ATTITUDES TOWARDS BISEXUAL MEN AND WOMEN: THE RELATIONSHIP BETWEEN RESPONDENTS’ ATTITUDES AND THEIR SEXUAL ORIENTATION

by

MARLENE ARNDT

THESIS

submitted in fulfilment of the requirements for the degree

DOCTOR LITTERARUM ET PHILOSOPHIAE

in

PSYCHOLOGY

in the

FACULTY OF HUMANITIES

at the

UNIVERSITY OF JOHANNESBURG

PROMOTER: DR K DE BRUIN

JANUARY 2009

ACKNOWLEDGEMENTS

Many people have helped me along the way:

My deepest gratitude to my supervisor, Dr. Karina de Bruin who tirelessly and meticulously assisted me during the writing of this thesis. Your kindness is appreciated.

Writing, like everything else in life, is easier with the support of a loving family. My mother and father have been a source of strength to me. They have given me the freedom and ability to enjoy life to the fullest. For this gift, I thank them. My grandmother who has always encouraged me to be all that I can be. My brother Andre, who supported me in the past and who will be there with me in the future. Thank you Charl, Meghan and Thora for your love and laughter.

To all my friends and relatives who remembered my name and continued to call despite my long social hibernation. Thank you.

Helena, I could not have continued to sit at my computer day after day without your generous spirit, healing energies, gutsy inspiration and laughter.

The financial assistance of the National Research Foundation (NRF) towards this research is acknowledged. Opinions expressed and conclusions arrived at are those of the author and are not necessarily to be attributed to the NRF.

I also remember all of those who have passed away.

ii ABSTRACT

This study posits that although the South African government has shown an unprecedented commitment to acknowledging and upholding the human rights of bisexual men and women, negative attitudes exist towards bisexual men and women. A mixed method research study was conducted, consisting of three phases. In the first phase of the study (quantitative), a biographical questionnaire, the Attitudes Regarding Bisexuality Scale (ARBS-FM) (Mohr & Rochlen, 1999), and the Sexual orientation scale of Berkey, Perelman-Hall and Kurdek (1990) was utilised. A total of undergraduate 1 459 students were selected by convenience sampling. However, the reliability of the scales of the ARBS-FM in this student sample appeared to be significantly lower than desired. Therefore, the researcher organised an expert group (N = 11). This second section of the study served as expert advice from cultural interpreters that brought about adaptation of the measures (qualitative) and was complementary to the quantitative study. The researcher endeavoured to seek elaboration and clarification of the results from one method with the results of the other method. In the third phase of the study a revised ARBS (consisting of two scales measuring attitudes towards bisexual men and women separately), a biographical questionnaire, and the sexual orientation scale was administered. A total of 578 undergraduate students were selected by convenience sampling. The psychometric properties of the revised ARBS instruments were investigated by means of confirmatory factor analyses. Results indicated a two-factor structure (Tolerance and Stability) for both of the ARBS-F and ARBS-M. Tucker’s coefficients of congruence showed that attitudes towards bisexual men and women may be measured by an integrated scale. Differential item functioning (DIF) and differential test functioning (DTF) analyses suggested the removal of some items that favoured either the black or the white group. The revised ARBS- F and revised ARBS-M were used for testing the stated hypotheses. The results indicated that participants’ attitudes towards bisexual men and women vary on a range of factors such as gender, religiosity, contact, and the sexual orientation of

iii the participants. Both heterosexual and homosexual students have more negative attitudes towards bisexual men and women than bisexual students. This double discrimination by heterosexuals and the gay and lesbian community is seldom recognised or acknowledged. Therefore, the current research findings elucidate that this oppression is real, and may have negative psychological effects on bisexual men and women. The results are discussed against the background of previous studies, and suggestions for future research are made.

Key words: bisexual men, bisexual women, biphobia, homophobia, heterosexism, sexual orientation, university community, survey, attitudes, students.

iv TABLE OF CONTENTS

Acknowledgements ii Abstract iii Table of contents v List of tables xii List of figures xv

CHAPTER 1 INTRODUCTION, RATIONALE AND AIM OF THE STUDY

1.1 Introduction 1 1.2 Sexuality in the South African context 2 1.3 Rationale of the study 5 1.4 Aim of the study and research questions 10 1.5 Terminology 13 1.6 Preview of the contents of the following chapters 16

CHAPTER 2 BISEXUALITY: A HISTORICAL AND CROSS-CULTURAL INQUIRY

2.1 Introduction 17 2.2 A historical inquiry 17 2.2.1 Classical Rome and Greece 18 2.3 The psychopathological and psychological debate regarding bisexuality 22 2.4 The gay liberation and women’s movement 29 2.5 A positive approach to sexuality 30 2.6 Bisexual identity development 32 2.7 Defining and measuring sexual orientation 40

v 2.8 A cross cultural inquiry 48 2.8.1 Cross-cultural studies and bisexuality 48 2.9 Chapter summary 51 2.10 Preview of the contents of the following chapter 51

CHAPTER 3 TWO LIVES TO LEAD: BISEXUAL MEN AND WOMEN

3.1 Introduction 52 3.2 The prevalence of bisexuality 52 3.3 Bisexual men and women 57 3.3.1 The basis of sexual attraction in bisexual men and women 57 3.3.2 Similarities, differences, and milestone events in the lives of bisexual men and women 59 3.4 Situational bisexuality 62 3.4.1 Introduction 62 3.4.2.1 Same-sex boarding schools 63 3.4.2.2 The prison environment 63 3.4.2.3 The military service 66 3.4.2.4 Prostitution and bisexuality 68 3.5 Chapter summary 69 3.6 Preview of the contents of the following chapter 70

CHAPTER 4 ATTITUDES TOWARDS BISEXUAL MEN AND WOMEN

4.1 Introduction 71 4.2 Attitudes and the Social Identity Theory 71 4.3 Attitudes and bisexual invisibility 75 4.3.1 Attitudes towards bisexual men and women 80

vi 4.3.2 Internalised biphobia 91 4.3.3 The psychological health of bisexual men and women 93 4.4 Chapter summary 98 4.5 Preview of the contents of the following chapter 99

CHAPTER 5 RESEARCH METHOD

5.1 Introduction 100 5.2 Aim of the study 100 5.3 A three phase research method 100 5.4 Research questions 101 5.5 Hypotheses 101 5.6 Phase one: Quantitative research method 102 5.6.1 Procedure 102 5.6.2 Participants 103 5.6.3 Instruments utilised in the first phase 103 5.6.3.1 Attitudes regarding Bisexuality Scale (ARBS-FM) 104 5.6.3.2 Sexual orientation instruments utilised in the first phase 106 5.6.3.3 Biographical questionnaire utilised in the first phase 108 5.7 Phase two: Expert advice 112 5.7.1 The aim of the expert group 114 5.7.2 The role of the researcher and ethical considerations 115 5.7.3 Participants of the expert group 115 5.7.4 Thematic analysis of the expert advise group discussion 116

5.8 Phase three: Quantitative research method 117 5.8.1 Procedure 117 5.8.2 Participants 117 5.8.3 Biographical questionnaire 118 5.8.4 Instruments administered in the third phase 118

vii 5.8.4.1 Revised version of the ARBS-FM 119 5.8.4.2 Revised version of the categories of sexual orientation 120 5.9 Statistical analyses 120 5.10 Chapter summary 122 5.11 Preview of the contents of the following chapter 122

CHAPTER 6 RESULTS

6.1 Introduction 123 6.2 Phase one: Statistical findings of the quantitative study 123 6.3 Phase two: Findings of the focus group 124 6.3.1 Discussing bisexuality 125 6.3.2 Discussing the ARBS-FM 126 6.3.3 Discussing the categories of sexual orientation 127 6.4 Phase three: Statistical findings of the quantitative study 128 6.4.1 Confirmatory factor analysis of the revised ARBS-F 128 6.4.2 Confirmatory factor analysis of the revised ARBS-M 133 6.4.3 Structural equivalence of factors 137 6.4.4 Reliabilities of the revised ARBS-F and ARBS-M subscales 137 6.4.5 Correlations between male and female subscale scores 138 6.5 Differential item functioning (DIF) and differential test functioning (DTF) 140 6.6 Hypothesis one 144 6.6.1 Investigating differences between black and white students regarding their performance on the revised ARBS-F 144 6.6.2 Investigating differences between black and white students regarding their performance on the revised ARBS-M 145 6.7 Hypothesis two 146 6.7.1 Investigating differences between male and female students regarding their performance on the revised ARBS-F 146

viii 6.7.2 Investigating differences between male and female students regarding their performance on the revised ARBS-M 147 6.7.3 Effect of ethnicity on gender differences 149 6.8 Hypothesis three 150 6.8.1 Investigating differences between students that know or do not know a bisexual man or woman regarding performance on the revised ARBS-F 150 6.8.2 Investigating differences between students that know or do not know a bisexual man or woman regarding performance on the revised ARBS-M 152 6.8.3 Effects of ethnicity on knowing and not knowing a bisexual man or woman 153 6.9 Hypothesis four 154 6.9.1 Investigating differences between students knowing and not knowing a lesbian woman or gay man regarding performance on the revised ARBS-F 154 6.9.2 Investigating differences between students knowing and not knowing a gay man or lesbian woman regarding performance on the revised ARBS-M 155 6.9.3 Effects of ethnicity on knowing and not knowing a gay man or a lesbian woman 156 6.10 Hypothesis five 157 6.10.1 Investigating differences between the religiosity groups regarding performance on the revised ARBS-F 157 6.10.2 Investigating differences between the religiosity groups regarding performance on the revised ARBS-M 158 6.10.3 Effects of ethnicity on religiosity 160 6.11 Hypothesis six 160 6.11.1 Investigating differences between the four sexual orientation categories’ performance on the revised ARBS-F 160 6.11.2 Investigating differences between the sexual orientation

ix categories’ performance on the revised ARBS-M 163 6.11.3 Effects of ethnicity on the four sexual orientation categories 165 6.12 Chapter summary 166 6.13 Preview of the contents of the following chapter 167

CHAPTER 7 DISCUSSION, LIMITATIONS AND RECOMMENDATIONS

7.1 Introduction 168 7.2 Phase one 168 7.3 Phase two 169 7.4 Phase three: Psychometric properties of the revised instrument 169 7.5 Hypotheses 170 7.5.1 Hypothesis one 170 7.5.2 Hypothesis two 171 7.5.3 Hypothesis three 174 7.5.4 Hypothesis four 176 7.5.5 Hypothesis five 177 7.5.6 Hypothesis six 178 7.6 Limitations 180 7.7 Recommendations 182 7.7.1 Measuring instruments 182 7.7.2 Qualitative research 183 7.7.3 Research participants 184 7.7.4 Improving the university environment for bisexual men and women 185 7.7.5 Counselling bisexual men and women 187 7.8 Conclusion 188 References 191 Appendices

x List of Tables

Table 5.1 Expert Group Participants 115 Table 6.1 Reliabilities of ARBS Subscales for the Total Group and the Black and White Groups 124 Table 6.2 Rotated Factor Pattern Matrix of the Revised ARBS-F for the Black Students 131 Table 6.3 Rotated Factor Matrix of the Revised ARBS-F for White Students 132 Table 6.4 Rotated Factor Pattern Matrix of the Revised ARBS-M for Black Students 135 Table 6.5 Rotated Factor Pattern Matrix of the Revised ARBS-M for White Students 136 Table 6.6 Coefficients of Congruence of the Factor Solution of the Revised ARBS-F and ARBS-M Subscales 137 Table 6.7 Reliabilities of Revised Scales for the Total Group and the Black and White Groups 138 Table 6.8 Correlations between Female and Male Subscales for the Black and White Groups 139 Table 6.9 DIF Statistics for the Female Tolerance Scale 140 Table 6.10 DIF Statistics for the Female Stability Scale 141

xi Table 6.11 DIF Statistics for the Male Tolerance Scale 142 Table 6.12 DIF Statistics for the Male Stability Scale 143 Table 6.13 Means and Standard Deviations for Ethnic Groups on the Revised ARBS-F 145 Table 6.14 Means and Standard Deviations for Black and White Students’ Responses on the revised ARBS-M 145 Table 6.15 Means and Standard Deviations for Male and Female Students’ Responses on the Revised ARBS-F 147 Table 6.16 Means and Standard Deviations for Male and Female Students’ Responses on the Revised ARBS-M 148 Table 6.17 Means and Standard Deviations for Students that Know or Do Not Know a Bisexual Man or Women on the Revised ARBS-F 151 Table 6.18 Means and Standard Deviations for Students that Know or Do Not Know a Bisexual Man or Women on the Revised ARBS-M 152 Table 6.19 Means and Standard Deviations for Students that Know or Students that Do Not Know a Gay Man or Lesbian Woman on the Revised ARBS-F 154 Table 6.20 Means and Standard Deviations for Students that Know or do not Know a Gay Man or Lesbian Woman on the Revised ARBS-M 155

xii Table 6.21 Means and Standard Deviations for Religiosity on the Revised ARBS-F 157 Table 6.22 Means and Standard Deviations for Students’ Responses for Religiosity on the Revised ARBS-M 159 Table 6.23 Means and Standard Deviations for Sexual Orientation Groups on the Revised ARBS-F 161 Table 6.24 Means and Standard Deviations for Sexual Orientation Groups on the Revised ARBS-M 163

xiii List of Figures

Figure 6.1 Revised ARBS-F scree plot for the black group. 129 Figure 6.2 Revised ARBS-F scree plot for the white group. 130 Figure 6.3 Revised ARBS-M scree plot for the black group. 134 Figure 6.4 Revised ARBS-M scree plot for the white group. 134 Figure 6.5 Interaction of ethnicity and gender on Female Tolerance. 150

xiv CHAPTER 1

INTRODUCTION, RATIONALE AND AIM OF THE STUDY

1.1 Introduction

As a society, we are fascinated by sexuality. Despite this fascination we have traditionally looked askance at people who collect information on sexuality. Historically, the study of sexuality has been stigmatised - anyone who undertook research in this area risked both their career and their personal reputation (Rust, 2000). As a result research on sexuality was limited and value laden. However, the field of research surrounding sexuality shifted in the 1960s, 1970s, and the 1980s.

The stigma surrounding research on sexuality abated as sexuality entered the political arena in the late 1960’s in the form of debates regarding issues such as birth control and abortion. In addition, the removal of “homosexuality” from the Diagnostic and Statistical Manual of the American Psychiatric Association in 1973 opened the door for research on sexual orientation.

The AIDS epidemic in the 1980’s forced researchers to realise the cost of sexual ignorance (Matteson, 1997). The AIDS epidemic also brought personal and political implications to the fore. Researchers not only theorised and wrote about how sexuality would change in the age of AIDS, they also worked with gay men to chart an alternative response to the AIDS epidemic - one based explicitly on gay liberation and implicitly on feminism. Together these researchers participated in a vociferous debate about the meanings and practices of health as well as the role that gender politics could play in the sexual liberation movement (Matteson, 1997).

According to Maddison (2000), the gay liberation movement is indebted to the women’s movement. The politics of sexuality was neither generated nor popularised by the gay liberation movement. It was the women’s movement of the 1970’s who initially challenged the notion of gender-as-conformity

1 (Hassim, 1999). The abovementioned was also evident in South Africa, as gay men and lesbian women commenced their struggle for equality and liberation.

1.2 Sexuality in the South African context

Lesbian women and gay men have historically been the victims of discriminatory laws and prejudices in South Africa. Because of this history of division and resistance, the experiences of homosexuals in South Africa are reported to be unique (Gevisser & Cameron, 1994). Until 1994 South Africa was ruled by a minority government, which founded a separatist ideology based on conservative theology, barring all expressions of sexuality outside a same-race heterosexual marriage. Therefore, prior to 1994 the law denied lesbians and gay men their basic human rights and reduced them to social outcasts and criminals (Gevisser & Cameron, 1994).

As late as 1977, the Criminal Procedure Act listed sodomy as a Schedule 1 offence, giving police broad powers to investigate cases and make arrests without warrants; allowing the state to intercept letters and other private communications in sodomy investigations; and disqualifying people convicted of sodomy - or their dependants - from receiving pensions (Retief, 1994). A 1987 parliamentary report defined homosexuality as an acquired behavioural pattern, a serious social deviation, and an evil. As a result of this, police would conduct random raids, bursting into a party or club, grabbing people who were kissing or dancing together, and bundling them into police vehicles. Photographers would line people up against the wall and “snap pictures of as many faces possible while policeman took down numbers of the cars parked outside; exposure could have meant unemployment, social isolation and vitriolic abuse wherever one went” (Retief, 1994, p.101). The unravelling of the legal and social stigma attached to homosexuality did not commence until 1993.

Cameron (1993), a South African judge on the Supreme Court of Appeal, listed the following potential consequences for lesbian and gay people in

2 obtaining protection in the final constitution: the decriminalisation of gay sex acts by abolishing the common law crimes of sodomy and “unnatural sexual offences” as well as the provisions of the Sexual Offences Act which also criminalised such acts; a uniform age of consent for lesbians and gay men and heterosexuals; guarantees of free speech and association; and laws against discrimination based on sexual orientation, including areas such as employment, tenancies, provision of public resources and services (Cameron, 1993).

The interim constitution, agreed upon by the main political parties, provided the basis for the 1994 elections - the first free vote in South Africa’s history. Following the election, the new parliament took on the dual functions of being both a constitutional assembly and having to draft a final constitution. Although all articles were subject to debate, the final constitution had to conform to thirty-three fundamental principles - including non-discrimination on the basis of sexual orientation (Gevisser & Cameron, 1994).

As part of this process in 1995, the AIDS Law Project at the University of the Witwatersrand commissioned the Human Sciences Research Council to complete an investigative report on the South African public’s attitudes on issues concerning lesbians and gay men (Reid & Dirsuweit, 2001). The survey took place at a time when the inclusion of “sexual orientation” in the constitution was a contested issue. Public opinion was split on giving homosexuals equal rights in the constitution, with 44% of the public against and 38% in favour thereof. On an attitude index, 48% of the public was rated anti-gay, 34% pro-gay and 18% undecided (Reid & Dirsuweit, 2001).

Regardless of the preceding discriminatory laws and divided public opinion, “sexual orientation” was included in the equality protections of the South African interim constitution in 1993, making South Africa the first country in the world to include that status in its Bill of Rights (Harper & Schneider, 2003).

The Equality Clause of article 9 of the constitution was adopted on the 8th of May 1996, and holds as follows:

3 (3) The state may not unfairly discriminate directly or indirectly against anyone on one or more grounds, including race, gender, sex, pregnancy, marital status, ethnic or social origin, colour, sexual orientation, age, disability, religion, conscience, belief, culture, language and birth - (4) No person may unfairly discriminate directly or indirectly against anyone on one or more grounds in terms of subsection (3). National legislation must be enacted to prevent or prohibit unfair discrimination.

Therefore, the South African government has shown unprecedented commitment to acknowledging and upholding the human rights of sexual minorities (Human Rights Watch, 2003). This commitment is evident by the fact that same-sex marriages became legal in South Africa on 30 November 2006. South Africa became the fifth country, and the first in Africa, to legalise same-sex marriage.

The Lesbian and Gay movement in South Africa appears to have progressed much faster, and made more progress with regards to equality and non- discriminative laws, than other Lesbian and Gay movements worldwide. Conversely, Sinclair (2005) who investigated homosexuality in the country from the 1960’s to 2000 maintains that South Africa’s Lesbian and Gay movement did not achieve much of what it set out to do, unlike the broader liberation struggle and to a lesser degree the women’s movement. Sinclair (2005, p. 320) states that the “in-fighting within gay organisations and their inability to unify, rendered the liberation movement in South Africa as ultimately unsuccessful”. This paradox is elucidated by Thoreson (2008) who purports that in the apartheid era; human rights were paramount as opposed to gay and women’s rights. The progress made by the gay liberation movement in South Africa is related to the fact that the South African Lesbian and Gay movement successfully made stable political alignments which allowed it to concentrate on lobbying and litigation, where it has compellingly argued that its own agenda of liberation and equality dovetails with that of the ruling party (Thoreson, 2008).

4 The recognition of homosexuality as a legitimate sexual orientation in South Africa can be viewed as progress from the perspective of both the scientific study of sexuality and of sexual politics. Even though research expertise in South Africa on sexuality is limited by a lack of interest in the field (Potgieter, 1997) various researchers have contributed in this field. In addition, the South African Lesbian, Gay, Bisexual and Transgender (LGBT) service organisations and their related programmes address and publish articles on the difficulties and injustices that sexual minorities faced in South Africa (cf. Epprecht, 2006; Gevisser & Cameron, 1994; Polders, 2007; Polders & Wells, 2004; Potgieter, 2003; Potgieter, 2005; Reddy, Potgieter & Mkhize, 2007; Reid & Dirsuweit, 2001; Stobie, 2004). However, according to Polders and Wells (2004, p. 2) research completed in South Africa has been largely of a “qualitative nature, with very few quantitative studies addressing issues faced by sexual minorities in South Africa”. Additionally, existing studies in South Africa have dealt mainly with white middle-class gay men (Polders & Wells, 2004).

1.3 Rationale of the study

Research over the past years has provided us with much information on the correlates of individuals’ attitudes toward and stereotypes regarding lesbian women and gay men (Arndt & De Bruin, 2006; Evans & D’Augelli, 1996; Herek, 1995; Lance, 2002). During the past 24 years, Herek and his colleagues have published 51 articles in peer-reviewed academic journals relating to attitudes towards lesbian women and gay men (Herek, 2008).

On the contrary, prior to 1980, bisexuality was neither a topic of contention nor of research interest (Rust, 2000). However, in the 1980’s, bisexuality became a “serious topic of discussion in the gay and lesbian press” (Rust, 1996, p. 12). In the early 1990’s, a wave of important work in feminist theory addressed the issue (George, 1992; Gerber, 1995; Rust, 1996; Weise, 1992). Along with increased visibility, however, came increased resistance from both gay and lesbian communities and the heterosexual mainstream. While the increasing recognition of homosexuality as a legitimate sexual orientation can

5 be viewed as progress, it is ironic that as gay men and lesbian women are becoming a visible minority, another invisible minority are taking their place in the struggle for recognition and legitimacy: people who identify themselves as bisexuals.

In comparison to the now substantial amount of literature on attitudes toward lesbian women and gay men, there are a limited number of studies outside the United States of America on attitudes towards bisexuality (Herek, 2002). When bisexuality is mentioned in research, the inclusion is often minimal. In most articles based on empirical research, the authors simply acknowledge that bisexuals were included in their samples of lesbian women and gay men by adding “and bisexual” to their sample descriptions and keyword list (cf. Holtzen, 1994; Ludwig & Brownell, 1999). For example, in a recent article titled “Relations among dimensions of feminism and internalized heterosexism in lesbians and bisexual women”, Szymanski (2004) stated that lesbians and bisexual women were combined for analyses in this study.

Another example of the neglect of bisexuality in research is that bisexuals are mentioned on the cover and in the introductory and concluding articles of the 1989 special issue of The Journal of Counseling and Development, “Gay, lesbian and bisexual issues in counseling”. Yet, of the 18 articles comprising this issue, only three mention bisexuals and not one article specifically addresses bisexuality. Similarly, the special issue of The Counseling Psychologist, “Counseling lesbian women and gay men” (1991) does not even include bisexuals. More recently, in the special issue of The Counseling Psychologist, “Internalised Heterosexism” (2008), the authors once again simply acknowledge that bisexuals were included in their samples of lesbian women and gay men by adding “and bisexual” to their sample descriptions and keyword list.

Those wishing to find research on bisexuality must sort through research and writings on gay and lesbian issues to find information relating specifically to bisexuality. The attitudes towards bisexual men and women, as well as their

6 needs and the difficulties that they are facing are often subsumed under larger discussions of gay and lesbian issues (Ryan, Brotman & Rowe, 2000).

Hansen and Evans (1985) ascribe the lack of acknowledgment and research on bisexuality not only to the hegemony of the dichotomous model (homosexual or heterosexual) of sexuality, but also to cultural erotophobia and the law of the excluded middle. Observing that bisexuals are stereotyped as promiscuous, Hansen and Evans (1985) claim that the mishandling of bisexuality stems partially from erotophobia, that is, cultural anxiety over sexual enjoyment unrestrained by cultural norms such as normative fidelity.

The law of the excluded middle describes the way in which individuals are classified sexually. Individuals are presumed heterosexual in the absence of evidence to the contrary, but an individual who engages in one homosexual act is classified as homosexual. Classification as a homosexual is maintained as any heterosexual behaviour subsequently is considered counterfeit (Hansen & Evans, 1985). This process of classification elucidates how multifaceted sexual experience is obliged to fit into dichotomous notions of sexuality, thereby rendering the coexistence of heterosexuality and homosexuality within an individual conceptually impossible (Rust, 2000).

The conceivability of bisexuality as a combination of hetero- and homosexuality, and the simultaneous inconceivability of such a combination, underlie contemporary attitudes toward bisexuality. The attitudes toward bisexual men and woman reflect the paradox of their roots. Foremost among these attitudes is the belief that bisexuality does not exist (Ochs, 1996). An article titled Straight, gay or lying? authored by Benedict Carey, appeared on the front page of the New York Times on the 5th of July 2005. The article was based on a study on bisexuality in males published in Psychological Science (Rieger, Chivers & Bailey, 2005). In this study a team of psychologists measured genital arousal patterns in responses to images of men and women. The psychologists found that men who identified themselves as bisexual were in fact exclusively aroused by either one sex or the other. The

7 article concludes that the study casts doubt on whether bisexuality exists, at least in men.

Esterberg (1997) posits that bisexual men and women are often assumed to be lesbian women or gay men who are denying their true sexuality because they are going through a transitional phase of coming out, because they are afraid to face their own and others’ homophobia, or because they are unwilling to shoulder the burden of being a member of an oppressed minority. Furthermore, individuals who engage in same-sex activity are often believed to be heterosexuals who are merely experimenting. According to Rust (1995), these invalidating beliefs are a cultural phenomenon which functions to preserve the belief that bisexuality does not exist, and as a result, limits the research on the attitudes that the heterosexual and homosexual communities have towards bisexual men and women.

Due to the lack of literature on attitudes towards bisexual men and women, the best available information comes from personal statements of bisexual men and women who described their experiences in anthologies or to researchers (Weinberg, Williams & Pryor, 1994). Bisexual individuals have reported lack of validation, isolation, and discrimination, lack of credibility, ostracism, negative attitudes and stereotyping from both heterosexual and lesbian and gay communities (Hutchins & Ka’ahumanu, 1991). Stereotypes abound in popular culture, with bisexuals widely perceived by heterosexual and lesbian and gay individuals as promiscuous, unable to have committed relationships or are likely to contract sexuality transmitted diseases (Eliason, 2001; Rust, 1995; Spalding & Peplaum, 1997). This was also evident in South Africa during the Zuma Rape Trial (Skeen, 2007). According to Skeen (2007), the most frequent method to deride a woman’s credibility is by proving the complainant’s promiscuity. During the trial, Ms. Khwezi’s HIV status, her sexual history, as well as her sexual orientation was widely publicised in the media. In addition, Judge van der Merwe proclaimed during the trial that according to him, Ms. Khwezi was not a lesbian, but that she was “at the very least bisexual” (Skeen, 2007, p. 35).

8 More recently, researchers have taken interest in the concomitant negative impact of negative attitudes on the lives of bisexual men and women (Firestein, 1996; Rust, 2000; Weber, 2008). The research findings confirm the information provided by bisexual men and women that report isolation, ostracism, stereotyping and discrimination from both the homosexual and heterosexual communities (Rust, 2000).

The Social Identity Theory (SIT; Tajfel & Turner, 1979) that underlies this study appears to be appropriate as it focuses on intergroup conflict and discrimination associated with group membership. According to Sidanius, Levin, Rabinowitz and Federico (1999), the Social Identity Theory is one of the dominant paradigms in investigations of human social interaction. The Social Identity Theory that originated in the discipline of psychology deals with the structure and function of identity as it relates to people’s membership in groups. This theory has been utilised extensively in the field of psychology (cf. Foley, Ngo & Loi, 2006; Lipponen & Leskin, 2006; Liss, Crawford & Popp, 2004;; Oswald & Lindstedt, 2006; Ostrove & Long, 2007; Ramsy et al., 2007; Susskind & Hodges, 2007; Vealey, 1997), and more specifically in the field of research on sexuality (cf. Clark & Maas, 1987; Cox & Gallois, 1996; Hegarty & Massey, 2006; Kristiansen, 1990; Stevens, 2004; Weber, 1990).

The Social Identity Theory (Tajfel & Turner, 1979) assumes that attitudes are shaped by an individual’s membership in social groups and the evaluations of one’s group in social comparison to specific other groups. In terms of this theory, the homosexual community views itself as superior to the heterosexual community, the heterosexual community views itself as superior to the homosexual community, and both these groups view themselves as superior to the bisexual community. Therefore, bisexuals see themselves as twice-rejected, once by the heterosexual community and again by the homosexual community.

This “double discrimination” is seldom recognised or acknowledged as a force of external oppression, yet this oppression is real and has many damaging effects on bisexual men and women (Firestein, 1996). Weber (2008) states

9 that bisexual individuals may experience minority stress complicated by marginality from both the straight and gay communities. “The marginalisation usually includes same-gender orientated friends urging them to adopt a gay lifestyle and heterosexual orientated friends pressuring them to conform to heterosexual standards” (Weber, 2008, p. 31). Shuster (1991, p. 267) states that due to this “double discrimination”, bisexual men and women frequently experience themselves as existing in two different worlds, and not fully fitting into either - “a feeling of political and personal homelessness”.

It is thus imperative to ascertain attitudes towards bisexual men and women in order for health workers to effectively and ethically serve bisexual clients (Firestein, 1996). Cochran (2001, p. 941) states that it would be “unconscionable for psychology as a profession to disregard the possibility of differences associated with sexual orientation in both mental health morbidity and treatment experience”. Davis and Wright (2001) conclude that bisexual men and women are a high need group and recommend that research be undertaken to investigate the needs of bisexuals as well as the difficulties that they experience. Lastly, the guarantees in the South African constitution have opened the opportunity for the discussion of sexual practices which were previously outlawed. In this climate, South Africa is well placed as a site to investigate attitudes towards bisexuality.

1.4 Aim of the study and research questions

The aim of this study was to measure attitudes towards bisexual men and women in the South African context. More specifically, the study aims to measure South African university students’ attitudes towards bisexual men and women. Although it was not the aim of this study to measure attitudes towards bisexual students per se, but rather bisexuality in general, the university setting appears to provide an appropriate context to investigate attitudes towards bisexual individuals. The reason for the aforesaid is that given the age that individuals engage in bisexual behaviour, have bisexual attractions and identify as bisexual, together with the findings of studies by Klein (1990), Herek (1995) and Evans and Broido (1999) that the university

10 years are a common time for individuals to come out and identify themselves as bisexual it appears to be an ideal time to conduct research into bisexuality. Hoburg, et al., (2004) in the United States of America, the results indicated that 29% to 32% of women university students and 12% to 19% of university men have reported having sexual feelings or preference for both men as well as women. In addition, during university years homosexual and heterosexual students often appear to have their initial contact with students that identify themselves as bisexual (Hoburg, Konik, Williams & Crawford, 2004). The results of this study would not only bear relevance to attitudes towards bisexual men and women in general, but also to bisexuality amongst the student population.

Given the prevalence of bisexuality amongst university students the psychosocial development of these students also needs to be taken into consideration. Chickering and Reissner’s (1993) theory of seven developmental vectors represents the most current, inclusive and widely used theory of student psychosocial development. The vectors include the developmental tasks of developing competence, managing emotions, moving through autonomy toward interdependence, developing mature interpersonal relations, establishing a sexual identity, developing purpose and integrity. A close examination of these vectors reveals theoretical assumptions that place bisexual students at considerable disadvantage in terms of achieving resolution of developmental tasks. According to Fassinger (1995), one might ask how an individual can possibly establish a secure identity in the face of marginalisation, invisibility and societal erasure. Moreover, one might expect the developing purpose - clarifying vocational, avocational and lifestyle plans and goals - to be ignored or stalled in the face of more pressing identity concerns related to safety, acceptance and belonging. Finally, one might ask how the process of developing a personally relevant set of beliefs and values by which to live - such as integrity - is influenced by invisibility and stigmatisation of one’s lifestyle.

As a result, this research was essentially exploratory, even though international research will be used to guide the building of hypotheses and the

11 interpretation of the data. In light of the above discussion, the primary purpose of this study was to explore the attitudes towards bisexual men and women. In brief, this study attempted to seek answers to the following research questions:

●Do homosexual, heterosexual and bisexual students have negative attitudes towards bisexual men and women?

●If negative attitudes exist, what are the most significant variables related to biphobia?

Hypotheses stated were not intended as a complex or exhaustive list of all possible hypotheses. Rather, they were used as heuristic or exploratory devices, in which the exploration of attitudes might profitably direct future research efforts and enhance our understanding of students’ attitudes towards bisexual men and women in South Africa. In view of the literature review, and the seemingly lack of research on attitudes towards bisexual men and women, the relationships between respondents’ attitudes toward bisexual individuals and various variables were explored. General hypotheses were made, avoiding the direction, due to the lack of supporting research regarding attitudes towards bisexual men and women in South Africa.

The initial statistical analyses of the data pertaining to these questions yielded unreliable results, with clear differences between how black and white students interpreted the items of the Attitudes Regarding Bisexuality Scale (ARBS-FM). Based on these findings the research was expanded to provide for the use of a more reliable instrument to measure the attitudes towards bisexual men and women.

Subsequently, a second phase in the research was introduced. An expert group (Nassar-McMillan & Borders, 2002) was convened to obtain more information concerning the black and white students’ understanding of the concept of bisexuality as well as their interpretation of the items of the ARBS- FM. After analysing the findings and recommendation of the group another

12 survey (quantitative phase) was conducted utilising a revised ARBS-F and ARBS-M in order to test the hypotheses.

1.5 Terminology

Although a list of terminology relevant to this study is provided, it should be noted that when describing another author’s theoretical or empirical work, the researcher utilises the terms used by that author in order to remain as true as possible to the original research. Furthermore, the exact definitions of many of the terms are themselves topics of intense debate (Bassow, 1992). However, as a guide, the following explanations will clarify how these terms are to be interpreted in the present study.

The construct attitude is an established term amongst scientists and psychologists. The central feature of all definitions of attitude, according to Allport (1954), is the idea of readiness for response. That is, an attitude is not behaviour, not something that a person does; rather it is a preparation for behaviour, a predisposition to respond in a particular way to the attitude object. The term attitude object is used to describe attitudes towards things, people, places, ideas, actions or situations, either singular or plural, and refers to bisexual men and women in the current study.

Sexual orientation is one component of a person's identity, which also comprises many other components, such as culture, ethnicity, gender, and personality traits. Sexual orientation is an enduring emotional, romantic, sexual, or affectional attraction that a person feels towards another individual (Le Vay, 1993). Sexual orientation falls along a continuum - someone does not have to be exclusively homosexual or heterosexual, but can feel varying degrees of attraction for both genders. In the current study, sexual orientation is defined as the direction of sexual feelings or behavior toward individuals of the opposite sex (heterosexuality), the same sex (homosexuality), or the combination of the two (bisexuality) (Le Vay, 1993). According to Sell and

13 Petrulio (1995), the terms heterosexual, homosexual and bisexual are commonly used terms by researchers to describe sexual orientation.

Throughout the thesis the researcher utilises the word bisexual as an adjective to describe sexual behaviour when it involves sex acts with both same sex and other sex partners during a given period of time. When the researcher uses the terms bisexual women, bisexual men, or bisexuals, the researcher may be writing both about people who identify with this label and those whose attractions are to persons of more than one sex or gender, whether or not they choose the bisexual label, or any label at all. The researcher also uses bisexuality to describe sexual attractions to both sexes, and to describe individuals who engage in bisexual behaviour, have bisexual attractions, or identify as bisexual (Firestein, 1996; Rust, 2000).

Biphobia encompasses a range of negative feelings geared towards and expressed in regards to either bisexual persons or the bisexual community as a whole. It refers to the fear of, discrimination against, or hatred of bisexuals. It needs not include homophobia, because there appear to be stereotypes that are specific to bisexual men and women. These include, but are not limited to, promiscuity, polygamy, and being confused (Rust, 2000).

Homophobia means negative attitudes towards lesbian women and gay men (Firestein, 1996). Homophobia is “any belief system, which supports negative myths and stereotypes and any of the varieties of negative attitudes which arise from fear or dislike of lesbians or gay men” (Martin, 1982, p. 341).

The word homosexual refers to sexual behaviour or attraction when it occurs exclusively with or toward people of the same sex. In general, except when the author is conveying another author’s research, the terms lesbian woman and gay man are used as far as possible rather than homosexual. This is intended to reduce the perpetuation of negative attitudes associated with the word homosexual and to clarify its ambiguous nature, as it has been used both to refer to men exclusively and lesbian women collectively.

14 The term coming out is used to signify acknowledging one’s homosexual attractions, contacting and participating in the gay and lesbian or bisexual community, and identifying oneself to others as being gay, lesbian or bisexual. Closeted refers to individuals that do not disclose their sexual orientation to others (Dworkin, 2000).

Heterosexism is an ideological system that denies, denigrates, and stigmatises any non-heterosexual form of behaviour, relationship or community. It is the belief that heterosexuality is the only valid and acceptable form of sexual identity and expression (Rust, 2000),

Sex refers not only to biological sex (i.e. male or female) but also to the act of having sex (i.e. sexual intercourse). Gender is generally taken to refer to the biological distinction between males and females. Unless clearly indicated by either the meaning of the sentence or section it appears in, gender can be taken to be interchangeable with biological gender. The terms male and female are taken to refer to individuals who belong to the social-biological category of either male or female, and are used interchangeably with the terms man and woman, respectively. This is a necessary simplification of the study’s terminology, though it is acknowledged that much social influence, variance, and construction attends the identification of individuals as either male or female (Howard & Hollander, 1997).

Religiosity is nominally defined as the belief in and reverence for a supreme, supernatural being or beings, often expressed by worship, ritual, and a certain way of life. Religiosity involves religious identity (religious group affiliation, religious self-identity), behaviour (attendance at religious services or individual religious practices), attitudes (importance of religion), perceptions (religion’s negative sanctions against certain behaviours), practices (adherence to rules and sanctions) (Hackney & Sanders, 2003).

15 1.6 Preview of the contents of the following chapters

In Chapter Two a literature review pertaining to the historical perspective of bisexuality, theoretical models regarding bisexuality, and cross-cultural evidence of bisexuality is provided. This is followed by a review of literature of the experiences of bisexual men and women in Chapter Three.

In Chapter Four the underlying theoretical approach utilised in this study Social Identity Theory, is discussed. This is followed by research findings related to homosexual and heterosexual attitudes towards bisexual men and women.

Chapter Five provides a description of the research method. This description includes details of the participants and the instruments used in the first phase (quantitative research method), the second phase (expert advice group) and the third phase of the study (quantitative research method).

In Chapter Six the statistical findings of the three phases of the study are provided. This is followed by a discussion of the results in Chapter Seven, and suggestions for future directions and interventions in this area are explored. The limitations of the study are given, and these are followed by a conclusion.

16 CHAPTER 2

BISEXUALITY: A HISTORICAL AND CROSS-CULTURAL INQUIRY

2.1 Introduction

As with all sexualities, bisexuality has a history. Yet, this history has barely even begun to be told (Eadie, 1996). Expressions of sexuality have been shown to vary according to historical period, as well as between and within different societies and cultures. Culture shapes sexual beings, and what is perceived as normal, natural, true, good, bad, right or wrong is connected to culture specific norms, rules, values and expectations. Data from historical and cross-cultural research clearly indicates that bisexuality, defined as sexual attraction or sexual behaviour with persons of both genders, has existed throughout history in diverse cultures (Ford & Beach, 1951; Mead, 1975).

This chapter examines the historical inquiry into bisexuality. The pathological and psychological debate regarding bisexuality, as well as evidence of bisexuality in cross-cultural research findings are also discussed.

2.2 A historical inquiry

Homosexuality and bisexuality have been the focus of significant interest for historians, classical scholars, anthropologists and sociologists. There is extensive information about bisexual behaviour and relationships in classical Greece and Rome (Cantarella, 2002; Dover, 1978). However, the use of the terms homosexuality, heterosexuality and bisexuality are modern words for modern concepts that the Greeks and the Romans would not have understood. In this regard Cantarella (2002, p. viii) points out that for the Greeks and Romans “homosexuality was not an option falling outside the norm, a different or deviant decision. It was just one part of the existence of

17 life during one’s lifetime that would alternate and interweave (sometimes simultaneously) with the love of a woman”.

2.2.1 Classical Rome and Greece

According to Himelhoch (1988), the refutation of the bisexual potential has prevailed since the Christianisation of the Roman Empire. Prior to this, an individual’s sexual attraction to women and men was not a subject for comment. The bisexual potential existed unobtrusively. Among the celebrated achievements of Athens was its ethical ideal for balancing matrimony and same-sex friendships.

In Greece, from the earliest times that the sources allow researchers to examine, homosexual relationships were part of a life experience regulated by a series of social norms, which laid down the time scales, and etiquette of these relationships, and their alternation of heterosexual relationships. This evidently means that this experience was legitimate, both socially and legally (Cantarella, 2002).

The Greek view of female homosexuality differed from that of male homosexuality. The Greeks had very strong underlying reasons for denying female homosexuality. Firstly, female pleasure was something that had to be controlled. Secondly, women were believed to experience more pleasure than men in sexual intercourse, and lastly, they were seen as being unable to control themselves. Thus, female’s sexual lives required controlling, more specifically, the kind that would have led them to experience pleasure that was not at the service of men (Cantarella, 2002).

The Greeks were undoubtedly aware that men, apart from their heterosexual impulses, also had homosexual leanings and this understanding underpinned the culture of bisexual divinities in many areas of Greece (Delcourt, 1966). However, this did not mean that the Greeks were entitled to free expression of their bisexuality. The Greek man had to undergo his homosexual experience at the correct time, according to the right rules. In reaching the

18 age of majority, the Greek boy had to take on the active role of a lover, with both women and men. In other words, the Greek male, in the space of a few short years, underwent two divergent sexual initiations. The first of these was an educational sexual relationship with older men and as a passive partner in the relationship he was taught to learn and assume a role which the second initiation a few years later - a dominant sexual relationship with women - enjoined him to forget (Dover, 1978).

According to Cantarella (2002) the Greeks were bisexual in the sense that, when they were boys, a man loved them; while in the first years of their own adulthood, they preferred to make love to adolescent boys. Later in life, they chose women, and even when they were married, they were allowed to have male lovers. According to Foucault (1985), the Greeks were bisexual if we mean by this that a Greek could desire a man or a woman, and it was ordinary for a male to change to a preference for a woman after loving a boy in his youth.

However, if we turn our attention to how they conceived this dual practice, we need to be aware of the fact that they did not recognise two kinds of desire. According to Foucault (1985), we can talk about Greek bisexuality, thinking of the free choice they allowed themselves between the two sexes, and what made it possible to desire a man or a woman as simply the appetite that nature had implanted in man’s heart for beautiful human beings, whatever their sex might be.

Similar to the Greeks, the Romans also felt that it was normal for a man to have sexual relationships with other men as well as with women. However, unlike the Greeks, the Romans did not believe that it was educational for boys to be passive partners in homosexual relationships. The sexual mentality of the Roman male was that of an aggressive dominator (Cantarella, 2002).

The Roman male had to be able to express his sexual exuberance, which could not be satisfied solely by intercourse with women, whenever and

19 however he wished. To this end, he had two categories of available men: prostitutes and slaves. Male prostitution was not forbidden, and as for slaves, it was understood that they were obliged to satisfy their owner’s desires (Cantarella, 2002).

In Greece, anyone who loved a boy had to court him, flatter him, prove his love for him, and convince him of the seriousness of his intentions. For a Roman, this would show a lack of virility, as his psychology was that of a conqueror. This is evident with regards to Caesar, who had been the lover of Nicomedes, King of Bithymia. Yet, Caesar was the invincible soldier who conquered the world and seduced women (Highet & Highet, 1953).

For many centuries, homosexuality had not only been tolerated, but was considered a sign of virility. The pagan world considered sexual relations between men to be an important part of sexuality that did not exclude relationships with women, and considered them necessary, dutiful and desirable. However, this same model of sexuality considered sexual relationships between women to be a foul coupling. This could be because women were accorded a relatively inconsequential role in society and their sexuality was scarcely recognised when it was recognised, it was viewed as something that had to be controlled (Gagnon, 1977).

In the early 400’s A.D., Christianity began to introduce a new sexual code, focusing on religious concepts of “holiness” and “purity”. Christianity introduced the principle of “naturalness”, which was exclusive to heterosexual intercourse. The distinction between passivity and activity, which identified manliness with the assumption of the active sex role with either men or women, was contrary to the new state religion, which condemned homosexuality in all its manifestations (Cantarella, 2002).

Initially, homosexuality was harshly condemned. From 342 A.D. onwards, with Constantius and his brother Constans, the repression of homosexuality intensified. Theodosius I in 390 A.D., returned to the topic, and in 438 A.D., at

20 the order of Theodosius II, all passive homosexuals were destined to be burned alive. In the Theodosian Code, the text reads:

“All those who are accustomed to condemn their own manly body transformed into a womanly one, to undergo sexual practices reserved for the other sex, and who have nothing different from women, will pay for his crime among the avenging flames, in front of the people” (cited in Cantarella, 2002, p. 181).

However, Christian morality demanded more. Homosexuals, active or passive, committed the unforgivable, unnameable sin. It was crucial to ensure that all those who abandoned themselves to practices against nature should be punished. In 533 A.D., Justinian did precisely that: all homosexuals, regardless of which role they assumed, were condemned to death by him (Cantarella, 2002).

What were the causes underlying the escalation of repression? According to Cantarella (2002), the answer lies within the Christian teachings. By Justinian’s time, the patristic writers had worked out a theology of sexuality that condemned relations between persons of the same sex as being against nature. This illustrates the deep and radical changes that pagan ethics underwent on coming into contact with Christian culture. The early church, which placed a premium on asceticism and celibacy, condemned all sexual behaviour not directly connected to procreation (Szasz, 1970).

Imperial policy in the area of homosexuality was orientated and determined by the desire to impose Christian morality. This was first implemented with a degree of caution, by laying down penalties for passive homosexuals that safeguarded the ancient principles of morality and thereafter “wiping out every manifestation of homosexuality, always and everywhere, as intercourse against nature offended the Lord” (Cantarella, 2002, p. 210). Therefore, nature no longer permitted them any alternatives. The only sexual act in accordance with nature was heterosexual intercourse.

21 The 19th century saw the transformation of homosexuality from a vice (moral standards) into a “condition” - one that shifted from the act itself to the actor, from a specific point in time to a life history (Paul, 1985). The focus was fixed upon the person. The transformation of socially condemned sexual acts into signs of medical illness or pervasive biological abnormality led to the emergence of the “homosexual person” who differed constitutionally from the “normal” population (Paul, 1985). This initiated the psychological and psychopathological debate regarding sexual orientation.

2.3 The psychopathological and the psychological debate regarding bisexuality

Early psychoanalytic theorists believed that the human species evolved from a primitive hermaphroditic state to the gender-differentiated physical form of today and that individual physiological and psychological development parallels this evolutionary process (Ritvo, 1990). Bisexuality was termed “psychic hermaphroditism”, while those actively homosexual were labelled “inverts”, the “third sex” and the “intersexes” (Paul, 1985).

Krafft-Ebing (1831/1965), credited as the founder of modern sexual pathology, and the author of Psychopathia Sexualis, followed the medical orthodoxy of the day in attributing psychopathology in general, and sexual perversion in particular, to degeneracy. However, unlike Tissot (1832/1965), he did not attribute degeneracy to the loss of vital fluid by masturbation. The causal explanation that Krafft-Ebing preferred was hereditary taintedness in the family pedigree (Krafft-Ebing, 1831/1965).

Karl Ulrich initiated the first challenge to the claim of perversion by choice in 1864. Based on the self-knowledge of being homosexual, Ulrich (1864, cited in Haeberle, 1998) postulated an inborn causality, which he located in the brain. Although Krafft-Ebing accepted this proposal, he paid little attention to neuropathological sexology. Alfred Binet (1871, cited in Money, 2003) also challenged the use of the principle of hereditary taint. Binet propounded a theory of associative learning to elucidate fetishism, to which Krafft-Ebing’s

22 response was that only those with a hereditary taint would be vulnerable. Not until the turn of the century would the emphasis shift from one’s active or passive role to sexual object choice.

Arguably, the most influential theorist of sexuality in the twentieth century, Freud (1905/1962), strove to formulate a psychological theory of sexuality. Freud (1905/1962) formulated an exclusively endopsychic explanation for perversion and sexuality in general, and in imagery and ideation, conscious and unconscious. The founder of psychoanalysis entertained complex and inconsistent views about homosexuality and bisexuality. The idea of bisexuality was most often used in explaining the origins of homosexuality. However, it also arose in conjunction with several other topics: the causes of symptoms associated with hysteria; the explanation of some traits uncovered in dream analysis (Freud, 1916/1953); the ambivalence of the child toward the father in the Oedipal situation (Freud, 1920/1961); and the choice of females or males as love-objects.

Freud (1910/1957) alleged that all persons have bisexual capacities, and that patients would cling to claims of bisexuality in order to avoid coming to terms with their homosexuality. As Freud (1910/1957, p. 472) stated: “A man’s heterosexuality will not tolerate homosexuality”. According to the Freudian viewpoint, the ultimate sexual adjustment of the patient was either heterosexual or homosexual. Furthermore, Freud (1905/1962) posited that heterosexuality constitutes of normal behaviour and homosexuality of deviant behaviour.

In addition, Freud (1925/1963) found the theory of bisexuality helpful in accounting for homosexuality, which he sees as an indication of arrested psychosexual development. During the pre-Oedipal stage, the primary focus for both girls and boys is the mother. At around the age of five or six, children enter the Oedipus complex. Desiring the mother for the boy means fearing castration by the father. However, girls discover that they have already been castrated and hold their mothers responsible for their lack of a penis. To resolve the Oedipus complex successfully, a girl has to transmute her desire

23 for a penis to a desire for a baby with her father. Girls have to change their love object from female to male and their centre of sexuality from clitoris to vagina in order to develop into healthy sexual adults. Because of the way girls have to switch object choice from female to male, Freud believed that bisexuality was more likely to occur in women. Later, Freud (1937/1963, pp. 261-262) used the term bisexual to describe persons with both homosexual and heterosexual attractions and behaviour:

It is well known that at all times there have been, as there still are, human beings who can take as their sexual objects persons of either sex without the one trend interfering with the other. We call these people bisexual and accept the fact of their existence without wondering much at it. But we have come to know that all human beings are bisexual in this sense and that their libido is distributed between objects of both sexes, either in a manifest or a latent for.

Similarly, others like Ellis (1914/1933) accepted the fact that no individual is purely male or female. As Ellis (1914/1933, p. 311) pointed out:

Not only a large proportion of persons who may fairly be considered heterosexual have at some time in their lives experienced a feeling which may be termed sexual toward individuals of their own sex, but a very large proportion of persons who are definitely and markedly homosexual have had relationships with persons of the opposite sex. The social pressure, urging all persons into the normal sexual channel, suffices to develop such slight germs of heterosexuality as homosexual persons may possess, and so render them bisexual.

In terms of the complexity of human sexuality Freud, unlike most of his predecessors, offered a more complicated analysis through the inclusion of the forces of repression, the unconscious and fantasy. Freud claimed to make

24 no moral judgment about sexuality and was involved in the movement to decriminalise homosexuality in Germany and Austria. However, his unyielding notions of what constitutes mature sexual behaviour as opposed to sexual perversion strongly echo his culture’s sexual mores (Paul, 1984). Following Freud’s death in 1939, the crisis of meaning posed by bisexuality for theories of sexual identity was sharpened in psychoanalytic discourse, and later, a radical shift in psychoanalytic discourse sought to challenge the concept of bisexuality.

In the United States of America during World War II and the post-war years, the majority of psychoanalytic writings about homosexuality and bisexuality utilised Freud’s model of childhood development of sexuality: oral, anal, phallic, latenty and genital (Sulloway, 1979). The outcome of either arrested development or regression to an earlier stage without repression was said to be perversion, and with repression, neurosis.

The emphasis in most articles at the time was on putting forth a model directed at elucidating the aetiology of homosexuality and linking this to specific types of psychopathology in homosexual or bisexual patients (Bayer, 1981). This line of conjecture rested on the notion that “normal” development always resulted in heterosexuality. Many of Freud’s contemporaries formed their own schools of psychology that gave weight to and developed different aspects of his theories. Such theories are evident in the writings of analysts such as Stekel (1922/1946), Bergler (1956) and Socarides (1953/1978).

Stekel (1922/1946, p. 39) believed that bisexuality is normal during childhood and that adult sexual orientation results from repression that occurs during the developmental process:

All persons originally are bisexual in their predisposition. There is no exception to this rule. Normal persons show a distinct bisexual period up to the age of puberty. The heterosexual then represses his homosexuality. If the heterosexuality is repressed, homosexuality comes to the forefront.

25 This is similar to Freud’s (1925/1963, pp. 71-72) view that everyone has some homosexual feelings: “The most important perversion, homosexuality, can be traced back to the constitutional bisexuality of all human beings. Psychoanalysis enables us to point to some trace or other of a homosexual object-choice in everyone”.

Stekel (1922/1946, p. 244) stated that “the proper therapeutic course would be to remove the inhibitions which stand between him and women, to make him heterosexual for all practical purposes”. The inconsistencies inherent in his theory seem to arise from two beliefs: that bisexuality may be natural for the human species, but that heterosexuality is the norm for the individual within society. Stekel (1922/1946) was explicit in his rendition of social norms into standards of psychological health. His postulation was that the healthy individual would attempt to conform as closely as possible to established values and traditions.

Hirschfeld (1946), who also believed that everyone is bisexual at a certain stage of development, also expressed the idea of bisexuality as a stage, frequently found in adolescence, but that people normally grow out of it. According to Socarides (1953/1978), derailment from the “normal” developmental track is most serious when it occurs during early childhood years. Such derailment, due to traumatic experiences of various types, not only produces obligatory homosexuality but also leads to global impairment in personality functioning.

The psycho-dynamically orientated practitioners offer no way of conceptualising homosexuality or bisexuality, except in the framework of developmental derailment, character pathology, and sexual pathology (Friedman & Downey, 1998). Furthermore, the goal of psychoanalytic psychotherapy is an exclusive heterosexual orientation, which is seen as more easily attainable by individuals with some heterosexual attractions or behaviour (Bieber, 1962; Socarides, 1953/1978; Stekel, 1922/1946).

26 The theories of bisexuality that were later developed focused mainly on the supposed impossibility of being sexually attracted to both men and women. For example, many psychiatrists believed that if any homosexual behaviour or attraction existed, then the individual concerned must be homosexual. Bergler (1956) viewed sexual orientation in strictly dichotomous terms and rejected the term bisexual as a descriptive category. Bergler (1956, pp. 80- 81) believed that those who consider themselves bisexual are denying a homosexual orientation:

Bisexuality - a state that has no existence beyond the word itself - is an out-and-out fraud. The theory claims that a man cannot be - alternatively or concomitantly - homosexual and heterosexual. Nobody can dance at two different weddings at the same time. These so-called bisexuals are really homosexuals with an occasional heterosexual excuse.

Bergler (1956) was by no means alone in his dismissal of the very existence of bisexuality. Stern (1961) purported that individuals claiming to be bisexual are highly disturbed individuals with identity problems and conflicts coupled with guilt feelings. Cory and Le Roy (1963, p. 89) described bisexuals as “overgrown adolescents who are confused, lack a sense of group identity, and whose ability to differentiate one form of sexuality from another have never developed”.

The accumulated efforts of theorists led to the official classification by the American Psychiatric Association of homosexuality as a mental disorder. This way of thinking about homosexuality influenced the architects of the DSM-I (1957), in which homosexuality was listed as a sociopathic personality disorder and the DSM-II (1968) in which it was listed as a sexual deviation such as sexual sadism or fetishism.

These views of homosexuality and bisexuality were not universally accepted. Well-known authors in the field disputed these theories based on data from surveys of sexual behaviour; clinical assessment and treatment based on the

27 illness model (Hoffman, 1968; Hooker, 1957; Kinsey et al., 1953; Marmor, 1965; Szasz, 1965; Weinberg, 1972).

As a result of the research of Kinsey et al. (1948, 1953) the distinction between normal and abnormal was seriously questioned. The Kinsey Report provided important empirical findings with which researchers and activists could mount a challenge to the domination of psycho-medical discourses of sexuality. Kinsey et al. (1953, p. 660) suggested that: “The capacity of an individual to respond erotically to any sort of stimulus, whether it is provided by another person of the same or of the opposite sex, is basic to the species”.

In 1963 Hooker published research that clearly contradicted the psychiatric association of homosexuality and pathology. Hooker (1963) was convinced that a full picture of homosexuality could not be derived from the investigation of clinicians who saw only homosexuals seeking psychological help, or from those who studied homosexuals in prisons and mental hospitals. Instead, broader based research was needed. It was Hooker (1963) who undertook the task and pioneered research that would in later years prove a rich source of material for those who challenged the assumption that homosexuality was a pathological condition. Hooker (1963, p. 159) argued that homosexuality is synonymous with social maladjustment, however, he goes on to say, “What is difficult to accept is that some homosexuals may be very ordinary individuals, indistinguishable, except in sexual pattern, from ordinary individuals who are heterosexual”.

Hooker (1956) emphasised the pressures on the homosexual - pressures that could be traced to heterosexual hostility. Relying on Allport’s discussion of the behaviour of stigmatised minorities in The Nature of Prejudice (1954), Hooker (1956, p. 219) stated:

It would be strange indeed if all the traits due to victimisation in minority groups in the homosexual were produced by inner dynamics of the personality, since he is also a member of an out-group, which is subject to extreme penalties involving,

28 according to Kinsey, cruelties which have not often been matched except in religious and social persecutions .

The challenge to the psycho-medical domination initiated by Kinsey and associates (1948, 1953) and Hooker (1956) was substantially revivified by Szasz in the 1970s. Szasz (1970), a professor of psychiatry, provided the most profound theoretical attack on the epistemic underpinnings of the discipline in a way explicitly relevant to questions of homosexuality. Szasz (1970, pp. 170-171) argued that:

My contention is that the psychiatric perspective on homosexuality is but a thinly disguised replica of the religious perspective which it displaces, and that efforts to ‘treat’ this kind of conduct medically are but thinly disguised methods for suppressing it..

In the late 1960s Western societies also witnessed the emergence, not only of a powerful anti-psychiatry movement, but also of the gay liberation and the women’s-movements.

2.4 The gay liberation and women’s movements

The first major recent contribution to a more balanced understanding of homosexuality and bisexuality came about through the “shift in attitudes and advocacy regarding homosexuality in society and through the development of more visible and vocal gay and lesbian communities” (Fox, 2004, p. 3). In addition to the homosexual community, the women’s movement argued that the overwhelming oppression faced by women, including bisexual and lesbian women, is male supremacy, that the oppression of women is a fundamental oppression, and that heterosexist oppression is dependent upon the oppression of women.

The women’s liberation theory of sexism and the practice of consciousness- raising enabled the release of a gay revolutionary consciousness. This

29 revolutionary consciousness was infused with anti-psychiatry rhetoric (Angelides, 2001). According to Firestein (1996), the paradigm shift involving bisexuality is clearer when viewed against the backdrop of a previous important shift in psychology’s approach to sexuality - the transition from the illness model of homosexuality to affirmative models of gay and lesbian psychology. The classification and ensuing declassification of homosexuality as pathology in psychiatry and psychology constitute the single most important shift to occur within the mental health fields with respect to sexual orientation to date. This allowed for the emergence of a more positive approach to sexuality in the late 1970s and the 1980s and also saw the emergence of gay and lesbian identity theory.

2.5 A positive approach to sexuality

Gay and lesbian identity theory characterises a significant shift in emphasis in developmental theory, away from illness models with aetiology and psychopathology characteristics toward an articulation of the factors involved in the construction of an affirmative model (Fox, 1996). This has led to a new set of ideas about sexual orientation, broadening the concept to that of sexual identity (Richardson, 1984).

Sexual identity is defined as a complex concept representing enduring, consistent self-recognition of the meanings that sexual behaviour and sexual orientation have for a person (Dworkin, 2000). Sexual identity became the fundamental issue that shaped the direction of the liberation’s future. For the majority of people whose only desire is and has been for the opposite sex, sexual identity is not a concern - they automatically regard themselves as normal, without even having to speculate about what normal means. Lesbian women and gay men, by contrast, have needed to find their own identities in order to challenge heterosexual assumptions of normality (Dworkin, 2000).

Research has shown that lesbian women and gay men undergo a process that involves “shifting one’s identity from the socially accepted heterosexual identity to the socially denigrated non-heterosexual gay and lesbian identity”

30 (Dworkin, 2000, p. 165). The gay and lesbian community use the term “coming out” to indicate acknowledging one’s homosexual attractions, contacting and participating in the gay and lesbian community, and identifying oneself to others as being gay or lesbian.

The development of comprehensive models of gay and lesbian identity development followed the elimination of homosexuality as a diagnostic category in the DSM (Cass, 1979; Coleman, 1981; Troiden, 1988). According to Cass’s model (1979), gay and lesbian identity development is a linear progression through the following six stages: identity confusion, identity comparison, identity tolerance, identity acceptance, identity pride, and identity synthesis. Cass (1979) sees bisexual self-identification as an example of identity foreclosure, postponing or preventing the formation of a positive homosexual identity. According to this model, persistent heterosexual attractions or behaviour are only transitional phenomena that some individuals experience as they move in the direction of permanent gay and lesbian identities. However, Cass (1990) later acknowledges that while sexual identities may be long lasting, they are not necessary fixed. An individual might minimise the significance of experiences that imply a different sexual identity to maintain an already established homosexual, heterosexual, or bisexual identity or might question this identity, potentially initiating the development of an alternative sexual identity. Coleman (1982) suggests stages centring on self-awareness, self-labelling, self-disclosure, stabilisation of gay and lesbian identity and active involvement in a gay and lesbian community.

The advent of the Gay and Lesbian Liberation Movement has produced pressure to identify with one or the other side. The coming out experience is often very stressful for lesbian women and gay men. However, the gay liberationists urge all homosexuals to “come out”. Coming out is a strategy that aims to facilitate the overt confrontation of society with homosexuality. Avoiding this is to “adopt the attitudes of their oppressors” (heterosexual men and women) (Angelides, 2001, p. 128). Furthermore, there was a “strong tendency within the gay liberation movement to reject all heterosexuals and to

31 demand a self-affirmation that denies any contact with non-gays” (Altman, 1979, p. 147). However, the strategy of coming out has made particularly obvious the awkward position of bisexual men and women. The coming out experience for bisexual men and women often meets with resistance from both the homosexual and heterosexual worlds, as well as conflict with their own internalised cultural values, which hardly ever see bisexuality as a valid option (Rust, 2001; Williams, 1999).

2.6 Bisexual identity development

Bisexual identity development has not received the attention that gay and lesbian identity models have (Collins, 2000). The attempts to model coming out as bisexual by adding bisexual identity to existing models of gay and lesbian identity development have shown that these models do not provide an accurate picture of identity development for bisexual men and women. There are a number of reasons why bisexuality might not fit these models. Troiden (1988) believes that the general lack of recognition given to bisexuality and the lack of a supportive community make it difficult for an individual to sustain a bisexual identity. Many bisexual men and women come out twice, once as lesbian woman or gay man in a heterosexual community and then again as bisexual in a gay and lesbian community, when they revalidate their attractions to the other sex (Schwartz & Blumstein, 1998).

Bisexual men and women face many issues when coming out, some similar to those faced by lesbian women and gay men, and others unique. However, for the most part, social science researchers such as Schafer (1976) explicitly discount the possibility of a bisexual identity. Ross’s (1979) article “Bisexuality: Fact or Fallacy” concludes that the label of bisexual is a fallacy. According to Ross (1979), it is either defence or guilt for an individual’s homosexuality, or in some cases, a slow adaptation to one’s homosexuality by first defining oneself as bisexual, then as homosexual. Others such as Chapman and Brannock (1987) acknowledge the possibility of bisexual identity, however, it is characterised as a transitional phase on the way to a lesbian or gay identity.

32 Many bisexuals choose to integrate themselves into either the homosexual or the heterosexual community, while others oscillate between the two communities in different social milieus or periods of their lives, shifting labels when appropriate (Wittstock, 1990). Bisexual men and women that come out first as lesbian or gay, and have learned to rely on the lesbian/gay community for support, also face rejection from this community when coming out as bisexual (McLean, 2001). As Blasingame (1992, p. 53) explains:

To come out as bisexual in the gay community for many means loss of support in the face of heterosexism. This fear kept me in the closet as a bisexual, because I was so fearful of losing my community, the gay community, my support system and grounding.

Newly self-identified bisexuals have a more difficult time finding a supportive community of people like themselves compared to the highly organised and visible gay and lesbian community. According to Bailey, Gurevich and Mathieson (2000, p. 2) “community is defined as a collective of people who experience their lives through a common social structure” and notes that emotional health is affected by sense of community.

The contemporary gay and lesbian community is highly organised and visible, and therefore in the face of rejection by the gay and lesbian community, newly identified bisexuals have a more difficult time finding a supportive community (Fox, 1996; Hutchins, 1996; Williams, 1999). In the face of the relative invisibility of positive connotations for bisexual men and women and the potential lack of community identification, the coming out process may be particularly difficult. The coming out process for a bisexual man or women does not imply merely the “crossing over of sides” from a heterosexual to a homosexual existence, it may put an individual in an uncomfortable “middle ground” that far too often is neither acknowledged nor given space within society’s dualistic limits.

33 Many individuals who come out as bisexual continue to switch back and forth between gay/lesbian and heterosexual identities. These individuals sometimes switch identities depending on the genders of their current sex partners. According to Schwartz and Blumstein (1998), bisexual identity once achieved is often not consistently maintained and cannot be appropriately understood as the static end stage of a developmental process of self- discovery. Consequently, coming out for the bisexual person can be a “continual existential crisis” in which the integration of private and public personas is never fully achieved (De Monteflores & Schultz, 1978, p. 63). Identity synthesis predicated on coming out may be an unrealistic and unattainable developmental goal for bisexual people. Both homosexual and bisexual people wrestle with discrimination and prejudice. However, bisexual identity development is further complicated by the experience of marginality in both the homosexual and heterosexual communities.

For bisexuals to arrive at a bisexual identity, only to experience continued demands to decide whether they are homosexual or heterosexual, might be followed by a process of self-denial as they attempt to fit into the socially acceptable dichotomous model. The social invalidation resulting from the dichotomous model potentially undermines the bisexual’s ability to answer basic developmental questions such as “Who am I?” and “Where do I fit in?” or to develop any sense of identity pride (Smiley, 1997, p. 376). Matteson (1997, p. 98) reports “that the move of some bisexuals to solely homosexual or heterosexual relationships was not necessarily an indication that bisexuality was not real for these men, but that social support and personal resources to handle the complexity were lacking”.

Taking into account the abovementioned difficulties and unique experiences, researchers have developed several models of bisexual identity formation (Bradford, 2004; Paul, 1985; Richardson, 1984; Suppe, 1984; Weinberg et al, 1994). Weinberg et al. (1994) developed a stage sequential model of bisexual self-identity based on research with bisexual men and women. Their model follows a trajectory beginning with heterosexual identity, and proceeding through four stages: initial confusion, finding and applying the bisexual label,

34 settling into the identity, and continued uncertainty. During the stage of initial confusion, many of the respondents have reported confusion and uncertainty regarding their sexual identity. For some, the experience of having strong sexual feelings for both sexes is disconcerting and sometimes frightening. Others are confused because they think strong sexual feelings for, or sexual behaviour with, the same sex means an end to their longstanding heterosexuality. Unaware of the term “bisexual”, some have tried to organise their sexuality by using the readily available labels “heterosexual” or “homosexual” - but these do not fit. Finally, others suggest that they experience a great deal of confusion because of internal “homophobia” (Weinberg et al., 1994).

According to Lourea (1985) this confusion experienced by bisexual men or women, should be seen as a sign of mental health, not an indication of neurosis, a fact which needs to be pointed out to bisexuals who often blame themselves. The confusion is largely due to what sociologists call marginality, the anomalous social position of individuals who are unable to find any clear group membership role because of their straddling conventional social categories or boundaries (Paul, 1988).

In the second stage, according to Weinberg et al. (1994), for many the discovery that the category “bisexual” in fact exists, is a turning point. For others, the experience of their first homosexual or heterosexual experience coupled with the recognition that the sex was pleasurable, is the defining moment. Settling into the identity is characterised by a more complete transition in self-labelling. An important factor in this stage is becoming more self-accepting. The bisexual man or woman becomes less concerned with the negative attitudes of others about their sexual preference.

Finally, confusion appears to be literally a built-in feature of being bisexual. Weinberg et al. (1994) attended to this issue directly in interviews with bisexual men and women. Two questions were asked: “Do you presently feel confused about your bisexuality?” and “Have you ever felt confused?” Eighty four percent of the men answered yes and approximately 56% of the women

35 answered yes. The primary response was that even after having discovered and applied the label bisexual to themselves, and having come to the point of apparent self-acceptance, they still experienced continued intermittent periods of uncertainty regarding their sexual identity (Weinberg et al., 1994). This is confirmed by many bisexual people reporting episodes of identity confusion and uncertainty throughout their adult lives. This confusion may be self-perpetuated or linked to social and environmental pressures such as managing a bisexual lifestyle in a community (Wolf, 1992). Because of societal mistrust and because of a lack of a supportive community, many bisexual individuals feel isolated, depressed and secretive.

Bradford (2004) suggests that bisexual identity involves a process of questioning one’s own reality, inventing one’s own identity, and maintaining that identity through encounters with cultural prejudice, denial, and personal invalidation, which carry continuous threats of isolation and invisibility. In the first stage, the refutation or invisibility of bisexuality in society results in doubting one’s own experience of both same- and other-gendered attraction. Under demands to self-define as either heterosexual or homosexual, bisexual people come to trust their own reality, despite cultural influences, in order to transcend this stage. Those who are able to affirm the reality of both attractions must then structure their reality and give meaning by creating a definition for it (Bradford, 2004). This process involves denunciation of those definitions offered by the culture that are founded on present relationship status and partner’s gender. Once they come to terms with who they are, they are met with the challenge of protecting their identity against constant lack of recognition. This stage, which may be long-term, is marked by marginalisation by both the heterosexual as well as the homosexual community, and requires bisexual men and women to establish their own sense of community. Furthermore, the final stage of transforming the adversity involves social action, leadership and participation in the community at large as bisexually identified individuals (Bradford, 2004).

Although a bisexual identity is stable for a substantial proportion of the bisexual population, those who identify as bisexual are more likely to have

36 changed their sexual identities more than once. Rust (1993) has found that self-identified bisexual women are more likely than lesbian women to change their sexual identities multiple times. Correspondingly, Weinberg et al., (1994), as well as Pattatucci and Hamer’s (1995) studies reveal greater variability over time in the sexualities of bisexual women than lesbian women or heterosexual women. Stokes, Damon and McKirnan (1997) re-interviewed 216 bisexual men one year after the initial interview, and found that 34% had shifted toward homosexuality and 17% had experienced a shift toward heterosexuality. The identity changes that some individuals experience are interpreted by some researchers and theorists as a sign of psychological maladjustment, developmental immaturity, and an indication that the individual has not yet achieved his or her final identity (Golden, 1987). However, changes in identity do not necessarily mean immaturity or that one has not yet achieved one’s identity.

Weinberg et al. (1994) argues that some changes in identity may be because of a deepening relationship with either a man or a woman, whilst others might move in the direction of heterosexuality because of the risk of HIV infection. In addition, Rust (2000) maintains that even though changes occur, this does not happen because they are engaged in a constant struggle to find their true identity, but rather find different sexual identities to be satisfying or true at different times. These changes therefore do not reflect immaturity but rather the restriction imposed on bisexual men and women by mono-sexist concepts of sexuality.

Research on heterosexuals who have become lesbian and on lesbian women who have become heterosexual (Golden, 1987) emphasises the fluidity and change in sexual identity formation. Indeed, many self-identified lesbian women report that they have engaged in heterosexual sex. Bright (1992) found that 46% of the lesbian women in his sample had engaged in heterosexual sex. In addition, a third of these women reported that they had sex with gay or bisexual men, suggesting that even for men the models of sexual identity are seriously problematic. The idea that homosexual and

37 heterosexual identities are fixed is therefore refuted by the ability of some individuals to change their sexual identity over time.

In recognising self-identity change, especially among bisexual men and women, several alternative models of bisexual identity formation have been suggested. For example, Twining (1983) suggests that a task model would be more appropriate, and would focus for example on self-acceptance as a task facing bisexuals rather than as a stage to be passed through. However, the most promising framework for understanding bisexual identity development is a social constructivist one (Paul, 1985; Richardson, 1984; Suppe, 1984). Rather than charting development as a linear progression through stages, the constructivist model emphasises the shifting, emergent nature of sexuality. Instead of being determined by a fixed sexual orientation, individuals are shapers of their fluid sexual selves across their lifespan. The role of context is central to this approach (Herdt, 1984). The bisexual man or woman develops a set of erotic feelings, sexual preferences and relational behaviours that are responsive to changing social, environmental and cultural influences. By placing the personal meaning-making system at the centre of identity development, the constructivist model reframes developmental failures as meaningful responses to interpreted reality (Ross, 1984).

It is imperative to note that people come to bisexuality in diverse ways. While the category bisexual is in itself diverse, the roads that lead to its adoption as a self-label are much more so. Not only are the paths to bisexuality many and varied, but getting there is a process that is dynamic, involving sometimes sudden changes and at other times gradual shifts from one stage to another (Blumstein & Schwartz, 1976).

Similar to lesbian women and gay men, bisexual women and men need to acknowledge and validate their homosexual attractions and relationships to achieve positive and integrated sexual identities. Bisexual men and women, however, need to acknowledge and validate both the homosexual and the heterosexual components of their identities, regardless of the degree to which both of these are actualised in their sexual behaviour or relationships. In view

38 of this dual task, it is necessary to explore what it takes for someone not only to self-identity as bisexual, but to do so with a sense of pride and self- affirmation.

According to Fox (1991), the essential ingredients to self-identify as bisexual with a sense of pride and self-affirmation are permission, recognition, validation, support and community acceptance. Permission refers to the bisexual man or women allowing himself or herself to experience sexual and emotional feelings in response to a given person or situation. Permission also means that those important in the life of the bisexual man or woman allow the individual to experience these feelings without demeaning or discounting their experiences (Fox, 1991). Recognition means naming the feelings and experiences for what they are, and then placing them in context of inclusion. According to Fox (1991) validation is probably the single most important factor determining whether an achieved bisexual identity can be positive. Validation from others affirms the bisexual man or woman’s feelings and confirms that the experience is good, valuable and acceptable. Support is the key to the maintenance and development of this identity over time (Fox, 1991). Support diminishes isolation and frequently takes the form of sharing experience and helps to offset difficulties that the bisexual man or woman might experience. Community support consists of the extension of goodwill and understanding. It is a broadened experience of validation and support wherein individuals sharing similar values and difficulties experience a sense of belonging and caring from the larger social grouping within which they operate (Fox, 1991).

Bode (1976) and Galland (1975) found that self-identified bisexuals are characterised by high self-esteem, self-confidence, autonomy, positive self- concept independent of social norms, assertiveness and cognitive flexibility. Just as research has confirmed that homosexuality is not associated with psychopathology, research has revealed no indication of psychopathology or maladjustment in bisexual men and women (Masters & Johnson, 1979; Nurius, 1983; Weinberg & Williams, 1974).

39 2.7 Defining and measuring sexual orientation

Throughout the latter half of the 20th century social scientists who criticised research for its neglect of bisexuality presented alternative ways of conceptualising sexuality, paving the way for research on bisexuality. The best-known alternative to the dichotomous view of sexual orientation is the Kinsey scale (1953) that departs from the traditional thinking about sexual orientation, and which describe the diversity of human sexual experience.

Prior to this, a dualistic paradigm of sexuality remained firmly in place until challenged by the groundbreaking and binary-breaking work of Kinsey and his team. Over 45 years ago the Kinsey studies provided information on the broad range of sexual behaviour. According to Kinsey et al. (1948, p. 661):

The problem of sexual preference is, after all, part of the broader problem of choices in general: the choice of the road one takes, of the clothes that one wears, of the food that one eats, of the place in which one sleeps, and of the endless other things one is constantly choosing. A choice of a partner in a sexual relation becomes significant only because society demands that there be a particular choice in this matter, and does not so often dictate one’s choice of food or of clothing.

Kinsey et al. (1953) argued that although some people are exclusively homosexual, whereas others are exclusively heterosexual, there are individuals who incorporate both homosexual and heterosexual responses and/or activities in their histories. Kinsey and his associates (1953) were among the first to depart from traditional thinking about sexual orientation in their emphasis on the inadequacy of a dichotomous model for describing the diversity of human sexual experience. To illustrate the inadequacy of the dichotomous model, Kinsey et al. (1948, p. 639) purport that:

The world is not divided into sheep and goats. Not all things are black nor all things white. It is a fundamental of taxonomy that

40 nature rarely deals with discrete categories. Only the human mind invents categories and tries to force facts into separated pigeonholes. The living world is a continuum in each and every one of its aspects. The sooner we learn this concerning human sexual behaviour the sooner we shall reach a sound understanding of the realities of sex.

The findings of the Kinsey studies (1948) were that 37% of the total male population had had at least some overt homosexual experiences to the point of orgasm between adolescence and old age. In addition, between 8% and 20% of females had made at least incidental homosexual responses or contacts in each of the years between 20 and 35 years of age. The publication of Kinsey et al.’s (1948; 1953) findings of male and female sexual behaviour, posed a major definitional crisis for the scientific formulations of homosexuality. The data suggested that sexual difference is a matter of degree and that degrees of bisexual and erotic capacity are much more common than exclusively same-gender sexual behaviour and erotic capacity (Richardson, 1984).

The results from the Kinsey Report - particularly those findings that challenged the dichotomous view of sexuality - were not unanimously accepted. Bergler and Kroger (1954) extrapolated Kinsey’s figures to show that if his sample were representative, between 30 and 35 million people in the United States of America would be on the homosexual side of the scale, and homosexuality would therefore be the national disease. They considered bisexuality impossible, believing that the way a homosexual unconsciously views sexuality and sexual objects, is entirely different from the way a heterosexual does. Furthermore, Bergler and Kroger (1954) blamed Kinsey for allowing people who behave bisexually to think they are normal when they are not.

Kinsey and associates (1948) were aware that the term bisexual was used to refer to individuals with both heterosexual and homosexual attractions and behaviour. The Kinsey team however believed that the use of three

41 categories (heterosexual, homosexual and bisexual) rather than two categories (heterosexual and homosexual) failed to represent the continuum of human sexual behaviour.

To address these problems, Kinsey et al. (1948) developed the Kinsey Heterosexual-Homosexual Scale (KHHS). The scale is an equal interval scale with continuous graduations between heterosexuality and homosexuality. An individual’s rating was based on relative amounts of heterosexual and homosexual responses. Kinsey used the scale to rate individuals on overt experiences and psychological reactions.

The ratings presented on the scale are as follow:

(0) Exclusively heterosexual (1) Predominantly heterosexual, only incidental homosexual (2) Predominantly heterosexual, but more than incidentally homosexual (3) Equally heterosexual and homosexual (4) Predominantly homosexual, but more than incidentally heterosexual (5) Predominantly homosexual, but incidentally heterosexual (6) Exclusively homosexual

Being equally heterosexual and homosexual on the KHHS indicated bisexuality. An individual receiving such a rating on the KHHS was referred to as a Kinsey three. However, some theorists challenged this scale. Altshuler (1984) argues that if sexuality were thus scaled, then perfect Kinsey three’s should exist. If such people do not exist, sexuality is dichotomous and not scaled. According to Altshuler (1984), none of his patients or interviewees indicated that they obtained equal pleasure from men and from women, but displayed a random pattern of male and female partners, and were psychologically unhealthy. He therefore concluded that Kinsey three’s - and therefore bisexuality - does not exist. Altshuler (1984) further stated that when bisexuality appears in an otherwise heterosexual person, it gradually replaces heterosexuality, and that calling oneself bisexual is a matter of status, face- saving and denial of conflict.

42 However, other researchers disagree with Altshuler (1984). According to Rust (2000), Kinsey threes constitute approximately 20% of all bisexuals. Also commenting on Altshuler’s (1984) research for perfect Kinsey threes as evidence of the existence of bisexuality, McConaghy (1987, p. 419) writes: “the naivety of this viewpoint is astonishing. It is equivalent to arguing that for a person to like herrings and caviar equally he or she must eat equal quantities of both”.

Although not rejecting the concept of a continuum in general, Pillard and Bailey (1995) implicitly attribute greater authenticity to the KHHS endpoints (exclusively heterosexual and exclusively homosexual). They report that survey respondents who presently label themselves “bisexual” tend to be relatively few. The researchers further purport that these bisexual individuals are adolescents or younger adults who move toward either end of the Kinsey continuum as they get older. According to Pillard and Bailey (1995), bisexuality is an immature phase and as individuals grow older, they gravitate towards either homosexuality or heterosexuality.

While the KHHS scale displays a less polarised view of sexual orientation, it fails to account for specific life situations, particularly those that changed over time. According to Klein (1978) in this sense the scheme still views sexual orientation in a static fashion. The complexity of labelling a person based on a Kinsey-like rating, and failing to account for changes over time can be exemplified by the following examples: A married man who feels he is heterosexual is sexually involved with a male lover; a girl who breaks up with her male lover, lives with a woman and then returns to the man; a woman who is in prison engages in sex with other women for several years but returns to a heterosexual lifestyle once she is released; a teenager who has sex with his male friend in the locker room and several hours later has sex with his girlfriend (Klein, Sepekoff & Wolf, 1985).

Given the above examples, the confusion surrounding sexual orientation is further complicated when genital activity is used as the only deciding factor in identifying a person’s sexual orientation while completely ignoring factors

43 such as affection, attraction, affiliation, emotional preference and fantasy (Blumstein & Schwartz, 1977). When these covert, non-genital factors, are not considered an entire aspect of human sexuality is overlooked. While the two-dimensional and continuous views of sexual orientation represented an improvement in assessment of sexual orientation, several clinicians and researchers have recommended additional dimensions.

Bell and Weinberg (1978) rated their subjects on two scales: one for sexual behaviour, and one for erotic fantasies. Their research revealed discrepancies between the two ratings, as one-third of the homosexual sample saw their behaviour as more exclusively homosexual than their erotic feelings. As another critique, Shively and De Cecco (1977) questioned the assumption implicit in unidimensional bipolar Kinsey-type scales that heterosexuality and homosexuality are opposing forms of sexuality. They argued that sexual orientation is composed of two independent continua, representing degree of heterosexuality and degree of homosexuality. As a result, Shively and De Cecco (1977) suggest the conceptualisation of sexual orientation by embracing physical, interpersonal and intrapsychic factors. The physical and intrapsychic factors are similar to the notions of Bell and Weinberg (1978) that describe the components of sexual behaviour (physical) and erotic fantasy (intrapsychic factors). The additional component, interpersonal affection, refers to associations involving love or affection that may or may not include genital contact. Again, discrepancies among these three scales within a given individual have been reported.

In addition to scales describing sexual behaviour, fantasies and emotional preference, Klein et al. (1985) define other dimensions, including sexual attraction, social preference, self-identification and heterosexual/homosexual lifestyle preference. Respondents rate themselves on a seven-point scale ranging from heterosexual only, to gay and lesbian only. Furthermore, Klein et al. (1985) contend that sexual orientation is not fixed or permanent, and provided ratings for the respondent’s past and present and the individual’s ideal choice, therefore taking into account the meaningful dimension of time. This view of sexual orientation as not fixed or permanent is elucidated by

44 Blumstein and Schwartz (1977, pp. 35-36). After studying 156 bisexual men and women, the researchers reported:

Perhaps the most interesting finding was that many respondents, who had once seemed well along the road to a life of exclusive heterosexuality or exclusive homosexuality, made major changes in sex-object choices. A very large number of both male and female respondents had made at least one full circle - an affair with a man, then one with a woman, and finally back to a man, or vice versa.

Klein (1978) maintains that bisexuality - which he sees as openness to sexual and emotional feelings towards both men and women - offers the greatest potential for intimacy. People who are not open in this way, Klein (1978) believes, put boundaries on their behaviour that cut down the possibilities for getting close to others. Similarly, in an article “Identity Conflict or Adaptive Flexibility?” by Zinik (1985), bisexuality is presented as flexibility, rather than conflict, and bisexuals are seen as recognising a co-existence and integration of homosexual and heterosexual feelings and behaviours. According to Zinik (1985) this does not mean that there is no ambivalence, though it is posited that much of the difficulty is created by society. Psychological confusion may accompany bisexuality, but is not inevitably. Some individuals can maintain a psychologically harmonious bisexual orientation.

According to Berkey et al. (1990) the Klein Sexual Orientation Grid takes a general approach to sexual orientation and does not focus on bisexuality in particular. Berkey, et al. (1990) pose the following question: is the individual who currently identifies him/herself as homosexual, but who has been involved in a long-term heterosexual relationship in the past, and who may even be a parent, homosexual or is he or she bisexual? In order to address such questions Berkey and associates (1990) developed a scale to reflect six categories of bisexuality in addition to heterosexuality, homosexuality and asexuality. The six categories of bisexuality are:

45 (1) Homosexual orientation prior to exclusive heterosexual orientation (2) Heterosexual orientation prior to exclusive homosexual orientation (3) Predominant homosexual orientation (frequent homosexual desires and/or sexual contacts) with infrequent heterosexual desires and/or sexual contacts (4) Predominant heterosexual orientation (frequent heterosexual desires and/or sexual contacts) with infrequent homosexual desires and/or sexual contacts (5) Equal orientation toward members of both sexes occur on a fairly regular basis (concurrent bisexual) (6) Equal orientation toward members of both sexes, where exclusive homosexual orientation is followed by exclusive heterosexual orientation (or vice versa), on an on-going basis (sequential bisexual)

Coleman (1990) offers another variation on the concept of sexuality as a set of continua by combining ideas from Klein (1990) and Shively and De Cecco (1977) with his own ideas about the structure of sexual orientation to form a multi-dimensional model. Current and ideal sexual self-identification are assessed using five-point scales ranging from exclusively homosexual to exclusively heterosexual, plus a sixth response category for individuals who are unsure of their self-identity. Physical, gender, sex-role and sexual orientation identities are assessed via blank circles instead of scales. Participants are asked to transform each circle into a two-area pie chart to indicate the relative proportions of male and female elements they perceive themselves as having or wish to have. Two other dimensions assess facets of sexuality that were not included in previous models: comfort with one’s sexual orientation and one’s current relationship status, the latter is measured using an extended version of the conventional marital status question. This scale allows complexity to be taken into account and is a clinical interview tool (Shively & De Cecco, 1977).

At the opposite extreme from the categories of the dichotomous model of sexuality, are individualised models of sexuality such as Money and Lamacz’s (1989) concept of love maps. Money and Lamacz (1989, p. 43) describe a

46 love map as analogous to a native language, “a developmental representation or template, synchronously functional in the mind and the brain, depicting the idealised lover, the idealised love affair, and the idealised program of sexuoerotic activity with that lover, projected in imagery and ideation, or in actual performance”. Each individual has his or her own unique love map resulting from the interaction of congenital and environmental influences.

In a similar way, Bierly (1984) recommends viewing sexuality as an integrated system in which several psychological and behavioural dimensions tend toward individualised homeostasis in the well-integrated personality. The relationships between identity, behaviour and various other dimensions are therefore the result of systematic adaptation producing highly individualised solutions. Models like Money and Lamacz’s (1989) and Bierly’s (1984) provide freedom in sexual orientation, because they allow for individualised variation. However, these models have limited potential as a basis for large- scale scientific research because they cannot be used to classify or rank subjects for the purpose of comparison.

In 1953 Kinsey and his associates departed from traditional thinking concerning sexual orientation in their emphasis on the inadequacy of a dichotomous model and initiated alternative definitions and measurement of sexual orientation. The multidimensional models of sexual orientation have developed out of the need to represent the various factors involved in the development and expression of human sexuality more accurately. This in turn has led to the development of more comprehensive models of assessing sexual orientation, and a broader acknowledgement of bisexuality as a distinct sexual orientation and identity. Alternative models such as those suggested by Klein (1985), Bierly (1984), Berkey et al. (1990) and Money and Lamacz (1990) have all contributed significantly to the increase of research on sexuality and more specifically bisexuality. The recognition of bisexuality and its emerging position in the evolving discourse on sexual orientation provides evidence that a major shift of paradigms has been initiated.

47 Data from cross-cultural research clearly indicates that bisexuality has existed throughout history in diverse cultures around the world (Ford & Beach, 1951; Mead, 1975). In the next section evidence of bisexuality in cross-cultural research findings is discussed.

2.8 A cross-cultural inquiry

In several cultures some individuals with homosexual behaviour also exhibit heterosexual behaviour, either concomitantly or at different life stages (Davenport, 1977; Mead, 1968). Anthropologists have provided comprehensive descriptions of how bisexuality has been integrated into such cultures, and provide clear evidence of the existence and integration of bisexual behaviour in diverse cultures around the world.

2.8.1 Cross-cultural studies and bisexuality

Three years after the publication of the Kinsey Report, Ford and Beach’s (1951) Patterns of Sexual Behaviour appeared. Although it is less explicit in its disagreement with standard psychiatric understandings of homosexuality than some of the other texts discussed, it presents material that subvents the conventional view of normal sexual behaviour.

Ford and Beach (1951) relied upon cross-cultural analysis and investigation and found that homosexual behaviour of one sort or another is considered normal and socially acceptable in members of the community in 64% of the 76 societies for which data was available. Ford and Beach (1951, p. 236) noted the challenge that dichotomous thinking presents:

When it is realised that 100 percent of the males in certain societies engage in homosexual as well as heterosexual alliances, and when it is understood that many men and women in our own society are equally capable of relations with partners of the same or opposite sex, then it should be clear that one

48 cannot classify homosexual and heterosexual tendencies as being mutually exclusive or even opposed to each other.

According to Herdt (1984), in certain Melanesian societies there is no great concern with classifying people into the dichotomous categories of heterosexual or homosexual. During the course of their life cycle people may engage in sexual contact with the members of the same or opposite sex. Homosexual contacts start in adolescent initiation rites. Many of these societies prescribe homosexual activity among young adult males and these practices are supported by the society as a whole.

Herdt’s (1980, 1981, 1990) studies of New Guinean culture illustrate the existence of bisexuality. The Sambia tribe’s boys are removed from women between the age of seven years and ten years and are communally trained to become warriors. It is believed that boys must have semen to develop into men. Therefore, to develop into men, the boys must obtain semen from their elders. The young boys consequently form homosexual relationships with post pubescent males. After the young boys have achieved puberty, they form homosexual relationships with younger boys, inseminating them to assist in their male development. Following ten years of training, the Sambia male is then permitted to marry a woman and produce offspring. According to Herdt (1984) the surveys clearly show that ritualised homosexuality in Melanaesia is more common than ever, even by contemporary surveys.

The Batak of Lake Toba in Sumatra are an ethnic group among whom homosexuality and heterosexuality are both universal customs. In the first phase, all males are homosexual among themselves between the ages of nine and nineteen. During this phase, they sleep in a community house. The second phase begins around age nineteen, after the families have negotiated a bride price. Thereafter, a man and his wife begin their heterosexual lives. Subsequently, men may have transient homosexual relationships (Money, 1977).

49 Epprecht (2002) draws attention to the existence of gay and bisexual Basotho men in Lesotho. According to Epprecht (2002), this gives testimony to the ability of men to say one thing but to do another, that is, to perform and maintain a public heterosexual persona while engaging in a private sexuality that contradicts that person. Epprecht’s (2002) informants appear to be heterosexual men and youths who are apparently untroubled by having sex with males. The apparent frequency with which self-identified heterosexual men have sex with homosexual men suggest that there is less internalised stigma to the act than typically encountered in Western societies. In the West, compulsory heterosexuality is not only expected as the dominant public performance but is enforced through internalised homophobia.

The abovementioned is not only evident amongst boys. Gay (1979) examined institutionalised friendship among adolescent girls and young women in Southern Africa. Lesotho’s economy is based on migrant male labour, which leaves the women dependent on male earnings, and creates unstable marital relations. Young girls develop close relationships, called “mummy-baby”, with slightly older girls. Sexual intimacy is an important aspect of these relationships. Mummy-baby relationships not only provide emotional support prior to marriage, but also support for married and unmarried women, either replacing or accompanying heterosexual bonds. These relationships point to the normality of adolescent homosexuality. The fact that close emotional relationships between women often have a significant place within other cultures even after heterosexual relationships have begun, suggests that the growing recognition of bisexuality may find support in studies of non-Western societies (Gay, 1979).

It is thus evident that many societies do not conceptualise diversity in sexual behaviour along the dimensions of heterosexual or homosexual at all (Churchill, 1967). In some cultural groups same-sex behaviour is not seen as homosexual orientation, and cultures vary in the existence and degree of negative sanctions associated with same-sex behaviour or interest (Weinberg & Williams, 1974). Societies also differ with regards to their sexual restrictiveness or permissiveness.

50 The preponderance of historical and anthropological data points to the existence of bisexual attractions and behaviour in many cultures. Bisexuality has been and remains normative in many cultures in which homosexual and heterosexual behaviour exists. However, specific ways of experiencing sexual attraction and behaviour are bound up with the specific cultural milieu (D’Emilio, 1983).

2.9 Chapter summary

This chapter provided a historical view on the nature of bisexuality. The historical prevalence and acceptance of bisexuality in ancient Greece and Rome was discussed, along with the subsequent suppression and abhorrence of bisexuality and homosexuality following the introduction of Christianity into these societies.

Bisexual identity formation, as well as the measuring of sexual orientation for research purposes was discussed. Furthermore, the changing view of bisexual as confused or undifferentiated homosexuals who are afraid to come out in the homosexual community was also discussed. Alternative views to the dichotomous model of sexuality were given, and the difficulties with arriving at a clear conceptualisation of bisexuality were acknowledged.

Lastly, the integration of bisexuality in non-Western cultures was described. It appears that, in many societies, bisexuality is seen as a normal rite of passage to adulthood, and is often acceptable later in life. However, these societies differ in the degree of permissiveness granted and the restrictions placed on bisexual behaviour.

2.10 Preview of the contents of the following chapter

In the following chapter, the prevalence of bisexuality, the basis for sexual attraction in bisexual men and women, similarities, differences and milestones in the lives of bisexual men and women are investigated. This is followed by a discussion of situational bisexuality.

51 CHAPTER 3

TWO LIVES TO LEAD: BISEXUAL MEN AND WOMEN

3.1 Introduction

The Kinsey Report in 1948 astounded the sexually conservative society at the time by reporting that same-gender sexual behaviour was much more common than formerly believed (Gagnon, 1990). The results from the Kinsey studies that “ten percent of the population is gay”, were exploited by gay and lesbian activists to give emphasis to the size and influence of the gay and lesbian community (Rust, 2000, p. 133).

However, unnoticed amid the reaction to this finding was the fact that the Kinsey findings proposed that sexual differences are a matter of degree and that degrees of bisexual behaviour and erotic capacity are more common than exclusive homosexuality (Kinsey et al., 1948; Kinsey et al., 1953).

In this chapter, the prevalence of bisexuality, the basis for sexual attraction in bisexual men and women, similarities, differences and milestones in the lives of bisexual men and women are investigated. This is followed by a discussion of situational bisexuality.

3.2 The prevalence of bisexuality

Evidence of the occurance of bisexual behaviour in men and woman has existed since the publication of the first Kinsey Report in 1948. The inescapable conclusion from the Kinsey studies was that a blend of homosexual and heterosexual behaviour in a person’s erotic life history is a common occurrence. Furthermore, it is entirely possible to engage in anywhere from a little to a great deal of homosexual behaviour. In the

52 following section, the researcher reviews the statistical findings on the prevalence of bisexuality.

Kinsey and his associates (1948, 1953) reported that 0.2% to 3.0% of women and 4% of post-adolescent men were exclusively homosexual, whereas 28% of women responded erotically to another woman and 50% of men had either had sexual experience with another man to the point of orgasm, after the onset of adolescence (37%) or responded erotically to another man (13%). These figures suggested that the capacity for bisexual response is more common than the capacity for exclusively homosexual response.

The Masters and Johnson study in 1979 found that the incidence of bisexuality was comparable to the incidence of homosexuality. For men, the percentage of bisexuals in the sample was 46.8%, while the percentage of homosexuals was 48.9%. For women, the percentage of bisexuals in the sample was 59.8%, while the percentage of homosexuals in the sample was 36.6%. Thus, while the ratio of bisexual men to gay men was lower than in the Kinsey Report, the study still found that bisexual men existed in comparable numbers to gay men, and the ratio of bisexual women to lesbian women was greater than what was found in the Kinsey study (Masters & Johnson, 1979).

The Janus and Janus Report (1993) was based on a sample of 2 765 adults living in 48 states in the United States of America. The researchers reported that 22% of the men and 17% of the women answered “yes” to the question “Have you had homosexual experiences?”, and that 5% of the women and 9% of the men described these experiences as frequent and ongoing. Furthermore, 2% of the women and 4% of the men identified themselves as homosexual whereas 3% of the women and 5% of the men identified themselves as bisexual. Thus, in this study, bisexual self-identity appears to be marginally more common than homosexual self-identity. The Janus Report (1993) also caused extensive political turmoil within the homosexual community because it was the first well-published evidence that the politically useful belief that ten percent of the population is gay may be incorrect.

53 Michael, Gagnon, Laumann and Kolata published the findings of the National Health and Social Life Survey (NHSLS) in 1994. The researchers reported that 4.3% of the women and 9.1% of the men in the sample of 3 432 women and men had engaged in one or more specific sexual activity with a member of the same gender since puberty. Incidentally, lower rates of bisexuality were found in a comparable study done in France in 1991-1992. Only 2.6% of the women and 4.1% of the men reported having had same-gendered sexual experiences (Spira, Bajos, Bejin & Beltzer, 1992). In a study in Britain, Wellings, Wadsworth and Johnson (1994) found bisexuals to exist in numbers greater than or comparable to the number of homosexuals. This was also evident in the Lauman study (1994). The researchers found that 4.1% of women reported attractions to both men and women and 3.9% of men were attracted to both men and women. In this study, 0.9% of the women reported attraction to women only and 3.1% of the men reported attractions to men only. Other non-representative studies offering estimates of the prevalence of bisexuality include a study commissioned by Playboy (Hunt, 1974). Hunt (1974) found that after the onset of adolescence, 17% to 18% of the men in this study had had homosexual experiences. A figure of 10% to 12% for single women was derived from Hunt’s study. The percentage of bisexuals in these studies is again greater than or comparable to that of homosexuals. However, the findings from international research necessitate careful interpretation because of potential cross-cultural differences in the definition of bisexuality or bisexual behavior.

The report of the National Centre for Health Statistics (2005) provides reliable national estimates of some basic statistics on certain types of sexual behaviour, sexual orientation and sexual attraction for men and women, 15 to 44 years of age, in the United States of America. The data presented here has been taken from the 2002 National Survey of Family Growth, and is based on 12 571 personal interviews with men and women. The results indicated that 3% of males had oral or anal sex with another male in the last 12 months. Four percent of females had a sexual experience with another female in the last 12 months. The proportions who have had same-sex contact in their lifetimes are 6% for males and 11% for females. One percent

54 of men 3% of women have had both male and female sexual partners in the last twelve months. In response to the question, “Do you think of yourself as heterosexual, homosexual, bisexual, or something else?” 90% of the men responded that they think of themselves as heterosexual, 2.3% of the men answered homosexual, 1.8% bisexual, 3.9% “something else” and 1.8% did not answer the question. Percentages for women were similar to those for men (National Centre for Health Statistics, 2005).

The survey participants were asked if they were sexually attracted to males, to females, or to both. Among the men, 92% said they were attracted “only to females” and 3.9% “mostly” to females. Among the women, 86% said they were attracted only to males and 10% “mostly” to males. Given the United States of America’s population of 260 million, the percentage may be small, but point to a large number of bisexual men and women (National Centre for Health Statistics, 2005).

The studies considered above all come to more or less the same conclusion about the relative incidence of bisexuality, namely that it is comparable to the incidence of homosexuality. However, problems with these studies merit discussion, yet, none of the concerns are serious enough to damage the validity of the research (Yoshino, 2000). The first problem concerns the flawed methodology of some of the studies. The Kinsey studies, the Masters and Johnson (1979) study, as well as the Janus and Janus (1993) studies all failed to engage in probability sampling (Lauman et al., 1994). Of the studies canvassed above, only the Lauman and the Wellings studies used such sampling.

The reason for this may be the difficulty in finding truly representative samples of hard-to-identify populations such as bisexuals and homosexuals. Instead, studies have relied on non-probability convenience samples, such as members of accessible organisations, persons who frequently visit public places for sexual contact, volunteer respondents and other publicly announced surveys (Turner, Miller & Moses, 1989). The studies therefore share the serious flaw that hampers practically all studies of human sexuality:

55 the samples were not representative of the United States of America’s population and therefore sample means cannot be generalised to that population to estimate the prevalence of bisexuality.

A second problem and one endemic to all sexuality studies, is the problem of having to rely on individual subjective accounts to determine sexual orientation. This raises the concern that individuals may have given dishonest responses. The most damaging way in which participants could have dissembled would have been for homosexual participants to have said that they were bisexual. Because of stigmatising behaviour, homosexuality is likely to be underreported (Warren, 1974) and therefore could result in under reporting of homosexuality and over reporting of bisexuality in the studies (Yoshino, 2000). Furthermore, bisexual men and women could have indicated that they were heterosexual. The combination of negatively sanctioned and easily concealed behaviour could lead to the situation where a more desirable social identity, heterosexuality, is assumed. However, the confidentiality assured in all the studies probably reduce the motive to dissemble in this way.

Researchers (Lauman, Gagnon, Michael & Michael, 1994; Lever, Kanouse, Rogers, Carson & Hertz, 2000; Rust, 1992) have also found that identification as homosexual or heterosexual is not necessarily an indicator of exclusive same- or other-sex attractions and behaviour, and therefore, self-identification does not reflect the entire range of sexual feelings and/or behaviour.

In addition, studies on the prevalence of bisexuality might be examined not to determine the actual number of bisexuals but rather to underscore their prevalence. Yoshino (2000) poses the question of why academic knowledge about the number of homosexuals has flowed so rapidly into mainstream culture, while academic knowledge about the number of bisexuals has not. According to Yoshino (2000), this is in itself indicative of the erasure of the conceptual category of bisexuality. According to Gonsiorek, Sell and Weinrich (1995), there has recently been an increase in disinformation about sexual orientation in the service of some political agendas. This work might depict

56 bisexuality as rare, in an apparent effort to suggest that the public policy of political disenfranchisement of a numerically small group is trivial. However, according to Rust (2000) the real point is that the significance of bisexuality lies not in its prevalence but in the knowledge we can gain about the nature of sexuality in general by studying bisexuality.

3.3 Bisexual men and women

The word bisexuality gives a deceptive sense of fixedness, suggesting a person in the middle, halfway between heterosexuality and homosexuality. In reality, research has shown that exceptionally few people come so neatly packaged. For example, it is possible for a woman who identifies herself as a lesbian to feel attracted to both men and women, or for a heterosexual married man to identify himself as heterosexual even though he has sexual contact with men. Moreover, men and women’s sexual behaviour often changes in the course of their lives, as well as in diverse situations.

The construct of sexual orientation has typically emphasised a partner’s gender as the most important single feature. Yet, bisexual men and women have both same and other gender sexual partners. Kaplan and Rogers (1984) argues that concepts of sexuality based on the physical sex of partners limit the way in which human sexuality is conceptualised and investigated. Researchers therefore suggest that preference might be determined by other variables such as emotional and personality-based features (Ross, Rogers & McCulloch, 1978).

3.3.1 The basis of sexual attraction in bisexual men and women

Vidal-Ortiz (2001, p. 272) asks the following questions:

“Why is a person’s sexual attraction defined by the genitalia of his or her partner? Could it be that one finds attractive someone else’s skin colour, their body shape, some specific conception of

57 beauty, their age, their socio-economic status, or their pleasure preferences?”

Ross and Paul (1992) investigated the basis of sexual attraction in bisexual men and women. The sample consisted of six respondents (three bisexual women and three bisexual men). Kelly’s (1955) repertory grid technique used in the research enabled the researchers to elucidate the cognitive categories used by an individual to organise personal perceptions as well as formal mathematical procedures for relating these constructs to individuals or categories. By examining the construct content of each independent component, it was possible to measure the relevance of each construct in an individual’s classification system for a sexual partner. Results from the study revealed the following information. Firstly, even with the provided construct pair of masculine-feminine, the majority of the dimensions which distinguished these individuals, were based on interactive rather than gender- based characteristics. This suggests that for bisexual men and women the dimensions on which male and female partners were conceptualised was non-gender based. Secondly, the research found that people who are suitable as sexual partners are those who are seen as being like oneself on a number of personality dimensions. Finally, some bisexual people make choices based on personality and physical attributes not necessarily associated with gender.

Concurring with the above findings, Rust (2000) found that her respondents spoke of being attracted to another person because of particular personality traits, conduct, interests, intellect, looks and style. Engel and Saracino (1986) also found that overall bisexuals, heterosexuals, lesbian women and gay men consider many of the same things important: the majority want partners who share their interests, values, religious beliefs and who are intelligent, affectionate, physically attractive, and dependable, with whom they will be able to have a committed relationship involving emotional intimacy and intensity, physical intimacy, fun and respect. However, they have found that bisexuals (63%) were less likely than heterosexuals (92%), lesbian women

58 and gay men (91%) to say that it is important for a relationship to be free of jealous feelings.

It appears that bisexual men and women rely more on gender-free choices related to sexual partners when compared to exclusively heterosexual or exclusively homosexual men and women, for whom gender is a key dimension in choice. This is illustrated in a study by Rust (1995). When asked to use a seven-point scale to rate the importance of a person’s biological sex in determining how strong their attraction was towards that person, only 12% of bisexual women and 28% of bisexual men said that sex is very or extremely important compared to 64% of lesbian women and 86% of gay men.

3.3.2 Similarities, differences, and milestone events in the lives of bisexual men and women

Bisexual men and women are in a unique position to compare the experiences of sexual contact and romance with men and with women. Both autobiographical comments by some bisexuals and social scientific findings suggest that for many bisexuals these experiences are different. For example, Clausen (1990, p. 454) writes: “My desire is for a specifically sexed individual. When I am with a woman, I love as a woman loves a woman, and when I am with a man, I love as a woman loves a man”.

Zinik (1985) has found that bisexual men and women report that they are similarly sexually aroused by men and by women, but that they receive more emotional satisfaction from their female partners and fall in love with women more often than with men. In support of this, Wolff (1977) has found that bisexual men’s same-sex relationships are shorter and more sexual than their other-sex relationships, whereas bisexual women’s same-sex relationships are fewer and longer lasting than their other-sex relationships. Similarly, Taywaditep and Stokes (1998, p. 35) conclude that “sex with men seemed to be less bound to relationships than sex with women”.

59 According to Gagnon and Simon (1973), differences in the bisexual behaviour, in physique and body texture, and in the emotional tone of the experience of men and women respondents, are consistent with general patterns of male and female sexuality. In the most comprehensive sociological study of bisexuality to date, Weinberg et al. (1994) found that the most frequently noted differences concerning sexual with men as opposed to women were about the kinds of behaviour reported. Both men and women have emphasised that there is more touching, hugging and caressing with female partners, along with a gentler, slower and softer quality to the interaction. Others have also felt that a person of one’s own gender tends to be more familiar with their sexual responsiveness than an other gendered individual.

Concerning bodily differences, the result is consistent with the differences for behaviour. When referring to female partners, both men and women have remarked on the smoothness and softness of the female body. When referring to men, both men and women have referred to the muscularity and hardness of the male body (Weinberg et al., 1994).

Two-thirds of the bisexuals that were interviewed reported that there were definite differences in the emotional tone with women when compared to men. Firstly, the most evident difference within relationships is that men and women find that it is more difficult for men to disclose their feelings. Secondly, the difference mentioned most frequently by both men and women points to the disparity between men and women. Both men and women have reported feelings of social inequality and a power imbalance. However, women believe that their relationships with other women are more egalitarian. Finally, bisexual men and women have emphasised that intimacy is improved and is easier to achieve in a relationship with a member of the same sex, and have pointed out that people of the same sex have more in common (Weinberg et al., 1994).

With reference to milestones, research indicates that bisexual men and women have their first heterosexual attractions in their early teens. Bisexual

60 women have their first heterosexual experiences and relationships with men in their middle teens. This is earlier than bisexual men who on average have their first sexual experiences and relationships with women in their late teens (Weinberg et al., 1994).

With regards to homosexual attractions, research findings indicate that on average, bisexual men have their first attractions in their early to middle teens, while bisexual women experience their first attractions in their middle to late teens (Fox, 1995; George, 1993; Rust, 1993). The results also signify that, on average, bisexual men and women first self-identify as bisexual in their early to middle twenties. Bisexual men and women are also more likely to disclose their sexual orientation to friends and partners rather than to family members or colleagues. Furthermore, bisexual women appear to adopt a bisexual identity sooner after their first homosexual experience than bisexual men (Fox, 1995).

Women generally find their first homosexual experience much less distressing than men do. Women have reported that such activities seem to be a natural extension of female affectionate behaviour and do not always have implications for their sexuality. Men are far more concerned with what the experience means for their masculinity and have reported much more difficulty in coping with homosexual behaviour and developing a bisexual identity than women (Gagnon & Simon, 1973).

Rust (2001) has also found that many bisexual men and women do not feel the need to be equally attracted to men and women and do not need to be sexually involved with both men and women in order to own a bisexual identity. Bisexuals may prefer one sex to the other, but are attracted to both men and women and remain open to sexual involvement with both (Rubenstein, 1982; Saliba, 1982). Weinberg et al. (1994) found that only 2% of the self-identified bisexual men in their research and 17% of the self- identified bisexual women are equally sexually and romantically attracted to and involved with women and men. In addition, the realisation that they are in love with a person (same or opposite sex) is often a requirement for sexual

61 attraction, sexual behaviour or a change in sexual identity. For men, both first heterosexual and homosexual experiences are likely to be with strangers, whilst the pattern for the women has been for sex to occur with a close companion.

As Gagnon (1977, p. 274) writes: “The question that needs to be asked is what kind of histories and contexts offer scripts to people at various moments in the life history which make sex with both genders possible”. As a result, researchers have studied bisexual behaviour in men and women in unusual social situations such as prisons, same-sex boarding schools, in the military and among prostitutes.

3.4 Situational bisexuality

3.4.1 Introduction

Prior to 1980, several researchers studied bisexuality under the pretext of situational homosexuality. The individuals in the studies were considered heterosexual men and women who were forced by situational constraints to engage in same-sex sexual behaviour. In addition, most people did not differentiate between individuals in terms of the motives and circumstances which underpinned their homosexual behaviour, rather, all persons were uniformly categorised as homosexual.

The incidence of same-sex behaviour in the following contexts cannot tell us whether the context in which same-sex behaviour occurred actually caused the behaviour, or whether the environment merely caused these individuals to realise sexual dispositions they had already had but would not have recognised had it not been for the influence of a single-sex environment. Furthermore, mere participation in one or more sexual acts with a same-sex partner by a heterosexual person did not signify that the person had changed his or her sexual orientation. However, their experiences with homosexuality gave them “bisexual lifetime biographies” that they would not otherwise have had (Rust, 2000, p. 239).

62 3.4.2 Same-sex boarding schools

There is support for the belief that same-sex behaviour is more common among the residents of boarding schools than among their peers (Hickson, 1995; Weinberg, 1967). However, due to the age of the individuals involved as well as the sensitivity of the topic, research on this issue has been minimal (Feldman, 1984). Nonetheless, a study of 1 873 adolescents (Schofield, 1965) found that a higher proportion of individuals who attended single-sex boarding schools reported homosexual activities (28%) than individuals who attended co-educational non-residential schools (3%).

It cannot be ascertained what proportion of former boarding school students continue their same-sex behaviour or what proportion live heterosexual lives. However, it can be hypothesised that the majority of individuals who attended boarding school live primarily heterosexual lives after leaving boarding school and amid these are undeniably a number of the individuals who did have same-sex experiences.

3.4.3 The prison environment

Most of the research proposes that rates of same-sex sexual activity among male prisoners are higher than rates of same-sex activity among men in free society. Data from the Institute for Sex Research indicated that 35% to 40% of inmates have had a same-sex sexual experience (Gagnon & Simon, 1973). Wooden and Parker (1982) found even higher rates. Sixty five percent of their participants had engaged in same-sex activity while in prison, including 55% of self-identified heterosexual men. However Gebhard, Gagnon, Pomeroy and Christenson (1965, p. 625) argued that the comparison between inmates and the general population reflects a difference between the “delinquent and non-delinquent lower socio-economic” population. Furthermore, according to the Gebhard team (1965) the men that ended up in prison were more likely to have engaged in same-sex sexual activity prior to incarceration than men who would not be in prison. The researchers established that approximately 88.2% of those who engaged in same-sex sexual activities in prison had done

63 so beforehand and that only 7% of the inmates had had their initial same-sex experience in prison.

In a more recent study Hensley, Tewksbury and Wright (2001) explored same-sex activity within a male maximum-security prison in the United States of America. When asked about their sexual orientation before incarceration, 79% reported they identified as exclusively heterosexual, 15% reported they were bisexual, and 6% identified as homosexual. However, when asked about their sexual orientation during incarceration, 69% identified as exclusively heterosexual with an additional 23% reporting that they were bisexual and 7% revealing that they were homosexual. These findings support previous studies on the prevalence of bisexuality in the prison population.

As mentioned in Chapter 2, different cultural groups have different sexual cultures and patterns of sexual behaviour. This is further supported by Wooden and Parker’s (1982) findings of cultural differences in the sexual behaviour of inmates. In this study, inmates were classified as heterosexual, homosexual or bisexual based on their self-identities. The findings were as follows: prior to incarceration, 62% of black heterosexuals, 30% of white heterosexuals and 23% of Chicano heterosexuals had had same-sex experiences. While in prison 81% of black, 55% of Chicano and 38% of white heterosexuals had engaged in same-sex activity. All bisexual and homosexual inmates had engaged in same-sex activity prior to incarceration, and all bisexual and homosexual inmates, regardless of ethnicity, had engaged in same-sex sexual behaviour whilst in prison. This also illustrates the fact that people’s sexual identities do not always reflect the entire range of their sexual feelings and/or behaviour (Lauman, et al., 1994).

There are fewer findings regarding rates of same-sex activity among women inmates. According to Michael et al. (1994), 4.3% of women in the United States of America’s population have engaged in a sexual act with women since puberty. According to Giallombardo (1966), the majority of the women in the prison system are not only involved in same-sex relationships but

64 participate in same-sex sexual activity within those relationships. Giallombardo (1966, p. 136) found that “the vast majority of inmates adjust to the prison world by establishing a homosexual alliance with a compatible partner as a marriage unit”.

Ward and Kassebaum (1965) reported that half of the inmates of the California Institution for Women had engaged in sexual activity with other women inmates. Furthermore, 90% of the respondents had had their first same-sex experience while in prison. Ward and Kassebaum’s (1965) participants, who identified themselves as bisexual whilst in prison, asserted that many women become strongly emotionally attached to their partners in prison and are certainly conflicted about giving up these same-sex relationships. However, the researcher’s emphasised that the majority of the women would most likely return to a heterosexual life style upon their release from prison for various reasons. These include conditions of parole and the continuation of family ties, including marital and parental ones.

Propper (1981) found that the case might be different for younger female offenders and reported that of the thirteen offenders who were involved in pleasurable homosexual behaviour while incarcerated, 69% expected to continue homosexual activity on release. Magura, O’Day and Rosenblum (1992) reported that most of the homosexual women in their study did not intend to become heterosexually involved after their release even though only four identified themselves as bisexual and none as lesbian. This could be an indication that the influence of external social structures is weaker for young female offenders.

Data suggests that there may be somewhat greater flexibility and fluidity in the sexual histories of women than men. This has been historically evident in literature, but is often submerged in research data. McIntosh (1981, p. 154) points out that “it is interesting to notice” that although at the age of 20 far more men than women have homosexual and bisexual patterns (27% as against 11%), by the age of 35 the figures are the same for both (13%).

65 Women seem to broaden their sexual experience as they get older, whereas most men narrow theirs and become more specialised.

The greater fluidity and flexibility in women’s sexuality can also be illustrated in the percentage of homosexual or bisexual men and women who have previously been married. Masters and Johnson (1979) found that 23% of homosexual men and women had previously been married. Similarly, Saghir and Robbins (1973) found that 25% of the women and 18% of the men in their study had previously been married. Bell and Weinberg (1978) found that 35% of the white women and 47% of the black women in their sample had previously been married, compared to 20% of the white men and 13% of the black men. Pertaining to married bisexual women, Coleman (1985) reported that 60% of the women had their first homosexual experience after they were married. While some did have homosexual experiences prior to marriage, very few accepted themselves as to be bisexual before the marriage.

Although some of the aforementioned studies conducted in the prison environment are 15 to 30 years old, they may not be outdated. Burkhart (1996) reported that her results indicate that little change has occurred in prison situation.

3.4.4 The military service

Another sex-segregated institution is the military service. It should be kept in mind that the majority of the research on homosexuality in the military service in the United States of America has been centred on the question of whether gay men and lesbian women should be excluded from military service. The gay and lesbian community has challenged this exclusion based on the Kinsey Reports (1948, 1953). These reports have become the source of the number (10% of the population is homosexual) that society agrees on when talking about sexuality. This has enabled gay men and women to produce powerful arguments in the form of slogans. If you were to accept this figure for prevalence of homosexuality in both men and women, you may have 20 million potential voters: “The voting lobby is the world’s safest closet”. Every

66 family has gays: “We are your children” and “How could 10% be unfit for military service?” (Voeler, 1990, pp. 32-38).

The men in the armed forces are typecast as masculine and unquestionably heterosexual. However, Gagnon and Simon (1973) suggested that groups of military personnel, after unsuccessfully looking for heterosexual outlets, might under the right circumstances agree to the proposal of a male homosexual. According to D’Emilio (1983, pp. 25-26), the sex-segregated nature of the armed forces “raised homosexuality closer to the surface for all military personnel. Living in close quarters, not knowing whether they would make it through the war, depending on one another for survival, men of whatever sexual persuasion formed intense emotional attachments”. Donald Vining (1979) who was exempted from military service because of his acknowledged homosexuality elucidated this by reporting to have had sexual relations with both self-acknowledged homosexual as well as heterosexual service members.

With regards to the perception that military personnel have been seen as include a higher proportion of lesbian women than the general population, D’Emilio (1983, p. 27 cited in Treadwell, 1954) ascribed this to “military policy that contributed to a situation that it took pains to deny”. Enrolment focused on a population group statistically likely to include a disproportionate number of lesbian women and heterosexual women with felxible sexuality In 1943, 70% of the women in military service were single, 83% were childless, 40% were under 25 years of age, and 67% were under 30 years of age (Berube, 1981; War Department, 1943).

Anxious to counteract its reputation of moral leniency, the military sought to avoid unwanted pregnancies by keeping its female personnel segregated. The instruction manuals for officers commended the desire for intense comradeship in service as one of the finest relationships possible for women. However, with emotional attachment serving as a powerful stimulus to female eroticism, such bonding led toward unintended results (D’Emilio, 1983).

67 While prison and the military service both provide milieus in which men and women might turn to their own sex because of lack of sexual contact with the opposite sex, in prostitution, men and women have sexual access to the opposite sex.

3.4.5 Prostitution and bisexuality

McCaghy and Skipper (1969) gathered data through in-depth interviews with 35 strippers. The girls were all white and ranged in age from 19 to 45, with 60% between the ages of 20 and 30, and 69% of the strippers were or had been married. According to the researchers, 50% to 75% of the girls were bisexual.

Based on a study of 136 female prostitutes (James,1976) indicated that 35 had experienced a lesbian relationship, seven reported frequent same-sex activity and six were exclusively lesbian. These same-sex experiences were for personal pleasure and were not engaged in acts of prostitution. According to James (1976), the high incidence of same-sex experience might reflect their greater willingness to admit their same-sex experience. These results imply that situational homosexuality need not be a negative response to conditions of heterosexual deprivation, but a positive choice in favour of same-sex activities.

The literature on same-sex male prostitution is relevant to bisexuality because many male prostitutes, regardless of their sexual identities, also engage in heterosexual behaviour for either money or pleasure. Before the HIV-epidemic, male-male prostitution received little attention from researchers. Prior to that time research on prostitution was concerned almost exclusively with female prostitutes (Allen, 1980).

Boles and Elifson’s (1994) study among 224 male prostitutes indicated that 35% of the sample identified as bisexual, 17% as homosexual and 46% as heterosexual. Morse, Simon, Balson and Osofsky (1992) established that in their sample of 211 male prostitutes, 41% identified as bisexual, 21% as

68 homosexual and 38% as heterosexual. Boles, Sweat and Elifson (1989) found that 13% of the male prostitutes had engaged in recreational sex with both male and female partners. Visano (1990) conducted research on the other participant in male prostitution, namely the customer. The results indicated that of the 60 customers that were interviewed, 36% identified as homosexual. Morse, Simon and their colleagues (1992) found that of the 15 participants that were recruited for the purpose of their study, 14 identified themselves as bisexual or heterosexual. Reports from male prostitutes estimated that approximately one-third of their customers were heterosexual or bisexual, and that 40% of them were married (Morse, Simon, Osofsky, Balson & Gaumer, 1991). These results confirm previous studies that indicate that there appears to be greater fluidity and flexibility in women and men, and may also illuminate the prevalence of homosexual or bisexual men and women that are married.

Certain situations present opportunities for sexual activity with members of one’s own sex and increase its attractiveness relative to the available heterosexual options, even for individuals who might arguably otherwise be heterosexual. According to Rust (2000), such situational homosexuality is more than technical bisexuality - it is true bisexuality, motivated either by a disregard for the sex of one’s partner or a desire for sexual and emotional intimacy with a member of one’s own sex.

3.5 Chapter summary

This chapter laid out studies aimed at determining the prevalence of bisexuality in society. Research evidence suggests that the incidence of bisexuality is higher than previously assumed, and is comparable to, or even higher than that of homosexuality. Methodological flaws in studies on the prevalence of bisexuality were indicated, although these flaws were not necessarily detrimental to the value of the research.

The nature of sexual attraction in bisexual men and women was also discussed. Research has shown that the nature of sexual attraction is

69 relatively similar across bisexual, homosexual and heterosexual men and women, although bisexual men and women make more gender-free choices of sexual partners.

In addition, the process of forming a bisexual identity was considered, along with the difficulties and dilemmas that bisexual individuals face when coming to terms with their sexuality. The differences and similarities between bisexual men and bisexual women’s sexual experiences and milestones in identity development were also discussed.

Finally, the incidence of bisexuality in certain situations was researched, with particular focus on prison, same-sex schools, military service and prostitution.

3.6 Preview of the contents of the following chapter

The following chapter focuses on attitudes and stereotypes regarding bisexual men and women, internalised homophobia and the psychological health of bisexual men and women.

70 CHAPTER 4

ATTITUDES TOWARDS BISEXUAL MEN AND WOMEN

4.1 Introduction

In attempting to understand the dilemma of the bisexual - an individual without legitimate social status or social identity, and somehow outside the conventional sexual categories - it helps to consider the sociological concepts of marginality. According to Stonequist (1937/1961), marginality refers to the set of circumstances faced by people who have not found, or cannot accept, a clear group-membership role. Common external threat often causes those under siege to cohere. This coherence is evident in the gay and lesbian community and in the heterosexual community.

The 1970s saw the emergence of a tremendously vital gay and lesbian community, which has fought its battles for civil rights based on being a disadvantaged minority. As sexual minorities who have themselves been rendered invisible, one might expect lesbian women, gay men, bisexual men and women to form a strong alliance because of their common marginalisation in a heterosexist society. However, it appears that lesbian women and gay men, and heterosexual men and women have common interests in denying the existence of bisexuality (Yoshino, 2000) and as a result they discriminate against bisexual men and women. The Social Identity Theory (Tajfel & Turner, 1979) that underlies this study appears to be appropriate as it focuses on intergroup conflict and discrimination associated with group membership.

4.2 Attitudes and the Social Identity Theory

The constructs of attitude, stereotype, self-concept, and self-esteem are very popular among social psychologists. At least one of these concepts was mentioned in 90% of the 601 articles published between 1996 and 1998 in the Journal of Personality and Social Psychology. An indication of the value of the

71 theory that inter-relates the four constructs is suggested by observing that two or more of the four constructs were mentioned in 72% of the 601 articles, three or more in 40%, and all four were mentioned in 12% of the articles (Greenwald et al., 2002).

The well-established body of research and theory on social identity and in particular Tajfel’s Social Identity Theory (Tajfel, 1982; Tajfel & Turner, 1986) and Turner’s self-categorisation theory (Turner, Hogg, Oakes, Reicher & Wetherell, 1987) relate the four constructs of attitude, stereotype, self-concept and self-esteem. Of particular relevance to the current study, is the fact that the Social Identity Theory’s focus’s on intergroup conflict and discrimination associated with group identification. In addition, this theory has been utilised in numerous studies relating to homosexuality (Clark & Maas, 1987; Hajek & Giles, 2006; Hammer, 1992; Hodson, Harry & Mitchell, 2008; Sakalli, 2003).

In a recent article, Hornsey (2008, p. 217) states that it is almost impossible to think or write about group processes today without reflecting on the core constructs of the Social Identity Theory, and that its “emergence has played a critical role in the resuscitation of interest in group processes both within and outside social psychology, and (somewhat unusually for a theory that is over 30 years old) interest in the theory seems to be only accelerating with time”. According to Hornsey (2008), the social identity approach is ambitious in scope and ultimately rests on simple, elegant, testable, and usable principles.

The Social Identity Theory is intended to be a social psychological theory of intergroup relations, group processes and the social self. It has its origins in early work in Britain by Tajfel (1959) on social factors involved in perception and on cognitive and social beliefs aspects of racism, prejudice and discrimination. The theory was then developed and fully formulated in collaboration with Turner and others in the mid to late 1970’s (Tajfel, 1974; Tajfel & Turner, 1979, Turner, 1982). During the early to mid 1980’s, Turner (1985) initiated an important theoretical development of the Social Identity Theory to produce the Self Categorisation Theory. The Self Categorisation Theory incorporates and extends The Social Identity Theory to a larger

72 collection of social phenomena by placing a social-cognitive account of the self at its theoretical center (Turner, 1987). Although distinct from Social Identity Theory in some respects, it is related closely enough to be considered part of the same theoretical and meta-theoretical enterprise as the Social Identity Theory (Hogg & McGarty, 1990).

According to the Social Identity Theory, in-group favouritism results from attempts to achieve positive group distinctiveness, or favourable evaluation of one’s own group relative to other groups. The process is driven by an individual’s social identity needs, or desires to enhance those aspects of one’s self-image that relate to group membership. In accordance with the Social Identity Theory, individuals’ self-concepts are partially dependent on how they perceive their in-group. The self-concept is enhanced when their in-group is viewed positively. As people desire to see themselves positively, they are motivated to perceive their in-groups as having positive distinctiveness from other groups (Tajfel & Turner, 1979). A tenet of the Social Identity Theory is that people make social comparisons between their in-group and out-groups in order to view the in-group favourably. Thus, perceivers could view the in- group as having more positive attributes than an out-group, and perceive the out-group as having more negative attributes than the in-group.

Therefore, the basic idea of the Social Identity Theory is that a social category (in this instance, sexual orientation) into which one falls, and to which one feels one belongs, provides a definition of who one is in terms of the defining characteristics of that category – a self-definition that is a part of the self- concept. This membership to a category is represented in the individual members’ mind as a social identity that both describes and prescribes their attributes as a member of that group – that is, what they should think and feel, and how they should behave. Thus, when a specific social identity becomes the salient basis for self-regulation in a particular context, self-perception and conduct becomes in-group stereotypical and normative, perceptions of the out-group become out-group stereotypical. Intergroup behaviour acquires competitive and discriminative properties to varying degrees depending on the nature of relations between the groups. In this regard the heterosexual and

73 homosexual communities have several out-group stereotypes regarding bisexuality. These include: bisexuals are indiscriminate about whom they have sex with, all bisexuals are swingers, bisexuals are unable to commit and bisexuality is just a phase (Burleson, 2005). These stereotypes have discriminative properties that ensure that the homosexual and heterosexual communities view themselves as distinct from the bisexual community.

The Social Identity Theory and Self Categorisation Theory models of group processes have a number of important features. They are general theories of the social group, they are not constrained by group size and distribution, they incorporate the role of both the immediate and the more enduring intergroup context in group behaviour, they account for a range of group behaviours (e.g. stereotyping, discrimination, attitudes), they are socio-cognitive and they do not construct group processes from interpersonal processes. Therefore, behavior is influenced by the categorical structure of society via the mediation of social identity and the accompanying process of self categorization. The contextual salience of specific social identities rests on the extent to which they render maximally meaningful a particular context, and contextual factors influence the form taken by identity-contingent cognitions and behaviours. Because social identity dynamics exist in which groups vie for positive social identity, intergroup relations and social identity are dynamically intertwined (Hogg, 2006; Hogg, Terry & White, 1995; Oskamp, 2002).

In both these theories there is recognition of Otherness, a very phenomenological sense of group situations and of group members’ interpretive capabilities. There is also a critical edge to these approaches, for the most basic insights of social identity not only include that recognising oneself as a member of a group is important for the purposes of self- identification but also that society is composed of social groups that have relationships of power and status to one another. When dominant groups in society have power and status to impose dominant values and ideologies that serve to legitimise the status quo, subordinate groups become involved in social change using several strategies (Hogg & Abrams, 1987; Tajfel, 1978). With reference to this research, it is the heterosexual community that has

74 historically held power and dominated values and ideologies. The subordinate homosexual community then became involved in changing their status and power in society in order to obtain equality. Another minority group is currently becoming involved in orchestrating social change to obtain recognition and legitimacy: people who identify themselves as bisexuals.

Work on the Social Identity Theory proceeds at a breathtaking pace within social psychology (Turner, 1999). Research has taken several interesting and new directions. In this respect, research on the impact of identity strength is particularly welcome (Branscombe, Wann, Noel & Coleman, 1993; Schmitt, Branscombe, Kobrynowicz & Owen, 2002). As an example, one should consider research by Branscombe et al. (1999) on the interaction between threat and group identification. The researcher uncovered evidence that highly identified group members may react to a threat with a show of solidarity, whereas low identifiers distance themselves from the group. This finding could have important implications for the current study. It appears that the highly identified heterosexual and homosexual communities may have reacted with a show of solidarity against the threat of bisexuality. The homosexual and heterosexual communities view their communities as having more positive attributes than the bisexual out-group. Consequently, the homosexual and heterosexual community portray bisexual men and women as promiscuous, who cannot have committed relationships or are likely to contract sexually transmitted diseases (Eliason, 2001; Rust, 1995; Spalding & Peplaum, 1997).

4.3 Attitudes and bisexual invisibility

Some individuals believe that bisexuals are actually people who are exclusively heterosexual or exclusively homosexual who are temporarily experimenting with both sexes (MacDonald, 1981). This experimentation usually occurs in sexually permissive subcultures such as college campuses, women’s consciousness raising groups or swinging (Dixon, 1984). They expect that the bisexual man or woman will return to their preferred monosexual orientation after their period of experimentation (Ochs, 1996). Furthermore, bisexuality is seen as a transitional orientation, moving toward

75 exclusively homo- or heterosexuality (Harry & Lovely, 1979). According to Bell (1975) all bisexuals are in reality homosexual and the heterosexual activities and interests shown by such people are used to cover up their genuine homosexuality.

According to Yoshino (2000), bisexual men and women remain invisible because heterosexual men and women, and lesbian women and gay men have overlapping political interests in bisexual erasure. It is as if heterosexual and homosexual individuals have concluded that whatever their disagreements, they will agree that bisexuality does not exist.

Bisexuality calls into question the sexual orientation of the self. To illustrate this compare the ease of proving one is heterosexual or homosexual in a world in which bisexuals do not exist, with the difficulty of proving the same thing in a world in which bisexuals are recognised. In a world that denies bisexual existence, cross-sex desire and same-sex desire are mutually exclusive. Bisexuality destabilises sexual orientation by making it logically impossible to prove that one has a monosexual identity. Both homosexual and heterosexual individuals have shared investments in stabilising their identities, as group members draw some comfort from social ordering, and knowing their own place in the social order (Yoshino, 2000). Ochs (1996, p. 232) provides anecdotal support for this investment:

Coming out as gay was the hardest and most painful thing I have ever done in my life. Now I am finally at a place where I have a solid identity, a community, a place to call home. Bisexuals make me uncomfortable because their existence raises the possibility that I might be bisexual myself. And coming to terms with my identity was so hard for me the first time around; I cringe at the thought of having to go through such a long, hard, painful process a second time.

76 Hence, homosexual individuals share an interest in identity stabilisation, since such stabilisation roots them in a community and relieves them of the anxious work of identity interrogation.

The process of bisexual erasure is also of interest in ensuring that the battle lines are clearly drawn between sexual minorities and those who oppress them. The bisexual is seen as less committed to fighting heterosexual privilege because of his or her ability to partake in those privileges. The bisexual is also viewed as inherently traitorous, insofar as solidarity requires not just supporting a group but also fighting and not consorting with the group’s enemies (Newsweek, 1974). For example, lesbian women see bisexual women as traitors ready to leave a lesbian relationship or leave the gay or lesbian activist movement for the heterosexual privilege they can gain through a relationship with a man (Dworkin, 1996).

In addition, lesbian women and gay men also fear being unable to compete with the benefits accorded by our culture to those in heterosexual relationships, and therefore believe that those who have a choice will ultimately choose heterosexuality. The reason for choosing a heterosexual relationship is that these relationships are privileged, and many bisexual men and women, as well as lesbian women and gay men adopt at least a public front of heterosexuality to avoid rejection by family members, develop their careers and raise children with societal approval (Ochs, 1996).

Bower (1995) concurs with the above statements by arguing that there is no question that, with an opposite sex lover, there are privileges such as legal marriage, and the safety to express affection in public. However, for women in particular, this “privilege” becomes a more complicated matter. Heterosexism is a problem not only because it limits sexuality, but also because it bolsters male dominance and power. Heterosexual privilege cannot be divorced from male privilege, and male privilege never benefits women unambiguously. Bower (1995) questions these privileges in view of the condition of most women in heterosexual relationships. The researcher states that whether married or not, whether housewives or working full-time or

77 part-time, the majority of women have primary responsibility for housework and the raising of children, are partially or totally financially dependent on their husbands or lover, are under pressure to keep themselves physically attractive in men’s terms, and are often physically or emotionally abused by husbands or lovers. Therefore, for women, heterosexual privilege is a double- edged sword. The cost of the privileges that a woman receives for heterosexuality often involves being dependent on men.

Additionally, the bisexual is threatening to lesbian women and gay men because they blur the division between heterosexual and lesbian/gay, thus undermining the basis for gay and lesbian identities and gay and lesbian political movements (Blumstein & Schwartz, 1977). The danger of bisexuality to gay and lesbian politics is most effectively counterbalanced by reinforcing the authenticity and clarity of the distinction between heterosexuality and homosexuality, and this approach perpetuates the invisibility of bisexuality (Rust, 2000). Michel (1996) employs the Aesop’s fable of the bat, the birds, and the beast in order to describe the abovementioned situation. The bat, which refused to take part in the war between the birds and the beasts, was finally left isolated and alone, being neither bird nor beast.

The hostility of lesbian women and gay men toward bisexual men and women is best understood in historical context. The lesbian feminist movement that began in the 1970s has changed and politicised the word “lesbian”. The lesbian label came to embody the concept of resistance to sexism and patriarchy. This is seen as the very embodiment of feminism. According to Bower (1995), men have always had more sexual freedom than women, and the repression of men’s sexuality has varied according to the interest of a particular ruling class. An example is the acceptance of male homosexuality in ancient Greece and Rome. These societies justified male homosexuality on the grounds of a male citizen’s right to use his inferiors - boys, slaves and women - for his own pleasure. Even when both people were male, sex was still an expression of dominance (Foucault, 1984).

78 In the lesbian feminist movement, women who do not sever all ties with men are seen as colluding with patriarchy. Young (1992, p. 80) illustrates this:

When I came out as lesbian, I learned from other lesbian women that bisexuality was a cop-out: it was a label used by women who were really lesbian women but wanted to maintain heterosexual privilege or by women who liked to experiment with lesbian women but were really heterosexual and when push came to shove, would run back to men and leave their lesbian sisters in the lurch.

At the same time, lesbian women were denied participation in organisations such as the National Organisation for Women, on the ground that their “visible presence would hinder the struggle for women’s rights by frightening off potential recruits to feminism from the general population and would feed into stereotypes of feminists as man haters” (Ochs, 1996, p. 230).

Heterosexism and homosexism both continue to privilege a particular sexual orientation, that is, each makes its own sexual orientation normative. Despite the difference in power between heterosexism and homosexism, the lesbian community has its own internal power relations (O’Conner, 1997). Lesbian women have been highly critical of, and to a large degree instrumental in the undermining of, normative heterosexuality. O’Conner (1997, p. 198) is concerned that in “undermining heterosexism, lesbian women have replicated a normative sexuality - a sexuality that excludes bisexuals on the basis of their desires for both men and women”.

Heterosexual individuals who do not feel as if they must prove their sexual orientation may have an interest in not having to think about their sexual orientation; however, bisexuality, more than homosexuality, requires them to do precisely that. It is far less threatening to the dominant heterosexual culture to perpetuate the illusion that homosexuals are “that category, way over there”, and are very different from heterosexuals. If “they” are so different, then heterosexuals do not have to face up to the possibility of

79 acknowledging same-sex attractions within themselves. According to Udis- Kessler (1991, p. 350), there is considerable anxiety in being forced to acknowledge that the “other” is not quite as different from you as you might like, creating a “crisis of meaning” for heterosexuals.

Abbott and Love (1972) acknowledge this crisis. According to them, the most significant group yet to speak up in the women’s movement on the whole matter of sexuality may well prove to be the bisexual women. One reason for the exclusion may be that bisexual women bring out fears of homosexuality in heterosexual women, and fears of heterosexuality in women who live and identify as lesbian women.

4.3.1 Attitudes towards bisexual men and women

The potential aversion to, and negative attitudes towards bisexual men and women can be viewed in classical anthropological terms. Mary Douglas’s (1966) influential interpretation links rituals and rules of social impurity to the rejection of human beings or animals that seem to lack socially approved characteristics, in particular when this lack blurs elementary consensual definitions. Most societies have invented elaborate rituals and ceremonies in order to socialise their young and secure their survival through reproduction in strictly formalised kinship, family and household arrangements based on clear norms of sexual mating. However, according to Shokeid (2001), bisexual men and women have apparently never successfully completed this process. They seem stuck, somewhere in a never-finishing process, unlike homosexuals, who have also failed, but who nevertheless demonstrate clear attributes of sexuality, and whose “position in the landscape of social life is less upsetting because of its stable demarcation” (Shokeid, 2001, p. 2).

Due to the invisibility of bisexual men and women, and because bisexuality is not considered as a valid sexual identity, research on biphobia (negative attitudes towards bisexuality and/or bisexual men and women) has rarely been conducted (Eliason, 1997). This contrasts with attitudes towards lesbian women and gay men. As gay and lesbian people became more visible in the

80 1970s, studies of homophobia (negative attitudes towards homosexuality and/or homosexual people) began to appear in scientific literature.

According to MacDonald (1981), bisexual men and women are less open about their sexuality than any other group. The lack of openness may be due to attitudes towards and stereotypes regarding bisexual men and women that exist in the homosexual and heterosexual communities. Furthermore, Burleson (2005) has identified certain barriers that prevent bisexual men and women from being open about their sexuality.

Firstly, we live in a homophobic society; some people are deterred by the threats of violence and discrimination from identifying as bisexual. The second barrier to identifying as bisexual could be that the person cannot find a bisexual community. Another barrier may be a person’s relationships. For example, people who are bisexual and happily married and are content not to explore their same-sex feelings. One’s faith may also be a barrier as most religions do not accept bisexuality (Burleson, 2005). Therefore, to follow a religion that teaches that bisexuality is wrong and to identify as bisexual may be problematic. Lastly, lack of education can be a barrier. For example, one bisexual man in his fifties who grew up in rural Minnesota started having sex with men in his early teens and with women a few years later. He maintained this pattern all his life until discovering the term bisexuality. According to Burleson (2005), there are probably many other barriers, including the workplace, family and friends.

An autobiographical account by a Muslim woman, Zubeida, to be found in Defiant desire: Gay and lesbian lives in South Africa (Gevisser & Cameron, 1995, p. 191) provides a personal account of the hiding of bisexuality:

I guess I feel oppressed as a bisexual person. Most gay and lesbian organisations don’t really cater for bisexuals - I think largely because bisexuals are even less visible than homosexuals. There is also so much distrust of bisexuals in the homosexual community. Sometimes we are seen as sitting on

81 the fence and enjoying the best of both worlds; usually we are seen as being unable to come out of the closet. I am not out at work. Even though the people I work with would call themselves progressive, I find them incredibly homophobic. Can you imagine how they would react to a bisexual in their midst? And I am not out to my family. They would not be able to cope..

This lack of openness is evident in a study by McLean (2001), where the majority of the 22 bisexual participants kept their sexuality a secret, regardless of the personal costs. Secrecy and lies are seen as preferable to the discrimination they may potentially face from others. McLean (2001) concludes that the secrecy and lies have an enormous impact on self-esteem and on individuals’ ability to accept themselves as valid, worthy individuals.

Research results have confirmed that there are negative attitudes towards bisexual men and women and that bisexual men and women are the targets of prejudicial actions and attitudes. In a community-based study of bias crime, Herek, Gillis and Cogan (1999) found that 15% of bisexual women (n = 190) and 27% of bisexual men (n = 191) have experienced a crime against their person or property because of their sexual orientation. According to the Kaiser Family Foundation’s (2001) telephone survey, 60% of bisexual respondents reported that they had experienced discrimination, 52% had been the target of verbal abuse, and 26% had not been accepted by their families of origin because of their sexual orientation. Furthermore, according to the Kaiser Family Foundation (2001) survey, bisexuals are less likely than gay men and lesbian women to report experiences with prejudice and discrimination.

Herek (2002) examined heterosexual adults’ attitudes toward bisexual men and women. The results of the study (N = 1 335) reveal that respondents’ attitudes toward bisexual men and women were more negative than for all other groups assessed - including religious, racial, ethnic and political groups, except injecting drug users. More negative attitudes toward bisexual women and men are associated with higher age, less education, lower income,

82 higher religiosity, political conservatism, traditional values regarding gender and sexual behaviour, authoritarianism, and lack of contact with bisexual men and women. In a similar vain, Eliason (1996) found that 1130 heterosexual respondents rated bisexuals as less acceptable than lesbian women or gay men. Of the heterosexual sample, 30% rated bisexuals as somewhat to very unacceptable, whereas 22% rated lesbian women and 23% rated gay men as somewhat to very unacceptable. Of the gay and lesbian respondents, 5% rated bisexuals as somewhat to very unacceptable, whereas only 1% rated lesbian women or gay men as unacceptable. Eliason (1997) found that heterosexual students are divided on their attitudes about the acceptability of bisexual women. Fifty percent rated bisexual women as acceptable while 50% rated them as unacceptable. Attitudes towards bisexual men were more negative, with 61% finding bisexual men unacceptable. Heterosexual students were less likely to rate lesbian women (38%) or gay men (43%) as unacceptable. The results of a recent study in Germany (Steffens & Wagner, 2004) are in line with these findings. In a sample of 2 006 self-identified heterosexual men and women, attitudes toward bisexual men and women were less favourable than those toward lesbian women and gay men.

Negative attitudes towards bisexual men and women abound. Firstly, there is the belief that bisexual men and women are promiscuous people who desire concurrent sexual relationships with both men and women and who pursue several relationships at the same time (Hutchins & Ka’ahumanu, 1991). According to one bisexual woman, her partner worried that having “a bisexual identity meant that I was declaring my need to have my cake and eat it too, so the hardest part was convincing her that I did not need a man and a woman in my life to be bisexual” (Hutchins & Ka’ahumanu, 1991, p. 171).

Secondly, bisexuals are believed to be likely to contract sexually transmitted diseases through unsafe sexual practices and to spread these diseases to their heterosexual partners (Sumpter, 1991). McCornick (1994, p. 65) suggests, “prejudiced heterosexuals falsely blame bisexuals for being members of a promiscuous group that is responsible for spreading the AIDS epidemic and other sexually transmitted diseases to the heterosexual

83 population”. The arrival of HIV and AIDS has resulted in increased attention to bisexuality, and added to heterosexuals’ discomfort with bisexuality. Bisexual men and women are regarded as vectors of HIV infection or other sexually transmitted diseases, as bisexuality if conflated with non-monogamy and sexual promiscuity. Articles in the media with titles such as “Is there a man in your man’s life?” (Heller, 1987), “The secret life of a bisexual husband” (Davidowitz, 1993), “No woman is safe” (Avery, 1991), “A perilous double love life” (Gelman et al., 1987) and “The risky business of bisexual love” (Gerrard & Haplin, 1989) perpetuate these stereotypes.

Even though there may be an element of truth in the articles, the media’s narrow focus sensationalises the stereotypes, and dehumanise the people behind and beyond the stereotypes. Worth (2003) points out that it is overly simplistic to look only at the sexual behaviour of bisexual men when there appear to be other equally important factors that codetermine the spread of HIV/AIDS in women. Worth (2003, pp. 73 - 74) relies on social, epidemiological and virological evidence to make her argument by saying that “the rate of new HIV diagnoses amongst men who have sex with men is now in decline in countries worldwide, and the number of women with HIV are rapidly increasing”. According to Worth (2003), the source of 40% of the new HIV diagnoses in the United States of America is sex between men and women, and HIV is now primarily transmitted through drug use and sex between men and women. In addition, according to Rust (1992), we usually hear about bisexuality only in the context of complicated uncomfortable situations. For example, a woman leaves her husband for another woman, a married man contracts HIV from sex with another man or a woman deserts her lesbian partner for a male lover (Ochs, 1996).

Thirdly, bisexuals are seen as likely to deceive a partner about their other romantic affairs, and unlikely to make a lasting commitment (George, 1993). Fourthly, it is believed that bisexuals are sexually talented lovers who enjoy an active sex life, who are highly knowledgeable and open-minded about sexuality, who have expert knowledge about sex and make use of special

84 sexual techniques because of their experiences with both men and women (Hutchins & Ka’ahumanu, 1991; Spalding & Peplaum, 1997).

Lastly, bisexuals are believed to reject the cultural ideal of a monogamous and committed relationship. Bisexuals’ romantic relationships are seen as generally poor in quality - lower in love, satisfaction and intimacy, and higher in conflict than heterosexual relationships (Testa, Kinder & Ironson, 1987). Therefore, bisexual men and women are “destined to become unfaithful lovers who will destroy the hearts (and possibly the health) of their devoted lesbian, gay or heterosexual partners” (McCornick, 1994, p. 65).

Spalding and Peplaum (1997) confirmed the abovementioned negative attitudes mentioned above. Their research results indicated that, compared to heterosexuals, bisexuals are viewed as proponents of sexual openness who are more likely to cheat on a partner, and are viewed as a high-risk group for the transmission of sexually transmitted diseases. Furthermore, bisexual men and women are considered fence sitters who want the best of both worlds or are too indecisive or confused to figure out whether they are lesbian/gay or heterosexual. They are also accused of being unable to practice monogamy, usually because they are thought to require two lovers - a man and a woman - for satisfaction, or because they are believed to be indecisive or noncommittal toward their lovers. However, several research studies have disputed the allegations (Stokes, McKirnan & Burzette, 1993; Rust, 1992) whilst other studies (Einhorn & Polgar, 1994; Lever et al., 2000; McLean, 2004) have confirmed the allegations made against bisexual men and women.

With regards to the perception that bisexual men and women require simultaneous involvement with both male and female partners, research indicates that only a small proportion of self-identified bisexuals are simultaneously involved. Rust (1992) found that of the 527 self-identified lesbian and bisexual women, only 17% of the 45 self-identified bisexual women reported simultaneous involvement. Stokes, McKirnan and Burzette

85 (1993) found that of the 105 bisexual men in their study, only 4% reported simultaneous involvement.

In contrast to the belief that bisexual men and women are unable to have committed relationships, Bressler and Lavender (1986) found that, on average, self-identified bisexual women (3%) have actually had more relationships lasting over six months than either heterosexual (2%) or homosexual (2%) women. However, Weinberg et al. (1994) found that among both men and women, heterosexuals, followed by bisexuals, and then lesbian women and gay men, report the longest relationships. In their longitudinal study, Weinberg et al. (1994) found that many of the bisexual men and women in their sample experienced monogamous relationships, and therefore Weinberg et al. (1994) concluded that many bisexual individuals experience stable relationships.

Conversely, Einhorn and Polgar (1994) reported that bisexuals are significantly less likely than lesbian women or gay men to be involved in a monogamous relationship (40% versus 62%). Correspondingly, Rust (1996) found that those who have identified as bisexual are less likely to be in a monogamous relationship than those who identified as lesbian or gay (16% versus 28%). Similarly, Weinberg et al. (1994), in their sample of bisexual men and women, found that non-monogamy is a common factor in bisexual intimate relationships and takes on a variety of forms: swinging, sexual triads, group sex parties, multiple involved partners and casual sex with friends. Swinging generally involves two or more pair-bound couples who jointly make a decision to switch sexual partners or engage in group sex. Singles may be included either through temporary coupling with another individual specifically for swinging or as part of a triadic or larger group sexual experience (Ramey, 1997). McLean (2004) reports that of 60 participants, 60% of the bisexual men and 52% of the bisexual women indicated that their relationships fall into a category that can be described as open. Twenty-five percent of the men and 35% of the women were in an exclusive relationship. A small number of the sample indicated that the issue of monogamy is still being negotiated, or has not yet been discussed.

86 Being bisexual does not necessarily imply that a person desires an open relationship, nonetheless, Trnka (1992, p. 106) claims that “bisexuality often brings non-monogamy up as an issue”. It is clear from her findings that to be bisexual brings with it a need to reconcile one’s attractions to both men and women with the desire for a committed relationship, and this often results in numerous bisexual men and women choosing to be in some type of open relationship. However, the negative attitudes regarding the relationships of bisexual people are influenced by the powerful Western cultural ideal of monogamy (McLean, 2004). In addition, Robinson (1997, p. 145) states that the ideology of monogamy forces individuals to “fit into neat, well-defined categories which don’t allow for the complexity and reality of the diverse ways in which human beings relate”.

In effect, this implies that relationships falling outside the heterosexual normatively coupled arrangement are not accepted. This has implications for bisexual men and women, many of whom form a variety of relationship types, of which a number are not monogamous, and as a result fail to conform to the accepted monogamous relationship structures of society (McLean, 2004).

In a study by Bradford (2004), 25% of the participants chose non- monogamous relationships. The participants reported having a difficult time with the cultural bias against non-monogamy, which is described as an issue even more unacceptable than bisexuality. They described suffering losses of friends and family due to their choice of non-traditional relationships and spoke of the constant stress of social prejudice, misunderstanding and invalidation.

According to Robinson (1997), not only must bisexual men and women fit into an explicitly heterosexual model of relationships, but also a model that defines a proper relationship as an intimate and closed union between one person and another. Individuals in open relationships find themselves accused of infidelity and polyamory (Robinson, 1997). Polyamory is sometimes also termed non-monogamy or responsible non-monogamy (Anapol, 1997). It refers to a relationship strategy rather than to the specific

87 sexual behaviour within any given relationship. Polyamory is characterised by one or both partners’ interest in carrying on more than one sexual and/or romantic relationship at a time, openly and with the other partner’s consent. Rust (1996) holds that bisexuals’ lower rates of monogamy are not due to inability to commit or practice fidelity, as is stereotypically assumed, but to positive choices in favour of other types of romantic and sexual relationships that might be more effective and stable ways of fulfilling sexual and emotional needs for some people.

McLean (2004) concludes that regardless of the stereotypes that claim that bisexual men and women are adulterous, unreliable and deceitful, the majority of the 60 bisexual participants in the study demonstrated a significant commitment to the principles of trust, truthfulness and communication in their close relationships. Most bisexual individuals in the study tend to prefer, and form significant primary relationships with, one partner and then arrange an open relationship in which they can explore a variety of sexual options. Others had monogamous or exclusive relationships in which having an outside sexual experience was not an option (McLean, 2004).

Bisexuals are also typecast as having more active sex lives than homosexuals or heterosexuals. It is important to remember that research findings regarding the sexual activity levels of bisexuals are based on means and medians. Similar to heterosexuals and homosexuals, bisexual men and women vary greatly in their levels of sexual activity.

Eliason (2001, p. 146) found that the majority of the 229 participants in her study endorsed the stereotype that “bisexuals have more flexible attitudes about sex than heterosexuals”. In a non-random sample of 2135 participants (18 - 49 years) the results indicated that, compared to respondents who have had exclusively heterosexual contact, respondents reporting bisexual contacts have had a significantly higher number of lifetime sex partners, higher frequency of anal and oral sex and masturbation, and lower age at orgasm and masturbation debut (Traeen, Stigum & Sorensen, 2002).

88 Rosario, et al. (1996) found that self-identified bisexual women have had lifetime medians of five male and one female sex partner, and self-identified bisexual men have had lifetime medians of five male and three female partners. Wood, Krueger, Pearlman and Goldbaum (1993) found that those men who identified as bisexual have had an average of eight male and three female sexual partners.

Weinberg et al. (1994) conducted a survey with 96 bisexual women, 116 bisexual men, 105 heterosexual women, 85 heterosexual men, 94 lesbian women and 186 gay men. They found that bisexual men had a median of three male partners in the previous year and 30 in their lifetimes, compared to six in the past year and 100 in their lifetime among gay men. Both bisexual and heterosexual men had lifetime medians of 20 female partners. Bisexual women had a lifetime median of 30 male partners as compared to 25 for female heterosexuals. This suggests that bisexual men and women have more male than female partners. Comments made by respondents in Weinberg et al.’s study (1994) indicate that the prevalence of male sex partners occurs because male partners are easier to find.

Stokes et al. (1993) found a positive correlation between lifetime numbers of male and female partners, indicating that some bisexual men and women are very active with both men and women, whereas others are not active with either. Ekstrand et al. (1994) reported that 13% of the self-identified bisexual men in his study were celibate, and Rust (1992) found that 21% of bisexual women were not involved in any kind of relationship, even a casual dating relationship.

Given that bisexual men and women are perceived to be greater risk for sexually transmitted diseases (STDs) and HIV, some research indicates results that are consistent with the negative perceptions. In 1993, 15 women infected through heterosexual contact were infected through intercourse with bisexual men (Centre for Disease Control, 1990). Weinberg et al. (1994) found that approximately 60% of both the male and the female respondents contracted an STD at least once in their lives. Shafer and Delgado (2002)

89 found that when compared with women who had sex exclusively with men, women who had sex with both men and women are significantly more likely to report past and recent high-risk sexual behaviour. These included sex with an HIV-positive man, multiple male sexual partners, sex with a man who has sex with men, sex with an injection drug user, and anal sex. This behaviour placed this group at a potentially higher risk of HIV and STDs infection than women who are exclusively sexual with either men or women. Similarly, Lever et al. (2000) found that bisexual men reported STDs more frequently (18%) than heterosexual men (10%) but less than exclusively homosexual men (35%). In 1996, Stokes, Taywaditep, Vanable and McKirnan interviewed 536 men who reported that they have had sex with men and women in the past three years. Twenty-seven percent of bisexual men in the sample engaged in unprotected intercourse with both male and female partners.

Information regarding HIV and bisexual men and women is usually lumped in with the homosexual samples. Doll, Myers, Kennedy and Allman (1997) found that in 166 articles mentioning bisexual men over a ten-year period, only 21 pointed out any differences between bisexual men and gay men, and only eight gave information exclusively about bisexual men. In the same ten years, the researchers found only 61 articles mentioning bisexual women, 22 of which compared bisexuals with lesbian women, and only three of which talked about bisexual women exclusively. Therefore, Doll et al. (1997) hold that the prevalence of HIV in the bisexual community is unknown, and it would seem there are more assumptions about the role of bisexuals in the HIV/AIDS epidemic than there are facts.

Declaring an open, bisexual identity in the heterosexual or homosexual communities often results in experiences of discrimination, antagonism and invalidation. This “double discrimination” by the heterosexual and homosexual community is hardly ever recognised or acknowledged. However, biphobia does not come only from the outside.

90 4.3.2 Internalised biphobia

Internalised oppression “refers to the acceptance and internalisation by members of oppressed groups of negative stereotypes and images of their group, beliefs in their inferiority, and concomitant beliefs in the superiority of the dominant group” (Smith, 1997, p. 289). For bisexual men and women, this manifests as internalised biphobia.

Internalised biphobia can be powerful, sometimes overwhelming, and the experience of loneliness, illegitimacy, embarrassment and confusion felt by many bisexuals can be disempowering, even disabling. Ochs (1996) attributes this to the fact that there are few role models available to bisexual individuals. An individual coming to terms with a bisexual identity is likely to feel a sense of isolation.

Due to bisexual invisibility, and the lack of role models or a bisexual community, most bisexuals develop and maintain their bisexual identities in isolation. While homosexuals have been able to identify with a community “that receives social and emotional support, bisexuals do not have a community in which they can belong” (Deacon, Reinke & Viers, 1996, p. 293). Without social recognition, there can be no bisexual community comparable to the gay community with its networks and institutions. Moreover, without community, there are no reference groups, supportive norms, or available symbols to counter the pull toward the two extremes of the continuum.

Therefore, many bisexual men and women have reported that they feel the pressure not to identify as bisexual, but to lie and hide their identities (McLean, 2001). In effect, the bisexual learns to hide twice, once from the mainstream society that rewards the establishment of heterosexuality, but also from the gay and lesbian community where announcements of bisexuality may be met with suspicion and distrust. In the light of the ungenerous reception accorded to bisexuals by the heterosexual and homosexual communities, it is hardly surprising that individuals’ sexual

91 biographies and sexual-identification often prove to be imperfect (Blumstein & Schwartz, 1977).

Because of the absence of a bisexual community, bisexuals spend the most of their time in the community that corresponds with the sex of their romantic partner. This can result in a sense of discontinuity if partners are changed and the next partner is of a different sex. As a result, the bisexual has to shift back and forth between two differing communities over time (Ochs, 1996). The consequences of attempting to maintain two independent social lives and identities (that is, in heterosexual society and in the gay and lesbian community) were recognised as early as 1937. According to Stonequist (1937/1961, pp. 145-146) the bisexual is vulnerable to the confusion arising from a dual sense of self:

We develop an idea of ourselves through imagining how we appear to other persons, and imagining their judgment of that appearance. Thence arises a self-feeling ranging from pride to mortification. In the case of the marginal man, it is as if he was placed simultaneously between two looking glasses, each presenting a sharply different image of himself. The clash in the image gives rise to a mental conflict as well as to a dual self- consciousness and identification.

Many people privately identify as bisexual, but to avoid conflict and to preserve their ties to a treasured community, choose to label themselves publicly as lesbian women, gay men or heterosexual, further contributing to bisexual invisibility. These men and women feel terror at the thought of being cast out or ostracised from the community from which they derive their support. Because an individual of an oppressed group is frequently seen as representative of all the members of that group, a bisexual-identified person may feel embarrassment when any bisexual person behaves in such a way as to strengthen negative stereotypes of bisexual people.

92 Studies further indicate that bisexual men and women are more “closeted” (non-disclosure of sexual orientation) and less likely to disclose their bisexuality to relatives, friends and colleagues. Smith, Johnson and Guenther (1985) found that bisexual women are more likely than lesbian women not to disclose their sexual orientation to a physician. Doll, Peterson, Magana and Cortier (1991) found that 24% of self-identified bisexual respondents, compared to 7% of self-identified homosexual respondents, have never self- disclosed their same-sex activity to others.

Stokes et al. (1993) reported that 19% of the bisexuals in their study said that they believed that none of the ten most important people in their lives knew about their bisexuality. This non-disclosure could reflect internalised biphobia. Stokes et al. (1993) found that internalised biphobia and the perceived reaction from friends and family members are among the best predictors for non-disclosure. Non-disclosure may have significant impact on the psychological health of bisexual men and women.

4.3.3 The psychological health of bisexual men and women

Health is a state of complete physical, mental and social well-being and not merely the absence of disease or infirmity. It is the extent to which an individual or group is able, on the one hand, to realise aspirations and satisfy needs and, on the other hand, to change or cope with the environment (National Centre for Health Statistics, 2005). According to Lockwood and Bursey (2001), being healthy is about having the skills and resources to meet your needs and goals, and to change or cope with what life brings. It is about you as an individual and the broader environment in which you live.

A long tradition of research literature has examined the extent to which negative attitudes towards homosexuality have affected the psychological health of lesbian women and gay men. How much more damaging might the denial of one’s very existence be for bisexuals?

93 Research indicates that bisexual men and women may have greater mental health challenges than those of other sexual orientations. Findings indicate that bisexual responses in surveys are higher for anxiety, depression, suicide and negative affect than lesbian, gay and heterosexual responses (Balsam & Rothblum, 2002; Jorm, Korten, Rodgers, Jacomb & Christenson, 2002). In addition, Jorm et al. (2002, p. 423) have found that, compared to homosexual and heterosexual participants, bisexual men and women experience more “current adverse life events, greater childhood adversity, less positive support from family and more negative support from friends”.

Page (2004) uses the following experiences to categorise a participant as having more serious mental health issues: a major illness; a major mental illness, a suicide attempt, participation in residential, rehabilitation, or day treatment services or a psychiatric hospitalisation. The results indicate that of the 217 self-identified bisexual men and women, 18% of the participants had serious mental issues. Two-thirds of the participants experienced moderate levels of stress related to their sexual orientation.

The literature specifies several factors associated with stress for bisexual men: isolation, loneliness, lack of community and reluctance to disclose their sexual orientation to family and friends (Rust, 2001; Steinman, 2001). Bisexual men experience pressure based on heterosexism, conflict with class and family norms (Appleby, 2001) and the assumption that bisexual men transmit AIDS to heterosexuals (Stokes et al., 1993).

Studies by Cochran and Mays (2000) and Cochran, Sullivan and Mays (2003) also describe bisexual men as experiencing a variety of symptoms, including drinking, depression, panic attacks, psychological distress and suicidal ideation related to sexual orientation. In an earlier study by Rotheram-Borus, Hunter and Rosario (1994), the researchers have found that, compared to rates found in general youth populations, gay and bisexual youths do not differ appreciably in drinking, drug use, conduct problems or emotional distress, yet they are more likely to attempt suicide.

94 McNair, Kavanagh, Agius and Tong (2005) compared the mental health status of early adult and mid-life women according to sexual orientation. The results indicated that younger, mainly heterosexual, bisexual and lesbian women had poorer mental health outcomes than exclusively heterosexual women. In the mid-age sample mainly heterosexual women experienced poorer mental health on all outcomes except for medically diagnosed anxiety. In addition, bisexual women had significantly higher odds of self-harm than exclusively heterosexual women. All non-heterosexual women in the sample reported higher levels of stress and lifetime abuse.

Jorm et al. (2002) presented findings from their research on mental health issues for bisexual men and women at the seventh International Bisexual Conference in Sydney, Australia, in 2002. The researchers established that a significant proportion of lesbian women, gay men and bisexuals in their sample reported what they considered risk factors for mental health issues, compared to their heterosexual counterparts. Furthermore, bisexual men and women reported a higher incidence of the same risk factors as the lesbian women and gay men in their sample. The risk factors identified include adverse childhood experiences, current adverse life events, poor social support and financial difficulties.

The impact of negative attitudes towards bisexuality and the importance of acknowledgement and validation are addressed in the American Psychological Association’s (APA) Guidelines for Psychotherapy with Lesbian, Gay and Bisexual Clients (Division 44/Committee on Lesbian, Gay and Bisexuals Concerns Joint Task Force on Guidelines for Psychotherapy with Lesbian, Gay and Bisexual Clients, 2000, p. 1445). This separate section acknowledges and addresses the impact that negative stereotypes and attitudes about bisexuality in society and within psychology have on bisexual women and bisexual men.

Bisexual adults and youth may experience a variety of stressors in addition to the societal prejudice due to same-sex attractions. One such stressor is that the polarisation of sexual orientation

95 into heterosexual and homosexual categories invalidates bisexuality. This view has influenced psychological theory and practice as well as societal attitudes and institutions.

In addition, a set of guidelines for psychotherapy with lesbian, gay and bisexual clients was published in 2000, reaffirming the profession’s position that homosexuality/bisexuality is not a mental illness (Division 44/Committee on Lesbian, Gay and Bisexuals Concerns Joint Task Force on Guidelines for Psychotherapy with Lesbian, Gay and Bisexual Clients, 2000). The guidelines strongly recommend that therapists proactively learn about bisexual issues and bisexual community resources in order to enable them to provide competent, and effective mental health services to this population. This approach is supported by Page’s (2004) research findings on the psychotherapy experiences of bisexual women and bisexual men. The researcher found that negative attitudes towards bisexual men and women do in fact affect bisexual clients in terms of their experiences of psychotherapy. An overwhelming proportion of the research participants advocated for greater therapist knowledge and education about bisexuality and bisexual issues.

Regardless of the above statements by the APA, in that same year the organisation published these guiding principles, articles were published in psychological journals that labelled homosexuality a form of psychopathology (Stone, 2000), or supported the practice of conversion therapy (Nicolosi, Byrd & Potts, 2000), a therapeutic approach previously condemned by the APA (APA, 2000). In addition, in 2004, the board of directors of the major accrediting body for marriage and family therapists and their supervisors, the American Association of Marriage and Family Therapists (AAMFT), took the stance that: “while prepared to declare homosexuality is not, in and of itself, a diagnosable mental illness, they were not yet ready to say the same about bisexuality” (Keppel, 2006, p. 87). The contradictions embodied by these facts are a reflection of the persistent social ambivalence engendered by the topic.

96 Page (2004) questioned the bisexual participants in her study about the most important issue or problem they faced in being a mental health consumer and a bisexual. The most frequent themes concerning this were invalidation of bisexuality, lack of knowledge about bisexual issues, interpretation of bisexual attractions and behaviours as unhealthy, lack of skill in working with bisexual issues and lack of proactive interventions.

According to Zipkin (1992, p. 59), many bisexual men and women feel that their bisexual desire is a burden. As one woman said: “being bisexual is a major conflict in my life. It involves more pain than pleasure; I would prefer to be one or the other. I do not care which. I would just like one clear identity. That would be a lot simpler”. Anderson and Randlet (1994) asked their respondents a series of five questions regarding their satisfaction with their lives as bisexuals, lesbian women and gay men, and ascertained that bisexuals were significantly less satisfied than lesbian women and gay men.

Evidently, bisexual men and women face many difficulties in attempting to form a positive, well-integrated bisexual identity. Christina (1995, p. 166) suggests that bisexual men and women should look at the way that lesbian women and gay men have stopped defending homosexuality and are now attacking homophobia. They have turned the debate around, away from “There’s nothing wrong with us, please accept us” to “What’s wrong with you that you don’t accept us?” Bisexual men and women must learn to do this, not only about their sexual orientation, but also about sexuality itself.

Bisexual men and women know what it is to have the sex fears of mainstream heterosexual culture projected onto them. However, bisexuals cannot defend themselves by embracing these fears as their own. Bisexual men and women should confront sexual stereotyping - not defensively, but with honesty and pride (Christina, 1995). However, Ochs (1996, p. 235) argues that, “Our conditioning, invisibility, and the negative images that surround us make it extremely difficult to feel an unqualified sense of pride in our bisexuality”.

97 Paul (1984) believes that there is potential for constructive social change to emerge from the current difficult position of bisexual men and women in our society. To resolve the experience of being associated with two opposing camps, while being an outsider in relation to both, requires that the bisexual individual discards the common frame of reference and adopts a broader, enhanced and integrated perspective on human sexuality and social relationships (Paul, 1984).

According to Loewenstein (2000), academics, social workers, psychologists and students of human development need to heighten their awareness of the great diversity of human behaviour and human experience. The more we learn about sexuality, the more we realise that textbook concepts of normality need to be expanded and revised. The first and last task of all humanistic research is always to clarify and enhance human experience (Hansen & Evans, 1985).

4.4 Chapter summary

The stigma around and negative attitudes towards bisexual men and women were discussed in this chapter. The marginalization of the bisexual community by heterosexual, gay and lesbian people that results in the relative invisibility of bisexual men and women was explained. Stereotypes regarding bisexuals were presented, along with research indicating that there was little truth attached to the stereotypes.

The issue of internalised biphobia in bisexual individuals was investigated. Due to the negative attitudes towards bisexual men and women, bisexuals tend to feel isolated and keep their identity hidden. A fear of rejection by both heterosexual and homosexual communities, along with a lack of a visible bisexual community often causes a bisexual man or women to remain closeted, or to simply associate with the community that corresponds with the sexual orientation of their current partner.

98 Research also shows that bisexual individuals may be more prone to mental health concerns compared to their heterosexual or gay and lesbian counterparts. Considerations for therapy and reducing the causes of these mental health issues were discussed.

4.5 Preview of the contents of the following chapter

The research method chapter is divided into three subsections: namely the first phase (quantitative method), the second phase (expert group advice) and the third phase (quantitative method). In the first section, The Attitudes Regarding Bisexuality Scale (ARBS) (Mohr & Rochlen, 1999), the self- identification instrument indicating the participants’ sexual orientation (Berkey et al., 1990), the biographical questionnaire and the procedures followed are discussed. In the second section, the mixed research method (expert group advice) will be discussed in terms of procedures and participants. This is followed by a discussion of the procedures, participants and instruments utilised in the third phase.

99 CHAPTER 5

RESEARCH METHOD

5.1 Introduction

This chapter is divided into five subsections. The aim of the study, the research questions and the hypotheses are discussed first. This is followed by a discussion of the research method, procedures, instruments and participants of the first phase (quantitative method). Thirdly, the research method, participants and procedures of the second phase (expert advice) are presented. Subsequently, the research method, procedures and participants of the third phase (quantitative), are discussed. Finally, an overview of the statistical analyses used in the study is presented.

5.2 Aim of the study

Researchers have noted the invisibility of bisexuality in public and academic spheres, as well as the negative impact of negative attitudes on the lives of bisexual men and women (Firestein, 1996; Ochs, 1996). Despite this, there appears to be an absence of interest in this sexual minority in South Africa. Given this absence of interest as well as the fragmentary state of knowledge on bisexuality, the aim of the research was to investigate attitudes toward bisexual men and women in a university setting in South Africa. Knowledge of the socio-political context that bisexual men and women face is a prerequisite to effective counselling with this population (Queen, 1996).

5.3 A three phase research method

A mixed research method was utilised to assess attitudes towards bisexual men and women in a university community. The research was conducted in three phases: a quantitative phase, employing a survey; a qualitative phase,

100 utilising an expert advice group; and a third phase utilising, once again, a survey.

The first phase yielded unreliable results, with clear differences between the black and white students’ interpretation of the items of the ARBS-FM. To gain understanding of and possible explanations for these differences, an expert group was organised. Therefore, in the current study, a qualitative phase was implemented to inform the quantitative phase sequentially. After analysing the findings of the expert advice group, and revising the instruments accordingly, another survey (quantitative phase) was conducted.

5.4 Research questions

As a result of the literature review, the following research questions were formulated:

●Do homosexual and heterosexual students have negative attitudes towards bisexual men and women?

●If negative attitudes exist, what are the most significant variables related to biphobia? (Research variables include: ethnicity, gender, knowing a bisexual man or woman, knowing a gay man or lesbian woman and religiosity).

5.5 Hypotheses

In view of the literature review, the relationships between respondents’ attitudes toward bisexual men and women and various variables were explored.

Hypothesis one: Black and white students have different attitudes towards bisexual men and women.

Hypothesis two: Male participants and female participants have different attitudes towards bisexual men and women.

101

Hypothesis three: Students who know a bisexual man or woman have different attitudes towards bisexual men and women when compare to students who do not know a bisexual man or woman.

Hypothesis four: Students who know a lesbian woman or gay man have different attitudes towards bisexual men and women when compare to students who do not know a lesbian woman or gay man.

Hypothesis five: Students who are more religious have different attitudes towards bisexual men and women than students that are less religious.

Hypothesis six: Homosexual and heterosexual students have different attitudes towards bisexual men and women than bisexual students.

5.6 Phase one: Quantitative research method

Phase one of the research employed a survey design. Research into the implications of sexual orientation in communities is often difficult due to the highly sensitive nature of the topic and consequent need for strict confidentiality and anonymity. Surveys are a commonly used method of eliciting information on sensitive topics. Surveys have been used successfully in studies ranging from sexual behaviour and attitudes towards lesbian women and gay men (Arndt & De Bruin, 2006; Herek, 1995; Tourangeau, Rasinski, Jobe, Smith & Pratt, 1997; Tourangeau & Smith, 1996) to drug use (Turner, Lessler & Gfroerer, 1992) and abortion (London & Williams, 1990). Surveys improve reporting of sensitive information by increasing privacy and reducing social desirability effects associated with interviewer administration (Schober, Fe Caces, Pergamit & Branden, 1992).

5.6.1 Procedure

Prior to commencing the study, the research proposal was approved by the Higher Degrees Committee of the relevant faculty and the ethical implications

102 approved. The researcher obtained permission from lecturers to administer the questionnaires in the classes during lecture times. The cover page of the questionnaire provided the participants with information regarding the purpose of the study, the right to voluntary participation and confidentiality. The instruments required between 10 and 15 minutes to be completed.

5.6.2 Participants

A total of 1 459 undergraduate university students (681 men, 932 women, 6 missing data) were selected by convenience sampling (Gravetter & Forzano, 2006). With regard to ethnicity, 926 students were black, 533 white, 92 coloured and 65 Asian (3 missing data).

The field of psychology has been criticised for an over-reliance on university samples - introductory psychology students in particular - and the practice of the generalisation of these findings to the larger population. However, as the literature on coming out indicates (Herek, 1995; Klein, 1990) the university years are a common time for individuals to begin to come out, either as bisexual or as gay or lesbian. These individuals often undergo identity transformation during their university years; they acknowledge their sexual orientation to themselves and others.

5.6.3 Instruments utilised in the first phase

The aim of the current study is to measure attitudes towards bisexual men and women in a university setting. A survey of 880 heterosexual students in a university community was conducted by Arndt and de Bruin (2005) utilising the Attitudes towards Lesbian and Gay Male Scale (ATLG). The results indicated that the ATLG was culturally appropriate and statistically sound for the South African context. As a result the researcher utilised a similar instrument, Attitudes Regarding Bisexuality Scale, to measure attitudes towards bisexual men and woman.

103 5.6.3.1 Attitudes Regarding Bisexuality Scale (ARBS-FM)

Until the development of the ARBS, various attempts were made to assess attitudes towards bisexual men and women. This was because of the increasingly visible and vocal community of self-identified bisexual men and women who helped to amplify awareness of the negative attitudes and stereotypes that bisexual men and women face (Hutchins, 1996). Furthermore, research findings (Eliason, 1997; Rust, 1995; Udis-Kessler, 1996) suggested that negative attitudes towards bisexuality are prevalent among heterosexual men and women, as well as lesbian women and gay men, and these attitudes towards bisexual men and women are not identical to those towards lesbian women and gay men (Mayfield & Carrubba, 1996; Queen, 1996; Steffens & Wagner, 2004).

Rust (1995) conducted a large-scale study assessing lesbian women’s attitudes towards bisexual women using a questionnaire developed specifically for that study. However, the reliability and validity of the instrument is limited because of the reliance on single-item measures of attitudes (Nunnally, 1978).

Eliason (1997) also developed a single-item scale measuring tolerance towards bisexuality. This scale exhibited good test-retest reliability (Pearson r = 0.77) over two weeks. The strength of this scale is that it can be used to compare attitudes toward individuals of different sexual orientation and gender. However, its exclusive focus on tolerance may limit the types of research questions that it is capable of addressing (Mohr & Rochlen, 1999) and reliability and validity may be limited in comparison to scales composed of multiple items (Nunnally, 1978).

Mayfield, Carruba and Louie (1995) developed the Attitudes toward Bisexuality Inventory (ATBI), a 24-item self-report inventory that measures the degree to which bisexuality is viewed as a legitimate and acceptable sexual orientation. The two subscales: normality/abnormality of bisexuality and non- existence of bisexuality, exhibited high internal consistency and validity. A

104 limitation of this Inventory is that it does not specify the gender of the bisexual target. Therefore, the ATBI does not allow researchers to address questions regarding differences in attitudes toward bisexual men and women. In addition, data on the psychometric properties of the ATBI have not been provided for lesbian woman and gay male populations (Mohr & Rochlen, 1999).

Mohr and Rochlen (1999) reviewed the empirical and theoretical literature in this area and attempted to form a conceptual framework that would provide a foundation for item generation. Although there is a growing acceptance for bisexuality and homosexuality, societal beliefs in the moral superiority of heterosexuality still prevails (Fox, 1996). Mohr and Rochlen (1999) therefore hypothesised that one aspect of attitudes towards bisexuality would relate to the degree to which individuals view bisexuality as a morally tolerable sexual orientation. In addition, discussions of attitudes regarding bisexuality also include attitudes that are distinct from the domain of tolerance. These attitudes include ideas that bisexual men and women are indecisive, untrustworthy partners, who are incapable of monogamy, confused individuals who are in transition from one sexual orientation to another or lesbian women and gay men who are in denial about their actual sexual orientation (Eliason, 1997; Udis-Kessler, 1996; Weise, 1992).

Mohr and Rochlen (1999) developed three scales: the ARBS-F, the ARBS-M and the ARBS-FM. The ARBS-F and ARBS-M measure attitudes regarding bisexual females and males on separate scales. The ARBS-FM measure attitudes towards bisexual females and males simultaneously. The ARBS-FM (Mohr & Rochlen, 1999) consists of two subscales. The first subscale, Stability (10 items), looks at the degree to which bisexuality is viewed as a legitimate, stable sexual orientation and bisexual people are seen as capable of forming committed romantic relationships. Tolerance (8 items), the second subscale, considers the degree to which bisexuality is viewed as moral, tolerable, and not harmful to society and attitudes reflect acceptance rather than disdain for bisexual men and woman. Items are rated on a 5-point Likert scale ranging from one (strongly disagree) to five (strongly agree).

105 The results of various studies (Mohr, Israel & Sedlacek, 2001; Mohr & Rochlen, 1999) indicated that items on the ARBS-FM evidenced moderate to high internal consistency reliability in several lesbian, gay and heterosexual samples. The factor structure of the scale was also supported in heterosexual, lesbian, and gay samples. In the first phase, one of the ARBS-FM items that read American values, was adjusted to provide for the South African population and therefore substituted with South African values.

5.6.3.2 Sexual orientation instrument utilised in the first phase

As scientists commence to treat sexual orientation as a demographic variable similar to religion or race, it is crucial to clarify the definition of sexual orientation in terms of the current study. Different definitions have been proposed and used to develop samples since the 1860s when sexual orientation first gained research interest. In addition, several literature reviews have found that researchers rarely included conceptual definitions in reports of their research (Sell, 1997).

The term sexual orientation has a wide variety of definitions in the literature, but generally comprises either a psychological or a behavioural component, or both. For the purpose of the current study, Le Vay’s (1993) concept has been used, which includes both psychological and behavioural components. According to Le Vay (1993) sexual orientation is the direction of sexual feelings and/or behaviour towards individuals of the opposite sex, the same sex or some combination of the two.

The research reviewed in this study used inclusion criteria that relied on respondents’ self-identification as lesbian, gay or bisexual. Other studieshave used reports of a history of same-gender sexual partners, using different time frames to classify individuals. Still others have used sampling location such as recruitment from a venue frequented by lesbian women, gay men or bisexual men and women and heterosexual men and women. Unfortunately, the definitions and measurement instruments used have resulted in the selection of divergent and often incomparable samples. Despite much debate around

106 essentialist and social constructionist views of sexual orientation, most studies continue to determine sexual orientation by means of self-identification (Broido, 2000). The results of a longitudinal study indicated that self- identification strongly corresponds with actual behaviour (McKirnan, Stokes, Doll & Burzette, 1995).

According to Gonsiorek et al. (1995) categories to indicate sexual orientation should capture complexities, yet maintain interpretability. In view of the literature mentioned previously, for the purpose of the current study the researcher utilised the categories of Berkey et al. (1990). Berkey and associates (1990) generated nine categories of sexuality, which take into account changes over time and reflect the diversity of potential sexual orientations. In addition, descriptions for each of the categories were provided.

Based on this, the participants in the first phase of the research were required to indicate which one of the sexual orientations applies to them:

Exclusively heterosexual: I have always considered myself strictly heterosexual.

Heterosexual with some homosexuality: I currently consider myself heterosexual, but occasionally I am attracted to, desire, or have sexual contact with members of my same sex.

Concurrent bisexual: I consider myself bisexual because I have equal attraction towards, desires for or sexual contact with members of my same and the opposite sex on a regular basis.

Sequential bisexual: I consider myself bisexual because I am equally attracted to, desire, or have sexual contact with members of both my same and the opposite sex. I tend to feel attracted to, desire, or have sexual contact with members of one sex for a period of time and then switch and feel attracted to,

107 desire, or have sexual contact with members of the other sex for a period of time.

Homosexual with some heterosexuality: I currently consider myself homosexual, but occasionally I am attracted to, desire, or have sexual contact with members of the opposite sex.

Past heterosexual, current homosexual: In the past, I considered myself heterosexual, but now I consider myself strictly homosexual.

Exclusively homosexual: I have always considered myself strictly homosexual.

Past homosexual, current heterosexual: In the past, I considered myself homosexual, but now I consider myself strictly heterosexual.

Asexual: I do not consider myself homosexual, heterosexual or bisexual. I have never had any attraction to, desire for, or sexual contact with members of either my same or the opposite sex.

5.6.3.3 Biographical questionnaire utilised in the first phase

A biographical questionnaire was constructed to obtain relevant personal information regarding individual respondents. Emphasis was placed on acquiring information that allowed for the testing of the hypotheses in respect of the variables of interest. The biographical questionnaire also served to identify certain features specific to the sample. The information pertained to age, gender, ethnic group, whether the participant knows a bisexual man or woman, whether the participant knows a gay man or lesbian woman, degree of religiosity and sexual orientation. The relationships between respondents’ attitudes toward bisexual men and women and the following demographic variables were explored:

108 Gender: Gender was reported to be one factor that influences attitudes towards bisexual men and women. Male participants tended to rate bisexual men and women less favourably than female participants (Herek, 2002).

Degree of commitment to religion: According to Herek (2002), deeply religious respondents demonstrate less favourable attitudes towards bisexual men and women.

Race: The first phase yielded unreliable results, with clear differences between the black and white students’ interpretation of the items of the ARBS- FM. Given the lack of research on race and bisexuality, the researcher subsequently included studies that looked at race and homosexuality.

There is a commonly held belief within South Africa that homophobia is more prevalent in the black community than in society at large. This homophobia has been cited as a contributing factor in slowing mobilisation against AIDS, as an obstacle to black lesbians and gay men coming to terms with their sexuality and as a challenge to the legitimacy of the gay rights movement (Brandt, 1999).

Research evidence suggesting that attitudes held by black and white heterosexuals toward lesbian women and gay men differ, is inconclusive. Levitt and Klassen (1974) found that white heterosexuals have significantly more negative attitudes towards lesbians and gay men than do black heterosexuals. Hudson and Ricketts (1980), Schneider and Lewis (1984) and Waldo (1998) found the opposite result. Research conducted by Waldo (1998) at the University of Illinois found that black students were more supportive of policies affirming lesbian and gay male students and more open to interpersonal contact with lesbian and gay male students. According to Waldo (1998) this was a result of the black students’ own status as minorities in the university community.

Recent research in the USA by Davis and Brown (2000), Herek and Capitanio (1996) and Sawyer (2000) reported that a higher percentage of

109 heterosexual black women (compared to their white counterparts) said that AIDS will help society by reducing the number of homosexuals, but no significant difference was found between heterosexual black and white men. Davis and Brown (2000) and Sawyer (2000) found greater acceptance of homosexuals among black women than black men. These results reiterate the role that gender plays in attitudes towards lesbians and gay men. However, these studies only included moderate numbers of black participants, 391 for Herek and Capitanio (1996), 800 for Davis and Brown (2000) and Sawyer (2000) included no white participants. Therefore, the above studies do not provide conclusive proof of ethnic differences.

Research has typically addressed racial differences only in passing, in studies focused on other issues, based on samples with small numbers of black participants (Lewis, 2003). In an attempt to rectify this, Lewis (2003) used responses from 7000 black participants and 43000 white participants in 31 national surveys conducted between 1973 and 2000 in order to give more definite answers on racial attitudinal differences towards lesbians and gay men. The findings suggest that black heterosexuals are 11 percentage points more likely than white heterosexuals to condemn homosexual relations as “always wrong” and 14 percentage points more likely to see them warranting Gods punishment in the form of AIDS, but no more likely to favour criminalising homosexuality. More black than white heterosexuals would remove a pro-gay book from their library (by 6 percentage points) and would not allow an admitted homosexual to give a speech in their community (by 4 percentage points). When the issue turns to employment, ethnic differences nearly disappear. Strikingly, black heterosexuals are 10 percentage points more likely than white heterosexuals to support a law prohibiting anti-gay employment discrimination (Lewis, 2003).

These differences in attitude are thought to result in greater stigmatisation of lesbians and gay men in black communities, causing more closeted behaviours and subsequently producing more stress (Rose, 1998). The greater stigmatisation of homosexuality is thought to cause black homosexuals to have difficulties coping with their generally more traumatic

110 life histories, rejection by black churches and to some extent, differential treatment by the white homosexual community (Battle & Lemelle, 2002). They also experience more disapproval from their families than do similar white homosexuals (Lewis, 2003).

Potgieter (1997; in Greene, 1997) conducted research on sexual orientation in the South African context, and interviewed homosexuals in the black community. According to the research, the family’s reaction to black lesbian women and gay men’s disclosure of their sexual orientation was to get them healed, and they were taken to traditional healers. The discourse of the family seems to be that the homosexual person is sick and in need of healing. Family members felt that it was their responsibility to take care of the sick relative, whether they approve of the sick relative’s conduct or not. The problematic assumption of the homosexuality as sickness within the black South African context ironically leads to a mandate for family members to maintain contact and support lesbian and gay male family members rather than reject them. The findings also suggest that the women in the families were more understanding than the men, that homosexuality was not spoken about, and that when members of the community have been educated, they have shown more understanding.

All the above findings demonstrate the inherent dangers of extrapolating from American samples to the rest of the world, and encourage the growth of cross-cultural psychology in non-Western countries. Additional research focusing on black South African attitudes toward homosexuality may be necessary to develop an effective, culture-specific campaign against homophobia.

According to Lewis (2003) racial religious differences may contribute to attitudinal differences. Beliefs about homosexuality vary substantially by religion, and by intensity of religious feeling (disapproval is highest among black heterosexuals who attend religious services frequently, who pray frequently, and who say that religion is very important in their lives).

111 Schulte and Battle (2003) found that racial differences in attitudes towards lesbian and gay male students disappear once religion is controlled.

In studies measuring attitudes towards lesbian women and gay men (Battle & Bennett, 2000; Lewis, 2003) the results indicated that black respondents had more negative attitudes than white respondents. However, recent research amongst the South African population (Arndt & De Bruin, 2006) has found that race did not have a statistically significant effect on attitudes towards lesbian women and gay men.

Contact with bisexual men, bisexual women, lesbian women and gay men: According to research findings, the relationship between contact with bisexuals and favourable attitudes towards bisexuals was stronger when respondents reported contact with lesbian women, gay men, and bisexual men and women (Arndt & De Bruin, 2006; 2004; Herek, 2002).

5.7 Phase two: Expert advice

The first phase yielded unreliable results, with clear differences between how black and white students interpreted the items of the ARBS-FM. To gain an understanding of this phenomenon, a qualitative component was added to the research in the form of an expert advice group. Morgan (1997) holds that survey researchers have noted the value of combining their work with expert groups. Regardless of the scant body of literature that exists in this area, this combination can lead to thorough mixed methods research (Morgan, 1997).

Santos (1999) suggests that if a scale shows poor reliability, individual items within the scale should be re-examined and modified or completely changed as needed (Santos, 1999). Based on this, the researcher decided to organise an expert group to determine possible reasons for the significantly low internal consistency reliability of the ARBS-FM. The purpose of the group was complementary to the quantitative study. The researcher sought elaboration and clarification of the results of one method through the results of another method (mixed method research) (Greene & Caracelli, 1997).

112

Mixed method research refers to the combination of qualitative and quantitative methods in research. According to Johnson and Onwuegbuzie (2004, p. 17), mixed methods research is formally defined as “the class of research where the researcher mixes or combines qualitative and quantitative research techniques, methods, approaches, concepts or language, into a single study”.

The majority of mixed method research designs are developed from two major types of mixed methods research: mixed model approaches (mixing qualitative and quantitative approaches within or across the stages of the research process) and mixed method approaches (the inclusion of a quantitative and a qualitative phase in an overall research study) (Johnson & Onwuegbuzie, 2004). Particularly applicable within the current research was the fact that a mixed methods research design emerged during the study, as a result of information that was obtained during the research process. To construct a mixed-method design, the researcher needs to state whether he or she operates largely within one dominant paradigm or not, and whether he or she conducts the phases concurrently or sequentially (Johnson & Onwuegbuzie, 2004).

The sequential explanatory designs do not use an explicit advocacy lens. In these designs, quantitative data are gathered and analysed, followed by qualitative data. Priority is unequal and is given to the quantitative data. According to Hansen, Creswell, Clark, Petska and Creswell (2005) these designs are functional for explaining study findings, especially when they are unexpected. In addition, quantitative researchers are becoming increasingly aware that some of their data may not be accurate and valid, because some survey respondents may not understand the meaning of questions to which they respond. It is also increasingly recognised that all data collection— quantitative and qualitative—operates within a cultural context. Patton (1990) recommends that when investigating human behaviour and attitudes, it is most fruitful to use a variety of data collection methods.

113 A mixed method approach may also lead researchers to modify the evaluation design or the data collection method. Therefore, the mixed method approach may guide researchers to uncover inconsistencies and discrepancies that alert the researcher to the need for re-examining the data collection and analysis procedures used (National Science Foundation, 1993). In addition, adaptation of existing instruments for use with different populations necessitates identifying appropriate items for inclusion. According to Nassar- McMillan and Borders (2002), engaging these populations as expert group participants can provide an efficient means for item refinement.

5.7.1 The aim of the expert advice group

Expert advice groups are becoming an important method used in adjunct to quantitative research. Expert advice groups can be used at the exploratory stages of a study (Krueger, 1988), during the study to evaluate a particular activity (Race, Hotch & Parker, 1994), or after a study to assess an activity. Expert advice groups can also be used as a method in its own right, or as a complement to other methods, and more specifically for validity checking (Morgan, 1988). In addition, the expert advice group elicit information in a way which allows researchers to ascertain why an issue is salient or not. Participants of the expert advice group reveal multiple understanding and meanings, and multiple explanations and attitudes are articulated in a relatively short period of time (Morgan, 1988).

According to Knobel (1993), expert advice groups make it easier to detect when participants’ failure to understand a questionnaire item. Of particular importance, according to Morgan (1997), is that at later stages of a survey, once the data are in and the analysis completed, expert groups can be valuable for follow-up data collection that pursues aspects of the analysis. This is especially important when the results are puzzling to the researcher.

In view of the previously mentioned research, the researcher assembled an expert advice group to determine possible reasons for the significantly low internal consistency reliability of the ARBS-FM. Therefore, in the current

114 study, a qualitative phase was conducted to inform the quantitative phase sequentially.

5.7.2 The role of the researcher and ethical considerations

Based on Krueger’s (1988) recommendations for facilitating a group the researcher promoted debate by asking open-ended questions; challenging participants to tease out a diverse range of meanings on the topic under discussion; probed for details; kept the session focused and ensured that all participants got a chance to speak. Participants were encouraged to keep the group discussion confidential. The researcher undertook to anonymise data obtained from the group participants (Holman, 1991).

5.7.3 Participants of the expert advice group

Table 5.1 Expert Group Participants

Black men 1

Black women 4 Coloured women 2 White women 1 Asian men 1 Asian women 2 Total 11

According to Hambleton, Merenda and Spielberger (2004) there is now a growing awareness that the adaptation of items requires expertise in the language and culture of the target group as well as in item writing. Therefore, the new trend in adapting items is to rely on a team effort, aimed at maximising the quality of the use of the instrument in a cross-cultural setting. The group consisted of nine intern psychologists, a counselling psychologist specialising in cross-cultural psychology, and the supervisor of the current study (Table 5.1). Although this group does not necessarily represent all the different class and racial groups in the South African context, care were taken to convene a heterogeneous group. The recommended number of people for

115 a group is usually between six and ten (MacIntosh, 1993). The session lasted three hours and was video recorded.

Calderon, Baker and Wolf (2000) combine qualitative (group) and quantitative (survey) research to guide the design of culturally appropriate research protocols. By administering newly developed or previously validated surveys to a group of people, the researchers can obtain information that ensures the survey’s cultural appropriateness, readability and comprehensibility. In accordance with these principles, the current expert group was heterogeneous in terms of gender, race and culture in order to obtain diverse opinions and views on bisexuality.

Based on the group dialogue it appeared that variables such as language proficiency, culture, education, socio-economic status and test-wiseness could have influenced the statistical findings from the first quantitative study (Foxcroft & Roodt, 2001). The results and contributions of the group are provided in the next chapter.

After analysing the findings and recommendation of the group another survey (quantitative phase) was conducted. The revised instruments, participants and procedures are discussed in the following section.

5.7.4 Thematic analysis of the expert advise group discussion

Thematic analysis is a method for identifying, analysing and reporting patterns (themes) within data. It minimally organises and describes the date, and clarifies various aspects of the research topic (Boyatzis, 1998). The researcher followed the six phases of analysis (Braun & Clarke, 2006) namely familiarising oneself with the data, generating initial codes, searching for themes, reviewing themes, defining and naming themes and producing the findings.

116 5.8 Phase three: Quantitative research method

5.8.1 Procedure

The researcher made contact with various faculty members. Permission was obtained to distribute the revised survey in the classes during lecture times. Different classes was utilised than in the first phase. The cover page of the questionnaire provided the participants with information regarding the purpose of the study, right to voluntary participation, and confidentiality. The revised instrument required between 15 and 20 minutes to complete.

5.8.2 Participants

In general, it is agreed that larger sample sizes are likely to result in more stable correlations among variables. According to Worthington and Whittaker (2006) and Comrey (1973), sample sizes of at least 300 are generally sufficient in most cases. In addition, Gorsuch’s (1983) suggested guidelines for minimum ratios of participants to items (5:1 or 10:1), have been widely cited in psychology research. The third phase sample met these criteria.

The total sample consisted of 646 students. Because the sample included significantly fewer coloured students 33 (5%) and Asian students 35 (5%) than black and white students, it was decided to focus only on the results of the latter two groups. Furthermore, the information obtained from the expert group clearly indicated that the various groups view the issues surrounding bisexuality in a different manner. Therefore it was decided not to include the Coloured and Asian participants as part of the Black group. The White and Black groups represent a significant proportion of the university population from which the sample had been drawn as well as being representative of the South African population.

The final sample consisted of 578 university students, 157 (27%) men and 421 (73%) women. With regard to ethnicity, 323 (56%) students were black and 255 (44%) were white. In terms of knowing a bisexual man or woman,

117 321 (56%) participants indicated that they know a bisexual man or woman. Five hundred and nine (88%) of the participants indicated that they know a gay man or a lesbian woman. Fifty-two (9%) of the participants indicated that they are not religious at all, 350 (61%) participants indicated that they are moderately religious and 176 (30%) indicated that they are deeply religious. In terms of sexual orientation, 512 (89%) of the participants indicated that they are heterosexual, 20 (3%) indicated that they are homosexual, 32 (6%) indicated that they are bisexual and 14 (2%) of the participants indicated that they are asexual.

5.8.3 Biographical questionnaire

The same biographical questionnaire utilised in phase one was included in the third phase of the study. This included information that pertained to the age, gender, ethnic group, degree of religiosity and the sexual orientation of the participant as well as whether the participants knew a bisexual man or woman or a gay man or lesbian woman,

5.8.4 Instruments administered in the third phase

As an outcome of the group discussion, the need had arisen to adapt the items of the ARBS-FM to provide for the varied interpretations of the items by the culturally diverse student population. According to Van de Vijver (2003), adapting items offer several advantages, such as increasing their sensitivity to the cultural context in which they are used, the ease and relatively low cost of adaptation and their flexibility in dealing with sources of bias. The aim of the adaptation was to render items that were culturally and linguistically appropriate in a cross-cultural context. The South African society is heterogeneous in terms of factors that moderate the reliability and validity of instruments.

118 5.8.4.1 Revised version of the ARBS-FM

The revised instrument was based on the ARBS-FM (Mohr & Rochlen, 1999). The revised version also consisted of two subscales. The first subscale is Stability (10 items): bisexuality is viewed as a legitimate, stable sexual orientation and bisexual people are seen as capable of forming committed romantic relationships. The second subscale is Tolerance (13 items): bisexuality is viewed as moral, tolerable, and not harmful to society and attitudes reflect acceptance rather than disdain for bisexual men and woman. Items are rated on a 5-point Likert scale ranging from one (strongly disagree) to five (strongly agree). The minimum score for the Tolerance subscale is 13 and the maximum score is 65. On the Stability subscale, the minimum score is 10 and the maximum score is 50. Higher scores indicate more negative attitudes. Attitudes towards bisexual men and women were measured separately. This was in line with the separate scales developed by Mohr and Rochlen (1999), namely the ARBS-F and ARBS-M.

Taking into account the recommendations of the group, the researcher generated more items utilising existing measures and findings from the literature review (Eliason, 1997; Herek, 2002; Mohr & Rochlen, 1999; Spalding & Peplau, 1997). Because of language proficiency, cultural and educational differences in the student population, the researcher wrote items that were simple, clear, concise, readable and distinct and which reflect the purpose of the scale. The research supervisor evaluated the items for clarity, conciseness, grammar, reading level and face validity (DeVellis, 2003; Worthington & Whittaker, 2006).

The researcher removed the terms “bisexual” and “bisexuality” from the survey, because there appear to be language and cultural differences concerning the meaning of the term “bisexual”. Vignettes were included that explained bisexual behaviour and attraction. The researcher based the wording of the vignettes on the definition of bisexuality provided by Firestein (1996).

119 The term “bisexual’ was substituted with male (John and David) and female (Mary and Susan) names. The vignette measuring attitudes towards bisexual women read as follows; Mary is a 27-year-old woman. Mary was in a long- term relationship with Peter. Currently, she is in a relationship with Susan. Mary has been sexually attracted to, and had sexual contact with men and women in the past. The vignette measuring attitudes towards bisexual men was worded as follows: John is a 27-year-old man. John was in a long-term relationship with Nancy. Currently, he is in a relationship with David. John has been sexually attracted to, and had sexual contact with men and women in the past.

5.8.4.2 Revised version of the categories of sexual orientation

Consensus was reached that this section of the survey should be simplified. Therefore, only four sexual orientation categories were utilised. In Chapter 2, the researcher discussed the dichotomous view (McLean, 2001) of sexuality as well as alternative views on categorising sexual orientation (Kinsey, 1953; Klein et al., 1985; Shively & De Cecco, 1977). Therefore, the researcher acknowledges that the use of only four categories (heterosexual, homosexual, bisexual and asexual) may fail to represent the continuum of human sexual behaviour.

Furthermore, only descriptions of the sexual orientation categories were provided, for example, I am only attracted to members of the same sex. The terms utilised in the nine categories of sexual orientation were problematic in terms of cultural, language and educational differences. Therefore, categories such as heterosexual, concurrent bisexual, sequential bisexual, past heterosexual and current homosexual were omitted.

5.9 Statistical analysis

Based on the research problem, the general purpose of the study was to determine attitudes of university students towards bisexuals. The statistical analysis of the data was undertaken with the aid of the Statistical Package for

120 Social Sciences (SPSS, version 15; Kinnear & Gray, 2007). Descriptive and inferential statistics were employed in the analysis of the data. The statistical methods selected in the analysis of the data are discussed below.

Descriptive statistics (mean scores and standard deviations) for all the subscales were calculated. Confirmatory factor analyses were performed to investigate whether the postulated two-factor model emerged for both the black and white groups. To determine the strength of the inter-correlations among the items, Tabachnick and Fidell (2001) recommended that the number of coefficients greater than 0.30 be considered to ascertain whether a factor analysis is appropriate. To further evaluate the factorability of the data, the Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy was calculated. The KMO index ranges from 0 to 1, with 0.6 suggested as the minimum value for a factor analysis (Tabachnick & Fidell, 2001). In determining the number of factors to be extracted, both the Kaiser eigenvalues-greater-than-one criterion and Cattell’s scree test were considered. According to the Kaiser criterion, only factors with eigenvalues of 1.0 or more should be retained (Cramer, 2003). Cattell’s scree test indicates the number of factors to be extracted by plotting the eigenvalues of the factors on a graph. All factors above the break in the plot are considered to be salient (Pallant, 2005). Within the broad spectrum of factor analysis, this study made use of Principal Axis factor analysis. Principal Axis factor analysis is a multivariate procedure which rotates the data so that maximum variabilities are projected onto the axes. Following the guidelines of Cohen (1988) correlations of about 0.10 may be regarded as small, correlations of about 0.30 as moderate, and correlations of 0.50 and higher as large.

Once the number of factors was determined the resultant factor matrix was considered for interpretation. To assist in the process of interpretation, the factors were rotated. Diekhoff (1992) states that the factors which explain the most variance are rotated to make their meaning clearer. For the purposes of this study the factors were rotated according to the oblique Direct Oblimin criteria. To enhance interpretation of the factor matrix, it was decided to focus on variables with loadings greater than 0.30. This approach was in line with

121 Diekhoff’s (1992) recommendation that factor loadings of greater than or equal to 0.30 for oblique rotations are salient.

Cronbach’s alpha coefficients (Cronbach, 1951) were estimated to determine the internal consistency reliability of the scales. A series of multivariate analysis of variance were employed to investigate the relationships of various variables to attitudes towards bisexuals. Follow-up ANOVAs were used to help determine the nature of the relationship between the variables. The Wilks' lambda was the test statistic performed in this study to express the differences between the means of the grouping variables. Statistical significance was interpreted at p ≤ 0.05.

5.10 Chapter summary

In this chapter, the aim of the study, the research questions and the hypotheses were discussed. The current research was conducted in three phases, and mixed method research was employed. The research method, procedures, instruments and participants of the first phase (quantitative method) were discussed. This included the Attitudes Regarding Bisexuality Scale (ARBS-FM) (Mohr & Rochlen, 1999), the self-identification instrument indicating the participants’ sexual orientation (Berkey et al., 1990) as well as the biographical questionnaire. The second phase of the research, namely an expert advice group (qualitative), was discussed in terms of the participants and procedures. Lastly, the research method, revised instruments, procedures and participants of the third phase (quantitative), were discussed and an overview of the statistical analyses of the data provided.

5.11 Preview of the contents of the following chapter

In the following chapter, the statistical findings of the two quantitative phases of the study as well as the findings of the qualitative phase are provided.

122 CHAPTER 6

RESULTS

6.1 Introduction

In this chapter, the statistical results of the first phase are provided. The results of the second phase (qualitative method), which inform and provide motivation for the third phase, are also presented. Finally, the statistical findings of the third phase pertaining to the psychometric properties of the revised instruments and the hypotheses formulated in the first chapter are presented.

6.2 Phase one: Statistical findings of the quantitative study

The Cronbach’s alpha coefficient (Cronbach, 1951) for the total sample was 0.68 for the Stability subscale and 0.80 for the Tolerance subscale. The Tolerance subscale yielded acceptable internal consistency reliability above 0.80 and compared well to findings of Mohr and Rochlen (1999). The Stability subscale did not meet the same level of reliability. Bearing in mind the unique cultural situation in South Africa, and the fact that instruments developed outside of South Africa are not necessarily reliable for the South African population, it was decided to investigate the reliabilities of the two subscales separately for the different ethnic groups.

The reliabilities of the subscales for the total group (including only the black and white participants) and for the black and white groups separately, are presented in Table 6.1

123

Table 6.1 Reliabilities of ARBS-FM Subscales for the Total Group and the Black and White Groups Cronbach’s alpha Subscales Total group Black group White group N = 1 459 n = 926 n = 533 Stability 0.68 0.61 0.80 Tolerance 0.80 0.75 0.87

Inspection of Table 6.1 shows significant differences between the reliabilities of the subscales obtained for the black and the white groups, especially with regards to the Stability subscale. These results indicate that the two groups probably did not have the same understanding of the items or did not attach the same meanings to the individual items.

Based on these results, a second phase in the research was introduced. A expert group was convened to obtain more information about black and white students’ understanding of the concept of bisexuality as well as their interpretation of the items of the ARBS-FM and their understanding of sexual orientation categories.

6.3 Phase two: Findings of the expert advice group

The findings of the group in terms of bisexuality, the ARBS-FM and the categories of sexual orientation were incorporated into the revised instruments that were utilised in the third phase of the research study. The aim of revising the instruments was to render items that were culturally, educationally and linguistically appropriate for the diverse South African student population. The following results from the group guided the revision process of the instruments.

124 6.3.1 Discussing bisexuality

The researcher prompted the participants to express and discuss their understanding of the term “bisexuality” as a sexual orientation as well as their perceptions of bisexual men and women. The terms “bisexual” and “bisexuality” appeared to be problematic. Participants differed about the definitions as well as their understanding of bisexuality as a sexual orientation. Participants also noted that there might be different terminology in other cultures describing bisexual behaviour.

Various themes emerged during the course of the group discussion. The themes are indicated below, followed by excerpts representing the views of the participants.

Theme 1: Confusion about the term bisexual:

“I always thought it was men and women that have both … both sexual organs?” / “I think that people that … do not know about sexual orientations might confuse bisexuality with bestiality. It sounds similar, doesn’t it?” / “Bi… that means two? So they have two partners?”/ “For me bisexual men and women have sex with a man or a woman at the same time. They do not mind”.

Theme 2: The role of language:

“We know about men that are married, they have a normal life and children. But at night, they go out to see men, to have sex. We call them “after-niners” because they go out late at night”. / “My opinion would be that sexuality is closely tied to cultural norms and values. The term bisexual could be confusing in terms of the different languages in South Africa. Language will play an important role if you want to change this scale”.

125 Theme 3: The recognition that bisexuality is a taboo in many cultures:

“We do not really talk about this … sex…in our culture, but when we do, we only say that someone is gay or straight. I do not think that we know what bisexual means”. / “The students that have completed the survey are young, they have just left home. People at home do not talk about sex. So they may not have heard about bisexuality”. / “Our religion condemns people that have sex outside of marriage. It is a sin. It is not something that we discuss with our friends or family. So something like bisexuality would never be discussed”. / “In our culture, we do not talk about sex; we would not discuss sexual orientation. It is not acceptable”.

Theme 4: Bisexual men and women are unfaithful, at risk for HIV/AIDS, and promiscuous:

“They make me uncomfortable. I know a man, and he sleeps with all of them. I don’t think I can be in relationship with someone that can not make up his mind”. / “I remember that a long time ago, the gay men were responsible for HIV/AIDS. Surely these bisexuals are at risk?” / “I know that some men and women have sex with other men and women. But they are prostitutes”.

Theme 5: Bisexuality is not a stable sexual orientation:

“Sometimes students get drunk, and have sex with people, but I think they are only experimenting, or curious”. / “I think that these people are just too scared to be gay”.

6.3.2 Discussing the ARBS-FM

The group participants also completed the ARBS-FM (Mohr & Rochlen, 1999). The following suggestions were obtained from the group concerning the instrument:

126 “Perhaps you must remove the term bisexual, as the students may not understand this term”.

“After this discussion, I understand bisexuality better. Maybe you can provide them with an example of the behaviour of someone that is bisexual. Or explain this before you hand out the survey?”

“Make the bisexual survey more user-friendly for the students...and measure the attitudes towards bisexual men and women separately. Perhaps this will be less confusing for the students”.

“Perhaps if you explain what bisexual men and women are all about, they may realise that they know such an individual”.

“I suggest that you add more items to the survey. Easier items that everybody can understand. There are some words that they may not understand”.

“There are items that mention homosexual and heterosexuality... I would use the terms gay and straight”.

“The negative items may be confusing for the participants”.

6.3.3 Discussing the categories of sexual orientation

Each participant also completed the nine categories of sexuality of Berkey et al. (1990). The participants experienced difficulty understanding these categories. The group made the following valuable contributions and suggestions:

“This is very confusing”.

“I did not realise that there are so many sexual orientations?”

127 “The words such as concurrent, sequential and exclusive… perhaps the students do not understand this?

“Once again, I think that the student sample do not understand the terms homosexual, heterosexual and bisexuality’.

“I would not know how to choose…it will take me a long time to read this and then make a decision”

From the above comments it was clear that the terms “bisexual/bisexuality”, the items on the ARBS-FM and the categories of sexual orientation were somewhat confusing to the group participants. The group consisted of senior psychologists, intern psychologists and professionals in the field of psychology. The participants in the quantitative phases of the study were younger and it was assumed that they would experience even more confusion when confronted with these terms, items and categories. in terms of the above-mentioned. The information obtained from the expert group was therefore used to revise the instructions pertaining to the completion of the instruments, as well as the items themselves.

6.4 Phase three: Statistical findings of the quantitative study

In this section the results related to the factor analyses of the revised ARBS-F and ARBS-M is provided. The second part of this section describes and summarises the statistical findings with regards to the research hypotheses presented in chapter one.

6.4.1 Confirmatory factor analysis of the revised ARBS-F

The 23 items of the ARBS-F were subjected to two separate confirmatory factor analyses for black and white students respectively, to determine whether the postulated two-factor structure emerged for the sample. Prior to performing the analyses, the suitability of the data for factor analysis was assessed. Inspection of the correlation matrixes (for black and white students

128 separately) revealed many coefficients of 0.30 and above. The Kaiser-Meyer- Oklin value was 0.89 for the black group and 0.90 for the white group, which is higher that the recommended value of 0.60 (Kaiser, 1970; Tabachnick & Fidell, 2001). The Bartlett’s Test of Sphericity (Bartlett, 1950; Tabachnick & Fidell, 2001) was statistically significant for both groups, providing further support for the factorability of the correlation matrixes.

For the black group, principal axis analysis revealed the presence of five factors with eigenvalues exceeding 1, explaining 29.02%, 9.76%, 5.90%, 5.14% and 4.99% of the variance respectively. For the white group, four factors with eigenvalues greater than 1 emerged, explaining 33.36%, 13.28%, 5.91% and 4.71% of the variance respectively. An inspection of the scree plots for both groups revealed a clear break after the second factor. The scree plots for the black and white groups are presented in Figure 6.1 and Figure 6.2 respectively. Using Cattell’s scree test (1966), it was decided to retain two factors for further investigation. The two factor structure is also supported by Mohr and Rochlen’s (1999) development of the original ARBS-F.

6

4 Eigenvalue 2

0

987654321 1110 12 1413 15 1716 18 212019 22 23 Factor Number Figure 6.1 Revised ARBS-F scree plot for the black group.

129 8

6

4 Eigenvalue

2

0

987654321 10 14131211 21201918171615 2322 Factor Number Figure 6.2 Revised ARBS-F scree plot for the white group.

To aid the interpretation of these two factors, Direct Oblimin rotation with Kaiser normalisation was performed. The rotated factor pattern matrix of the ARBS-F for the black group is presented in Table 6.2 and for the white group in Table 6.3. The rotated solution revealed both factors to be showing a number of strong loadings. The interpretation of the two factors was consistent with the results reported by Mohr and Rochlen (1999). The results of this analysis supported the inclusion of Tolerance and Stability as separate subscales to measure attitudes towards bisexual women, as suggested by Mohr and Rochlen (1999).

130 Table 6.2 Rotated Factor Pattern Matrix of the Revised ARBS-F for the Black Students Factor 1 2 1 0.70 -0.07 2 0.61 -0.13 3 0.58 -0.11 4 -0.05 0.27 5 0.63 -0.13 6 0.39 0.11 7 0.01 0.35 8 0.79 0.03 9 0.41 0.18 10 0.32 0.66 11 0.26 0.26 12 0.58 0.06 13 -0.03 0.52 14 0.60 -0.01 15 0.49 0.25 16 0.49 0.20 17 0.62 0.13 18 0.69 -0.14 19 0.69 -0.01 20 0.62 -0.19 21 0.67 -0.02 22 0.14 0.59 23 -0.15 0.25 Note. Salient factor loadings >0.30 are underlined.

Inspection of Table 6.2 reveals that for the black group, all the Tolerance- items loaded on Factor 1. Items 9, 15 and 16 also loaded on Factor 1, and not as expected on Factor 2. Although item 10 also loaded on Factor 1

131 (Tolerance), it showed a much stronger loading on Factor 2 (Stability). Only four of the Stability items loaded on Factor 2. Items 4, 11 and 23 (postulated to measure Stability) did not show any significant factor loadings.

Table 6.3 Rotated Factor Matrix of the Revised ARBS-F for White Students Factor 1 2 1 0.81 -0.03 2 0.62 -0.07 3 0.61 0.04 4 -0.21 0.52 5 0.73 -0.08 6 0.42 0.03 7 -0.20 0.41 8 0.83 -0.09 9 0.33 0.38 10 0.33 0.60 11 0.29 0.35 12 0.57 0.18 13 0.07 0.70 14 0.66 -0.02 15 0.31 0.40 16 0.47 0.40 17 0.59 0.01 18 0.61 -0.01 19 0.78 -0.06 20 0.75 -0.13 21 0.65 0.09 22 0.11 0.72 23 -0.05 0.62 Note. Salient factor loadings >0.30 are underlined.

132 Inspection of Table 6.3 reveals that for the white group, all the Tolerance- items loaded on Factor 1. Items 9, 10, 15 and 16 loaded on both factors, with items 10 and 15 loading stronger on Factor 2 (Stability). In total, ten items showed salient loadings on Factor 2.

6.4.2 Confirmatory factor analysis of the revised ARBS-M

The 23 items of the revised ARBS-M were also subjected to two separate confirmatory factor analyses for black and white students to determine whether the postulated factor structure emerged for the sample. Prior to performing the analyses the suitability of the data for factor analysis was assessed. Inspection of the correlation matrixes (for black and white students separately) revealed many coefficients of 0.30 and above. The Kaiser-Meyer- Oklin value was 0.91 for the black group and 0.92 for the white group. These values were substantially higher than the minimum value of 0.60 recommended by Tabachnick and Fidell (2001). The Bartlett’s Test of Sphericity (Bartlett, 1950) was statistically significant for both groups. All these results suggest that a factor analysis of the 23 items of the revised ARBS-M was justified.

For the black group, principal axis analysis revealed the presence of five factors with eigenvalues exceeding 1, explaining 33.76%, 10.69%, 5.92%, 5.56% and 4.43% of the variance respectively. For the white group, once again four factors with eigenvalues greater than 1 emerged, explaining 36.42%, 13.82%, 5.65% and 4.47% of the variance respectively. An inspection of the scree plots for both groups revealed a clear break after the second factor. The scree plots for the black and white groups are presented in Figure 6.3 and Figure 6.4 respectively. Using Cattell’s scree test (1966), it was decided to retain two factors for further investigation. As was the case with the revised ARBS-F, the two factor structure was in line with Mohr and Rochlen’s (1999) suggestion.

133

8

6

4 Eigenvalue

2

0

987654321 16151413121110 17 201918 21 22 23 Factor Number

Figure 6.3 Revised ARBS-M scree plot for the black group.

10

8

6

4 Eigenvalue

2

0

987654321 10 1514131211 1716 2221201918 23 Factor Number

Figure 6.4 Revised ARBS-M scree plot for the white group.

To aid the interpretation of these two factors, Direct Oblimin rotation was performed. The rotated factor pattern matrix of the revised ARBS-M for the black group is presented in Table 6.4 and for the white group in Table 6.5. The rotated solution revealed both factors to be showing a number of strong factor loadings. Again, the interpretation of the two factors was consistent with

134 the results reported by Mohr and Rochlen (1999). These results supported the inclusion of Tolerance and Stability as separate subscales to measure attitudes towards bisexual women, as suggested by Mohr and Rochlen (1999).

Table 6.4 Rotated Factor Pattern Matrix of the Revised ARBS-M for Black Students Factor 1 2 1 0.78 0.02 2 0.72 0.08 3 0.14 0.59 4 0.27 0.32 5 0.47 0.17 6 -0.03 0.39 7 0.24 0.38 8 -0.06 0.43 9 0.32 0.21 10 0.78 -0.05 11 0.64 -0.03 12 0.42 0.35 13 0.35 0.39 14 0.78 -0.04 15 0.78 -0.14 16 0.60 0.12 17 0.74 -0.11 18 0.71 0.04 19 0.71 -0.13 20 0.69 0.06 21 0.20 0.54 22 -0.01 0.59 23 -0.26 0.37 Note. Salient factor loadings >0.30 are underlined.

135 Inspection of Table 6.4 reveals that for the black group, all the Tolerance- items loaded on Factor 1. Items 12 and 13 (Stability items) loaded on Factor 1 and Factor 2. Apart from these two items, the remaining eight Stability items loaded on Factor 2, as expected.

Table 6.5 Rotated Factor Pattern Matrix of the Revised ARBS-M for White Students Factor 1 2 1 0.80 0.06 2 0.63 0.16 3 0.07 0.64 4 0.26 0.52 5 0.64 0.13 6 -0.09 0.49 7 0.26 0.44 8 -0.25 0.64 9 0.30 0.14 10 0.76 -0.04 11 0.69 -0.05 12 0.27 0.56 13 0.23 0.49 14 0.82 0.00 15 0.77 -0.09 16 0.59 0.07 17 0.69 -0.15 18 0.70 0.12 19 0.77 -0.14 20 0.68 0.13 21 0.19 0.64 22 0.08 0.75 23 -0.10 0.57 Note. Salient factor loadings >0.30 are underlined.

136 Inspection of Table 6.5 indicates that for the white group, all 13 Tolerance- items had salient loadings on Factor 1 and all 10 Stability-items showed salient loadings on Factor 2.

6.4.3 Structural equivalence of factors

In comparing the results of the factor analyses for the black and white students it was clear that, although a stronger two-factor structure emerged for the white group, the black and white groups showed the same pattern of high and low factor loadings on the ARBS-F and the ARBS-M respectively. Agreement of the factors for the black group and the white group was examined in terms of Tucker’s coefficients of congruence which are reported in Table 6.6. Inspection of the table shows that the coefficients of congruence were all acceptable (> 0.90) suggesting that these factors had equal meaning for both black and white groups.

Table 6.6 Coefficients of Congruence of the Factor Solution of the Revised ARBS-F and ARBS-M Subscales Factors Tucker’s coefficients of congruence Female Stability 0.93 Female Tolerance 0.99 Male Stability 0.99 Male Tolerance 0.99

6.4.4 Reliabilities of the revised ARBS-F and ARBS-M subscales

The internal consistency reliabilities of the revised scales were estimated by using Cronbach’s alpha coefficient (Santos, 1999). These coefficients are presented in Table 6.7 for the total group as well as separately for the black and white groups.

137 Table 6.7 Reliabilities of Revised Scales for the Total Group and the Black and White Groups Cronbach’s alpha Subscales Total group Black group White group N = 578 n = 323 n = 255 Female Stability 0.74 0.67 0.82 Female Tolerance 0.90 0.89 0.91 Male Stability 0.80 0.76 0.85 Male Tolerance 0.92 0.92 0.92

Table 6.7 shows that the Tolerance subscales yielded acceptable reliabilities, all exceeding 0.90. Regarding Female Stability, the reliabilities for the total group and especially for the white group were also acceptable. Although the reliability for the black group was not above the desired 0.70 level, it increased from 0.61 in phase 1 of the study, to 0.67. With regards to Male Stability, the reliabilities for the total group and the black and white groups were acceptable (> 0.75).

6.4.5 Correlations between male and female subscale scores

Although the results from the factor analyses of the revised ARBS-F and ARBS-M showed that the factors manifested in a similar way for the black and white students, the reliabilities reported in Table 6.7 were not comparable for the black and white groups with respect to the Stability subscales. This may be due to cultural, educational and linguistic differences between the black and white students. The correlations between the total scores of the Female Tolerance and Male Tolerance as well as between the Female Stability and Male Stability subscales were obtained separately for the black and white groups. Correcting for measurement error, the true correlations were also estimated. The correlation coefficients are presented in Table 6.8.

138 Table 6.8 Correlations between Female and Male Subscales for the Black and White Groups Correlation coefficients Subscales Black group White group r Corrected r r Corrected r Tolerance 0.88 0.97 0.92 1.00 Stability 0.74 1.00 0.84 1.00

The results presented in Table 6.8 show very strong true correlations between the female and male orientations for both the Tolerance and Stability subscales (ranging from 0.97 to 1.00). Based on this, it appears that black students have the same attitudes towards bisexual women and men with respect to Tolerance and Stability. White students also do not seem to have different attitudes towards bisexual men and women. These results suggest that attitudes towards bisexual men and women can be measured by a general scale and not by two separate scales that differentiate between attitudes towards bisexual men and women.

Although it was clear that attitudes towards bisexual men and women do not have to be measured separately, the total scores of the Female and Male subscales were utilised separately for hypotheses testing. These subscales consist mostly of the same items, with the only difference the gender of the bisexual person. Combining the items into one overall Tolerance subscale and an overall Stability subscale at this stage would yield inflated internal consistency reliability coefficients and therefore unreliable results.

139 6.5 Differential item functioning (DIF) and Differential test functioning (DTF)

Factor analyses indicated that the revised ARBS-F and ARBS-M measured the same constructs (Tolerance and Stability) for the black and white students. As indicated in section 6.4.3, Tucker’s coefficients of congruence were above 0.90 for all the subscales. Investigating whether the items function in the same way for the two groups, each of the four scales (Female Tolerance, Female Stability, Male Tolerance and Male Stability) was subjected to separate differential item functioning analyses with the DIFAS 4.0 programme (Penfield, 2005). The results are summarised in Tables 6.9 to 6.12

Table 6.9 DIF Statistics for the Female Tolerance Scale Item Mantel Chi- Liu-Agresti LOR SE LOR Z square LOR 1 0.888 0.168 0.177 0.949 2 1.440 -0.203 0.172 -1.180 3 8.410 0.491 0.164 2.994 5 0.572 -0.162 0.213 -0.761 6 17.013 -0.781 0.192 -4.068 8 0.437 -0.124 0.188 -0.660 12 1.843 0.234 0.170 1.376 14 28.533 1.064 0.218 4.881 17 6.651 -0.460 0.178 -2.584 18 0.034 -0.037 0.200 -0.185 19 5.988 -0.442 0.179 -2.469 20 0.320 -0.127 0.227 -0.559 21 10.796 0.628 0.184 3.413

For Female Tolerance, five items with statistically significant uniform DIF (p < 0.05) were identified. Using the classification system suggested by Penfield

140 and Algina (2006), two of these items yielded large DIF effects (Liu-Agresti LOR > 0.64). Item 14 (“Mary is a pervert”) was relatively easier to agree with for the black group, whereas item 6 (“Mary’s sexual behaviour places her partners at risk for HIV”) was relatively easier to agree with for the white group. As a next step, differential test functioning (DTF) was investigated to see whether the subscale functions equivalently for the black and white groups. With all 13 items included, the variance estimator of DTF was Nu^2 = 0.192. Using the classification system suggested by Penfield and Algina (2006) this may be classified as large DTF (Nu^2 > 0.14). These results indicate that the Female Tolerance scale functioned differently for the black and white groups. Removing the two items with large DIF effects (items 6 and 14) produced an 11-item scale with small DTF (Nu^2 = 0.066). The reduced scale exhibited little cultural bias from a statistical perspective. The correlation between the full scale and the reduced scale was 0.99, suggesting that the reduced scale measures the same construct as the full scale. The reliability of the reduced scale was 0.89, which is only slightly lower than the reliability of the full scale (0.90).

Table 6.10 DIF Statistics for the Female Stability Scale Item Mantel Chi- Liu-Agresti LOR SE LOR Z square LOR 4 1.826 -0.217 0.159 -1.365 7 6.959 0.436 0.160 2.725 9 0.800 0.151 0.167 0.904 10 10.002 -0.552 0.176 -3.136 11 5.292 -0.375 0.163 -2.301 13 0.167 -0.069 0.169 -0.408 15 0.459 -0.113 0.166 -0.681 16 1.575 0.206 0.163 1.264 22 3.152 -0.304 0.171 -1.778 23 19.168 0.703 0.157 4.478

141 For Female Stability, four items with statistically significant uniform DIF (p < 0.05) were identified. Using the classification system suggested by Penfield and Algina (2006), one of these items yielded large DIF effects (Liu-Agresti LOR > 0.64). Item 23 (“One day Mary will admit her true sexual orientation”) were relatively easier to agree with for the black group. As a next step, differential test functioning (DTF) was investigated to see whether the subscale functions equivalently for the black and white group. With all 10 items included, the variance estimator of DTF was Nu^2 = 0.108. Although this value does not fit Penfield and Algina’s (2006) classification of large DTF (Nu^2 < 0.14), it was decided to remove item 23 to see whether the reduced scale could provide a smaller DTF. Removing item 23 produced a 9-item scale with small DTF (Nu^2 = 0.049). The reduced scale exhibited little cultural bias from a statistical perspective. The correlation between the full scale and the reduced scale was 0.99, suggesting that the reduced scale measures the same construct as the full scale. The reliability of the reduced scale was 0.73, which is only slightly lower than the reliability of the full scale (0.74).

Table 6.11 DIF Statistics for the Male Tolerance Scale Item Mantel Chi- Liu-Agresti LOR SE LOR Z square LOR 1 0.539 -0.141 0.191 -0.738 2 2.348 0.265 0.171 1.550 5 9.622 0.549 0.172 3.192 9 20.538 -0.855 0.190 -4.500 10 26.729 1.106 0.212 5.217 11 9.353 -0.531 0.176 -3.017 14 1.194 0.210 0.189 1.111 15 0.351 0.125 0.207 0.604 16 2.436 -0.277 0.173 -1.601 17 1.993 0.278 0.198 1.404 18 3.956 -0.369 0.186 -1.984

142 19 0.193 -0.097 0.219 -0.443 20 4.755 0.409 0.189 2.164

For Male Tolerance, five items with statistically significant uniform DIF (p < 0.05) were identified. Using the classification system suggested by Penfield and Algina (2006), two of these items yielded large DIF effects (Liu-Agresti LOR > 0.64). Item 10 (“John is a pervert”) was relatively easier to agree with for the black group, whereas item 9 (“John’s behaviour places his partners at risk for HIV”) was relatively easier to agree with for the white group. As a next step, differential test functioning (DTF) was investigated to see whether the subscale functions equivalently for the black and white group. With all 13 items included, the variance estimator of DTF was Nu^2 = 0.204. Using the classification system suggested by Penfield and Algina (2006) this may be classified as large DTF (Nu^2 > 0.14). These results indicated that the Male Tolerance scale functioned differently for the black and white groups. Removing the two items with large DIF effects (items 9 and 10) produced an 11-item scale with small DTF (Nu^2 = 0.069). The reduced scale exhibited little cultural bias from a statistical perspective. The correlation between the full scale and the reduced scale was 0.99, suggesting that the reduced scale measures the same construct as the full scale. The reliability of the reduced scale was 0.91, which is only slightly lower than the reliability of the full scale (0.92).

Table 6.12 DIF Statistics for the Male Stability Scale Item Mantel Chi- Liu-Agresti LOR SE LOR Z square LOR 3 2.051 -0.249 0.173 -1.439 4 0.447 0.112 0.167 0.671 6 2.749 0.277 0.163 1.699 7 0.054 -0.038 0.162 -0.235 8 6.288 -0.423 0.172 -2.459 12 0.944 0.168 0.170 0.988

143 13 0.043 -0.034 0.164 -0.207 21 0.109 -0.057 0.171 -0.333 22 2.532 -0.292 0.189 -1.545 23 8.363 0.483 0.163 2.963

For Male Stability, two items with statistically significant uniform DIF (p < 0.05) were identified. Using the classification system suggested by Penfield and Algina (2006), none of these items yielded large DIF effects (Liu-Agresti LOR > 0.64). As a next step, differential test functioning (DTF) was investigated to see whether the subscale functions equivalently for the black and white groups. With all 13 items included, the variance estimator of DTF was Nu^2 = 0.04. According to Penfield and Algina (2006), this may be classified as small DTF (Nu^2 > 0.14). These results indicated that the Male Stability scale functions similarly for the black and white groups and exhibited little cultural bias from a statistical perspective.

From the next section onwards, the results pertaining to the stated hypotheses are presented. The total scores on the reduced scales, as described in section 5.5, were used in these analyses. The results for each hypothesis are presented separately for the revised ARBS-F and the revised ARBS-M. After each hypothesis, the results of an investigation into the effect of ethnicity on the independent variables are briefly reported. This investigation was deemed important as a strong cross-cultural enquiry emerged in the course of the study. In instances where ethnicity did show an effect on the results, the means and standard deviations are reported for black and white male and black and white female students separately.

6.6 Hypothesis one

6.6.1 Investigating differences between black and white students regarding their performance on the revised ARBS-F

Table 6.13 shows the means and standard deviations of the scores obtained for the revised ARBS-F by the black and white students respectively.

144 Table 6.13 Means and Standard Deviations for Ethnic Groups on the Revised ARBS-F Std. Scale Ethnicity Mean Deviation N Female Black 26.75 10.35 323 Tolerance White 26.32 9.72 255 Total 26.56 10.07 578 Female Black 29.78 5.62 323 Stability White 29.16 5.84 255 Total 29.51 5.72 578

A one-way between-groups multivariate analysis of variance was performed to investigate the differences in the attitudes of black and white students towards bisexual women. Female Tolerance and Female Stability served as the dependent variables and ethnicity as the independent variable. There was no statistically significant difference between the white students and black students on the combined dependent variables: F(2, 575) = 0.85, p = 0.43; Wilks’ Lambda = 0.99; partial eta squared = 0.003.

6.6.2 Investigating differences between black and white students regarding their performance on the revised ARBS-M

Table 6.14 shows the means and standard deviations of the two ethnic groups, namely black and white, on the revised ARBS-M.

Table 6.14 Means and Standard Deviations for Black and White Students’ Responses on the revised ARBS-M Scale Ethnicity Mean Std. Deviation N Male Black 27.20 11.07 323 Tolerance White 26.06 10.14 255 Total 26.70 10.67 578

145 Male Stability Black 33.89 6.39 323 White 32.78 6.77 255 Total 33.40 6.58 578

A one-way between-groups multivariate analysis of variance was performed to investigate ethnic differences regarding attitudes towards bisexual men. Male Tolerance and Male Stability served as the dependent variables and ethnicity as the independent variable. There was no statistically significant difference between the white students and black students on the combined dependent variables with regards to their attitudes towards bisexual men: F(2, 575) = 2.12, p = 0.12; Wilks’ Lambda = 0.99; partial eta squared = 0.007.

6.7 Hypothesis two

6.7.1 Investigating differences between male and female students regarding their performance on the revised ARBS-F

The means and standard deviations of the male and female students’ responses to the revised ARBS-F are reported in Table 6.15. The means and standard deviations are reported separately for black and white male and female students.

A one-way between-groups multivariate analysis of variance was performed to investigate gender differences with regards to attitudes towards bisexual women. Female Tolerance and Female Stability served as the dependent variables and gender as the independent variable. There was a statistically significant difference between male students and female students on the combined dependent variables: F(2, 574) = 11.64, p < 0.001; Wilks’ Lambda = 0.96; partial eta squared = 0.04. Gender explained approximately 4% of the variance in the linear combination of the Tolerance and Stability scales of the revised ARBS-F. This may be described as a small effect and shows that the practical significance of the differences in the global attitudes of men and women towards bisexual women is relatively small, although noticeable.

146 Table 6.15 Means and Standard Deviations for Male and Female Students’ Responses on the Revised ARBS-F

Scale Gender

Male

Black (n = 87) White (n = 69) Total (N = 156)

Mean SD Mean SD Mean SD F-Tolerance 30.99 10.77 27.43 10.81 29.42 10.90 F-Stability 29.62 5.30 28.99 5.47 29.34 5.37 Female Black (n = 235) White (n = 186) Total (N = 421) Mean SD Mean SD Mean SD F-Tolerance 25.18 9.78 25.91 9.28 25.50 9.56 F-Stability 29.80 5.72 29.22 5.98 29.54 5.84 Note. F-Tolerance = Female Tolerance, F-Stability = Female Stability.

When the results for the dependent variables were considered separately, follow-up univariate ANOVA showed a statistically significant difference between male and female students only for the Female Tolerance scale: F(1,575) = 17.65, p < 0.001, partial eta squared = 0.03. For this scale, gender explained only approximately 3% of the variance. Inspection of the mean scores in Table 6.15 indicates that male students (M = 29.42, SD = 10.90) held more negative attitudes toward bisexual women than female students (M = 25.50, SD = 9.56).

6.7.2 Investigating differences between male and female students regarding their performance on the revised ARBS-M

The means and standard deviations of the male and female students’ responses to the revised ARBS-M are reported in Table 6.16. These are also subdivided in terms of ethnicity.

147 A one-way between-groups multivariate analysis of variance was performed to investigate gender differences concerning attitudes towards bisexual men. Male Tolerance and Male Stability served as the dependent variables and gender as the independent variable. There was a statistically significant difference between male students and female students on the combined dependent variables: F(2, 574) = 16.40, p < 0.001; Wilks’ Lambda = 0.95; partial eta squared = 0.05. Gender explained approximately 5% of the variance in the linear combination of the Tolerance and Stability scales of the revised ARBS-M. This may be described as a small effect and shows that the practical significance of the differences in the global attitudes of men and women towards bisexual women is relatively small, although noticeable.

Table 6.16 Means and Standard Deviations for Male and Female Students’ Responses on the Revised ARBS-M

Scale Gender

Male Black (n = 87) White (n = 69) Total (N = 156) Mean SD Mean SD Mean SD M-Tolerance 32.43 11.23 28.52 11.51 30.70 11.48 M-Stability 34.67 6.13 33.16 6.33 34.00 6.24 Female Black (n = 235) White (n = 186) Total (N = 421) Mean SD Mean SD Mean SD M-Tolerance 25.27 10.41 25.15 9.45 25.22 9.98 M-Stability 33.55 6.44 32.65 6.94 33.15 6.67 Note. M-Tolerance = Male Tolerance, M-Stability = Male Stability.

When the results for the dependent variables were considered separately, follow-up univariate ANOVA showed a statistically significant difference between male and female students only for the Male Tolerance scale:

148 F(1,575) = 31.59, p < 0.001, partial eta squared = 0.05. For this scale, gender explained approximately 5% of the variance. Inspection of the mean scores in Table 6.16 indicates that male students (M = 30.70, SD = 11.48) held more negative attitudes toward bisexual men than female students (M = 25.22, SD = 9.98).

6.7.3 Effect of ethnicity on gender differences

As a further step, the effect of ethnicity on gender differences was investigated. The Female Tolerance and Female Stability scores of the black men, white men, black women and white women and the interaction of gender and ethnicity were subjected to 2 X 2 MANOVA. There was a statistically significant difference between the scores of the four groups on the combined dependent variables (Female Tolerance and Female Stability): F(2, 572) = 3.13, p = 0.05; Wilks’ Lambda = 0.99; partial eta squared = 0.01. When the results for the dependent variables were considered separately, the effect of ethnicity on gender differences was statistically significant only for the Female Tolerance scale: F(1,573) = 5.24, p = 0,022, partial eta squared = 0.009. For this scale, ethnicity explained nearly 1% of the variance, hence a very small practical significant result. Inspection of the mean scores indicates that black men held more negative attitudes toward bisexual women than white men, black women and white women. These results are graphically represented in Figure 6.5.

149 31.00

30.00 Ethnic Group: Black White 29.00

28.00

27.00

Estimated Marginal Means Marginal Estimated 26.00

25.00

Male Female Gender

Figure 6.5 Interaction of ethnicity and gender on Female Tolerance.

The Male Tolerance and Male Stability scores of the black men, white men, black women and white women and the interaction of gender and ethnicity were also subjected to 2 X 2 MANOVA. There was no statistically significant difference between the scores of the four groups on the combined dependent variables (Male Tolerance and Male Stability): F(2, 572) = 1.92, p = 0.15; Wilks’ Lambda = 0.99; partial eta squared = 0.007.

6.8 Hypothesis three

6.8.1 Investigating differences between students that know and students that do not know a bisexual man or woman regarding performance on the revised ARBS-F

The means and standard deviations of students’ responses to whether they know a bisexual man or woman and those who do not, on the revised ARBS- F are provided in Table 6.17.

150 Table 6.17 Means and Standard Deviations for Students that Know or Do Not Know a Bisexual Man or Women on the Revised ARBS-F Knowing a bisexual man Std. Scale or woman Mean Deviation N Female Tolerance Yes 24.33 9.28 321 No 29.33 10.38 255 Total 26.54 10.08 576 Female Stability Yes 29.39 5.71 321 No 29.66 5.77 255 Total 29.51 5.73 576

A one-way between-groups multivariate analysis of variance was performed to investigate differences between knowing a bisexual man or woman and not knowing a bisexual man or woman with regards to attitudes towards bisexual women. Female Tolerance and Female Stability served as the dependent variables and knowing a bisexual man or woman as the independent variable. There was a statistically significant difference in attitudes between those who know a bisexual man or woman and those who do not know a bisexual man or woman: F(2, 573) = 20.91, p < 0.001; Wilks’ Lambda = 0.93; partial eta squared = 0.07. Knowing or not knowing a bisexual man or woman explained approximately 7% of the variance in the linear combination of the Female Tolerance and Female Stability scales of the revised ARBS-F.

When the results for the dependent variables were considered separately, follow-up univariate ANOVA showed a statistically significant difference between knowing a bisexual man or woman and not knowing a bisexual man or woman on the Female Tolerance scale: F(1,574) = 37.07, p < 0.001, partial eta squared = 0.06. For this scale, knowing or not knowing a bisexual man or woman explained approximately 6% of the variance. Inspection of the mean scores in Table 6.17 indicates that students who know a bisexual man or woman (M = 24.33, SD = 9.28) hold more positive attitudes towards bisexual women than students who do now know a bisexual man or woman (M = 29.33, SD = 10.38).

151 6.8.2 Investigating differences between students that know or do not know a bisexual man or woman regarding performance on the revised ARBS-M

Table 6.18 shows the means and standard deviation of students’ responses to whether they know a bisexual man or woman and those who do not on the revised ARBS-M.

Table 6.18 Means and Standard Deviations for Students that Know or Do Not Know a Bisexual Man or Women on the Revised ARBS-M Knowing a bisexual man or Std. Scale woman Mean Deviation N Male Tolerance Yes 24.33 9.93 321 No 29.63 10.89 255 Total 26.68 10.69 576 Male Stability Yes 33.10 6.72 321 No 33.78 6.41 255 Total 33.40 6.59 576

A one-way between-groups multivariate analysis of variance was performed to investigate differences between knowing a bisexual man or woman and not knowing a bisexual man or woman concerning attitudes towards bisexual men. Male Tolerance and Male Stability served as the dependent variables and knowing a bisexual man or woman as the independent variable. There was a statistically significant difference between knowing a bisexual man or woman and not knowing a bisexual man or woman: F(2, 573) = 19.74, p < 0.001; Wilks’ Lambda = 0.94; partial eta squared = 0.06. Knowing or not knowing a bisexual man or woman explained approximately 6% of the variance in the linear combination of the Male Tolerance and Male Stability scales of the revised ARBS.

152 When the results for the dependent variables were considered separately, follow-up univariate ANOVA showed a statistically significant difference between knowing a bisexual man or woman and not knowing a bisexual man or woman on the Male Tolerance scale: F(1,574) = 37.18, p < 0.001, partial eta squared = 0.06. For this scale, knowing or not knowing a bisexual man or woman explained approximately 6% of the variance. Inspection of the mean scores in Table 6.18 indicated that students who know a bisexual man or woman (M = 24.33, SD = 9.93) hold more positive attitudes towards bisexual men than students who do now know a bisexual man or woman (M = 29.63, SD = 10.89).

6.8.3 Effect of ethnicity on knowing or not knowing a bisexual man or woman

As a further step, the effect of ethnicity on differences regarding knowing or not knowing a bisexual man or woman was investigated. The Female Tolerance and Female Stability scores of the black men, white men, black women and white women and the interaction of knowing or not knowing a bisexual man or woman and ethnicity were subjected to 2 X 2 MANOVA. There was no statistically significant difference between the scores of the four groups on the combined dependent variables (Female Tolerance and Female Stability): F(2, 571) = 0.86, p = 0.42; Wilks’ Lambda = 0.99; partial eta squared = 0.003.

The Male Tolerance and Male Stability scores of the black men, white men, black women and white women and the interaction of knowing or not knowing a bisexual man or woman and ethnicity were also subjected to 2 X 2 MANOVA. There was no statistically significant difference between the scores of the four groups on the combined dependent variables (Male Tolerance and Male Stability): F(2, 571) = 0.19, p = 0.821; Wilks’ Lambda = 0.99; partial eta squared = 0.001.

153 6.9 Hypothesis four

6.9.1 Investigating differences between students that know and do not know a gay man or lesbian woman regarding performance on the revised ARBS-F

Table 6.19 shows the means and standard deviation of students who know a gay man or lesbian woman and those who do not on the revised ARBS-F. These are also further divided in terms of ethnicity.

Table 6.19 Means and Standard Deviations for Students that Know or Students that Do not Know a Gay Man or Lesbian Woman on the Revised ARBS-F Knowing a gay man Std. Scale or lesbian woman Mean Deviation N Female Tolerance Yes 25.91 9.86 509 No 31.36 10.37 69 Total 26.56 10.07 578 Female Stability Yes 29.53 5.75 509 No 29.35 5.57 69 Total 29.51 5.72 578

A one-way between-groups multivariate analysis of variance was performed to investigate differences between students who know a gay man or lesbian woman and those who do not with regards to attitudes towards bisexual women. Female Tolerance and Female Stability served as the dependent variables and knowing a gay man or lesbian woman as the independent variable. There was a statistically significant difference in attitudes between those who know a gay man or lesbian woman and those who do not: F(2, 575) = 11.70, p < 0.001; Wilks’ Lambda = 0.96; partial eta squared = 0.04. Knowing a gay man or lesbian woman explained approximately 4% of the variance in the linear combination of the Female Tolerance and Female Stability scales of the revised ARBS-F.

154 When the results for the dependent variables were considered separately, follow-up univariate ANOVA showed a statistically significant difference in attitudes between those who know a gay man or lesbian woman and those who do not on the Female Tolerance scale: F(1,576) = 18,36, p < 0.001, partial eta squared = 0.03. For this scale, knowing or not knowing a gay man or lesbian woman explained approximately 3% of the variance. Inspection of the mean scores in Table 6.19 indicates that students who know a gay man or lesbian woman (M = 25.91, SD = 9.86) hold more positive attitudes towards bisexual woman, than students who do now know a gay man or bisexual woman (M = 31.36, SD = 10.37).

6.9.2 Investigating differences between students that know and do not know a gay man or lesbian woman regarding performance on the revised ARBS-M

Table 6.20 shows the means and standard deviation of students who know a gay man or lesbian woman and those who do not on the revised ARBS-M.

Table 6.20 Means and Standard Deviations for Students that Know or do not Know a Gay Man or Lesbian Woman on the Revised ARBS-M Knowing a gay man or lesbian Std. Scale woman Mean Deviation N Male Tolerance Yes 25.95 10.49 509 No 32.17 10.46 69 Total 26.70 10.67 578 Male Stability Yes 33.47 6.58 509 No 32.93 6.60 69 Total 33.40 6.58 578

A one-way between-groups multivariate analysis of variance was performed to investigate differences between students who know a gay man or lesbian woman and those who do not concerning attitudes towards bisexual men. Male Tolerance and Male Stability served as the dependent variables and knowing a gay man or lesbian woman as the independent variable. There was

155 a statistically significant difference between attitudes of students who know a gay man or lesbian woman and those who do not know a gay man or lesbian woman F(2, 575) = 15.09, p < 0.001; Wilks’ Lambda = 0.95; partial eta squared = 0.05. Knowing or not knowing a gay man or lesbian woman explained approximately 5% of the variance in the linear combination of the Male Tolerance and Male Stability scales of the revised ARBS-M.

When the results for the dependent variables were considered separately, follow-up univariate ANOVA showed a statistically significant difference in attitudes between students who know a gay man or lesbian woman and those who do not on the Male Tolerance scale only: F(1,576) = 21.38, p <0.001, partial eta squared = 0.04. For this scale, knowing or not knowing a gay man or lesbian woman explained approximately 4% of the variance. Inspection of the mean scores in Table 6.20 indicates that students who know a gay man or lesbian woman (M = 25.95, SD = 10.49) hold more positive attitudes towards bisexual men, than students who do now know a gay man or lesbian woman (M = 32.17, SD = 10.46).

6.9.3 Effect of ethnicity on knowing and not knowing a gay man or a lesbian woman

As a further step, the effect of ethnicity on differences between the attitudes of those who know or do not know a gay man or lesbian woman was investigated. The Female Tolerance and Female Stability scores of the black men, white men, black women and white women and the interaction of knowing or not knowing a bisexual man or woman and ethnicity were subjected to 2 X 2 MANOVA. There was no statistically significant difference between the scores of the four groups on the combined dependent variables (Female Tolerance and Female Stability): F(2, 573) = 2.42, p = 0.09; Wilks’ Lambda = 0.99; partial eta squared = 0.008.

The Male Tolerance and Male Stability scores of the black men, white men, black women and white women and the interaction of knowing or not knowing a gay man or lesbian woman and ethnicity were also subjected to 2 X 2

156 MANOVA. There was no statistically significant difference between the scores of the four groups on the combined dependent variables (Male Tolerance and Male Stability): F(2, 573) = 2.76, p = 0.082; Wilks’ Lambda = 0.99; partial eta squared = 0.01.

6.10 Hypothesis five

6.10.1 Investigating religiosity differences on the revised ARBS-F

The means and standard deviations for religiosity scores on the ARBS-F are given in Table 6.21.

Table 6.21 Means and Standard Deviations for religiosity on the Revised ARBS-F Std. Scale Religiosity Mean Deviation N Female Tolerance Not at all religious 22.44 8.76 52 Moderately religious 25.38 10.05 350 Deep religious 30.12 9.48 176 Total 26.56 10.07 578 Female Stability Not at all religious 28.25 6.54 52 Moderately religious 29.35 5.54 350 Deep religious 30.19 5.78 176 Total 29.51 5.72 578

The differences in attitudes towards bisexual women were tested by means of a one-way between-groups multivariate analysis of variance, where religion served as the independent variable and Female Tolerance and Female Stability as the dependent variables. There was a statistically significant difference between the religious groups on the combined dependent variables: F(4, 1148) = 9.39, p < 0.001; Wilks’ Lambda = 0.94; partial eta squared = 0.03. Religion explained approximately 3% of the variance in the linear combination of the Female Tolerance and Female Stability scales of the

157 revised ARBS-F. This may be described as a small effect and shows that the practical significance of the differences in religiosity is relatively small, although noticeable.

When the results of the dependent variables were considered separately, follow-up univariate ANOVA showed a statistically significant difference between the religious groups for the Female Tolerance scale only: F(2, 575) = 18.84, p < 0.001, partial eta squared = 0.06. For this scale, religiosity explained approximately 6% of the variance. Post hoc Scheffe tests showed differences between the means of deeply religious students and moderately religious students as well as between deeply religious students and non- religious students. Inspection of the mean scores in Table 6.21 indicates that deeply religious student group held more negative attitudes towards bisexual woman (M = 30.12, SD = 9.48) than the moderately religious group (M = 25.38, SD = 10.05) and the non-religious group (M = 22.44, SD = 8.76). No statistically significant differences were shown between the attitudes towards bisexual women for the moderately religious group and the non-religious group.

6.10.2 Investigating the religiosity differences on the revised ARBS-M

The means and standard deviations for religiosity scores on the ARBS-F are given in Table 6.22.

158 Table 6.22 Means and Standard Deviations for Students’ Responses for religiosity on the Revised ARBS-M Std. Scale Religiosity Mean Deviation N Male Tolerance Not at all religious 32.21 7.60 52 Moderately 33.25 6.36 350 religious Deeply religious 34.05 6.45 176 Total 33.40 6.58 578 Male Stability Not at all religious 22.42 9.12 52 Moderately 25.50 10.76 350 religious Deep religious 30.34 9.94 176 Total 26.70 10.67 578

The differences in attitudes towards bisexual men were tested by means of one-way between-groups multivariate analysis of variance, where Male Tolerance and Male Stability served as the dependent variables and religion served as the independent variable. There was a statistically significant difference between the religious groups on the combined dependent variables, F(4, 1148) = 8.83, p < 0.001; Wilks’ Lambda = 0.94; partial eta squared = 0.03. Religion explained approximately 3% of the variance in the linear combination of the Tolerance and Stability scales of the revised ARBS- M. This may be described as a small effect and shows that the practical significance of the differences in religiosity is relatively small, although noticeable.

When the results of the dependent variables were considered separately, follow-up univariate ANOVA showed a statistically significant difference between the religious groups for the Male Tolerance scale only: F(2,575) = 17.55, p < 0.001, partial eta squared = 0.06. For this scale, religiosity explained approximately 6% of the variance. Post hoc Scheffe tests showed

159 differences between the means of deeply religious students and moderately religious students as well as between deeply religious students and non- religious students. Inspection of the mean scores in Table 6.22 indicates that deeply religious students held more negative attitudes towards bisexual men (M = 34.05, SD = 6.65) than the moderately religious group (M = 33.25, SD = 6.36) as well as the non-religious group (M = 32.21, SD = 7.60). No statistically significant differences in the attitudes towards bisexual men were present between the moderately religious group and the non-religious group.

6.10.3 Effect of ethnicity on religiosity

As a further step, the effects of ethnicity on differences regarding attitudes between the three religious groups were investigated. The Female Tolerance and Female Stability scores of the black men, white men, black women and white women and the interaction of religiosity and ethnicity were subjected to 2 X 2 MANOVA. There was no statistically significant difference between the scores of the groups on the combined dependent variables (Female Tolerance and Female Stability): F(4, 1142) = 1.94, p = 0.10; Wilks’ Lambda = 0.99; partial eta squared = 0.007

The Male Tolerance and Male Stability scores of the black men, white men, black women and white women and the interaction of religiosity and ethnicity were subjected to 2 X 2 MANOVA. There was no statistically significant difference between the scores of the four groups on the combined dependent variables (Male Tolerance and Male Stability): F(4, 1142) = 1.15, p = 0.33; Wilks’ Lambda = 0.99; partial eta squared = 0.004.

6.11 Hypothesis six

6.11.1 Investigating the differences between the sexual orientation categories’ performance on the revised ARBS-F

The means and standard deviations for the different sexual orientation groups are given in Table 6.23, for the black, white and total groups respectively.

160 Table 6.23 Means and Standard Deviations for Sexual Orientation Groups on the Revised ARBS-F

Scale Sexual Orientation

Heterosexual Black (n = 282) White (n = 230) Total (N = 512) Mean SD Mean SD Mean SD F-Tolerance 27.71 10.18 26.68 9.62 27.25 9.94 F-Stability 30.11 5.66 29.17 5.65 29.69 5.67 Homosexual Black (n = 7) White (n = 13) Total (N = 20) Mean SD Mean Mean SD Mean F-Tolerance 22.86 12.08 27.54 10.70 25.90 11.12 F-Stability 27.43 4.16 33.38 6.61 31.30 6.45 Bisexual Black (n = 20) White (n = 12) Total (N = 32) Mean SD Mean Mean SD Mean F-Tolerance 15.25 3.29 18.08 7.20 16.31 5.20 F-Stability 26.65 4.45 24.42 5.48 25.81 4.90 Asexual Black (n = 14) White (n = 0) Total (N = 14) Mean SD Mean Mean SD Mean F-Tolerance 25.71 9.87 00.00 00.00 25.71 9.87 F-Stability 28.86 5.56 00.00 00.00 28.86 5.56 Note. F-Tolerance = Female Tolerance, F-Stability = Female Stability.

A one-way between-groups multivariate analysis of variance was performed to investigate the differences in the attitudes towards bisexual women of students in the sexual orientation groups. Female Tolerance and Female Stability served as the dependent variables and sexual orientation as the independent variable. There was a statistically significant difference between

161 the sexual orientation groups, F(6, 1146) = 7.06, p < 0.001; Wilks’ Lambda = 0.93; partial eta squared = 0.04. Sexual orientation explained approximately 4% of the variance in the linear combination of the Female Tolerance and Female Stability scales of the revised ARBS-F. This may be described as a small effect and shows that the practical significance of the differences in terms of sexual orientation is relatively small, although noticeable.

When the results of the dependent variables were considered separately, follow-up univariate ANOVA showed a statistically significant difference between the sexual orientation groups for the Female Tolerance scale: F(3, 574) = 12.62, p < 0.001, partial eta squared = 0.06. For this scale, sexual orientation explained approximately 6% of the variance. Post hoc Scheffe tests showed differences between the means of the bisexual group and the heterosexual, homosexual and asexual groups respectively. Inspection of the mean scores in Table 6.23 indicates that heterosexual students (M = 27.25, SD = 9.94) homosexual students (M = 25.90, SD = 11.12), as well as asexual students (M = 25.71, SD = 9.87) hold more negative attitudes towards bisexual woman than bisexual students do (M = 16.31, SD = 5.20).

The univariate ANOVA also showed a statistically significant difference between the sexual orientation groups for the Female Stability scale: F(3, 574) = 5.45 p = 0.001, partial eta squared = 0.03. Sexual orientation explained approximately 3% of the variance. Post hoc Scheffe tests showed differences between the means of the bisexual group and the heterosexual, homosexual and asexual groups respectively. Table 6.23 reveals that heterosexual students (M = 29.69, SD = 5.67) homosexual students (M = 31.30, SD = 6.45) and asexual students’ (M = 28.86, SD = 5.56) attitudes towards bisexual women were more negative than bisexual students’ attitudes towards bisexual women (M = 25.81, SD = 4.90). There were no statistically significant differences between the mean scores of the heterosexual, homosexual and asexual students.

162 6.11.2 Investigating the differences between the sexual orientation categories’ performance on the revised ARBS-M

The means and standard deviations for the different sexual orientation groups are given in Table 6.24. These are also presented separately for black and white students.

Table 6.24 Means and Standard Deviations for Sexual Orientation Groups on the Revised ARBS-M

Scale Sexual Orientation

Heterosexual Black (n = 282) White (n = 230) Total (N = 512) Mean SD Mean SD Mean SD M-Tolerance 28.18 10.94 26.41 10.12 27.39 10.60 M-Stability 34.29 6.38 32.80 6.57 33.62 6.50 Homosexual Black (n = 7) White (n = 13) Total (N = 20) Mean SD Mean Mean SD Mean M-Tolerance 20.00 7.91 27.15 10.38 24.65 10.00 M-Stability 30.57 4.08 37.31 7.74 34.95 7.34 Bisexual Black (n = 20) White (n = 12) Total (N = 32) Mean SD Mean Mean SD Mean M-Tolerance 15.15 3.69 18.08 7.13 16.25 5.33 M-Stability 30.30 5.49 27.67 6.56 29.31 5.95 Asexual Black (n = 14) White (n = 0) Total (N = 14) Mean SD Mean Mean SD Mean M-Tolerance 28.14 11.12 00.00 00.00 28.14 11.12 M-Stability 32.57 6.93 00.00 00.00 32.57 6.93 Note. M-Tolerance = Male Tolerance, M-Stability = Male Stability.

163 The differences in attitudes towards bisexual men were tested by means of a one-way-between-groups multivariate analysis of variance, where sexual orientation served as the independent variable and Male Tolerance and Male Stability as the dependent variables. There was a statistically significant difference between the different sexual orientation groups on the combined dependent variables, F(6, 1146) = 6.67, p < 0.001; Wilks’ Lambda = 0.93; partial eta squared = 0.03. Sexual orientation explained approximately 3% of the variance in the linear combination of the Male Tolerance and Male Stability scales of the revised ARBS-M. This may be described as a small effect and shows that the practical significance of the differences in terms of sexual orientation is relatively small, although noticeable.

When the results of the dependent variables were considered separately, follow-up univariate ANOVA showed a statistically significant difference between the sexual orientation groups for the Male Tolerance scale: F(3, 574) = 11.91, p < 0.001, partial eta squared = 0.06. For this scale, sexual orientation explained approximately 6% of the variance. Post hoc Scheffe tests showed differences between the means of the bisexual group and the means of the other three groups respectively. Inspection of the mean scores in Table 6.24 indicated that heterosexual students (M = 27.39, SD = 10.60), homosexual students (M = 24.65, SD = 10.00) and asexual students (M = 28.14, SD = 11.12) hold more negative attitudes towards bisexual men than bisexual students do (M = 16.25, SD = 5.33).

Univariate ANOVA also showed a statistically significant difference between the sexual orientation groups for the Male Stability scale: F(3, 574) = 4.85, p = 0.002, partial eta squared = 0.03. For this scale, sexual orientation explained approximately 3% of the variance. Post hoc Scheffe tests showed differences between the means of the bisexual group and the means of the other three groups respectively. Inspection of the mean scores in Table 6.24 indicated that heterosexual students (M = 33.62, SD = 6.50), homosexual students (M = 34.95, SD = 7.34) and asexual students (M = 32.57, SD = 6.93) hold more negative attitudes towards bisexual men than bisexual students do (M =

164 29.31, SD = 5.95). There were no statistically significant differences between

the mean scores of the heterosexual, homosexual and asexual students.

6.11.3 Effect of ethnicity on the four sexual orientation categories

As a further step, the effect of ethnic differences between the sexual orientation groups was investigated. The Female Tolerance and Female Stability scores of the black men, white men, black women and white women and the interaction of sexual orientation and ethnicity were subjected to 2 X 2 MANOVA. There was no statistically significant difference between the scores of the groups on the combined dependent variables (Female Tolerance and Female Stability): F(4, 1140,00) = 2.274, p = 0.059; Wilks’ Lambda = 0.98; partial eta squared = 0.008.

The Male Tolerance and Male Stability scores of the black men, white men, black women and white women and the interaction of sexual orientation and ethnicity were subjected to 2 X 2 MANOVA. There was a statistically significant difference between the scores of the four groups on the combined dependent variables (Male Tolerance and Male Stability): F(4, 1140) = 2.58, p = 0.036; Wilks’ Lambda = 0.98; partial eta squared = 0.009. Sexual orientation explained approximately 1% of the variance in the linear combination of the Male Tolerance and Male Stability scales of the revised ARB-MS. This may be described as a small effect and shows that the practical significance of the differences in terms of sexual orientation is very small.

When the results of the dependent variables were considered separately, follow-up univariate ANOVA showed a statistically significant difference between the interaction of sexual orientation and ethnicity for the groups for the Male Stability scale only: F(2, 571) = 3.74, p = 0.024, partial eta squared = 0.01. For this scale, sexual orientation explained approximately 1% of the variance. Post hoc Scheffe tests showed differences between the means of the bisexual group and the means of the other three groups respectively. Inspection of the mean scores in Table 6.24 shows that white homosexual

165 students (M = 37.31, SD = 7.74) hold more negative attitudes towards bisexual men than black homosexual students do (M = 30.57, SD = 4.08).

6.12 Chapter summary

In view of the statistical results, it is evident that students’ attitudes towards bisexual men and women vary on a range of factors such as gender, religiosity, knowing a bisexual and the sexual orientation of the participants.

There was no statistically significant difference between the white students and the black students on the combined dependent variables on the ARBS-F and the ARBS-M. Gender explained approximately 4% of the variance in the linear combination of the Tolerance and Stability scales of the revised ARBS- F, and 5% on the revised ARBS-M. Female students were more positive towards bisexual men and women than male students. Knowing a bisexual man or woman explained approximately 6% of the variance in the linear combination on the revised ARBS-F and the revised ARBS-M. Students who know a bisexual man or woman hold more positive attitudes towards bisexual men and women. Knowing a gay man or lesbian women explained approximately 4% of the variance on the revised ARBS-F and 3% of the variance on the revised ARBS-M. Students who know a gay man or lesbian woman hold more positive attitudes towards bisexual men and women.

Religiosity explained approximately 3% of the variance on the ARBS-F and the ARBS-M. Deeply religious students held more negative attitudes towards bisexual men and women than moderately religious or students that are not religious at all. In terms of sexual orientation, the results indicated that sexual orientation explained approximately 4% of the variance on the ARBS-F and 3% of the variance on the ARBS-M. Post hoc Scheffe tests showed differences between the means of the bisexual group and the means of the other three groups (heterosexual, homosexual, and asexual).The results indicated that heterosexual, homosexual and asexual students held more negative attitudes towards bisexual men and women than bisexual students.

166 6.13 Preview of the contents of the following chapter

In the following chapter, the statistical findings presented in Chapter 6 are discussed. First, the results related to phase one and two of the study are discussed. Next, the psychometric properties of the revised instruments are discussed. Thereafter, the hypotheses and the relationships between respondents’ attitudes toward bisexual men and women and various demographic variables are discussed. This is followed by recommendations, and the limitations of the study.

167 CHAPTER 7

DISCUSSION, LIMITATIONS AND RECOMMENDATIONS

7.1 Introduction

This chapter discusses the statistical findings presented in Chapter 6. Firstly, brief discussions of the results of the first and second phases of the research are provided. Thereafter the results pertaining to the psychometric properties of the revised instrument are discussed. The hypotheses and the most significant variables related to attitudes towards bisexual men and women are then discussed. In some instances, given the lack of information on attitudes towards bisexual men and women, the researcher drew on the extensive body of research describing attitudes towards lesbian women and gay men. The present study is then evaluated in terms of its limitations. This is followed by suggestions concerning future studies and recommendations with the view to improving attitudes towards bisexual men and women.

7.2 Phase one

During this phase, the original ARBS-FM (Mohr & Rochlen, 1999) was administered to the university sample. The internal consistency reliability for the total sample was acceptable for the Tolerance subscale, comparing well to the reliability coefficients reported by Mohr and Rochlen (1999). The Stability subscale did not yield acceptable reliability coefficients. Instruments that have been developed outside of South Africa are not necessarily applicable to, and reliable for, the complex South African population (Foxcroft & Roodt, 2001). Investigating the reliabilities of the subscales separately for the black and white groups showed that these two groups did not attach the same meanings to the individual items of the ARBS-FM.

168 7.3 Phase two

The first phase yielded unreliable results, with clear differences between how the black and white students interpreted items of the ARBS-FM, as well as the nine categories of sexual orientation. The purpose of the expert advice group was therefore complementary to the quantitative study, as the researcher endeavoured to seek clarification for the results obtained in the first phase. The group was conducted to gain a better understanding of how black and white students view the concepts of bisexuality and sexual orientation as well as how they interpret the items of the ARBS-FM.

The expert advice group consisted of eleven participants. The researcher administered the survey to the participants, and obtained valuable feedback from them. From the group discourse it appeared that the term bisexual is problematic, and that different cultural groups may use different terminology to describe bisexuality. Furthermore, the group participants found the nine categories of sexual orientation difficult to comprehend. The participants indicated that variables such as language proficiency, culture, education and social-economic status may have influenced the statistical findings from the first quantitative study. Various recommendations were made by the group participants. The adaptation of the ARBS-FM, as well as the nine categories of sexual orientation, was considered necessary given the multicultural and multilingual nature of South African society.

7.4 Phase three: Psychometric properties of the revised instrument

Based on the results of phases one and two of the study, the original ARBS- FM was revised and developed into two instruments, one measuring attitudes towards bisexual women and the other towards bisexual men. The items of the revised ARBS-F and ARBS-M were subjected to two separate confirmatory factor analyses for black and white students respectively to determine whether the two-factor structure proposed by Mohr and Rochlen (1999) could also be found for these student groups. Based on the results of the scree test, two factors were retained for further investigation. The

169 interpretation of the two factors was consistent with those reported by Mohr and Rochlen (1999). Therefore, the subscales of Tolerance and Stability were used to measure attitudes towards bisexual women and men. Using Tucker’s coefficients of congruence it was found that these two factors have equal meaning for the black and white students.

In comparison to the internal consistency reliabilities of the original ARBS-FM obtained in phase one; the revised scales yielded significantly higher Cronbach’s alpha coefficients. With the exception of the Female Stability subscale for the black group, all the other subscales showed coefficients of higher than 0.70 for both groups. Correlation coefficients for female and male bisexual orientation for both the Tolerance and Stability subscales exceeded 0.97, indicating that black students and white students hold similar attitudes towards bisexual women and men with respect to Tolerance and Stability. This was in line with the findings of Mohr and Rochlen (1999). The ARBS-FM (Mohr & Rochlen, 1999) consists of an equal number of bisexual female and bisexual male items, and is used to assess attitudes about bisexuality without regard for attitude levels specific to bisexual women or bisexual men. Given the current research findings, attitudes towards bisexual men and women can be measured by an integrated scale, which provides for attitudes towards both bisexual men and women.

Analyses pertaining to differential item functioning and differential test functioning (Penfield, 2005) on both the revised ARBS-F and ARBS-M suggested the removal of various items that were biased in favour of either the white group or the black group. After removing these items, the reduced scales were used for the final analyses related to the stated hypotheses.

7.5 Hypotheses

7.5.1 Hypothesis one

Hypothesis one stated that students from different ethnic groups have different attitudes towards bisexual men and women. This hypothesis was

170 included because studies measuring attitudes towards lesbian women and gay men (Battle & Bennett, 2000; Lewis, 2003), have indicated that black respondents in the United States of America have more negative attitudes than their white counterparts. However, Arndt and De Bruin (2006) found that race does not have a statistically significant effect on attitudes towards homosexuals in South Africa.

In terms of attitudes towards bisexual men and women, it appears that black and white students do not have different attitudes towards bisexual men and women. There was no statistically significant difference between the white students and black students on the combined dependent variables, namely Tolerance and Stability, with regards to their attitudes. These findings support recent research results concerning attitudes towards lesbian women and gay men in South Africa (Arndt & De Bruin, 2006).

7.5.2 Hypothesis two

Hypothesis two stated that male and female students differ in terms of their attitudes towards bisexual men and woman. One of the most consistent findings to emerge from research is that males and females have different attitudes towards bisexual men and women. Previous studies (Eliason 1997; Herek, 2002) have focused on the relationship between gender and attitudes towards bisexual men and women. For example, Elliason (1997) found that heterosexual men rated bisexual men as very unacceptable, but were more tolerant or even accepting of bisexuality in women. Mohr and Rochlen (1999) also found that heterosexual men viewed female bisexuality as being more moral and tolerable than male bisexuality.

Over the last two decades, American research has focused on the relationship between gender and attitudes towards lesbian women and gay men. These research findings indicated that men exhibit more negative attitudes towards lesbians and gay men than women. For example, in a study, researchers found a higher prevalence of negative attitudes towards homosexuals in men than in women, This was also the case when attitudes

171 towards lesbians and gay men were considered separately (Herek & Capitanio, 1996; Wellings et al., 1994).

Similarly to other studies measuring attitudes towards lesbian women and gay men (Herek & Capitanio, 1996; Wellings et al., 1994), the current research findings indicate that the male participants were less tolerant towards bisexual men and women than the female respondents. Gender explained approximately 3% of the variance on the revised ARBS-F, and 5% on the revised ARBS-M, with male students having more negative attitudes towards bisexual men and bisexual woman than female students. Thus, it appears that the differences in the male and female participants’ attitudes towards bisexual men and women are similar to the differences in their attitudes towards lesbian women and gay men (Arndt & De Bruin, 2006). With regards to the effect of ethnicity on gender differences, the only result that was significant in the current research was on the revised ARBS-F. Ethnicity explained only approximately 1% of the variance, with black men having more negative attitudes towards bisexual women than white men, black women and white women. This finding has limited practical significance and offers support for the findings pertaining to hypothesis one, where no statistically significant differences were found between the attitudes of black and white students.

In previous research, bisexual men and women have also commented that heterosexuals appeared to regard them as homosexuals, which suggests that expressions of hostility toward bisexuals are often rooted in antigay attitudes (Rust, 2000; Weinberg et al., 1994). It appears that participants in the research study were focusing on the homosexual component of bisexual men and women, and disregarding the fact that bisexual men and women also have relationships with opposite sex partners. Even though bisexual men and women form heterosexual as well as same-sex relationships, it is the latter that are stigmatised (Herek, 2000).

172 These findings raise intriguing questions as to the factors underlying these differences in attitudes towards bisexual men and women. It appears that the factors that underlie these differences are similar to the factors underlying the differences in male and female participants’ attitudes towards lesbian women and gay men. One of the underlying factors could be gender differentiation, the belief that the natures of women and men require them to hold different social roles. Heterosexism, the belief that heterosexual relationships are necessary for a complete and happy life, especially for men could also be a factor (Glick & Fiske, 1997). From the perspective of a gender-role belief system, men hold more negative attitudes towards bisexual men and bisexual women because they place higher value on female dependence, gender differentiation, and heterosexuality. Homosexuality is viewed, in part, as a rejection of traditional gender-roles and therefore presents a strong threat to the self-concept of individuals who are highly invested in traditional gender roles.

The research findings of Whitley (2001), confirm that gender-role beliefs are closely linked to men’s attitudes towards homosexuality, accounting for 23% of the variance in attitudes in the study. Consequently, men have more negative attitudes towards lesbians and gay men than women, and may even engage in anti-gay behaviour as a form of self-reaffirmation, demonstrating the magnitude of their rejection of homosexuality (Kite & Whitley, 1998). Similarly, men may be more negative towards bisexual men and women as these individuals reject traditional gender roles and are a threat to the self- concept of individuals who are invested in traditional gender roles.

McCreary (1994) used a cultural construction of gender to explain the more negative attitudes of heterosexual men towards gay men. The male sex role emphasises the link between heterosexuality and masculinity, thus heterosexual men may need to affirm their masculinity by rejecting gay and bisexual men. According to Rich (1980), heterosexuality is not an aspect of femininity that must be affirmed and proven, but is instead an aspect of femininity that is assumed and compulsory. Therefore, women may not need to assert their heterosexuality as strongly as men do.

173 The results from the present study suggest that South African students’ negative attitudes towards bisexual men and women may be partially explained by factors related to gender-roles and heterosexuality. The violation of traditional gender roles may be particularly aversive to male participants because of the perceived threat to male power and privilege. The more positive attitudes of the female participants might be because feminism has expanded women’s roles in South Africa. In addition, women might display more empathy for the social status of bisexual men and women because of women’s greater care-giving roles, as well as their subjective experience of sexism (Potgieter, 1997 in Greene, 1997). Research further suggests that there may be somewhat greater flexibility and fluidity in the sexual histories of women as opposed to men and as a result female participants have less negative attitudes towards bisexual men and women (Bell & Weinberg, 1978; McIntosh, 1981; Rust, 2001; Saghir & Robbins,1973).

7.5.3 Hypothesis three

Hypothesis three stated that students who know a bisexual man or women have more positive attitudes towards bisexual men and women. Given the age that individuals first engage in bisexual behaviour, have bisexual attractions and identify as bisexual, together with research findings (Evans & Broido, 1999; Herek, 1995; Klein, 1990) that university year are a common time for individuals to come out, it is important to call attention to the prevalence of bisexuality among university students.

In the current study, 321 (56%) of the participants indicated that they know a bisexual man or woman. In addition, 32 (5%) of the participants indicated that they are bisexual, compared to the 20 (3%) students who indicated that they are homosexual. These percentages concur with previous research findings (Janus & Janus, 1993; Kinsey et al., 1948, 1953; Masters & Johnson, 1979) that indicated that the prevalence of bisexuality is greater than or comparable to that of homosexuality.

174 Thus, the invisibility of bisexuality does not reflect the incidence of bisexuality in the population. The current research findings concur with Yoshino’s (2000) research findings that suggest that bisexual invisibility is not a reflection of the fact that bisexual men and women do not exist, but that it is rather a product of social erasure.

In terms of the hypothesis, Allport (1954) proposed that regular contact with members from different groups is a necessary means for reducing negative attitudes. The few studies that have investigated friendships with lesbians and gay men (O’Boyle & Thomas, 1996) suggest that the sustained positive contact offered by friendships can help heterosexual women and men develop positive attitudes towards lesbians and gay men. Furthermore, visible friendships across sexual orientation lines can serve as powerful, positive models for university students who might otherwise forgo attempts to forge such friendships themselves. The attitudes of some respondents may be related to their opportunities (or lack of opportunities) to interact personally with bisexual men and women.

The results of the present study support the research hypothesis and Allport’s (1954) contact hypothesis, which indicate that negative attitudes can be reduced through contact between majority and minority group members. In the current study, knowing a bisexual man or woman explained approximately 6% of the variance on the revised ARBS-F and the revised ARBS-M, with students knowing a bisexual man or woman having more positive attitudes towards bisexual men and women.

Stereotypes abound in popular culture, with bisexuals being widely perceived as promiscuous, unable to have committed relationships and likely to contract sexually transmitted diseases (Eliason, 2001; Rust, 1995; Spalding & Peplaum, 1997). It appears that contact with bisexual men and women challenges negative stereotypes and leads to a more positive attitude towards bisexual men and women (Amir & Ben-Ari, 1985).

175 It appears that participants having had experiences with bisexual men and women that disconfirmed the negative stereotypes may have led to more positive attitudes towards bisexual men and women. Consequently, if contact does indeed reduce biphobia, then this would be a possible path to reducing negative attitudes towards bisexual men and women.

7.5.4 Hypothesis four

Hypothesis four stated that students who know a lesbian woman or gay man have more positive attitudes towards bisexual men and women. Of the participants in the current study 509 (88%) indicated that they know a lesbian woman or gay man. A large body of research on attitudes towards members of stigmatised minority groups has been guided by the previously explained contact hypothesis (Allport, 1954).

Among heterosexual individuals, contact with lesbian women and gay men has been found to be a correlate of stereotype reduction and more positive attitudes towards these individuals in college (Geller, 1991; Herek, 2002; Herek & Glunt, 1993). The current findings are consistent with previous research findings.

In the current study, knowing a gay man or lesbian woman explained approximately 4% of the variance on the revised ARBS-F and 5% on the revised ARBS-M, with students knowing a gay man or lesbian woman having less negative attitudes towards bisexual men and women. The results of the present study therefore once again support Allport’s (1954) contact hypothesis, which indicates that negative attitudes may be reduced by contact between majority and minority group members. The findings of the current study may be as a result of the participants having had experiences with sexual minorities that led to more positive attitudes towards bisexual men and women. Therefore, interpersonal contact with sexual minorities is a significant predictor of participants’ attitudes towards bisexual men and women.

176 7.5.5 Hypothesis five

Hypothesis five stated that students who are more religious have more negative attitudes towards bisexual men and women. More conservative views regarding sexuality have been associated with higher levels of biphobia. Societal values, especially as determined by religion, are important factors related to the formation of attitudes and stereotypes about specific groups. Previous studies have found that individuals who are more religious, have more conservative religious beliefs and attend church frequently are more negative towards sexual minorities (Glassner & Owen, 1976; Gray, Kramer, Minick, McGee, Thomas & Greiner, 1996; Herek & Glunt, 1993).

Waldo (1998) conducted research on students from 12 religious groups and their perception about the degree to which the campus is accepting, supportive (or otherwise) of sexual minorities and their concerns. Findings from the study provided support for the claim that heterosexual students with strong religious convictions demonstrate less favourable views toward sexual minorities.

The findings from the present study suggest that a higher level of negative attitudes exists when students are deeply religious. Religiosity explained approximately 6% of the variance on the revised ARBS-F and the revised ARBS-M. Deeply religious students held the most negative attitudes towards bisexual men and women, followed by the moderately religious group and the non-religious group.

The current research findings are consistent with previous findings (Herek, 2002; Waldo, 1998) that deeply religious people hold more negative attitudes towards moral tolerance of female and male bisexuality than people who are not at all religious or moderately religious. Therefore, negative attitudes toward bisexual men and women are probably part of a general belief system that includes a high level of religiosity and traditionalism regarding gender and sexuality.

177 7.5.6 Hypothesis six

Hypothesis six stated that heterosexual, homosexual, and asexual students have more negative attitudes towards bisexual men and women than bisexual students. The results indicated that heterosexual, homosexual and asexual students had more negative attitudes towards bisexual men and women than bisexual students. Sexual orientation explained approximately 6% of the variance on the revised ARBS-F and the revised ARBS-M. Heterosexual students held the most negative attitudes towards bisexual men and woman, followed by homosexual students, asexual students and bisexual students. These negative attitudes may stem from a lack of accurate information about sexuality and sexual orientation, from sexual taboos that permeate the South African culture, from lack of knowledge about bisexuality specifically, and articles in the media that perpetuate negative stereotypes.

The current findings substantiate previous research findings (Firestein, 1996; Hutchins & Ka’ahumanu, 1991) that bisexual men and women reported lack of validation, isolation, ostracism, negative attitudes and stereotyping from both the heterosexual and homosexual communities. The findings confirm the notion that bisexuals see themselves as twice rejected, once by heterosexuals, and again by homosexuals. This double discrimination by heterosexuals and the gay and lesbian community is seldom recognised or acknowledged. The current research findings illuminate that this oppression is real, and may have damaging effects on bisexual men and women (Firestein, 1996). Therefore, it appears that as a result of the double discrimination, bisexual men and women may experience themselves as existing in two different worlds, and not fully fitting into either - “a feeling of political and personal homelessness” (Shuster, 1991, p. 267). Thus, it appears that biphobia is not simply a variant of homophobia. Lesbian women and gay men’s negative attitudes towards bisexuality may be evidence for one major difference between homophobia and biphobia (Rust, 1995).

Biphobia has often been assumed to be identical to homophobia. Yet, only modest scientific attention has been paid to the relationship between the

178 respondents’ sexual orientation and their attitudes towards bisexual men and women. The current findings suggest that both homosexual and heterosexual participants have negative attitudes towards bisexual men and women. This double discrimination by heterosexual men and women and the gay and lesbian community is seldom recognised.

According to Fishbein and Ajzen’s (1975) theories of Reasoned Action and Planned Behaviour, one’s attitude toward a group will affect one’s behaviour toward that group. Therefore, students’ responses to a survey concerning attitudes toward bisexual men and women will reflect how the students will act when encountering an individual from that group. Given the current research finding, declaring a bisexual identity in the heterosexual or homosexual student community may result in experiences of discrimination, antagonism and invalidation.

Furthermore, the Social Identity Theory (Tajfel & Turner, 1979) utilises the human propensity to categorise individuals into groups and then offers a logical rationale for the development and maintenance of inter-group bias. In this instance, the homosexual community may view themselves as superior to the heterosexual community; the heterosexual community may view themselves as superior to the homosexual community; and both these aforementioned groups may view themselves as superior to the bisexual community.

Given the recent research findings, it appears that oppression by the heterosexual and homosexual student community is real and may have damaging effects on bisexual students. A fear of rejection by both heterosexual and homosexual communities, along with a lack of a visible bisexual community may cause bisexual students to remain closeted, or to simply associate with the community that corresponds with the sexual orientation of their current partner. Biphobia and homophobia should therefore be considered to be related, but distinct, phenomena. Given the relatively recent recognition of bisexuality as a category of sexual identity, it is not surprising that empirical research on attitudes towards bisexual individuals is

179 scant. This study represents a first step in collecting information regarding participants’ attitudes towards bisexual men and women. The findings in the present study contribute to the literature on the subject and provide results that may lead to new insights into the attitudes towards bisexual men and women in South Africa.

7.6 Limitations of the study

A limitation of the current study is that the majority of the literature and research findings presented are taken from American sources. Although the American research findings are relevant to the current study, caution needs to be taken not to imply that these studies are necessarily equally valid for South Africa context. Findings from the American studies can thus not be generalised to South Africa. This is evidenced in the clear differences in the manner in which the black and white students interpreted items of the ARBS- FM in the first phase of the study. Therefore, adaptation of instruments is vital, given the multicultural nature of South African society.

A similar questionnaire was utilised in a previous study (Arndt & De Bruin, 2005). This questionnaire included an assessment of the participants’ attitudes towards lesbians and gay men, operationalised by the respondent’s scores on the Attitudes towards Lesbians and Gay Men Scale (ATLG) (Herek, 1994). Assessment of reliability and validity of the ATLG was conducted with a sample of 368 undergraduate volunteers. Alpha coefficients indicated satisfactory levels of internal consistency for the ATLG scale (α = 0.90) and the subscales (α = 0.89 for the ATG, and α = 0.77 for the ATL). Despite the fact that a similar questionnaire was utilised in a previous study (Arndt & De Bruin, 2005) in South Africa, the ARBS-FM used in the current study did not proof to be reliable for use amongst the South African population. Therefore, a pilot study would have provided valuable information pertaining to the appropriateness of the ARBS-FM in South Africa.

Furthermore, the researcher measured the attitudes towards bisexual men and bisexual woman separately in the third phase of the study. The results

180 suggest that attitudes towards bisexual men and women can be measured by a general scale and not by two separate scales.

In Chapter two the researcher discussed alternative views on categorising sexual orientation. In phase one of the study, the researcher attempted to include nine different categories of sexual orientation (Berkey et al., 1990). However, as a result of the expert group’s recommendations, only four categories were included in the third phase of the study. This limits the current research and does not necessary reflect the diversity of potential sexual orientations.

Most studies cited in the literature review describe predominantly white participants. Therefore, it is recognised that bias may occur, as certain groups are often excluded from research. In the current study the researcher focused on the black and white students’ attitudes towards bisexual men and women and (because of a significantly lower number of participants) omitted the coloured and Asian participants’ responses. Building on current findings, further research with a variety of populations, educational levels and age groups in different settings and geographical areas is necessary to give a clearer picture of the extent of biphobia in South Africa.

Although the student sample was substantial (N = 1 459 for the first phase, and N = 578 for the third phase) it was drawn from a single university. Data from other universities and post-school settings may have yielded different results. The study also did not include professional university staff in its sample. Further investigation is therefore required concerning the attitudes of the academic and administrative staff towards bisexual men and women. According to Herek (2000), university students may be one of the most liberal subcultures and therefore the degree of biphobia in the general population might actually be underestimated by this sample. In this regard, Herek (2002) suggest that education could be a factor that influences attitudes towards bisexual men and women. Educational attainment has a significant impact on biphobia, with students having more positive attitudes towards bisexual men and women. Furthermore, younger respondents have also grown up in an era

181 characterised by increasingly tolerance for sexual minorities (Herdt, 2001) Younger respondents who have a higher level of education are thus likely to be more positive towards bisexual men and women. Thus, caution must be used in generalising the results of the present study to the general population

A further limitation is that the research relied on volunteer participants. Consequently, there are concerns regarding the representativeness of the research sample. This concern is particularly appropriate in sexuality research, where volunteers have been found to be more liberal, sex-positive and more permissive than those who choose not to participate (Strassberg & Lowe, 1998). However, according to Wiederman (1999), research suggests a minimal effect of volunteer bias on anonymous surveys of sexuality as virtually all college students express willingness to participate in anonymous sex surveys.

Another limitation relates to the role of social desirability in shaping individuals’ responses to bisexual attitude measures. Although some communities may accept or even encourage negative attitudes toward bisexual men and women, people who see themselves as liberal may be reluctant to express disapproval towards any marginalised individuals. Therefore, researchers should attempt to find ways of accounting for underreported negative attitudes. The measure of political correctness for sexual orientation developed by Brittan-Powel, Bashshur, Pak and Meyenburg (1999) could serve as a starting point for efforts to assess this type of social desirability in studies of attitudes regarding bisexuality.

Finally, the Social Identity Theory was used as a theoretical framework for measuring attitudes towards bisexuals. The researcher acknowledges that, even although this theory has often been used in research on attitudes, it has limitations and other theoretical approaches should be explored.

7.7 Recommendations

7.7.1 Measuring instruments

182

In view of the outcome of the first and third phases of the study, further research on appropriate measures to ascertain the attitudes towards bisexual men and women in South Africa is recommended. Additional research pertaining to the validity and reliability of the revised ARBS-F and ARBS-M utilised in the current study is further recommended.

The categories of sexual orientation included in surveys also require further examination. Although the use of self-labels (homosexual and heterosexual), is problematic, self-ascription is one of the most important aspects of an individual’s identity (Hansen & Evans, 1985). In this regard, Gonsiorek et al. (1995) advise that categories that indicate sexual orientation should capture complexities, yet maintain interpretability.

Furthermore it is evident that the dichotomous model of sexuality (heterosexual and homosexual) contains assumptions that introduce bias in research on sexual orientation. These categories of sexuality also mask diversity that can be crucial for generating accurate research results (Lorber, 1996). Bisexual men and women and gay men and lesbian women are often categorised as a single entity in social scientific research studies, resulting in the potential distortion of knowledge. The current research findings regarding the prevalence of bisexual men and women (5%) draw attention to the importance of the inclusion of bisexuality as a sexual orientation category in research studies.

7.7.2 Qualitative research

Given the contribution of the expert group in the current study, qualitative methods have much to offer researchers dealing with minority or marginalised groups. Lonborg and Phillips (1996, p. 187) describe qualitative methods as good techniques for generating information for description “when little is known about a particular population”. Croteau (1996, p. 228) agrees and added that open-ended inquiries “allow the concepts and models to emerge

183 from the unique experiences and perceptions of the socially marginalised groups”.

Future studies might consider comparing actual experiences of discrimination and harassment of bisexual men and women in their sample. More research needs to be conducted that specifically looks at the health of bisexual men and women. Where possible this research should take a community based participatory action research approach in order to best serve the communities and individuals being studied. As Ryan et al. (2000, p. 5) observes “documenting the experiences of bisexual people in accessing health care is essential to the project of adapting health services to meet the needs of these individuals, their families and communities”.

The nature of the subjective bisexual experience, including longitudinal change, needs to be further investigated. However, a qualitative study would not solve the problem of obtaining data from students who choose to keep their sexual orientation secret. Thus, while qualitative methods may offer improvements, the fact remains that, while an important and timely issue, the study of bisexual men and women’s experiences in the university community is a difficult task. In view of this difficulty, suggestions to increase participation of bisexual men and women in studies are offered below.

Finally, it should be mentioned that the discussion on theoretical issues on sexuality could have included issues of essentialism versus social constructionism. Further qualitative research in this field can explore a discussion on this issue.

7.7.3 Research participants

Researchers should make the relationship between a research project and benefits for the community explicit. For many bisexual men and women there is not a clear connection between participating in a study and promoting the well-being of the community. Participants may feel there is little point in

184 participating in research because they see no results or action following research recommendations. There is a need for wider dissemination of research leading to changes in policy and practice. This may encourage participation in research (Zea, Reisen & Diaz, 2003).

According to Orrrell-Valente, Pinderhughes, Valente and Laird (1999) individuals are also more likely to participate in research when they perceive racial and socio-economic similarities with researchers. Patton (1990) has asserted the necessity of indicating the researchers’ qualifications for the study and the paradigmatic orientation and assumptions underlying the study. The use of insiders from the community is critical to successfully recruit participants. Establishing a collaborative relationship with leaders within understudied segments of the larger bisexual community is a further tool for entry. These leaders often act as guardians and will not permit researchers whom they distrust into the community. Respect for the fact that many bisexual men and women are not out is also crucial, as many of them are cautious because of experiences with biphobia (Diaz, 1998). Given the personal nature of sexual orientation, there are significant ethical issues connected with research in this field. It is crucial, in order to not jeopardise future research, that research is carried out with sensitivity, that it does not breach privacy, and that subjects can be assured of confidentiality.

Given the prevalence of misinformation and negative attitudes toward bisexual men and women and the potential damaging effects of such attitudes, it may be time to shift the focus of research from attitudes toward bisexuality to the dissemination of the research findings so that appropriate intervention programmes can be instituted. Clausen (1997, pp. 148-149) challenged experts to focus on issues of justice and equity:

For over 100 years, anxious laypeople have turned to the experts hoping to be enlightened as to the “truth” of their sexual natures, or to be reassured that their desires are acceptable. Explanation becomes validation - or proscription, taboo… But the wounds will not be healed by advances in knowledge. We already know

185 enough. Instead, we need to work on enlarging the space for different kinds of people to live well with each other.

7.7.4 Improving the university environment for bisexual men and women

According to Evans (2000), both students and academic staff at universities need to take the lead in shifting the values of higher education to support all students, and in developing interventions to change negative attitudes towards sexual minorities. The development of formal policies is an important step for communicating and practicing inclusiveness. University policies should make it clear that harassment and bias directed at bisexual men and women is no more tolerable than racist behaviour and will be subject to the same procedure and penalties.

Based on extensive research (Bohan, 1996; Clark & Serovich, 1997; Evans & Broido, 1999; Keppel, 2002; Malaney, Williams & Geller, 1997; McCarthy, 2002; Meyer, 2004; Pearson, 2003; Rhoads, 1995; Smiley, 1997; Williams, 2002) the following suggestions are offered to challenge negative attitudes towards sexual minorities in university communities:

Training that addresses biphobia may be provided for staff members. Furthermore, this training ought to incorporate discussions of the environment that bisexual men and women face both in and out of class. Leaders of the faculties and administrative staff should be encouraged to provide visible support to sexual minorities. “Safe” places should be established for bisexual men and women on campus. Such places are offices or departments that any member of the campus community can go to, and be assured of confidentiality and of sensitivity to their sexual identity.

Professional staff and students at the university need to be educated during their professional programmes on issues specific to bisexual men and women. Furthermore, education for providers should emphasise the necessity of not making assumptions regarding people’s sexuality. Students’ awareness and sensitivity to differences in sexual preferences should be increased.

186 There could be specific academic courses and programmes on bisexual issues. Most likely, universities would find it easier to include discussions of bisexuality within existing courses such as social psychology, the psychology of women, the psychology of gender, or courses focusing on diversity.

Pearson’s (2003) seminar that incorporates popular songs, readings and class discussions might provide a useful framework when designing training methods for working with sexual minorities. A book by Bohan (1996), titled Psychology and sexual orientation: Coming to terms, provides students with information that is directly relevant to their future work as psychologists. The book also includes a description of courses, that provide a serious scholarly look at the experience of sexual minorities as well as information for sexual minorities, who rarely if ever see themselves reflected in the literature, particularly in a positive manner.

Instruction also needs to occur outside of the classroom, especially within residence halls and student organisations. Evans and Broido’s (1999) study provides recommendations for residence hall professionals, and also provides information that would assist them in creating an environment conducive to the well-being of sexual minorities. Student organisations play an especially critical role in identity development for students, allowing them to actively view the outcomes of their organised efforts. Student led organisations on campuses encourage action, and may be of assistance in change within the lesbian, gay and bisexual communities (McCarthy, 2002; Williams, 2002). According to Meyer (2004), identifying with others, connecting socially and coming to understand oneself as part of a larger community provides students with a sense of empowerment.

7.7.5 Counselling bisexual men and women

Bisexuality challenges traditional rules about sexual expression. In addition, bisexuality is virtually invisible in therapy literature. As a result, bisexual clients experience a range of internal and external crises that may bring them to counselling. Counselling services available on the university campus need

187 to be sensitive to the issues facing those identifying as bisexual. The importance of acknowledging the negative attitudes towards bisexual men and women has direct implications for community and social psychologists in addressing these negative attitudes, reducing victimisation and preventing future incidents. In this regard, is has been noted that although bisexual men and women encounter numerous difficulties that affect their mental health, they continue to be underserved by the counselling profession. Therefore, a counselling course on sexual orientation is recommended, as knowledge and training have been linked to more positive attitudes toward sexual minorities (Kim, D’Andrea, Poonam & Gaughen, 1998). The health profession needs to ensure that all people are supported and assisted in dealing with their sexuality - regardless of the label they choose.

Counselling bisexual clients effectively requires that mental health counsellors first examine their own values and perceptions regarding sexuality, and attempt to understand some of the complicated issues bisexual men and women face. Keppel (2002) comments on the fact that the oppression of bisexual men and women by gay and lesbian individuals is largely unrecognised and misunderstood by professionals. When professionals do not understand that bisexuals may be oppressed by or isolated from both the heterosexual and homosexual communities, they are unable to assess stress, anxiety, relationships or risks appropriately. It is also recommended that a range of bisexual specific programmes and services be developed. These could include information and resources around sexual health developed specifically for bisexual men and women, sex education, support groups for bisexuals and their families, bisexual positive counsellors and mental health services and an anonymous phone line or website. Community partnership in the creation of support systems for bisexuals and the building of a bisexual community is advised. Appropraite interventions include individual support, group counselling and psycho- education. Bisexual clients will benefit from active, non-judgmental affirmative counselling, which provides opportunities for determining individual pathways for sexual, emotional and social satisfaction (Smiley, 1997).

188 7.8 Conclusion

South Africa has one of the most progressive and inclusive constitutions in the world. It has extensive protection of human rights, including the protection of, and respect for diversity. This is something that the ideologies of many Southern African countries such as Zimbabwe, Namibia and Botswana still do not recognise. The amount of public attention paid to issues of sexual orientation has increased dramatically in the past decade. Government, corporations and other institutions are adopting policies that prohibit discrimination based on sexual orientation. The findings of this study combined with past research on sexual minorities, have significant implications concerning attitudes towards bisexual men and women in a university setting. Despite the South African constitution’s guarantee of protection for sexual minorities, it appears that litigation has not filtered down to the level of everyday life in South Africa, and negative attitudes toward sexual minorities persist.

The current research findings provide a first step for further investigation among researchers interested in research in the field of sexuality within the South African student community specifically, and the larger community in general. One of the greatest challenges for South African higher education is to develop a healthy campus climate where opportunities for learning include cultural appreciation and reciprocity among people of diverse backgrounds.

It is important to keep in mind that within the category “bisexual” there are a diverse array of individuals who could also be members of other oppressed or marginalised groups with varying levels of social power and influence; for example, women, people of colour, people living with chronic illness and people with disabilities. According to Greene and Croom (2000), even when people share membership in some of these multiple groups, each individual will have his or her own set of reactions and expectations.

Not only is individual assessment important, but each university campus and the whole of South African higher education should be examined on a

189 continual basis. Therefore, it is important that research be conducted that illuminates the experiences of bisexual men and women, as well as the attitudes towards bisexual men and women, in order to identify effective strategies for overcoming prejudice directed at bisexuals. This data would provide indicators for developing a climate that is supportive of learning, personal growth and professional development. Such knowledge, coupled with a firm belief in the importance of a healthy campus environment for bisexual men and women, may impact on the culture of each institution, by presenting South African higher education as an equitable and humanised social environment in an increasingly conflictual global community. D’Emilio (1990, p. 18) states that “having been granted the extraordinary privilege of thinking critically as a way of life, students should be astute enough to recognise when a group of people is being systematically mistreated”. The research findings of D’Emilio (1990) concluded that the university setting appears to be an environment that is conducive to change or a critical examination of attitudes.

However, there is no easy solution to combating negative attitudes towards sexual minorities on campus. For at least eighteen years before students arrive on campus, they have likely internalised at least some of the negative attitudes prevalent in the South African culture - a culture that has yet to shake its misconceptions and negative attitudes of persons who are lesbian, gay, or bisexual.

According to Fairyinton (2005, p. 4):

The lesson that bisexuality bares is a good one for the 21st century. We live in a world whose reality is more complicated than the simplified binaries of our language and understanding. Bisexuality as the synthesis, the middle ground between seemingly irreconcilable differences, is a form of thinking that has boundless possibilities for social progress”.

190 Hurtado (1999) challenged post secondary institutions by asking what kind of society we aspire to become. If post secondary institutions seek to create the educationally powerful learning environments for all students, it is necessary that they attend to issues of sexual difference as well as other dimensions of diversity. By further examining the experiences of bisexual men and women we can learn more about how to more effectively serve all students and society at large.

191 REFERENCES

Abbott, S. & Love, B. (1972). Sappho was a right-on woman: A liberated view of lesbianism. London: Stein and Day.

Adam, B.D., Sears, A. & Schellenberg, E.G. (2000). Accounting for unsafe sex: Interviews with men who have sex with men. The Journal of Sex Research, 37, 24-36.

Ajzen, I. (1988). Attitudes, personality and behavior. Chicago, IL: Dorsey.

Allen, D.M. (1980). Young male prostitutes: A psychosocial study. Archives of Sexual Behavior, 9 (5), 399-426.

Allport, G.W. (1954). The nature of prejudice. Cambridge, MA: Addison- Wesley.

Altman, D. (1979). Homosexual oppression and liberation. New York: Avon.

Altshuler, K.Z. (1984). On the question of bisexuality. American Journal of Psychotherapy, 38 (4), 484-493.

American Psychiatric Association (1957). Diagnostic and Statistical Manual of Mental Disorders. Washington, DC: American Psychiatric Association.

American Psychiatric Association (1968). Diagnostic and Statistical Manual of Mental Disorders. Washington, DC: American Psychiatric Association.

American Psychiatric Association (1973). Diagnostic and Statistical Manual of Mental Disorders. Washington, DC: American Psychiatric Association.

American Psychiatric Association (1980). Diagnostic and Statistical Manual of Mental Disorders. Washington, DC: American Psychiatric Association.

192 American Psychiatric Association (1987). Diagnostic and Statistical Manual of Mental Disorders. Washington, DC: American Psychiatric Association.

American Psychiatric Association (1994). Diagnostic and Statistical Manual of Mental Disorders. Washington, DC: American Psychiatric Association.

American Psychological Association Division 44/Committee on Lesbian, Gay and Bisexual Concerns Joint Task force on Guidelines for Psychotherapy with Lesbian, Gay and Bisexual Clients. (2000). Guidelines for psychotherapy with lesbian, gay and bisexual clients. American Psychologist, 55 (12), 1440-1451.

Amir, Y. & Ben-Ari, R. (1985). International tourism, ethnic contact, and attitude change. Journal of Social Issues, 41, 105-115.

Anapol, D.M. (1997). Polyamory: The new love without limits: Secrets of sustainable intimate relationships. California: Times Change.

Anderson, L.R. & Randlet, L. (1994). Self-monitoring, perceived control and satisfaction with self-disclosure of sexual orientation. Journal of Social Behavior and Personality, 9 (4), 789-800.

Angelides, S. (2001). A history of bisexuality. Chicago: The University of Chicago.

Appleby, G.A. (2001). Ethnographic study of twenty-six gay and bisexual working-class men in Australia and New Zealand. Journal of Gay and Lesbian Social Services: Issues in Practice, Policy and Research, 12 (3/4), 119-132.

Aranow, R. (1991). Potential lovers. In L. Hutchins and L. Ka’ahumanu (Eds.). Bi any other name: Bisexual people speak out (pp. 233-239). Boston: Alyson.

Arndt, M. & De Bruin, G.P. (2006). Attitudes toward lesbians and gay men: Relations with gender, race, religion among university students. Psychology in Society, 33, 16-30.

193 Avery, S. (1991, January). AIDS: Why no woman is safe. New Women, pp. 96-98.

Bailey, N., Gurevich, M. & Mathieson, C. (2000). Invoking community: Rethinking the health of lesbian and bisexual women. Ottawa: Pink Triangle Services.

Balsam, K.F. & Rothblum, E.D. (2002). Sexual orientation and mental health: A comparison of adult siblings. Paper presented at the 110th Annual Convention of the American Psychiatric Association: Chicago.

Bartlett, M.S. (1950). Tests of significance in factor analysis. British Journal of Psychology, 3, 77-85.

Basow, S.A. (1992). Gender: Stereotypes and roles. New York: Brooks Cole.

Battle, J. & Bennet, M. (2000). Research on lesbian and gay populations within the African American community: What have we learned? African American Research Perspectives, 6 (2), 35-47.

Battle, J. & Lemelle, A.J. (2002). Gender differences in African American attitudes toward gay males. Western Journal of Black Studies, 26, 134-140.

Bayer, R. (1981). Homosexuality and American Psychiatry: The political diagnosis. New York: Basic Books.

Bell, A.P. (1975). Research in homosexuality: Back to the drawing board. Archives of Sexual Behavior, 4 (4), 421-431.

Bell, A.P. & Weinberg, M.S. (1978). Homosexuality: A study of diversity among men and women. New York: Simon and Schuster.

Bergler, E. (1956). Homosexuality: Disease or way of life? New York: Hill and Wang.

194 Bergler, E. & Kroger, W. (1954). Kinsey’s myth on female sexuality. New York: Grove and Stratton.

Berkey, B.R., Perelman-Hall, T. & Kurdek, L.A. (1990). The multidimensional scale of sexuality. Journal of Homosexuality, 19 (4), 67-87.

Betz, N.E. (1991). Implications for counseling psychology-training programs: Reactions to the special issue. The Counseling Psychologist, 19, 248-252.

Bhugra, D. & De Silva, P. (1998). Dimensions of bisexuality: An exploratory study using focus groups of male and female bisexuals. Sexual and Marital Therapy, 13 (2), 145-157.

Bieber, I. (1962). Homosexuality: A psychoanalytic study of male homosexuals. New York: Basic Books.

Bierly, H. (1984). Gender identity and sexual behavior. In K. Howells (Ed.). The psychology of sexual diversity (pp. 63-88). New York: Basil Blackwell.

Blasingame, B.M. (1992). The roots of biphobia: Racism and internalized heterosexism. In E.R. Weise (Ed.). Closer to home: Bisexuality and feminism (pp.162-174). Seattle: Seal.

Blumstein P.W. & Schwartz, P. (1976). Bisexuality in men. Urban Life, 5 (3), 339-358.

Blumstein, P.W. & Schwartz, P. (1977). Bisexuality: Some social psychological issues. Journal of Social Issues, 33 (2), 30-45.

Bode, J. (1976). View from another closet: Exploring bisexuality in women. New York: Hawthorn.

Bohan, J. S. (1996). Psychology and sexual orientation: Coming to terms. New York: Routledge.

195 Boles, J.M. & Elifson, K.W. (1994). Sexual identity and HIV: The male prostitute. Journal of Sex Research, 31 (1), 39-46.

Boles, J.M., Sweat, M. & Elifson, K.W. (1989, March). Bisexuality among male prostitutes. Paper presented at CDC workshop of bisexuality and AIDS: Atlanta, Georgia.

Boulton, M. (1991). Review of the literature on bisexuality and HIV transmission. In R. Tielman, M. Carballo and A. Hendriks (Eds.). Bisexuality and HIV/AIDS: A global perspective (pp. 187-209). Buffalo, NY: Prometheus.

Bower, T. (1995). Bisexual women, feminist politics. In N. Tucker (Ed.). Bisexual politics: Theories, queries and visions (pp. 99-107). New York: Harrington Park.

Boyatzis, R.E. (1998). Transforming qualitative information: Thematic analysis and code development. New York: Sage.

Bradford, M. (2004). The bisexual experience: Living in a dichotomous culture. Journal of Bisexuality, 4 (1/2), 7-23.

Brandt, E. (Ed.). (1999). Dangerous liaisons: Blacks, gays and the struggle for equality. New York: New Press.

Branscombe, N.R., Wann, D.L., Noel, J.G. & Coleman, J. (1993). In-group or out-group extremity: Importance of the threatened social identity. Personality and Social Psychology Bulletin, 19, 381-388.

Braun, V. & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3, 77-101.

Bressler, L.C. & Lavender, A.D. (1986). Sexual fulfillment of heterosexual, bisexual and homosexual women. Journal of Homosexuality, 12 (2/3), 109- 122.

196 Bright, S. (1992). Susie Bright’s sexual reality: A virtual sex world reader. Pittsburgh, P.A: Cleis.

Brink, T.L. (1993). Religiosity: Measurement. In F.N. Magill (Ed.). A survey of social science: Psychology. (pp. 2096 – 2102). CA: Salem Press. Britton-Powel, C., Bashur, M., Pak, E.H. & Meyenburg, T. (1999, August). A measure of political correctness for issues of sexual orientation. Paper presented at the annual convention of the American Psychological Association, Boston, MA.

Broido, E.M. (2000). Constructing identity: The nature and meaning of lesbian, gay and bisexual identities. In R.M. Perez, K.A. DeBord and K.J. Bieschke (Eds.). Handbook of counseling and psychotherapy with lesbian, gay and bisexual clients (pp. 12-33). Washington, DC: American Psychological Association.

Bryant, W. (1991). Love, friendship and sex. In L. Hutchins and L. Ka’ahumanu (Eds.). Bi any other name: Bisexual people speak out (pp. 69- 73). Boston: Alyson.

Burkhart, K.W. (1996). Women in prison: Inside the concrete womb. Boston: Northeastern University.

Burleson, W.E. (2005). Bi America: Myths truths and struggles of an invisible community. New York: Harrington Park.

Burr, C. (1993, May 25). Homosexuality and biology. The Atlantic Monthly, pp. 47-65.

Calderon, J.L., Baker, R.S. & Wolf, K.E. (2000). Focus Groups: A qualitative method complementing quantitative research for studying culturally diverse groups. Education for Health, 13, 91-95.

197 Cameron, E. (1993). Sexual orientation and the constitution: A test case for Human Rights. South African Law Journal, 467-474.

Cantarella, E. (2002). Bisexuality in the ancient world (2nd ed.). New Haven: Yale University.

Carey, B. (2005, July 5). Straight, Gay or Lying? Bisexuality revisited. The New York Times, p. A1.

Cass, V.C. (1979). Homosexual identity formation: A theoretical model. Journal of Homosexuality, 4, 219-235.

Cass, V.C. (1990). The implications of homosexual identity formation for the Kinsey model and scale of sexual preference. In D.P. McWhirter, S.A. Sanders and J.M. Renisch (Eds.). Homosexuality/Heterosexuality: Concepts of sexual orientation (pp. 239-266). New York: Oxford University.

Cattell, R.B. (1966). The scree test for the number of factors. Multivariate Behavioral Research, 1, 245-276.

Centre for Disease Control (1990). HIV/AIDS Surveillance Report, 5, 4:8. (Cited by McKirnan, Stokes and Burzette 1995).

Chapman, B.E. & Brannock, J.C. (1987). Proposed model of lesbian identity development: An empirical examination. Journal of Homosexuality, 14 (3/4), 69-80.

Chickering, A.N. & Reissner, L. (1993). Education an identity. San Francisco: Jossey-Bass.

Christina, C. (1995). Bisexuality. In N. Tucker (Ed.). Bisexual politics: Theories, queries and visions (pp. 162-166). New York: Harrington Park.

198 Churchill, W. (1967). Homosexual behavior among males: A cross-cultural and cross-species investigation. New York: Hawthorn.

Clark, R.D. & Maas, A. (1987). Black South African Lesbians: Discourses on motherhood and women’s roles. Journal of Lesbian Studies, 7, 135-151.

Clark, W.M. & Serovich, J.M. (1997). Twenty years and still in the dark? Content analysis of articles pertaining to gay, lesbian and bisexual issues in marriage and family therapy journals. Journal of Marital and Family Therapy, 23 (3), 239-253.

Clausen, J. (1990). My interesting condition. Journal of Sex Research, 27 (3), 445-459.

Clausen, J. (1997). Beyond gay or straight: Understanding sexual orientation. Philadelphia: Chelsea House.

Cochran, S.D. (2001). Emerging issues in research on lesbians and gay men’s mental health: Does sexual orientation really matter? American Psychologist, 56, 931-947.

Cochran, S.D. & Mays, V.M. (2000). Lifetime prevalence of suicide symptoms and affective disorders among men reporting same-sex sexual partners. American Journal of Public Health, 90 (4), 573-578.

Cochran, S.D., Sullivan, J.G. & Mays, V.M. (2003). Prevalence of mental disorders, psychological distress, and mental services use among lesbians, gay and bisexual adults in the United States. Journal of Consulting and Clinical Psychology, 71 (1), 53-61.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences. Hillsdale, NJ: Lawrence Erlbaum.

199 Coleman, E. (1981). Bisexual and gay men in heterosexual marriage: Conflict and resolutions in therapy. Journal of Homosexuality, 7 (1/2), 93-103.

Coleman, E. (1982). Developmental stages of the coming out process. Journal of Homosexuality, 7, 31-43.

Coleman, E. (1985). Integration of male bisexuality and marriage. Journal of Homosexuality, 11 (1/2), 189-208.

Coleman, E. (1987). Assessment of sexual orientation. Journal of Homosexuality, 13 (1/2), 9-24.

Coleman, E. (1990). Toward a synthetic understanding of sexual orientation. In D.P. McWhirter, S.A. Sanders and J.M. Renisch (Eds.). Homosexuality/ Heterosexuality: Concepts of sexual orientation (pp. 267-276). New York: Oxford University.

Collins, J.F. (2000). The intersection of race and bisexuality: A critical overview of the literature and past, present and future directions of the borderlands. Journal of Bisexuality, 4 (1/2), 99-116.

Comrey, A.L. (1973). A first course in factor analysis. New York: Academic Press.

Cory, D.W. & Le Roy, J.P. (1963). The homosexual and his society. New York: Citadel.

Cox, W. & Gallois, C. (1996). Gay and lesbian identity development: A social identity perspective. Journal of Homosexuality, 4, 1-15.

Cramer, D. (2003). Advanced quantitative data analysis. New York: Open University Press.

200 Cronbach, L.J. (1951). Coefficient alpha and the internal structure of tests. Psychometrika, 16, 297-344.

Croteau,J.M. (1996). Research on the work experiences of lesbian, gay and bisexual people: An integrative review of methodology and findings. Journal of Vocational Behavior, 48, 195-209.

Davenport, W. (1977). Sex in cross-cultural perspectives. In F.A. Beach (Ed.). Human sexuality in four perspectives (pp. 115-163). Baltimore, MD: John Hopkins University.

Davidowitz, E. (1993). The secret life of bisexual husbands. Redbook, 181 (5), 114-117.

Davis, D.W. & Brown, R.E. (2000). The intolerance of Black Nationalism. Paper presented at the annual meeting of the American Political Science Association, Washington, DC.

Davis, C. & Wright, A. (2001). How well are we doing? A survey of the GLBT population of Ottawa. Ottawa: Pink Triangle Services.

Deacon, S.A., Reinke, L. & Viers, D. (1996). Cognitive-behavioral therapy for bisexual couples: Expanding the realms of therapy. The American Journal of Family Therapy, 24 (3), 293.

De Cecco, J.P. & Shively, M.G. (1983). From sexual identities to sexual relationships: A contextual shift. Journal of Homosexuality, 9 (2), 1-26.

D’Emilio, J. (1983). Sexual politics, sexual communities: The making of a homosexual minority in the United States, 1940-1970. Chicago: University of Chicago.

D’Emilio, J. (1990). The campus environment for gay and lesbian life. Academie, 76, 16-19.

201

D’Emilio, J. (1992). Making trouble: Essays on gay history, politics and the university. New York: Routledge.

Delcourt, M. (1966). Hermaphroditea. Brussels: Collection Latomus.

De Monteflores, C. & Shultz, S.J. (1978). Coming out: Similarities and differences for lesbians and gay men. Journal of Social Issues, 34 (3), 59-72.

DeVellis, R.F. (2003). Scale development: Theory and applications. Thousand Oaks, CA: Sage.

Diamond, M. (1998). Bisexuality: A biological perspective. In E.J. Haeberle and R. Gindorf (Eds.). Bisexualities: The ideology and practice of sexual contact with both men and women (pp. 53-80). New York: Continuum.

Diaz, R.M. (1998). Latino gay men and HIV: Culture, sexuality and risk behavior. New York: Routledge.

Diekhoff, G. (1992). Statistics for the social and behavioral sciences: Univariate, bivariate, multivariate. Iowa: Wm.C. Brown. Dixon, J.K. (1984). The commencement of bisexual activity in swinging married women over age thirty. Journal of Sex Research, 20 (1), 71-90.

Doll, L., Myers, T., Kennedy, M. & Allman, D. (1997). Bisexuality and HIV risk: Experiences in Canada and the United States. Annual Review of Sex Research, 8, 102.

Doll, L., Peterson, J., Magana, J.R. & Cortier, J.M. (1991). Male bisexuality and AIDS in the USA. In R.A. Tielman, M. Carballo and A.C. Hendriks (Eds.). Bisexuality and HIV/AIDS: A global perspective (pp. 27-40). Buffalo: Prometheus.

Douglas, M. (1966). Purity and danger. London: Routledge.

202 Dover, K.J. (1978). Greek homosexuality. New York: Vintage.

Dworkin, S.H. (1996). Bisexual women: Understanding sexual identity. Paper Presented at the 104th Annual meeting of the APA: Toronto, Canada.

Dworkin, S.H. (2000). Individual therapy with lesbian, gay and bisexual clients. In K. A. DeBord and K.J. Bieschke (Eds.). Handbook of counseling and psychotherapy for lesbian, gay and bisexual clients (pp. 157-181). Washington, D.C: American Association.

Eadie, J. (1996). Activating bisexuality: Towards a bisexual politics. In J. Bristow and A.R. Wilson (Eds.). Activating theory: Lesbian, gay and bisexual politics (pp. 139-170). London: Lawrence and Wishart.

Einhorn, L. & Polgar, M. (1994). HIV-risk behavior among lesbians and bisexual women. AIDS Education and Prevention, 6 (6), 514-523.

Ekstrand, M.L., Coates, T.J., Guydish, J.R., Hauck, W.W., Collette, L. & Hulley, S.B. (1994). Are bisexually identified men in San Francisco a common vector for spreading HIV infection to women? American Journal of Public Health, 84 (6), 915-919.

Eliason, M.J. (1996). Identity formation for lesbian, bisexual and gay persons: Beyond a minoritizing view. Journal of Homosexuality, 30 (3), 31-58.

Eliason, M.J. (1997). The prevalence and nature of biphobia in heterosexual undergraduate students. Archives of Sexual Behavior, 26, 317-326.

Eliason, M.J. (2001). Bi-negativity: The stigma facing bisexual men. Journal of Bisexuality, 1, 137-154.

Ellis, H. (1933). Psychology of sex: A manual for students. New York: Emerson. (Original work published 1914).

203 Engel, J.W. & Saracino, M. (1986). Love preferences and ideals: A comparison of homosexual, bisexual and heterosexual groups. Contemporary Family Therapy, 8 (3), 241-250.

Epprecht, M. (2002). Male-male sexuality in Lesotho: Two conversations. Journal of Men’s Studies, 10 (3), 373-386.

Esterberg, K. G. (1997). Lesbian and bisexual identities: Constructing communities, constructing selves. Philadelphia: Temple University.

Evans, N.J. (2000). The impact of LGBT safe zone project on campus climate. Journal of College Student Development, 43, 522-539.

Evans, N.J. & Broido, E.M. (1999). Coming out in college residence halls: Negotiating, meaning making, challenges, supports. Journal of College Student Development, 40 (6), 658-659.

Evans, N.J. & D’Augelli, A.R. (1996). Lesbians, gay men and bisexual people in College. In R.C. Savin-Williams and K.M. Cohen (Eds.). The lives of lesbians, gays and bisexuals: Children to adults (pp. 201-226). Fort Worth, TX: Harcourt Brace.

Fairyinton, S. (2005). Bisexuality and the case against dualism. The Gay and Lesbian Review, 12 (4), 32-37.

Fassinger, R.E. (1995). From invisibility to integration: Lesbian identity in the workplace. Career Development Quarterly, 44, 148-167.

Feldman, l. P. (1984). The homosexual preference. In K. Howells (Ed.). The psychology of sexual diversity (pp. 5-41). New York: Basil Blackwell.

Firestein, B.A. (Ed.). (1996). Bisexuality: The psychology and politics of an invisible minority. Thousand Oaks, Cal: Sage.

204 Fishbein, M. & Ajzen, I. (1975). Attitudes toward objects as predictors of single and multiple behavioral criteria. Psychological Review, 81, 59-74.

Foley,S., Ngo,H. & Loi, R. (2006). Antecedents and consequences of perceived personal gender discrimination: A study of solicitors in Hong Kong. Sex Roles, 55, 197-208.

Ford, C.S. & Beach, F.A. (1951). Patterns of sexual behavior. New York: Harper and Row.

Foucault, M. (1984, February). How we behave. Home and Garden, pp. 61- 69.

Foucault, M. (1985). The use of pleasure. (Volume 2). (R. Hurley, Trans). New York: Penguin Books.

Fox, A. (1991). Development of a bisexual identity. In L. Hutchins and L. Ka’ahumanu (Eds.). Bi any other name: Bisexual people speak out (pp. 34- 35). Boston: Alyson.

Fox, R.C. (1995). Bisexual identities. In A.R. D’Augelli and C.J. Patterson (Eds.). Lesbian, gay and bisexual identities over the lifespan: Psychological perspectives (pp. 48-86). New York: Oxford University.

Fox, R.C. (1996). Bisexuality in perspective: A review of theory and research. New York: Oxford University.

Fox, R.C. (Ed.). (2004). Current research on bisexuality. New York: Harrington Park.

Foxcroft, C. & Roodt, G. (2001). An introduction to psychological assessment in the South African context. Cape Town: Oxford University Press.

205 Freud, S. (1953). Introductory lecture on psychoanalysis. New York: Basic Books. (Original work published 1916).

Freud, S. (1957). Leonardo da Vinci and a memory of his childhood. Standard Edition, 11, 59-138 (Original work published 1910).

Freud (1961). The psychogenesis of a case of homosexuality in a women. New York: Basic Books. (Original published in 1920).

Freud, S. (1962). Three essays on the theory of sexuality. New York: Basic Books. (Original work published 1905).

Freud, S. (1963). Analysis terminable and interminable. In S. Rieff (Ed.). Therapy and technique (pp. 233-272). New York: Collier. (Original work published in 1925).

Freud, S. (1963). An autobiographical study. New York: Norton. (Original work published 1937).

Friedman, R.C. & Downey, J.I. (1998). Psychoanalysis and the model of homosexuality as psychopathology: A historical overview. American Journal of Psychoanalysis, 58 (3), 249-271.

Gagnon, J.H. (Ed.). (1977). Human sexuality in today’s world. Boston: Little Brown.

Gagnon, J.H. (1990). The explicit and implicit use of the scripting perspective in sex research. Annual Review of Sex Research, 1, 1-43.

Gagnon, J.H. & Simon, W. (1973). Sexual conduct: The social sources of human sexuality. Chicago: Aldine.

206 Galland, V.R. (1975). Bisexual women (Doctoral dissertation, California School of Professional Psychology, 1975). Dissertation Abstracts International, 36 (6), 3037B.

Gay, J. (1979). Mummies, babies, friends, and lovers in Lesotho. In E. Blackwood (Ed.). The many faces of homosexuality (pp. 97-116). New York: Harrington Park.

Gebhard, P.H., Gagnon, J.H., Pomeroy, W.B. & Christenson, C.V. (1965). Sex offenders: An analysis of types. New York: Harper and Row.

Gelman, D., Drew, L., Hager, M., Anderson, M., Graine, G. & Hutchinson, S. (1987, July). A perilous double love life. Newsweek, pp. 44-46.

George, S. (1993). Women and bisexuality. London: Scarlet.

Gerrard, S. & Haplin, J. (1989, October). The risky business of bisexual love. Cosmopolitan, pp. 203-205.

Gevisser, M. & Cameron, E. (Eds.). (1995). Defiant desire: Gay and lesbian lives in South Africa. London: Routledge.

Giallombardo, R. (1966). Society of women: A study of women’s prison. New York: Wiley and Sons.

Glassner, B. & Owen, C. (1976). Variations in attitudes toward homosexuality. Cornell Journal of Social Relations, 2, 161-176.

Glick, P. & Fiske, S.T. (1997). Hostile and benevolent sexism: Measuring ambivalent sexist attitudes toward women. Psychology of Women Quarterly, 21, 119-137.

207 Golden, C. (1987). Diversity in women’s sexual identities. In Boston Lesbian Psychologies Collective (Ed.). Lesbian psychologies: Explorations and challenges (pp. 19-34). Chicago: University of Illinois.

Gonsiorek, J.C., Sell, R.L. & Weinrich, J.D. (1995). Definition and measurement of sexual orientation. Suicide and Life-Threatening Behavior, 25, 40-53.

Gorsuch, R.L. (1983). Factor analysis. Hillsdale, NJ: Lawrence Erlbaum.

Gravetter, F.J. & Forzano, L.B. (2006). Research methods for the behavioural sciences. Australia: Thomson.

Gray, D.P., Kramer, M., Mimick, P., McGee, L., Thomas, D., & Greiner, D. (1996). Heterosexism in nursing education. Journal of Nursing Education, 35,204-210.

Greene, B. (1997). Ethnic and cultural diversity among lesbians and gay men. London: Sage.

Greene, J. & Caracelli, V. (1997). Advances in mixed methods evaluation: The challenge and benefits of integrating diverse paradigms. San Francisco: Jossey-Bass.

Greene, B. & Croom, G.L. (Eds.). (2000). Education, research and practice in lesbian, gay, bisexual and transgendered psychology. London: Sage.

Greenwald, A. G., Banaji, M. R., Rudman, L. A., Farnham, S. D., Nosek, B. A., & Mellott, D. S. (2002). A unified theory of implicit attitudes, stereotypes, self-esteem, and self-concept. Psychological Review, 109, 3-25.

208 Hackney, C.H. & Sanders, G.S. (2003). Religiosity and mental health: A meta- analysis of recent studies. Journal for the Scientific Study of Religion, 42, 55- 71.

Haeberle, E.J. (1998). Bisexuality: History and dimensions of a modern scientific problem. In E.J. Haeberle and R. Gindorf (Eds.). Bisexualities: The ideology and practice of sexual contact with both men and women (pp. 13- 51). New York: Continuum.

Hajek, C. & Giles, H. (2006). Inter-group communication schemas: Cognitive representations of talk with gay men. Language and Communication, 25, 161- 181.

Hambleton, R.K., Merenda, P.F., & Spielberger, C.D. (Eds.). (2004). Adapting educational tests and psychological tests for cross-cultural assessment. Mawhaw: Erlbaum.

Hammer, K.M. (1992). Gay-bashing: A Social Identity analysis of violence against lesbians and gay men. National Criminal Justice Reference Service, 1, 179-190.

Hansen, W.E., Creswell, J.W., Clark, V.L.P., Petska, K.S. & Creswell, J.D. (2005). Mixed methods research design and Counseling Psychology. Journal of Counseling Psychology, 52, 224-235.

Hansen, C.E. & Evans, A. (1985). Bisexuality reconsidered: An idea in pursuit of a definition. Journal of Homosexuality, 11, 1-6.

Harper, G.W. & Schneider, W.S. (2003). Oppression and discrimination among lesbian, gay, bisexual and transgendered people and communities: A challenge for community psychology. American Journal of Community Psychology, 31, 243-252.

209 Harris, D.A. (1977). Social-psychological characteristics of ambisexuals (Doctoral Dissertation, Univesity of Tennessee, 1977). Dissertation Abstracts International, 39 (2), 574A.

Harry, J. & Lovely, R. (1979). Gay marriages and communities of sexual orientation. Alternative Lifestyles, 2, 177-220.

Hassim, S. (1999). Gender social location and feminist politics in South Africa. Transformation.

Hegarty, P. & Massey, S. (2006). Anti-homosexual prejudice as opposed to what? Queer Theory and the Social Psychology of anti-homosexual attitudes. Journal of Homosexuality, 52, 47-58.

Heller, A.C. (1987, May). Is there a man in your man’s life? What every girl should know about the bisexual guy. Mademoiselle, 93, 134-154.

Hensley, C., Tewksbury, R. & Wright J. (2001). Exploring the dynamics of masturbation of consensual same-sex activity within a male maximum- security prison. Journal of Men’s Studies, 10 (1), 59-71.

Herdt, G.H. (1980). Semen depletion and the sense of maleness. Ethnopsychiatry, 3, 79-116.

Herdt, G.H. (1981). Guardians of the flutes: Idioms of masculinity. New York: McGraw-Hill.

Herdt, G. H. (1984). A comment on cultural attributes and fluidity of bisexuality. Journal of Homosexuality, 10 (3), 53-61.

Herdt, G.H. (1990). Developmental discontinuities and sexual orientation across cultures. In D.P. McWhirter, S.S. Sanders and J.M. Reinisch (Eds.). Homosexuality/Heterosexuality: Concepts of sexual orientation (pp. 208-236). New York: Oxford University.

210 Herdt, G.H. (2001). Social change, sexual diversity and tolerance for bisexuality in the United States. In A.R. D’Augelli and C.J. Patterson (Eds.). Lesbian, gay and bisexual identities and youth: Psychological perspectives (pp. 267-283). New York: Oxford University.

Herek, G.M. (1994). Assessing heterosexuals' attitudes toward lesbian and gay men: A review of empirical research with the ATLG scale. In Green, B & Herek, G.M (Eds.) Psychological perspectives on lesbian and gay issues: Lesbian and gay psychology: Theory, research, and clinical applications. Thousand Oaks, CA: Sage.

Herek, G.M. (1995). Stigma and sexual orientation: Understanding prejudice against lesbians, gay men and bisexuals: Psychological perspectives on lesbians and gay issues. Thousand Oaks, CA: Sage.

Herek, G.M. (2000). The psychology of sexual prejudice. Current Directions in Psychological Science, 9, 19-22.

Herek, G.M. (2002). Heterosexuals’ attitudes toward bisexual men and women in the United States. The Journal of Sex Research, 39, 264-275.

Herek, G.M. (2008).Selected abstracts. http:/psychology.ucdavis.edu/rainbow/ HTML/bibabs.html.

Herek G.M. & Capitanio, J.P. (1996). Some of my best friends: Intergroup contact, concealable stigma and heterosexual attitudes towards gay men and lesbians. Personality and Social Psychology Bulletin, 22, 412-424.

Herek, G.M., Gillis, J.R. & Cogan, J.C. (1999). Psychological sequelae of hate crime victimization among lesbian, gay and bisexual adults. Journal of Consulting and Clinical Psychology, 67, 945-951.

211 Herek, G.M. & Glunt, E.K. (1993). Interpersonal contact and heterosexuals attitudes towards gay men. Results from a national survey. Journal of Sex Research, 30, 239-244.

Hickson, A. (1995). The poisoned bowl: Sex, repression and the public school system. London: Constable.

Highet, G. & Highet, H. (1953). Sexual life in ancient Rome. London: Sage.

Himelhoch, B. (1988). The bisexual potential. In T. Geller (Ed.). Bisexuality: A reader and resource book. Ojai Cal: Times Change.

Hirschfeld, M. (1946). The sexual history of the World War. New York: Panurge.

Hoburg,R., Konik, J., Williams, M. & Crawford, M. (2004). Bisexuality among self-identified heterosexual college students. Journal of Bisexuality, 4 (1/2), 25-36.

Hodon, G., Harry, H., & Mitchell, A. (2008). Independent benefits of contact and friendships on attitudes toward homosexuals among authoritarians and highly identified heterosexual. European Journal of Social Psychology.

Hoffman, M. (1968). The gay world: Male homosexuality and the social creation of evil. New York: Basic Books.

Hogg, M.A. (2006). Social identity theory. In P.J. Burke (Ed.). Contemporary social psychological theories (pp. 111-136). Palo Alto, CA: Stanford University Press.

Hogg, M.A. & Abrams, D. (1988). Social identifications: A social psychology of intergroup relations and group processes. London: Routledge.

212 Hogg, M.A. & McGarty, C. (1990). Self categorization and social identity. In D. Abrams and M.A. Hogg (Eds.). Social Identity Theory: Constructive and critical advances (pp. 10-27). London: Harvester Wheatsheaf.

Hogg, M.A., Terry, D.J. & White, K.M. (1995). A tale of two theories: A critical comparison of identity theory with social identity theory. Social Psychology Quarterly, 58, 255-269.

Holman, R. (1991). Ethics in Social Research. Harlow: Longman.

Holtzen, D.W. (1994). Handedness and sexual orientation. Journal of Clinical and Experimental Neuropsychology, 16 (5), 702-712.

Hooker, E. (1956). A preliminary analysis of group behavior of homosexuals. Journal of Psychology, 41, 219-230.

Hooker, E. (1957). The adjustment of the male overt homosexual. Mattachine Review, 33-44.

Hooker, E. (1963). The problem of homosexuality in modern society. New York: E.P. Dutton.

Hornsey, M.J. (2008). Social identity theory and self-categorization theory: A Historical Review. Social and Personality Psychology Compass, 2, 204-222.

Howard, J. & Hollander, J.A. (1997). Gendered situations, gendered selves: A gender lens on Social Psychology. Thousand Oaks, CA: Sage.

Hudson, W.W. & Rickets, W.A. (1980). A strategy for the measurements of homophobia. Journal of Homosexuality, 5, 357-372.

Human Rights Watch (2003). More than a name: State-sponsored homophobia and its consequences in Southern Africa. New York: Human

213 Rights Watch and the International Gay and Lesbian Human Rights Commission.

Humphreys, L. (1970). Tearoom trade: Impersonal sex in public places. Chicago: Aldine.

Hunt, M. (1974). Sexual behavior in the 1970’s. Chicago: Playboy.

Hurtado, S. (1999). Reaffirming educators’ judgment: Educational value of diversity. Journal of Liberal Education, 85 (2), 24-31.

Hutchins, L. (1996). Bisexuality: Politics and community. In B.A. Firestein (Ed.). Bisexuality: The psychology and politics of an invisible minority (pp. 240-259). Thousand Oaks, Cal: Sage.

Hutchins, L. & Ka’ahumanu, L. (1991). Bi any other name: Bisexual people speak out. Boston: Alyson.

James, J. (1976). Motivations for entrance into prostitution. In L. Crites (Ed.). The female offender (pp. 177-198). Lexington, Mass: Lexington.

Janus, S.S. & Janus, C.L. (1993). Homosexuality. In S.S. Janus and C.L. Janus (Eds.). The Janus report on sexual behavior (pp. 69-71). New York: John Wiley and Sons.

Johnson, R.B. & Onwuegbuzie, A.J. (2004). Mixed methods research: A research paradigm whose time has come. Educational Researcher, 33, 14-27.

Jorm, A.F., Korten, A.E., Rodgers. B., Jacomb, P.A., & Christenson, H. (2002). Sexual orientation and mental health: Results from a community survey of young and middle-aged adults. British Journal of Psychiatry, 180, 423-427.

214 Kaiser, H.F. (1970). A second generation Little-Jiffy. Psychometrika, 35, 401- 415.

Kaiser Family Foundation (2001). Inside-out. A report on the experience of lesbians, gays and bisexual in America and the public’s view on issues and politics related to sexual orientation. Retrieved October 6, 2005, from http://www.kff.org.

Kaplan, G.T. & Rogers, L.J. (1984). Breaking out of the dominant paradigm: A new look at sexual orientation. Journal of Homosexuality, 10 (3/4), 71-75.

Kelly, G.A. (1955). The psychology of personal constructs. New York: Norton. Keppel, B. (2002). The challenges and rewards of life as an outspoken bisexual elder. Ottawa: Pink Triangle Services.

Keppel, B. (2006). Affirmative psychotherapy with older bisexual women and men. Journal of Bisexuality, 6 (1/2), 85-104.

Kim, B.S., D’Andrea, M.J., Poonam, K.S. & Gaughen, K.J. (1998). A multicultural study of university students’ knowledge of and attitudes toward homosexuality. Journal of Humanistic Education and Development, 36, 171- 182.

Kinsey, A.C., Pomeroy, W.B. & Martin, C.E. (1948). Sexual behavior in the human male. Philadelphia: Saunders.

Kinsey, A.C., Pomeroy, W.B., Martin, C.E. & Gebhard, P.H. (1953). Sexual behavior in the human female. Philadelphia: Saunders.

Kite, M.E. & Whitely, B.E. (1998). Do heterosexual women and men differ in their attitudes toward homosexuality. In G.M. Herek (Ed.). Stigma and sexual orientation: Understanidng prejudice against lesbians, gay men and bisexuals (pp. 39-61). Thousand Oaks, CA: Sage.

215 Klein, F. (Ed.). (1978). The bisexual option: A concept of one-hundred percent intimacy. New York: Arbor House.

Klein, F. (1990). The need to view sexual orientation as a multivariable dynamic process: A theoretical perspective. In D.P. McWhirter, S.A. Sanders and J.M. Renisch (Eds.). Homosexuality/Heterosexuality: Concepts of sexual orientation (pp. 239-266). New York: Oxford University.

Klein, F. (Ed.). (1993). The bisexual option. New York: Harrington.

Klein, F., Sepekoff, B. & Wolf, T.J. (1985). Sexual orientation: A multi-variable dynamic process. Journal of Homosexuality, 11 (1/2), 35-49.

Knobel, J. (1993). The design and analysis of focus group studies: A practical approach. In D.L. Morgan (Ed.). Successful focus groups: Advancing the state of the art (pp. 35-40). Newbury Park, CA: Sage.

Krafft-Ebing, R. (1965). Psychopathia sexualis: A medico forensic study. New York: G.P. Putnam’s Sons. (Originally published 1831).

Kristiansen, C.M. (1990). The symbolic/value-expressive functions of out- group attitudes among homosexuals. The Journal of Social Psychology, 130, 61-76.

Krueger, R.A. (1988). Focus groups: A practical guide for applied research. London: Sage.

Lance, L.M. (2002). Heterosexism and homophobia among college students. College Student Journal, 36, 410-415.

Lauman, E.O., Gagnon, J.H., Michael, R.T. & Michael, S. (1994). The social organizations of sexuality: Sexual practices in the United States. Chicago: University of Chicago.

216 Le Vay, S. (1993). The sexual brain. Cambridge, MA: MIT Press.

Lever, J., Kanouse, D.E., Rogers, W.H., Carson, S. & Hertz, R. (2000). Behavior patterns and sexual identity of bisexual males. Journal of Sex Research, 29 (2), 141-167.

Levin, S. & Sidanius, J. (1999). Social dominance and social identity in the United States and Israel: In-group favouritism or out-group derogation? Political Psychology, 20, 99-126.

Levitt, E.E. & Klassen, A.D. (1974). Public attitudes toward homosexuality: Part of the 1970 National survey by the Institute of Sex Research. Journal of Homosexuality, 1, 29-43.

Lewis, G.B. (2003). Black-white differences in attitudes toward homosexuality and gay rights. Public Opinion Quarterly, 67, 59-79.

Lipponen, J. & Leskin, J. (2006). Conditions of contact, common in-group identity, and in-group bias toward contingent workers. The Journal of Social Psychology, 6, 671-685.

Liss, M, M., Crawford, M. & Popp, D. (2004). Predictors and correlates of collective action. Sex Roles, 50, 771-783.

Lockwood, P. & Bursey, B. (2001). Your everyday wellness guide: A gay, lesbian, bisexual, transgender community resource guide. Ottawa: Pink Triangle Services.

Loewenstein, S.F. (2000). On the diversity of love objects orientations among women. In P.C. Rust (2000). Bisexuality in the United States (pp. 203-216). New York: Columbia University.

217 Lonborg, S.D. & Phillips, J.M. (1996). Investigating the career development of gays, lesbians and bisexuals. Methodological considerations and recommendations. Journal of Vocational Behavior, 48, 176-194.

London, K.A. & Williams, L.B. (1990, May). The use of self-administered questionnaire for improving abortion reporting in the 1988 National Survey of Family Growth. Paper presented at the annual meeting of the Population Association of America, Toronto, Canada.

Lorber, J. (1996). Beyond the binaries: Depolarizing the categories of sex, sexuality and gender. Sociological Inquiry, 66 (2), 143-159.

Lourea, D.N. (1985). Psychosocial issues related to counseling bisexuals. Journal of Homosexuality, 11, 51-63.

Ludwig, M.R. & Brownell, K.D. (1999). Lesbians, bisexual women, and body image: An investigation of gender roles and social group affiliation. International Journal of Eating Disorders, 25, 89-97.

MacDonald, A.P. (1981). Bisexuality: Some comments on research and theory. Journal of Homosexuality, 6 (3), 21-35.

MacDonald, A.P. (1983). A little bit of lavender goes a long way: A critique of research on sexual orientation. Journal of Sex Research, 19, 94-100.

MacIntosh, J. (1993). Focus groups in distance nursing education. Journal of Advanced Nursing, 18, 1981-1985.

Maddision, S. (2000). Fags, Hags and Queer Sisters. London: MacMillan.

Magura, S., O’Day, J. & Rosenblum, A (1992). Women usually take care of their girlfriends: Bisexuality and HIV risk among female intravenous drug users. Journal of Drug Issues, 22 (1), 179-190.

218 Malaney, G.D., Williams, E.A. & Geller, W.W. (1997). Assessing campus climate for gays, lesbians and bisexuals at two institutions. Journal of College Development, 38, 365-376.

Marmor, J. (Ed.). (1965). Sexual inversion: The multiple roots of homosexuality. New York: Basic Books.

Masters, W.H. & Johnson, V.E. (1979). Homosexuality in perspective. Boston: Little Brown.

Matteson, D.R. (1997). Bisexual and homosexual behavior and HIV risk among Chinese-, Filipino-, and Korean-American men. Journal of Sex Research, 34 (1), 93-104.

Mayfield, W.A. & Carrubba, M.D. (1996, August). Validation of the Attitudes Toward Bisexuality Inventory. Presented at the 104th Annual Convention of the American Psychological Association, Toronto, Ontario, Canada.

Mayfield, W.A., Carrubba, M.D. & Louie, C. (1995, August). Development of an inventory to measure attitudes toward bisexuality. In M.J. Patton (Chair). Performance-based outcomes of research in counseling psychology. Symposium conducted at the 103rd Annual Convention of the American Psychological Association, New York.

McCaghy, C.H. & Skipper, J.K. (1969). Lesbian behavior as an adaptation to the occupation of stripping. Social Problems, 17, 262-270.

McCarthy, R. (2002, May). Report assesses climate for gays, lesbians at UGA. The Atlanta Journal Sentinel, p. 8.

McConaghy, N. (1987). Heterosexual/Homosexual: Dichotomy or continuum. Archives of Sexual Behavior, 16 (5), 411-424.

219 McCornick, N. (1994). Sexual salvation: Affirming women’s sexual rights and pleasures. Westport, CT: Praeger.

McCreary, D.R. (1994). The male role and avoiding femininity. Sex Roles, 31, 517-531.

McIntosh, M. (1981). The homosexual role. Social Problems, 16 (2), 152-182.

McKirnan, D.J., Stokes, J.P., Doll, L. & Burzette, R.G. (1995). Bisexually active men: Social characteristics and sexual behavior. Journal of Sex Research, 32 (1), 65-81.

McLean, K. (2001). Living life in the double closet: Bisexual youth speak out. Hecate, 27 (1), 109-118.

McLean, K. (2004). Negotiating non-monogamy: Bisexuality and intimate relationships. Journal of Bisexuality, 4 (1/2), 83-97.

McNair, R., Kavanagh, A., Agius, P. & Tong, B. (2005). The mental health status of young adult and mid-life non-heterosexual Australian women. Australian and New Zealand Journal of Public Health, 29 (3), 265-277.

Martin, A. (1982). Some issues in the treatment of gay and lesbian patients. Psychotherapy, Research and Practice, 19, 341-348.

Mead, M. (1968). Growing up in New Guinea. Harmondsworth: Penguin Books.

Mead, M. (1975, January). Bisexuality: What’s it all about? Redbook, 29-31.

Meyer, I. (1995). Minority stress and mental health in gay men. Journal of Health and Social Behavior, 36, 38-56.

220 Meyer, M.D. (2004). We’re too afraid of these imaginary tensions: Student organizing in lesbian, gay, bisexual and transgender campus communities. Communication Studies, 55 (4), 499-515.

Michael, R.T., Gagnon, J.H., Laumann, E.O. & Kolata, G. (1994). Sex in America: A definitive survey. Boston: Little Brown and Company.

Michel, R. (1996). From performativity to interpretation: Toward a social semiotic account of bisexuality. In D.E. Hall and M. Pramaggione (Eds.). Representing bisexualities: Subject and cultures in fluid desire (pp. 55-69). New York: New York University.

Mohr, J.J., Israel, T. & Sedlacek, W.E. (2001). Counselors’ attitudes regarding bisexuality as predictors of counselor’s clinical responses: An analogue study of a female bisexual client. Journal of Counseling Psychology, 48, 212-222.

Mohr, J.J. & Rochlen, A.B. (1999). Measuring attitudes regarding bisexuality in lesbian, gay male and heterosexual populations. Journal of Counseling Psychology, 46, 353-369.

Money, J. (1977). Bisexual, homosexual and heterosexual: Society, law and medicine. Journal of Homosexuality, 2, 229-233.

Money, J. (2003). History, causality and sexology. The Journal of Sex Research, 40 (3), 237-242.

Money, J. & Lamacz, M. (1989). Vandalized lovemaps. Buffalo, N.Y: Prometheus.

Morgan, D.L. (1988). Focus groups as qualitative research. London: Sage.

Morgan, D.L. (1997). Focus groups as qualitative research (2nd ed.). London: Sage.

221 Morse, E.V., Simon, P.M., Balson, P.M. & Osofsky, H.J. (1992). Sexual behavior patterns of customers of male street prostitutes. Archives of Sexual Behavior, 21 (4), 347-357.

Morse, E., Simon, P.M., Osofsky, H.J., Balson, P.M. & Gaumer, H.R. (1991). The male prostitute: A vector for transmission of HIV infection into the heterosexual world. Social Science and Medicine, 32 (5), 535-539.

Nakayama, T. (1998). The impact of an LGBT safe zone project on campus climate. Journal of College Student Development, 43 (4), 522-539.

Nassar-McMillan, S.C. & Borders, L.D. (2002). Use of focus groups in Survey Item Development. The Qualitative Report, 7, 1-11.

National Center for Health Statistics (2005, 9 October). Sexual behavior and selected health measures: Men and women 15-44 years of age, United States, 2002. Retrieved October 9, 2005 from http://www.cdc.gov.

National Science Foundation (1993). User-friendly handbook for project evaluation: Science, Mathematics, Engineering and Technology. Arlington, VA: NSF.

Newsweek (1974, May) Bisexual chic: Anyone goes. Newsweek, 83, p. 90. Nicolosi, J., Byrd, A.D. & Potts, R.W. (2000). Retrospective self-reports of changes in homosexual orientation: A consumer survey of conversion therapy clients. Psychological Reports, 86, 1071-1088.

Nunnally, J.C. (1978). Psychometric theory. New York: McGraw-Hill.

Nurius, P.S. (1983). Mental health implications of sexual orientation. Journal of Sex Research, 19 (2), 119-136.

O’Boyle, C.G. & Thomas, M.D. (1996). Friendships between lesbian and heterosexual women. In J.S. Weinstock and E.D. Rothblum (Eds.). Lesbian

222 friendships: For ourselves and each other (pp. 240-248). New York: University Press.

Ochs, R. (1996). Biphobia: It goes more than two ways. In B.A. Firestein (Ed.). Bisexuality: The psychological and politics of an invisible minority. Thousand Oaks, Cal: Sage.

Ochs, R. & Deihl, M, (1992). Moving beyond binary thinking. In W. Blumenfeld (Ed.). Homophobia: How we all pay the price (pp. 67-75). Boston: Beacon.

O’Conner, P. (1997). Warning: Contents under heterosexual pressure. Hypatia, 12 (3), 193-198.

Orrell-Valente, J.K., Pinderhughes, E.E., Valente, E. & Laird, R.D. (1999). If it’s offered, will they come? Influences on parents’ participation in a community-based conduct problems prevention program. American Journal of Community Psychology, 27, 753-783.

Oskamp, S. (2002). Attitudes and opinions. New Jersey: Prentice Hall.

Ostrove, J.M. & Long, S.M. (2007). Social class and belonging: Implications for college adjustment. Review of Higher Education, 4, 363-387.

Oswald, D.L. & Lindstedt, K. (2006). The content and function of gender self- stereotypes: An exploratory investigation. Sex Roles, 54, 447-458.

Page, E.H. (2004). Mental health services experiences of bisexual women and bisexual men: An empirical study. Journal of Bisexuality, 4 (1/2), 135-160.

Pallant, J. (2005). SPSS survival manual: A step by step guide to data analysis using SPSS for Windows (Version 12). (2nd ed.). New York: Open University Press.

223 Pattatucci, A.M. & Hamer, D.H. (1995). Development and familiarity of sexual orientation in females. Behavior Genetics, 25 (5), 407-420.

Patton, M.Q. (1990). Qualitative evaluation and research methods. Newbury Park, CA: Sage.

Paul, J.P. (1984). The bisexual identity: An idea without social recognition. Journal of Homosexuality, 9 (2/3), 45-63.

Paul, J.P. (1985). Bisexuality: Reassessing our paradigms of sexuality. Journal of Homosexuality, 11 (2/3), 21-34.

Paul, J.P. (1988). Counseling issues in working with a bisexual population. In M. Shernof and W.A. Scott (Eds.). The sourcebook on lesbian/gay health care (pp. 142-150). Washington, DC: National Lesbian/Gay Health Foundation.

Pearson, Q.M. (2003). Breaking the silence in the counsellor education classroom: A training seminar on counseling sexual minority clients. Journal of Counselling and Development, 81 (3), 292-303.

Penfield, R.D. & Algina, J. (2006). A generalized DIF effect variance estimator for measuring unsigned differential test functioning in mixed format tests. Journal of Educational Measurement, 43 (3), 295-312.

Pillard, R.C. & Bailey, M. (1995). A biological perspective on sexual orientation. Clinical Sexuality, 8 (1), 71-84.

Polders, L.A. (2007). Factors affecting vulnerability to depression among gay men and lesbian women. Unpublished Masters Dissertation: UNISA.

Polders, L.A. & Wells, H. (2004). Overall research finings on levels of empowerment among LGBT people in Gauteng, South Africa. OUT Research, LGBT wel-being.

224 Potgieter, C. A. (1997). From apartheid to Mandela’s constitution. Black South African lesbians in the nineties. In B. Greene (Ed.), Ethnic and cultural diversity among lesbians and gay men (pp. 88 – 116). Calif: Sage Publications.

Potgieter, C. A. (2005). Sexualities? Hey, this is what black South African lesbians have to say about relationships with men, the family, heterosexual women and culture. In M. van Zyl and M. Steyn (Eds.), Performing queer: shaping sexualities 1994-2004 (pp. 177-192). Roggebaai: Kwela Books.

Potgieter, C.A. (2003). Black South African Lesbians: Discourses on motherhood and women’s roles. Journal of Lesbian Studies, 7 (3), 135-151.

Propper, M. (1981). Prison homosexuality: Myth and reality. Toronto: Lexington.

Queen, C. (1996). Bisexuality, sexual diversity and the sex-positive perspective. In B.A. Firestein (Ed.). Bisexuality: The psychology and politics of an invisible minority (pp. 103-126). Thousand Oaks, C.A: Sage.

Race, K.E., Hotch, D.F. & Parker, T. (1994). Rehabilitation program evaluation: Use of focus groups to empower clients. Evaluation Review, 18, 730-740.

Ramey, J.W. (1997). Emerging patterns of innovative behavior in marriage. In J.R. Smith and L.G. Smith (Eds.). Beyond monogamy (pp. 103-137). Baltimore: John Hopkins.

Ramsey,L.R., Haines, M.E., Hurt, M.M., Nelson, J.A. Turner, D.L. Liss, M. & Erchull, M.J. (2007). Thinking of others: Feminist identification and the perceptions of others’ beliefs. Sex Roles, 56, 611-616.

225 Reid, G. & Dirsuweit, T. (2001). Understanding systematic violence: Homophobic attacks in Johannesburg and the surrounds. Unpublished dissertation. Wits Institute of Social and Economic Research.

Reddy, V., Potgieter, C., & Mkhize, N. (2007). Cloud over the rainbow nation: Corrective rape and other hate crimes against black lesbians. HSRC Review, 5 (1), 9-11.

Retief, G. (1994). Keeping Sodom out of the lager. In M. Gevisser and E. Cameron (Eds.). Defiant desire, (p. 103). Johannesburg: Raven.

Rhoads, R.A. (1995). The campus climate for gay students who leave the closet. The Chronicle for Higher Education, 41, 56-67.

Richardson, D. (1984). The dilemma of essentiality in homosexual theory. Journal of Homosexuality, 9 (2/3), 79-90.

Rieger, G., Chivers, M.L. & Bailey, J.M. (2005). Sexual arousal patterns of bisexual men. Psychological Science, 16 (8), 579-584.

Ritvo, L.B. (1990). Darwin’s influence on Freud: A tale of two sciences. New York: Yale University.

Robinson, V. (1997). My baby just cares for me: Feminism, heterosexuality and non-monogamy. Journal of Gender Studies, 6 (2), 143-157.

Rosario, M., Meyer-Bahlburg, H.F., Huner, J., Exner, T.M., Gwadz, M. & Keller, A.M. (1996). The psychosexual development of urban lesbian, gay and bisexual youths. Journal of Sex Research, 33 (2), 113-126.

Rose, S. (1998). Searching for the meaning of AIDS: Issues affecting black gay men. In V. Derlega & A.P. Barabee (Eds.). HIV and social interaction (pp. 56-82). Thousand Oaks, CA: Sage.

226 Ross, M.W. (1979). Bisexuality: Fact or fallacy. British Journal of Sexual Medicine, 6, 49-50.

Ross, M.W. (1984). Beyond the biological model: New directions in bisexual and homosexual research. Journal of Homosexuality, 10 (3/4), 63-70.

Ross, M. W. (1991). A taxonomy of global behavior. In R. Tielman, M. Carballo and A. Hendriks (Eds.). Bisexuality and HIV/AIDS: A global perspective (pp. 21-26). Buffalo, NY: Prometheus.

Ross, M.W. & Paul, J.P. (1992). Beyond gender: The basis of sexual attraction in bisexual men and women. Psychological Reports, 71, 1283-1290.

Ross, M.W., Rogers, L.J. & McCulloch, H. (1978). Stigma, sex and society: A new look at gender differentiation and sexual variation. Journal of Homosexuality, 3, 113-132.

Rotheram-Borus, J.M., Hunter, J. & Rosario, M. (1994). Suicidal behavior and gay related stress among gay and bisexual male adolescents. Journal of Adolescent Research, 9 (4), 498-508.

Rubenstein, M. (1982). An in-depth study of bisexuality and its relation to self- esteem. (Doctoral dissertation cited by Zinik, 1985).

Rust, P.C. (1992). The politics of sexual identity: Sexual attraction and behavior among lesbian and bisexual women. Social Problem, 39 (4), 366- 386.

Rust, P.C. (1993). Neutralizing the political threat of the marginal women: Lesbian beliefs about bisexual women. Journal of Sex Research, 30 (3), 214- 228.

Rust, P.C. (1995). Bisexuality and the challenge to lesbian politics: Sex, loyalty and revolution. New York: University Press.

227 Rust, P.C. (1996). Monogamy and polyamory: Relationship issues for bisexuals. In B.A. Firestein (Ed.). Bisexuality: The psychology and politics of an invisible minority (pp. 127-148). Thousand Oaks, Cal: Sage.

Rust, P.C. (2000). Bisexuality in the United States: A social science reader. New York: Columbia University.

Rust, P.C. (2001). Two many and not enough: The meanings of bisexual identities. Journal of Bisexuality, 1 (1), 31-68.

Ryan, B., Brotman, S. & Rowe, B. (2000). Access to care: Exploring the health and well-being of gay, lesbian, bisexual and tow-spirit people in Canada. Montreal: McGill Centre for Applied Family Studies.

Sakalli, N. (2002). The effects of social contact with a lesbian person on the attitude change toward homosexuality in Turkey. Journal of Homosexuality, 44, 111-119.

Saghir, M.T. & Robbins, E. (1973). Male and female homosexuality: A comprehensive investigation. Baltimore: Williams and Wilkins.

Saliba, P. (1982). Research project on sexual orientation. Bi-monthly Newsletter of the Bisexual Center of San Francisco, 6 (5), 3-6.

Sanders, S.A., Reinisch,J.M. & McWhirter, D.P. (1990). Homosexuality/Heterosexuality: An overview. In D.P. McWhirter, S.S. Sanders and J.M. Reinisch (Eds.). Homosexuality/Heterosexuality: Concepts of sexual orientation (pp. 106-134). New York: Oxford University.

Santos, J.R.A. (1999). Cronbach’s alpha: A tool for assessing the reliability of scales. Journal of Extension, 37, (2), 30-35.

Savin-Williams, R.C. (2001). A critique of research on sexual-minority youths. Journal of Adolescence, 24, 5-13.

228 Sawyer, M. (2000). Talking dirty after dark: Themes of sexuality in black ideology. Paper presented at the annual meeting of the Western Political Association, March, San Jose, CA.

Schafer, S. (1976). Sexual and social problems of lesbians. Journal of Sex Research, 12, 50-69.

Schmitt, M.T., Branscombe, N.R., Kobrynowicz, D. & Owen, S. (2002). Perceiving discrimination against one’s gender group has different implications for well-being in women and men. Personality and Social Psychology Bulletin, 28, 197-210.

Schneider, W. & Lewis, I.A. (1984). The straight story on homosexuality and gay rights. Public Opinion, 7, 16-29.

Schober, S., Fe Caces, M., Pergamit, M. & Branden, L. (1992). Effects of mode of administration on reporting of drug use in the National Longitudinal Survey. In C.F. Turner, J.T. Lessler and J.C. Gfroerer (Eds.). Survey measurement of drug use: Methodological studies. Rockville, MD: National Institute on Drug Use.

Schofield, M. (1965). Sociological aspects of homosexuality: A comparative study of three types of homosexuals. Boston: Little Brown.

Schulte, L.J. & Battle, J. (2003). The relative importance of ethnicity and religion in prediction attitudes towards gay men and lesbians. Journal of Homosexuality, 50, 69-76.

Schwartz, P. & Blumstein, P. (1998). The acquisition of sexual identity: Bisexuality. In E.J. Haeberle and R. Grindorf (Eds.). Bisexualities: The ideology and practice of sexual contact with both men and women (pp. 182- 212). New York: Continuum.

229 Sell, R.L. (1997). Defining and measuring sexual orientation. Archives of Sexual Behavior, 26 (6), 643-659.

Shively, M.G. & De Cecco, J.P. (1977). Components of sexual identity. Journal of Homosexuality, 3 (1), 41-48.

Shokeid, M. (2001). Book review of bisexuality: A critical reader. New York: Routledge.

Shuster, R. (1991). Beyond defense: Considering next steps for bisexual liberation. In L. Hutchins and L. Ka’ahumanu (Eds.). Bi any other name: Bisexual people speak out (pp. 266-274). Boston: Alyson.

Sidanius, J., Levin, S., Rabinowitz, J.L. & Frederico, C.M. (1999). Peering into the jaws of the beast: The integrative dynamics of social identity, symbolic racism and social dominance. In D.A. Prentice and D.T. Miller (Eds.). Cultural divivdes: Understanding and overcoming group conflict (pp. 80-132). New York: Russel Sage.

Sinclair, R. (2005). The official treatment of white, South African, homosexual men and the consequent reaction of gay liberation from the 1960 to 2000. Unpublished doctoral thesis: University of Johannesburg.

Skeen, E. (2007). The rape trial: Jacob Zuma, AIDS, conspiracy, and tribalism in Neo-liberated Post-Apartheid South Africa. Unpublished thesis: Princeton University.

Smiley, E.B. (1997). Counseling bisexual clients. Journal of Mental Health Counseling, 19 (4), 373-382.

Smith, A. (1997). Cultural diversity and the coming out process: Implications for clinical practice. In B. Greene (Ed.). Ethnic and cultural diversity among lesbians and gay men (pp. 279-300). Thousand Oaks, CA: Sage.

230 Smith, E.M., Johnson, S.R. & Guenther, S.M. (1985). Health care attitudes and experiences during gynecological care among lesbians and bisexuals. American Journal of Public Health, 75 (9), 1085-1087.

Socarides, C.W. (1978). Homosexuality. New York: Jason Aronson. (Original work published 1953).

Spalding, L.R. & Peplaum, L.A. (1997). The unfaithful lover: Heterosexuals’ perceptions of bisexual and their relationships. Psychology of Women Quarterly, 21, 611-625.

Spira, A., Bajos, N., Bejin, A. & Beltzer, N. (1992, December). AIDS and sexual behavior in France. Nature, 3, 407-409.

Steffens, M.C. & Wagner, C. (2004). Attitudes toward lesbians, gay men, bisexual women and bisexual men in Germany. The Journal of Sex Research, 41 (2), 137-152.

Steinman, E. (2001). Interpreting the invisibility of male bisexuality: Theories, interactions, politics. Journal of Bisexuality, 1 (1/2), 15-45.

Stekel, W. (1946). Bi-sexual love. New York: Emerson. (Original work published 1922).

Stern, J. (1961). The sixth man. New York: Doubleday.

Stevens, R.A. (2004). Understanding gay identity development within the college environment. Journal of College Student Development, 45, 185-206.

Stobie, C. (2004). “Biopia” in biography: Representations of bisexuality in selected South African lives. Scrunity, 9 (2), 35-52.

231 Stokes, J.P., Damon, W. & McKirnan, D.J. (1997). Predictors of movement toward homosexuality: A longitudinal study of bisexual men. Journal of Sex Research, 34 (3), 304-312.

Stokes, J.P., McKirman, D.J. & Burzette, R.G. (1993). Sexual behavior and condom use, disclosure of sexuality and stability of sexual orientation in bisexual men. Journal of Sex Research, 30 (3), 203-213.

Stokes, J.P., Miller, R.L. & Mundhenk, R. (1998). Toward an understanding of behaviorally bisexual men: The influence of context and culture. The Canadian Journal of Human Sexuality, 7 (2), 118-133.

Stokes, J.P., Taywaditep, K., Vanable, P. & McKirnan, D.J. (1996). Bisexual men, sexual behavior, and HIV/AIDS. In B. Firestein (Ed.). Bisexuality: The psychology and politics of an invisible minority (pp. 149-168). Thousand Oaks, Cal: Sage.

Stone, M.H. (2000). Psychopathology: Biological and psychological correlates. Journal of American Academy of Psychoanalysis, 28, 203-235.

Stonequist, E.B. (1961). The marginal man: A study in personality and culture conflict. New York: Russel and Russel. (Original published in 1937).

Strassberg, D.S. & Lowe, K. (1998). Volunteer bias in sexuality research. Archives of Sexual Behavior, 24, 369-382.

Sulloway, F.J. (1979). Freud, biologist of the mind: Beyond the psychoanalytic legend. New York: Basic Books.

Sumpter, S.F. (1991). Myths/realities of bisexuality. In L. Hutchins and L. Ka’ahumanu (Eds.). Bi any other name: Bisexual people speak out (pp. 12- 13). Boston: Alyson Publications.

232 Suppe, J.B. (1984). In defense of a multidimensional approach to sexual identity. Journal of Homosexuality, 10 (3), 7-14.

Susskind, J.E. Hodges, C.H. (2007). Decoupling children’s gender-based in- group positivity from out-group negativity. Sex Roles, 56, 707-716.

Szasz, T.S. (1965). Legal and moral aspects of homosexuality. In J. Marmor (Ed.). Sexual Inversion: The multiple roots of homosexuality (pp. 124-139). New York: Basic Books.

Szasz, T.S. (1970). The manufacturer of madness: A comparative study of the inquisition and the mental health movement. New York: Dell.

Szymanski, D.M. (2004). Relations among dimensions of feminism and internalized heterosexism and bisexual women. Sex Roles, 51, 145-178.

Tabachnick, B.G. & Fidell, L.S. (2001). Using multivariate statistics. New York: Harper and Row.

Tajfel, H. (1959). Quantitative judgement in social perception. British Journal of Psychology, 50, 16-29.

Tajfel, H. (1974). Social identity and intergroup behavior. Social Science Information, 13, 65-93.

Tajfel, H. (1978). Differentiation between social groups. London: Academic Press.

Tajfel, H. (1982). Social psychology of intergroup relations. Annual Review of Psychology, 33, 1-30.

Tajfel, H. & Turner J.C. (1979). An integrative identity theory of intergroup behavior. In S. Worchel and W.G. Austin (Eds.). The Psychology of Intergroup Relations (pp. 7-24). Chicago: Nelson-Hall.

233 Tajfel, H. & Turner,, J. (1986). The social identity theory of intergroup behavior. In S. Worchel & W. G. Austin (Eds.). Psychology of intergroup relations (pp. 7-24). Chicago: Nelson-Hall.

Taywaditep, K.J. & Stokes, J.P. (1998). Male bisexualities: A cluster analysis of men with bisexual experience. Journal of Psychology and Human Sexuality, 10 (1), 15-41.

Testa, R.J., Kinder, B.N. & Ironson, G. (1987). Heterosexual bias in the perception of loving relationships of gay males and lesbians. Journal of Sex Research, 23, 163-172.

Thoreson, R.R. (2008). Somewhere over the rainbow nation: Gay, lesbian and bisexual activism in South Africa. Journal of Southern African Studies, 34, 679-697.

Tissot, S.A. (1965). A treatise on the diseased produced by orgasm. In C. Rosenberg and C. Smith-Rosenburg (Eds.). The secret vice exposed. Some arguments against masturbation. New York: Arno Press. (Original work published in 1832).

Tourangeau, R. & Smith, T.W. (1996). Asking sensitive questions: The impact of data collection mode, question format, and question context. Public Opinion Quarterly, 60, 275-304.

Tourangeau, R., Rasinski, K., Jobe, J.B., Smith, T.W. & Pratt, W.F. (1997). Sources of error in a survey on sexual behavior. Journal of Official Statistics, 13, 341-365.

Traeen, B., Stigum, H & Sorensen, D. (2002). Sexual diversity in urban Norwegians. The Journal of Sex Research, 39 (4), 249-259.

Treadwell, M.E. (1954). The women’s army corps. Washington, D.C. (Cited by D’Emilio, 1983).

234 Trnka, S. (1992). A pretty good bisexual kiss there. In E.R. Weise (Ed.). Closer to home: Bisexuality and feminism (pp. 103-113). Seattle: Seal.

Troiden, R.R. (1988). Gay and lesbian identity: A sociological analysis. New York: General Hill.

Turner, J.C. (1975). Social comparison and social identity: Some prospects for intergroup behavior. European Journal of Social Psychology, 5, 5-34.

Turner, J.C. (1982). Towards a cognitive redefinition of the group. In H. Tajfel (Ed.). Social identity and intergroup relations. New York: Cambridge Press.

Turner, J.C. (1978). Social categorization and social discrimination in the minimal group paradigm. In H. Tajfel (Ed.). Differentiation between social groups (pp. 101-140). London: Academic Press.

Turner, J.C. (1999). Some current issues in research on social identity and self-categorization theories. In N. Ellemers, R. Spears and B. Doosje (Eds.). Social Identity: Context, commitment, content (pp. 6-34). Oxford: Blackwell.

Turner, J.C., Hogg, M.A., Oakes, P.J., Reicher, S.D. & Wetherell, M.S. (1987). Rediscovering the social group: A self-categorization theory. Oxford: Blackwell.

Turner, C.F., Lessler, J.T. & Gfroerer, J.C. (Eds.). (1992). Survey measurement of drug use: Methodological studies. Washington, DC: U.S. Government Printing Office.

Turner, C.F., Miller, H.G. & Moses, L.E. (1989). Sexual behavior and AIDS. In C.F. Turner, H.G. Miller and L.E. Moses (Eds.). AIDS, sexual behavior and intravenous drug use (pp. 73-185). Washington, D.C: National Academy.

235 Twining, A. (1983). Bisexual women: Identity in adult development (Doctoral Dissertation, Boston University School of Education, 1983). Dissertation Abstracts International, 44 (5), 1340A

Udis-Kessler, A. (1991). Bisexuality and an essentialist world: Toward an understanding of Biphobia. In T. Geller (Ed.). Bisexuality: A reader and sourcebook (pp. 36-39). Ojai, Cal: Times Change.

Udis-Kessler, A. (1996). Challenging the stereotypes. In Off Pink Collective (Eds.). Bisexual horizons: Politics, histories, lives (pp. 45-57). London: Lawrence and Wishart.

Ulrich, K.H. (1864). Inclusa (Cited by Haeberle 1998).

Van de Vijver, F.J.R. (2003). Bias and equivalence: Cross-cultural perspectives. In J.A. Harkness, F.J. Van der Vijver, and P.P. Mohler (Eds.). Cross-cultural survey methods (pp. 143-155). New York: Wiley.

Vealey, R.S. (1997). Transforming the silence on lesbians in sport. Women in Sport and Physical Activity Journal, 6, 165-177.

Vidal-Ortiz, S. (2001). A world of its own – not the best of both worlds. Journal of Sex Research, 38 (3), 271-274.

Vining, D. (1979). A gay diary, 1933-1946. New York: Plenum.

Visano, L.A. (1990). The impact of age on paid sexual encounters. Journal of Homosexuality, 20 (2/3), 207-226.

Voeler, B. (1990). Some uses and abuses of the Kinsey Scale. In D.P. McWhirter, S.A. Sanders and J.M. Renisch (Eds.). Homosexuality/ Heterosexuality: Concepts of sexual orientation. New York: Oxford University.

236 Waldo, D.R. (1998). Out on campus: Sexual orientation and academic climate in a university context. American Journal of Community Psychology, 26, 745- 774.

Ward, D.A. & Kassebaum, G.G. (1965). Homosexuality: A mode of adaptation in a prison for women. Social Problems, 12 (2), 159-177.

War Department (1943). Sex hygiene course: Officers and officer candidates, A women’s army auxiliary corps. Pamphlet no 35. Washington, D.C. (Cited by D’Emilio, 1983).

Warren, C.A. (1974). Identity and community in the gay world. New York: Wiley.

Weber, G.N. (2008). Using to numb the pain: Substance use and abuse among lesbian, gay and bisexual individuals. Journal of Mental Health Counselling, 30, 31-49.

Weinberg, I. (1967). The English public schools: The sociology of elite education. New York: Atherton.

Weinberg, M.S. (1972). Society and the healthy homosexual. New York: St. Martin’s.

Weinberg, M.S. & Williams, C.J. (1974). Male homosexuals: Their problems and adaptations. New York: Oxford University.

Weinberg, M.S., Williams, C.J. & Pryor, D.W. (1994). Dual attraction: Understanding bisexuality. New York: Oxford University.

Weise, E.R. (1992). (Ed.). Closer to home: Bisexuality and feminism. Seattle, WA: Seal.

237 Wellings, K., Wadsworth, J. & Johnson, A. (1994). Sexual diversity and homosexual behavior. In A.M. Johnson, J. Wadsworth, K. Wellings and J. Field (Eds.). Sexual attitudes and lifestyles (pp. 183-224). Oxford: Blackwell.

Whitley, B.E. (2001). Gender-role variables and attitudes toward homosexuality. Sex Roles, 45, 691-701.

Wiederman, M.W. (1999). Volunteer bias in sexuality research using college participants. Journal of Sex Research, 36, 59-66.

Williams, M.J. (1999). Sexual pathways: Adapting to dual sexual attraction. Westport, Conn: Praeger.

Williams, S. (2002, March 29). Students hope their silence will make a statement for gay rights. The Milwaukee Journal Sentinel, p. 1.

Wittstock, M. (1990). The best of both worlds and still nothing: Bisexuals come out to talk. In T. Geller (Ed.). Bisexuality: A reader and sourcebook (pp. 26-33). California: Times Change.

Wolf, T.J. (1992). Bisexuality: A counseling perspective. In S.H. Dworkin and F.J. Gutierrez (Eds.). Counseling gay men and lesbians: Journey to the end of the rainbow (pp. 175-187). Alexandria, VA: American Association for Counseling and Development.

Wolff, C. (1977). Bisexuality: A study. London: Quartet.

Wood, R.W., Krueger, L.E., Pearlman, T. & Goldbaum, G. (1993). HIV transmission: Women’s risk from bisexual men. American Journal of Public Health, 83, 1757-1759.

Wooden, W.S. & Parker, J. (1982). Men behind bars: Sexual exploitation in prison. New York: Plenum.

238 Worth, H. (2003). The myth of the bisexual infector? HIV risk and men who have sex with men. Journal of Bisexuality, 3 (2), 69-88.

Worthington, R.L. & Whittaker, T.A. (2006). Scale Development Research: A content analysis and recommendations for best practices. The Counseling Psychologist, 34 (6), 806-838.

Yoshino, K. (2000). The epistemic contract of bisexual erasure. Stanford Law Review, 52, 353-459.

Young, S. (1992). Dichotomies and displacement: Bisexuality in queer theory and politics. In S. Phelan (Ed.). Playing with Fire: Queer Politics, Queer Theories (pp. 51-74). New York: Routledge.

Zea, M.C., Reisen, C.A. & Diaz, R.M. (2003). Methodological issues in research on sexual behavior with Latino gay and bisexual men. American Journal of Community Psychology, 31, 281-291.

Zinik, G.A. (1984). The relationship between sexual orientation and eroticism, cognitive flexibility and negative affect (Doctoral dissertation, University of California, 1984). Dissertation Abstracts International, 45 (8), 2707B.

Zinik, G.A. (1985). Identity conflict or adaptive flexibility? Bisexuality reconsidered. Journal of Homosexuality, 11 (1/2), 7-19.

Zipkin, D. (1992). Why bi? In E.R. Weise (Ed.). Closer to home: Bisexuality and feminism (pp.55-73). Seattle, Wash: Seal.

239

February 2006 -February 2007

Dear Participant,

Thank you very much for participating in this research project – we believe that the results will help us understand information relating to bisexuality. Please be assured that the results will be kept confidential and be used for academic purposes only. Please do NOT put your name on the survey. Please be honest with your responses.

Warm regards

Marlene Arndt M.A. (Psych)

For any information regarding this survey, please contact Dr. K. de Bruin at the Psychology Department at the University of Johannesburg: (011) 559-3131.

SECTION A: BIOGRAPHICAL QUESTIONNAIRE Please mark the appropriate block with an ‘X’

1. Age (in years):

2. Gender: 1 Man 2 Woman

3. In which faculty are you? 1 Education 2 Law 3 Engineering

4 Humanities 5 Science 6 Art, design, architecture

7 Management 8 Health 9 Economics and finance

1 2 3 4 4. Ethnic Group: Black White Coloured Asian

5. Do you know a bisexual man or woman? 1 Yes 2 No

6. Do you know a lesbian or gay man? 1 Yes 2 No

7. How religious are you? 1 Not at all religious 2 Moderately religious 3 Deeply religious

SCALE 1

Please read each of the following statements and rate them according to how accurately they describe your attitudes and beliefs. Please respond honestly and answer every question according to the rating scale below.

“Bisexual men and women are sexually attracted to or have sexual contacts with both men and women”

1 ------2 ------3 ------4 ------5 Strongly Strongly Disagree Agree

1. Men who claim to be bisexual are in denial about their true sexual orientation. 2. The growing acceptance of female bisexuality indicates a decline in South African values. 3. Women who call themselves bisexual are temporarily experimenting with their sexuality. 4. Bisexual men are sick. 5. Male bisexuals are afraid to commit to one lifestyle. 6. Bisexual women have a clear sense of their true sexual orientation. 7. I would not be upset if my sister were bisexual. 8. Bisexual women are confused about their sexual orientation. 9. Bisexual men should not be allowed to teach children in schools. 10. Female bisexuality is harmful to society because it breaks down the natural divisions between sexes. 11. Male bisexuality is not a phase, but a stable sexual orientation. 12. Male bisexuals have a fear of committed intimate relationships. 13. Bisexuality in men is immoral. 14. The only true sexual orientations for women are homosexuality and heterosexuality. 15. Female bisexuality is unnatural. 16. Bisexuality is a stable sexual orientation for men. 17. Male bisexuality is not a perversion. 18. Women who identify as bisexual have not yet discovered their actual sexual orientation.

SCALE 2

Carefully read each of the categories and descriptions listed below and mark the appropriate box with an “X”.

In terms of my sexual orientation, I identify myself as:

CHOOSE ONE OF THE FOLLOWING:

Category Description Mark one box with an “X” 1. Exclusively I have always considered myself strictly heterosexual. heterosexual 2. Heterosexual with I currently consider myself heterosexual, but occasionally I some homosexuality am attracted to, desire, or have sexual contact with members of my same sex. 3. Concurrent I consider myself bisexual because I have equal attraction bisexual towards, desires for, or sexual contact with members of my same and opposite sex on a fairly regular basis. 4. Sequential I consider myself bisexual because I am equally attracted bisexual to, desire, or have sexual contacts with members of both my same and opposite sex. I tend to feel attracted to, desire, or have sexual contact with members of one sex for a period of time and then switch and feel attracted to, desire, or have sexual contact with members of the other sex for a period of time. 5. Homosexual I currently consider myself homosexual, but occasionally I with some am attracted to, desire, or have sexual contact with heterosexuality members of the opposite sex. 6. Past heterosexual, In the past, I considered myself heterosexual, but now I current homosexual consider myself strictly homosexual. 7. Exclusively I have always considered myself strictly homosexual. homosexual 8. Past homosexual, In the past, I considered myself homosexual, but now I current heterosexual consider myself strictly heterosexual. 9. Asexual I do not consider myself homosexual, heterosexual or bisexual. I have never had any attraction to, desire for, or sexual contact with members of either my same or opposite sex

February 2008 Dear Participant,

Thank you very much for participating in this research project – we believe that the results will help us understand information relating to bisexuality. Please be assured that the results will be kept confidential and be used for academic purposes only. Please do NOT put your name on the survey. Please be honest with your responses.

Warm regards

Marlene Arndt M.A. (Psych)

For any information regarding this survey, please contact Dr. K. de Bruin at the Psychology Department at the University of Johannesburg: (011) 559-3131.

SECTION A: BIOGRAPHICAL QUESTIONNAIRE Please mark the appropriate block with an ‘X’

1. Age (in years):

2. Gender: 1 Man 2 Woman

3. In which faculty are you? 1 Education 2 Law 3 Engineering

4 Humanities 5 Science 6 Art, design, architecture 7 Management 8 Health 9 Economics and finance

4. Ethnic Group: 1 Black 2 White 3 Coloured 4 Asian

5. Do you know a bisexual man or woman? 1 Yes 2 No

6. Do you know a lesbian or gay man? 1 Yes 2 No

7. How religious are you? 1 Not at all religious 2 Moderately religious 3 Deeply religious

8. WHICH ONE OF THE FOLLOWING DESCRIBES YOUR SEXUAL ORIENTATION? DESCRIPTION Mark one box with an “X” 1. I am only attracted to, desire, or have sexual contact with members of the opposite sex. 2. I am only attracted to, desire, or have sexual contact with members of the same sex. 3. I am attracted to, desire, or have sexual contact with both men and women. 4. I have never had any attraction to, desire for, or sexual contact with members of my same or opposite sex.

Mary is a 27-year-old woman. Mary was in a long-term relationship with Peter. Currently, she is in a relationship with Susan. Mary has been sexually attracted, and had sexual contact with men and women in the past.

Read the following statements and mark the appropriate block with an “X” 1. Strongly disagree 2. Disagree somewhat 3. Unsure 4. Agree somewhat 5. Strongly agree 1. Mary has no values. 2. Mary’s sexual behavior is

unnatural. 3. Mary will always be unfaithful. 4. Mary is temporarily experimenting

with her sexuality. 5. Mary is sick. 6. Mary’s sexual behavior places her

partners at risk for HIV. 7. Mary is just going through a

phase. 8. Mary has no morals. 9. Mary is afraid to admit that she is

a lesbian. 10. Mary is confused about her

sexuality. 11. Mary is afraid to commit herself

to intimate relationships. 12. Mary is obsessed with sex. 13. Mary has not yet discovered her

true sexual orientation. 14. Mary is a pervert. 15. Mary does not want to commit to

one lifestyle. 16. Mary denies her true sexual

orientation. 17. I will be upset if Mary was my

sister. 18. Mary should not be allowed near

children. 19. Mary’s behavior is not

acceptable. 20. Mary should be punished.

21. Mary can not be trusted.

22. Mary is not sure about her sexual

orientation 23. One day Mary will admit her true

sexual orientation.

John is a 24 year old man. John was in a long term relationship with Nancy. Currently, he is in a relationship with David. John has been sexually attracted, and had sexual contact with men and women in the past.

Read the following statements and mark the appropriate block with an “X” 1. Strongly disagree 2. Disagree somewhat 3. Unsure 4. Agree somewhat 5. Strongly agree 1. John has no morals.

2. John will always be unfaithful.

3. John is confused about his

sexuality. 4. John is afraid to admit that he is

gay.

5. John is obsessed with sex.

6. John is just going through a

phase. 7. John is afraid to commit himself to

intimate relationships. 8. John is temporarily experimenting

with his sexuality. 9. John’s behavior places his

partners at risk for HIV.

10. John is a pervert.

11. John’s sexual behavior is

unnatural. 12. John denies his true sexual

orientation. 13. John does not want to commit to

one lifestyle.

14. John has no values.

15. John is sick.

16. I will be upset if John was my

brother. 17. John should not be allowed near

children. 18. John’s behavior is not

acceptable.

19. John should be punished.

20. John can not be trusted.

21. John is afraid to commit to one

lifestyle. 22. John is not sure about his sexual

orientation. 23. One day John will admit his true

sexual orientation.