300 (.279) .300 (.280) .299 (.279) .301 (.280) .268 (.249) .314 (.292) *
Afam -.112 (-.500) -.052 (-.233) .083 (.368) .016 (.072) .016 (.072) .055 (.244) *
Female -.026 (-.074) -.024 (-.069) -.025 (-.072) -.111 (-.316) -.054 (-.154) -.026 (-.074) *
Age .102(.009) .101 (.008) .100 (.008) .102 (.008) .100 (.008) .101 (.008) *
Education -.231 (-.108) -.217 (-.101) -.219(-.102) -.237 (-.111) -.218 (-.102) -.218 (-.102) *
Fundamental -.170 (-.317) -.183 (-.340) -.170 (-.317) -.172 (-.320) -.172 (-.320) -.171 (-.318) *
Pol. View .114(.116) .112(.114) .214(.127) .113 (.116)* .112(.115) .112(.114) *
RaceXeduc .132(.047) *
RaceXfund .074 (.200) *
RaceXpol. View -.072 (-.076) *
SexXeduc .090 (.019) *
SexXroleScale .051 (.054) *
RaceXroleScale -.056 (-.113) *


1977 .067 (.242) .066 (.240) .067 (.240) .068 (.243) .067 (.241) .067 (.242) *
1985 .114(.408) .114(.405) .113(.404) .114(.409) .114(.407) .115 (.410) *
1988 .050(.308) .050 (.307) .051(.310) .051(.311) .051(.311) .051(.312) *
1989 .038 (.226) .037 (.224) .037 (.224) .038 (.227) .038 (.228) .038 (.227) *
1990 .046 (.289) .046 (.288) .046 (.287) .046 (.291) .046 (.287) .047 (.292) *
1991 .035 (.208) .034 (.205) .034 (.206) .035 (.208) .034 (.204) .035 (.209) *
1993 .024 (.138) .023 (.132) .024 (.137) .024 (.139) .024 (.135) .0252 (.145) *
1994 .023 (.103) .023 (.104) .023 (.104) .024 (.104) .024 (.106) .0245 (.109)
1996 -.0006 (-.003) -.001 (-.006) -.0003 (-.001) -.0002 (-.001) -.0008 (-.003) -.0006 (-.003)


R-Squared .355 .354 .354 .354 .354 .355
F-Value 198 197 197 197 197 197
Dependent variable: HomoScale
N=6133
* p<.05
**Values correspond to standardized parameter estimates. Those in parentheses correspond to the parameter estimate.
Source General Social Survey 1977-1998









Race interactions

In analysis of the interaction of race and educationO1 (Model 11), we find that the

effect of having higher levels of education does not produce the same rate of predicted

decrease in homonegativity between whites and blacks. Additionally, we find that

African Americans with no education are expected to have lower levels of

homonegativity than whites. However, as the level of education increases, African

Americans will not experience the same rate of predicted decrease in homonegativity. At

approximately 10 years of education, controlling for other effects, we observe that the

two lines cross. In other words, beyond 10 years of education we find that whites' benefit

from higher levels of education predicts lower levels of homonegativity than African

Americans at the same level of education; whereas, before ten years of education, African

Americans are less homonegative than whites at the same level (Figure 4-1)1.

An interesting aspect of this interaction is the almost reverse effect education has on

African Americans (i.e., African Americans with lower levels of education are expected

to have lower homonegativity than whites with the same education, compared to reverse

that occurs when the two groups have higher levels). Finally, important in this

relationship, is to take into consideration that increases in education do indeed predict

lower levels of homonegativity for both groups, however the impact is greater for

Caucasians. This implication suggests whites homonegativity is more sensitive to

changes in education whereas African Americans' homonegativity is more immune.


10 Interactions were also assessed for Education and Religious Fundamentalism, and Gender and Religious
Fundamentalism, Gender and Race, however these interactions were not significant and will not be
discussed.

1 All figures shown are to scale however the scales shown differ in effort to visually demonstrate the
interactive relationships discussed interactive relationships discussed











3

2.5


SI-- Whites
1.5- Blacks




0.5
0 4 8 12 16 20
Education in Years

Figure 4-1 Interactive effects of education and race on homonegativity

In Model 12, we take into consideration the interaction between race and religious

fundamentalism. This interaction is highly similar to the race and education interaction.

African Americans expressing fundamental religious ideologies are expected to be less

homonegative than Caucasians at the same level of fundamentalism. However, as with

the interaction between education and race, the predicted lines cross as religious

fundamentalism "moves" toward more liberal religious ideologies (Figure 4-2). African

Americans expressing liberal religious ideologies are expected to be more homonegative

than Caucasians in the same category, controlling for other effects. In this situation, we

see that having religiously liberal ideologies "benefits" Caucasians more than African

Americans in their attitudes toward gays and lesbians. As with education, having

religiously liberal ideologies does decrease predicted values in homonegativity, but not at

the same rate across races. Finally, the evidence surrounding Model 10 (religious

fundamentalism better explaining race in predicted homonegativity) is quite interesting. It

was previously argued that racial differences in negative attitudes toward lesbians and










gays could be better attributed to differences in religious fundamentalism. Nonetheless,

Model 12 proposes that religious fundamentalism acts independent of races.

1.9
1.8
1.7
1.6
O 1.5 -*- Whites
S1.4 --- Blacks
1.3
1.2
1.1

1 2 3
Fundamental Moderate Liberal

Figure 4-2 Interactive effects of religious fundamentalism and race on homonegativity

Model 13 assesses the interaction between race and political ideologies. Contrary to

the interactions between race-education and race-fundamentalism, the interaction

between race and political ideologies predicts liberal African Americans to be more

homonegative than Caucasians with the same political ideology (controlling for other

factors) (Figure 4-3). However, having a more conservative political ideology has a

greater impact on homonegativity for Caucasians than African Americans, in that

Caucasians responding as conservative or extremely conservative are expected to display

more homonegativity than African Americans with the same ideology (controlling for

other factors).

In sum of interactions with race, we find an overall pattern of effect that is greater

for Caucasians than for African Americans. The slope of the parameter estimate for the

African American group has, in all circumstances, been closer to zero. In other words, we

find that African American homonegativity is more consistent across the variations of










political ideologies, religious fundamentalism, and education, whereas increases or

decreases in these variables presents greater impacts for Caucasians levels of

homonegativity.

7

1.5

1.3
1 -- Whites
S11 Blacks
0.9

0.7

0.5
1 2 3 4 5 6 7
Political Ideology

Figure 4-3 Interactive effects of race and political ideology on homonegativity.

Gender interactions

Model 14 takes into consideration the interactive effects of gender and education

(Figure 4-4). In this circumstance, we find that women overall display less

homonegativity. Additionally, we find that the benefit to higher levels of education on

decreasing predicted homonegativity has a more significant impact for men than women.

Women's lack of decrease in homonegativity for higher levels of education could

probably be a result of lower overall homonegativity to begin with.

Interactions with sex, race, and the gender role scale

Finally, analysis of interaction with the gender role scale was assessed. Model 15

analyzes interaction between gender and the gender role scale while Model 16 considers

racial interactions with the gender role scale. Interaction finds that women expressing

more egalitarian attitudes toward gender roles will subsequently display less










homonegative affect than males in the same category. However women who display

more traditional gender role ideologies are actually predicted to display higher levels of

homonegativity than males in the same category (controlling for other variables) (Figure

4-5).

3

2.5


SMales
Females
1.5




0.5
0 4 8 12 16 20
Education In Years

Figure 4-4 Interactive effects of education and gender on homonegativity

The final interactive effect to be discussed accounted for the interaction of race and

the gender role scale, and its impact on Homonegativity (Model 16). In this analysis we

find that African Americans holding egalitarian attitudes about gender roles are expected

to be more homonegative than Caucasians with similar gender role attitudes. However

the impact of having more traditional gender roles has a greater substantial impact for

Caucasians, to where predicted values in homonegativity actually surpass those of

African Americans ( Figure 4-6). Findings indicate that increases in the gender role scale

do correspond to increases in homonegativity for both Caucasians and African

Americans, however the impacts of having more traditional gender role ideologies is

much greater for Caucasians than for African Americans (at least in explaining negative

attitudes toward lesbians and gays).











2.5
2.3
2.1
1.9
C- 1.7
S-*-Males
1.5 -- Females
1.3
1.1
0.9
0.7
0.5
0 1 2 3 4 5
RoleScale
Figure 4-5 Interactive effects of gender and the RoleScale on homonegativity.


2.5


2

o -- s Whites
S1.5 Blacks


1


0.5
0 1 2 3 4 5
RoleScale
Figure 4-6 Interactive effects of race and the RoleScale on homonegativity.

All nonadditive models discussed underwent subsequent F-tests12 to determine if

the interactive models provided a significantly better model fit than Model 2 (the full

model). In all instances, interactive models did indeed provide significant F-values (p <

0.05). In other words, allowing for interaction actually provided more explanatory ability

than models ignoring this feature.


12 Not Shown














CHAPTER 5
DISCUSSION AND CONCLUSION

The purpose of this study was to discuss why gender and racial differences in

attitudes toward homosexuals might differ. Additionally, this study aimed to discuss

these differences as a result of the impact of gender role ideologies. Findings indicate that

gender differences in homonegativity can be hidden by normative differences in age,

education, gender role ideologies or religious fundamentalism. Furthermore, male

homonegativity is easier influenced by educational changes than females. In the case of

race, multivariate analysis finds that prominent racial differences between blacks and

whites homonegativity can be better explained by religious fundamentalism. In

considering interaction, the data indicate whites' attitudes toward homosexuals have a

tendency to be more sensitive to changes in education, religious fundamentalism and

gender role ideologies. Finally, in the consideration of the impacts of gender role

ideologies, we find confirmed expected relationships that traditional gender role

ideologies correlate with increases in homonegativity; even while controlling for the

effects of race, gender, education, religious fundamentalism, political ideology, and year

data were collected from. Substantial to this correlation was its strength these ideologies

have, over other variables, in uniquely explaining homonegative attitudes, in addition to

the sensitivity in homonegativity experienced by females as a result of changes in gender

role ideologies.

Previous research on homonegativity has identified correlates among individuals

that stimulate homonegativity, in addition to providing some reasoning why these









relationships may exist. This study has built a bridge in that it has done both. The data

have demonstrated linkages between gender role ideologies, race, gender, religious

fundamentalism, age and education with subsequent attitudes toward lesbians and gays.

Just like race and gender, it is assumed that gender role ideologies precede

homonegativity and not the reverse. In this circumstance we can now better understand

the effects that gender role schemas play in our every day lives. Barbara Risman in her

book Gender Vertigo (1998) has proposed that gender should be as insignificant as the

length of one's toes or the color of one's eyes. Nonetheless, gender is at the forefront of

everything we do. The elimination of gender inequality can only be accomplished

through the elimination of gender. Although there may be many factors influencing

homonegativity, the data have demonstrated that gender role schemas are still the best

predictor, even when controlling for these other influential factors. In essence, this study

helps expose the link between gender inequality and homonegativity. Although an

elimination of gender inequality could have a significant impact on reducing

homonegativity, the data also demonstrate that there are many other factors beyond

gender role ideologies that influence homonegativity. Future research could analyze

potential linkages between gender role ideologies and any number of the other

independent variables considered in this study. An altering of the structures of gender and

gender role ideologies will subsequently have effects on any number of other variables

(i.e., challenging gender role ideologies will not only effect homonegativity but also

institutions such as the government, religion, families).

Previous research have almost universally identified that men tend to express more

homonegative attitudes than women. This study has argued that this effect could largely









be attributed to the impacts of gender role ideologies. Importantly noted was the lack of

an apparent gender difference in bivariate analysis of homonegativity (i.e., men and

women did not significantly differ in homonegative attitudes). Multivariate analysis

helped to unveil this relationship and it was suggested that the initial relationship might

have been masked by gender differences in the sample (i.e., men were younger on

average and tended to have more education on average). Many of the previous

discussions on gender differences in homonegativity were able to identify this difference

through bivariate analysis and little attention was given toward unmaking this

relationship. Additionally, many previous researchers had a tendency to use more

convenient samples such as college students, medical students, etc. The General Social

Survey uses a probability-based sample that is far more accurate at portraying the

American population than a college sample. Although this research has suggested

bivariate differences were not discovered because of other factors, further research is

needed to analyze why or why not differences may exist.

This study also suggests the need for understanding African American

homonegativity. Research has continually demonstrated that racial discrepancies exist,

but has rarely offered as explanation for why this may be the case. Bivariate analysis

supported previous research in conclusions about racial differences. Nonetheless, this

relationship was better explained through religious fundamentalism. This acted as a

preliminary indication and explanation as to why these differences may exist, however

the incorporation of nonadditive effects in the multivariate models further complicated

any assumptions. In all circumstances of interaction discussed, we find that African

Americans are less sensitive to the changes in other variables; and although political









ideology, religious fundamentalism, and education do provide explanatory power,

changes across these variables for African Americans are minute compared to those of

Caucasians. Understandably, there may be influences not addressed by this study that

could better explain African American homonegativity. The lack of variability/sensitivity

could be attributed to the smaller sample size of this group in the sample or poor choice

of variables that may confound homonegativity for African Americans. Additionally the

African American sample is under representative of black men. Again, this provides

further reason to devote considerable attention to African American homonegativity. In

analyzing racial differences it is equally important to assess class differences. It could be

possible that class differences are what cause this plateau effect among African

Americans.

In analyzing gender interactions with education, a unique relationship was

discovered. As predicted, women display less homonegativity than men in similar

educational categories. However the benefit of having a higher level of education to

reduce homonegativity is more substantial in males. This could be due to several reasons.

First, it is possible that because female homonegativity is already low, there is little that

an increase in education can due to reduce this further. Second, is the idea that education

may not be as good a predictor for reducing homonegativity among females. This would

be similar to the discussion of interaction and African American homonegativity.

When considering interactions between gender role ideologies (the RoleScale) and

gender, a unique unexpected relationship was discovered. Egalitarian women were

expected to be less homonegative than men in the same category; however female

adherence to gender role ideologies greatly increases predicted homonegativity (to where









is surpasses the predicted homonegativity of males in the same category). Since

traditional gender role ideologies typically have more substantial negative impacts on

women, I argue that women who support these values have embedded themselves with a

much stricter gender role reality. In this study, women overall were expected to display

more egalitarian attitudes toward gender roles (especially those specifically concerning

the roles and behaviors of women). However, in the event that women display traditional

roles (contrary to normative expectations), we find individuals who embrace structural

norms more stringent than society imposes. These individuals, I propose, are classic

illustrations of the impacts that stringent gender role ideologies can have on attitudinal

structures (specifically toward gays and lesbians, or any individuals/groups who violate

supposed gender norms). These women have adopted these stringent traditional gender

role ideologies and allowed them to influence how they see the world. I call this

phenomenon a belief effect experienced by these women. The effect of these females

believing in strict traditional gender role ideologies, contrary what may be expected of

females, creates a strict gender role schema observed indirectly through homonegativity.

Future research on the belief effect observed with this data may help to confirm the deep-

rooted linkages between gender role ideologies and attitudes toward sexual minorities.














CHAPTER 6
LIMITATIONS AND RECOMMENDATIONS

There are several features of this study that could benefit from refinement in the

future. The first is unidimensionality of the gender role scale. Gender role ideologies

were operationalized by considering attitudes about the equality of women. Ideals of

gender and sex roles span much further than attitudes about women. Ideally, a scale of

gender role attitudes would encompass many dimensions that gender role ideologies are

demonstrated through (e.g., childrearing responsibilities, housework, attitudes about men,

attitudes about how to raise children, attitudes toward gender neutrality). Additionally,

the questions forming the homonegativity scale do not specify the gender of the

"homosexual" being described in addition to being rather one-dimensional. As discussed,

attitudes toward lesbians versus gays are not universal. It secondly omits bisexual

individuals along with transgender and inter-sexed individuals. There is substantial

research indicating that attitudes and behaviors may not necessarily match up.

Additionally, there is the stipulation that just because individuals may not agree with

homosexuality does not mean they would deny homosexuals equal rights. And although

both scales used do represent an array of questions, variability in the scales boarder what

a truly quantitative variable should be. Hence there may be some question as to whether

Ordinary Least Squares methodology was best used for the Mulivariate analysis.

Nonetheless, univariate analysis did provide relatively strong support that the variables

and scales used do approximate normal distributions.









Additionally, this study gave little focus to the effect of changes over time. A

substantial body of research in the area of studying homonegativity has been devoted to

observed changes over time. In an effort to increase sample size and statistical power,

twenty years of data were conglomerated. Although the year the data were collected from

was controlled for, the data taken from 1985 does implicate a possible period effect that

was only briefly addressed by this study. Either the use of a single large cross-sectional

dataset, or further analysis of potential age, period, or cohort effects should be considered

in future research.

The General Social Survey, although a very strong research tool, in itself presents

issues of validity and accuracy. This study is primarily assessing attitudes and omits

behaviors. It is generally accepted that attitudes compliment behaviors (and vice versa)

however it is firmly established that this may not be true all the time (i.e., front-stage

versus back-stage behavior). Additionally, it is taken for granted that survey respondents

are completely honest. The General Social Survey is given face to face; hence it is

possible that respondents may respond dishonestly on a variety of questions in effort to

avoid disapproval from the interviewer.

The sexual orientation of the respondent could have significant influence over their

responses. Gay respondents would be far more likely to respond in favor of attitudes

toward homosexuals independent of many other control variables. There are many ways

to assess homosexual behavior including self-identification, feelings and behaviors.

There is a behavioral measure used in the General Social Survey that assess whether

respondents have had sexual relations with someone of the same or opposite sex. Future

research could consider this variable and assess its potential influence.









As discussed, considerable attention has been given to analyzing gender differences

in homonegativity. This study has suggested that these differences would largely be as a

result of differing attitudes about gender roles. Nonetheless, this study finds that gender

differences still prevail even while controlling for gender role ideologies. In essence,

what variables could better explain the gender differences between male and female

homonegative attitudes?

Finally, considerable attention was devoted to racial differences among African

Americans and Caucasians, however no attention was allotted to any other racial or

ethnic categories. This is a huge limitation in that a significant portion of the United

States is comprised of a variety of ethnic minorities. Future research must consider the

plethora of ethnic diversity that should exist in a representative sample of the United

States population.















LIST OF REFERENCES


American Psychiatric Association (2000) Diagnostic and Statistical Manual of Mental
Disorders : DSM-IV-TR. Washington DC.: American Psychiatric Association.

Batson, C. D., Flink, C. H., Schoenrade, P. A., Fultz, J., & Pych, V. (1986). Religious
orientation and overt versus covert racial prejudice. Journal of Personality and
Social Psychology, 40, 290-302.

Batson, C. D., Schoenrade, P. A., & Pych, V. (1985). Brotherly love or self-concern?
Behavioural consequences of religion. In L. B. Bron (Ed.), Advances in the
psychology of religion. Oxford: Pergamon Press.

Baumrind, D. (1995). Commentary on sexual orientation: Research and social
Implications. Developmental Psychology, 31, 130-136.

Bem, S. L. (1993). The lenses of gender: Transforming the debate on sexual inequality.
New Haven, CT: Yale University Press.

Beran, N. J., Claybaker, C., Dillon, C., & Haverkamp, R. J. (1992). Attitudes toward
minorities: A comparison of homosexuals and the general population. Journal of
Homosexuality, 23, 65-83.

Bierly, M. M. (1985). Prejudice toward contemporary outgroups as a generalized attitude.
Journal of Applied Social Psychology, 15, 189-199.

Black, B., Oles, T. P., and Moore, L., (1998). The relationship between attitudes;
homophobia and sexism among social work students. Affila Journal of Women and
Social work; 13, 2, 166-190.

Blee, K., & Tickamyer, A. (1995). Racial differences in men's attitudes about women's
gender roles. Journal of Marriage and the Family, 57, 21-30.

Bowman, R. (1979). Public attitudes toward homosexuality in New Zealand.
International Review of Modem Sociology, 9, 229-238.

Chng, C. & Moore, A. (1991). College students' beliefs and behaviors about AIDS:
Implication for family life educators. Family Relations, 40, 258-263.









D'Augelli, A. R., & Rose, M. L. (1990). Homophobia in a university community:
Attitudes and experiences of heterosexual freshman. Journal of College Student
Development, 31, 484-491.

Davis, J.A. and T. W. Smith. 1998. General Social Surveys, 1972-2000. Cumulative
online codebook and data file (www.icpsr.umich.edu/gss). Chicago, accessed
November 16, 2002.

Deaux, K., & Kite, M. E. (1987). Thinking about gender. In B. B. Hess &M. M Ferree
(Eds.), Analyzing Gender: A handbook of social science research (p. 920117).
Newbury Park, CA: Sage.

Ernst, F., Francis, R., Nevels, H. & Lemen, C. (1991). Condemnation of homosexuality
in the black community: A gender-specific phenomenon? Archives of Behavior, 20,
579-585.

Estrada, A. X. & Weiss, D. J. (1999). Attitudes of military personnel towards
homosexuals. Journal of Homosexuality. 37, 4, 83-97.

Faderman, L. (1991). Odd girls and twilight lovers: A history of lesbian life in Twentieth-
Century America. New York, Penguin.

Ferree, M. M. (1990). Beyond Separate Spheres: Feminism and Family Research. Journal
of Marriage and the Family, 52, 866-884.

Fishbein, H.D. (1996). Peer prejudice and discrimination. Boulder, CO. Westview Press.

Freudenberg, N. (1989). Preventing AIDS: A guide to effective education for the
prevention of HIV infection. American Public Health Association, Washington DC.

Glenn, N., & Weaver, C. N. (1979). Attitudes towards premarital, extramarital, and
homosexual relations in the U.S. in the 1970s. Journal of Sex Research, 15, 108-
118.

Gramick, J. (1983). Homophobia. A new challenge. Social Work, 28, 137-141.

Greaves, W. L. (1987). The black community. In H. Dalton & S. Burris (Eds.). AIDS and
the law (pp. 281-289). New Haven: Yale University Press.

Hansen, G. L. (1982). Androgyny, sex-role orientation, and homosexism. Journal of
Psychology, 112, 39-45.

Harper, P. B. (1991). Eloquence and epitaph: Black Nationalism and the homophobic
impulse in response to the death of Max Robinson. Social Text, 28, 68-86.









Heaven, P. & Oxman, L. N. (1999). Human values, conservatism and stereotypes of
homosexuals. Personality and Individual Differences. 27,1, 109-118.

Herek, G. M. (1984). Beyond homophobia: A social psychological perceptive on
attitudes toward lesbians and gay men. Journal of Homosexuality, 10, 1-21.

Herek, G. M. (1986). On heterosexual masculinity: some psychical consequences of the
social construction of gender and sexuality. American Behavioral Scientist, 29, 563-
577.

Herek, G. M. (1988). Heterosexuals' attitudes toward lesbians and gay men: Correlates
and gender differences. Journal of Sex Research, 25, 451-477.

Herek, G. M., & Glunt, E. K. (1993). Interpersonal contact and heterosexual's attitudes
toward gay men: Results from a national survey. The Journal of Sex Research, 30,
239-244.

Herek, G. M. (1994). Assessing heterosexuals' attitudes towards lesbian and gay men: a
review of empirical research with the ATLG scale. In b. Green and G. M. Herek
(Eds.), Contemporary Perspectives on lesbian and Gay Issues in Psychology (p.
206-228) Newbury Park, CA: Sage.

Herek, G. M. (2000). The Psychology of sexual prejudice. Current Directories in
Psychological Science, 9, 19-22.

Herek, G. M. (2002) Gender gaps in public opinion about lesbians and gay men. Public
Opinion Quarterly. 66, 40-66.

Hudson, W. W. & Ricketts, W. A. (1980). A strategy for measurement of homophobia.
Journal of Homosexuality, 5, 357-372

Irwin, P., & Thompson, N. L. (1977). Acceptance of the rights of homosexuals: A social
profile. Journal of Homosexuality, 3, 107-121.

Johnson, M. E., Brems, C., and Alford-Keating, P. (1997) Personality correlates of
homophobia. Journal of Homosexuality. 34, 57-69.

Kaiser Family Foundation (2001). Inside-Out: a report on the experiences of lesbians,
gays, and bisexuals in America an the public's views on issues and policies related
to sexual orientations. #3193 www.kff.org. Kaiser Family Foundation, accessed
April 8, 2003.

Kite, M. E. (1984). Sex differences in attitudes toward homosexuals: A meta-analytic
review. Journal of Homosexuality, 10, 69-81.









Kite, M. E. and Whitley, B. E. (1996) Sex differences in attitudes toward homosexual
persons, behaviors and civil rights: a meta-analysis. Personality and Social
Psychology Bulletin, 22, 336-353.

Klamen, D. L., Grossman, L. S., Kopacz, D. R. (1999) Medical student homophobia
Journal of Homosexuality, 37, 53-63.

Lance, L. M. (1992). Changes in homophobic views as related to interaction with gay
persons: a study in the reduction of tensions. International Journal of Group
Tension, 22, 291-299.

Laner, M. R., & Laner, R.H., (1979). Personal style or sexual preference: Why gay men
are disliked. International Review of Modern Sociology, 9, 215-228.

Laner, M. R., & Laner, R.H., (1980). Sexual preference or personal style? Why lesbians
are disliked. Journal of Homosexuality, 5, 339-356.

Logan, C. R. (1996) Homphobia? No, Homoprejudice. Journal of Homosexuality. 31,
31-53.

Morin, S., & Garfinkle, E. (1978). Male Homophobia. Journal of Social Issues, 34, 29-
47.

Nyberg, K. L., & Alston, J. P. (1976). Analysis of public attitudes toward homosexual
behavior. Journal of Homosexuality, 2, 99-107.

Oliver, M. B., & Hyde, J. S. (1993). Gender differences in sexuality: A meta-analysis.
Psychological Bulletin, 114, 29-51.

Pratte, T. (1993). A comparative study of attitudes toward homosexuality: 1986 and
1991. Journal of Homosexuality, 26, 77-83.

Price, J. H. (1982). High school students' attitudes toward homosexuality. Journal of
School Health, 52, 469-474.

Price, V. & Hsu, M. (1992). Public opinions about AIDS policies: The role of
misinformation and attitudes towards homosexuals. Public Opinion Quarterly, 56,
29-52.

Proulx, R. (1997) Homophobia in Northeastern Brazilian University Students. Journal of
Homosexuality. 34, 47-57.

Risman, B. (1998) Gender Vertigo: American Families in Transistion. New Haven, CT.
Yale University Press.









Seltzer, R. (1992) The social location of those holding antihomosexual attitudes. Sex
Roles, 26, 391-398.

Staples, R. (1982). Black masculinity: The Black male's role in American society. San
Francisco: Black Scholar Press.

Thompson, E.H., Grisanti, C., & Pleck, J.H. (1985). Attitudes toward the male role and
their correlates. Sex Roles, 13, 413-427.

Waldner, L. K., Sikka, A., Baig, S. (1999). Ethnicity and sex differences in university
students knowledge of AIDS, fear of AIDS, and homophobia. Journal of
Homosexuality, 37, 117-133.

Walters, A. S., & Hayes, D. M. (1998). Homophobia within schools: Challenging the
culturally sanctioned dismissal of gay students and colleagues. Journal of
Homosexuality, 35, 1-23.

Weinberg, G. (1973). Society and healthy homosexual. New York: Anchor books.

Whitcomb, D. H. (2001). Development, implementation, and applications of an
instrument to measure relationships among sexual orientation, political orientation,
and socioeconomic status in an undergraduate population. Dissertation Abstracts
international. 61, 9-B, 5062.

Whitley, B. E., Jr. (1987). The relationship of sex-role orientation to heterosexuals'
attitudes toward homosexuals. Sex Roles, 17, 103-113.

Whitley, B. E., Jr., & Kite, M. E. (1995). Sex differences in attitudes toward
homosexuality: A comment on Oliver and Hyde (1993). Psychological Bulletin,
117, 146-154.

Wills, G., & Crawford, R. (2000). Attitudes toward homosexuality in Shreveport-Bossier
City, Louisiana. Journal of Homosexuality, 38, 97-115.

Young, M., & Whertvine, J. (1982). Attitudes of heterosexual students toward
homosexual behavior. Psychological Reports, 51, 673-674.















BIOGRAPHICAL SKETCH

Christian Grov, born in 1980, was the second child of Sharon and Kjell Grov. Both

Sharon and Kjell were foreign immigrants to the United States in search of a better life

for themselves and their children. Raised in south Florida, Christian graduated

salutatorian of Stranahan High School, Ft. Lauderdale in 1998 and immediately moved

into the honors program at Broward Community College. After only two semesters,

Christian finished his associate's degree and moved to Gainesville, Florida in pursuit of a

bachelor's degree in psychology. Upon enrolling in one course in sociology, Christian

decided to double major in both psychology and sociology. In May of 2001, Christian

graduated with honors in both areas. That August, he began the masters program in

sociology at the University of Florida, where he focused on the study of gender, families,

and human sexuality. Two years later, Christian graduated with a Master of Arts degree.

In the fall of 2003, Christian will be beginning the doctoral program in sociology at the

City University of New York, Graduate Center (CUNY) in New York City.