A new approach to the study of sexual violence: development and psychometric properties of a preliminary multi-item research instrument assessing rape-resistant attitudes

Sexual violence is a significant problem in the United States. Recent social and political events, such as the #MeToo movement, have brought discussions about sexual violence, victim-blaming, and predatory behavior to the fore of public dialogue in important ways. Understanding, and thus eliminating, sexual violence is a concern of most social science disciplines, with much research set within the conceptual frameworks of rape myth acceptance and rape-supportive attitudes. This important work has informed study of sexual violence to a great degree, but there has been a call in the literature to consider rape-resistant attitudes to gain a more nuanced understanding of sexual violence. Psychometric tools are commonly used to assess unobservable attitudes, and the psychometric properties of these tools are important indicators of reliability and validity. This mixed-methods project takes seriously the call to identify rape-resistant attitudes and develop instruments to test them. A search of relevant literature and interviews with experts in the field informed the conceptualization of an attitudinal dimension that may be considered rape-resistant. These attitudes indicate not only low rape proclivity but also reluctance to participate in victimblaming. A psychometric instrument containing statements intended to measure raperesistant attitudes was developed and tested in both a pilot study and full sample of college students from the University of Utah. Statistical analyses through principal component factor analysis, structural equation modeling, and multivariate regression confirmed that the instrument measuring rape-resistant attitudes was indeed both reliable and valid with data from this sample, and is ready to be tested in other populations. In addition to the statistically and conceptually refined 25-item instrument, the Abbreviated Rape-Resistant Attitudes (ARRA) scale, analysis shows that a set of 10 attitudinal dimensions informs a broader index of rape-resistant attitudes. Sixty items tested in this sample are appropriate for domain sampling to answer specific research questions related to rape resistance and sexual violence. The process of developing the ARRA is detailed here. Psychometric properties are discussed, and rape resistance as a conceptual framework is considered. Usage notes for the ARRA and/or domain sampling from the index of tested items are offered.

A NEW APPROACH TO THE STUDY OF SEXUAL VIOLENCE: DEVELOPMENT AND PSYCHOMETRIC PROPERTIES OF A PRELIMINARY MULTI-ITEM RESEARCH INSTRUMENT ASSESSING RAPE-RESISTANT ATTITUDES by Colleen Marie James A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Sociology The University of Utah August 2019 Copyright © Colleen Marie James 2019 All Rights Reserved The University of Utah Graduate School STATEMENT OF DISSERTATION APPROVAL The dissertation of Colleen Marie James has been approved by the following supervisory committee members: Claudia Geist , Chair May 29th 2019 Date Approved Kim Korinek , Member May 29th 2019 Date Approved Marie Sarita Gaytan , Member Date Approved Heather C. Melton , Member May 29th 2019 Date Approved Wanda S. Pillow , Member May 29th 2019 Date Approved and by the Department of Ming Wen , Chair of Sociology and by David B. Kieda, Dean of The Graduate School. ABSTRACT Sexual violence is a significant problem in the United States. Recent social and political events, such as the #MeToo movement, have brought discussions about sexual violence, victim-blaming, and predatory behavior to the fore of public dialogue in important ways. Understanding, and thus eliminating, sexual violence is a concern of most social science disciplines, with much research set within the conceptual frameworks of rape myth acceptance and rape-supportive attitudes. This important work has informed study of sexual violence to a great degree, but there has been a call in the literature to consider rape-resistant attitudes to gain a more nuanced understanding of sexual violence. Psychometric tools are commonly used to assess unobservable attitudes, and the psychometric properties of these tools are important indicators of reliability and validity. This mixed-methods project takes seriously the call to identify rape-resistant attitudes and develop instruments to test them. A search of relevant literature and interviews with experts in the field informed the conceptualization of an attitudinal dimension that may be considered rape-resistant. These attitudes indicate not only low rape proclivity but also reluctance to participate in victimblaming. A psychometric instrument containing statements intended to measure raperesistant attitudes was developed and tested in both a pilot study and full sample of college students from the University of Utah. Statistical analyses through principal component factor analysis, structural equation modeling, and multivariate regression confirmed that the instrument measuring rape-resistant attitudes was indeed both reliable and valid with data from this sample, and is ready to be tested in other populations. In addition to the statistically and conceptually refined 25-item instrument, the Abbreviated Rape-Resistant Attitudes (ARRA) scale, analysis shows that a set of 10 attitudinal dimensions informs a broader index of rape-resistant attitudes. Sixty items tested in this sample are appropriate for domain sampling to answer specific research questions related to rape resistance and sexual violence. The process of developing the ARRA is detailed here. Psychometric properties are discussed, and rape resistance as a conceptual framework is considered. Usage notes for the ARRA and/or domain sampling from the index of tested items are offered. iv This work is dedicated to my cherished husband, Kelly, and our fantastic sons, Brock and Josh, whose unwavering faith and endless support made this possible. TABLE OF CONTENTS ABSTRACT .......................................................................................................................................... iii LIST OF FIGURES .......................................................................................................................... viii ACKNOWLEDGMENTS ................................................................................................................ ix Chapters 1 INTRODUCTION........................................................................................................................... 1 Cultural Context.......................................................................................................................... 5 Interdisciplinary Study ............................................................................................................... 8 Theories of Sexual Violence ................................................................................................... 10 The Function of Rape-Myth Acceptance and Rape-Supportive Attitudes ..................... 19 Rape Myth Acceptance ............................................................................................................ 20 Rape-Supportive Attitudes ...................................................................................................... 26 Literature-Informed Assumptions ......................................................................................... 35 2 METHODS ...................................................................................................................................... 39 Theory ........................................................................................................................................ 40 Nine Steps From Conceptualization to Statistical Analyses .............................................. 42 Summary .................................................................................................................................... 60 3 QUALITATIVE RESULTS.......................................................................................................... 62 Content Analysis ....................................................................................................................... 62 Semistructured Interviews ....................................................................................................... 74 Focus Group ............................................................................................................................. 83 Pilot Study .................................................................................................................................. 84 Summary .................................................................................................................................... 84 4 QUANTITATIVE RESULTS ...................................................................................................... 86 Description of Participants ..................................................................................................... 86 Ten Dimensions Underlying Rape Resistance ..................................................................... 88 Creation of the Abbreviated Rape-Resistant Attitudes (ARRA) Scale ...........................100 Summary ..................................................................................................................................106 5 REGRESSION ANALYSES ......................................................................................................141 Mean Differences ...................................................................................................................142 Multivariate Regression..........................................................................................................143 Correlations With Established Scales ..................................................................................144 6 CONCLUSIONS ..........................................................................................................................148 Usage Notes.............................................................................................................................148 Rape Resistance as an Ideal Type.........................................................................................150 Limitations, Strengths, Future Research, and Applied Possibilities ...............................154 Reflexive Assessment. ............................................................................................................163 Conclusion ...............................................................................................................................167 Appendices A: COMPLETE LIST OF RAPE-RESISTANT ATTITUDES ..............................................168 B: FULL INDEX OF ITEMS FOR DOMAIN SAMPLING .................................................175 C: SURVEY QUESTIONNAIRE ................................................................................................178 REFERENCES.................................................................................................................................194 vii LIST OF FIGURES Figures 4.1 Communication SEM ................................................................................................................109 4.2 Understanding Patriarchy SEM ................................................................................................111 4.3 Understanding Privilege and Entitlement SEM ....................................................................114 4.4 Sex Positivity SEM .....................................................................................................................116 4.5 Understanding Perpetrators and Survivors SEM ..................................................................118 4.6 Self-Governance SEM ...............................................................................................................120 4.7 Consent SEM ..............................................................................................................................122 4.8 Knowledge About Sexual Violence SEM ...............................................................................124 4.9 Personal Values – Sex SEM ......................................................................................................127 4.10 Personal Values – General SEM ............................................................................................129 4.11 Reluctance to Blame Victims SEM........................................................................................132 4.12 Self-Monitoring for Perpetration SEM .................................................................................135 4.13 Abbreviated Rape-Resistant Attitudes SEM ........................................................................139 ACKNOWLEDGMENTS I would like to express my gratitude to my committee. To Dr. Korinek, who approached me while I was an undergraduate student in her Gender and Sexuality class and asked what I planned to do after graduation, then gently steered me along the path to this PhD program. Kim, I would never have started on this path without you, and I wish grad school had all been like my first semester as your TA. To Dr. Melton, who taught me how to bravely speak about sexual violence in academic settings without wavering. Heather, I learned this was possible from you. To Dr. Gaytán, who always asked, “Are you okay?” after any change or delay (and there have been several). Sarita, I always knew you cared about me, not just my work and progress. To Dr. Pillow, whose influence has changed me as both a scholar and a person. Pillow, I feel incredibly lucky you misremembered which class I was in and serendipitously put me on a different, but exactly right, path. I am transformed because of you. To Dr. Geist… how do you thank someone when even $10 words do not suffice? Claudia, as my mentor, cheerleader, and occasional taskmaster, you have pushed me hard and hugged me hard, neither because you particularly wanted to but because both were what I needed. You scared me to death when I needed that, and you lifted me up when I needed that. I know I was unruly at times, and I cannot sufficiently express my gratitude that you met me where I was and pushed me where I needed to be. Thank you for helping me believe my data, learn fuzzy statistical witchcraft, and for being patient with my self-doubt and personal upheavals. I have been in awe of you for years, and I have learned more from you than I can describe. Now for the interpretive dance! CHAPTER 1 INTRODUCTION Sexual violence is a significant social problem in the United States. The search for understanding, and thus prevention, has inspired decades of research. This research utilizes various frameworks and theories, all of which evolve over time to address weaknesses in previous methods and analyses. The sociological imagination asks us to consider the social context in which an individual functions. Thus, broad themes relating to both societal- and individual-level contributors to sexual violence have been identified, measured, and their correlates studied extensively. The United States has been identified as a “rape culture,” or a culture that expects, misunderstands, and excuses sexual violence. Individuals within a rape culture learn attitudes and beliefs about sexual violence that excuse and exonerate perpetrators, minimize their crimes and the effects, and discredit survivors. Scholars have identified these as rape-supportive attitudes (RSA), and holding these attitudes has been shown to contribute to individual rape proclivity, victim blaming, and the perpetuation of rape culture. Acceptance of “false beliefs about rape, rape victims, and rapists” (Burt, 1980), known as rape myth acceptance (RMA), has been linked to many problematic outcomes that interfere with survivor healing and the pursuit of criminal justice. Studying RSA and RMA has led to a great deal of understanding about the associations between various social contexts and dominance orientations that are conducive to sexual violence. Despite the breadth and quality of work that has been done using RSA and RMA as 2 conceptual frameworks, opportunities for better understanding remain. Just as RSA have been specifically identified and measured, it is possible that more rape-resistant attitudes (RRA) may be present in individuals who live within a rape culture. In every study of sexual violence in these frameworks, some individuals in each sample held RSA and RMA more lightly than others in the sample, regardless of their positionality within the same or similar social contexts. With no psychometric assessment tool to measure attitudes related to this inverse phenomenon, there has been a dearth of understanding of what attitudes may exist alongside RSA and RMA that encourage resistance rather than acceptance of dominant rapesupportive narratives supplied by a rape culture. Consequently, we are limited to what we can infer from measuring RSA, both with regard to rape proclivity and propensity to blame victims. Nonnormative experiences of sexual violence have also been neglected through the heteronormativity of most research regarding sexual violence. Both RSA and RMA principally position victims of sexual violence as disempowered women who are preyed upon by aggressive men, and fail to sufficiently consider male victims, female perpetrators, or individuals who experience or perpetrate same-sex sexual violence. Studying sexual violence in this manner limits knowledge production. It restrains our conceptualization, and thus our understanding, of perpetrators and survivors, as well as the general social climate toward survivors who do not fit the assumed gender configuration as they seek social and criminal justice. Narratives of seemingly inevitable oppression of fragile women by hegemonic men also fail to recognize survivors as diverse, empowered individuals of all genders who often experience posttraumatic growth. With no work looking specifically at RRA and no tools to measure them, it is challenging to consider potential rape-resistant attitudes in conjunction with other contexts 3 that resist dominant rape-supportive narratives. Social scientists do not know what individual, social, familial, economic, political, or demographic factors may contribute to that resistance, or how RRA may impact understanding of sexual violence, self-reported rape proclivity, tendency toward victim blaming, or empathy for survivors. We do not know if and how holding RRA would influence survivor’s healing trajectories, or if and how holding RRA would impact the pursuit of social and criminal justice. Relatedly, we know that holding low RMA is associated with better outcomes for survivors, but we do not understand what other attitudes may contribute to the positive healing many survivors experience post assault. The need to study these phenomena is evident, and having new psychometric tools promises to further expand knowledge to the benefit of all. For these reasons, there has also been a specific call in the literature to deliberately extend the lens of interrogation into RRA: Interestingly, no work in either the rape victim literature or the rape survivor literature has examined correlates of individuals’ rape-resistant attitudes, beliefs, and perceptions. Although this lack of attention to rape-resistant psychology is unfortunate, it offers many unexplored avenues for future research.... We recommend the development of reliable and valid psychometric assessment tools to assess individuals’ rape-supportive and rape-resistant beliefs, attitudes, and perceptions, including identifying how such components are related to each other and to other relevant constructs. Although a number of validated and unvalidated measures exist in the rape victim literature to examine individuals’ rape-related beliefs, attitudes, and perceptions, as previously suggested, these measures focus on individuals’ rape-supportive psychology. (Hockett & Saucier, 2015) The existence of such an assessment tool holds a great deal of potential. Not only will all facets of victim blaming and perpetration be better understood with the ability to measure both RSA and RRA, there are likely benefits to the pursuit of social and criminal justice. The possibilities of thoroughly exploring sexual violence from a position of resistance rather than acceptance of the unacceptable status quo. As feminist theorist of color María Lugones observes, theorizing only from a position of oppression does little to suggest how that 4 oppression might be eradicated. We stand to learn a great deal by theorizing from empowerment as Lugones envisions (Lugones, 2003). This shift in perspective could lead to previously understudied areas of focus with new lenses for understanding. Having the ability to measure both RRA and RSA could conceivably allow much greater precision in sexual violence research, and encourage theorizing from a position of empowerment. In addition to this, the disruption of gendered assumptions about sexual violence would benefit the work in general. The relatively smaller body of literature studying same-sex rape, male survivors, and female perpetration has not only extended our knowledge of sexual violence itself, it has led to better survivor care for everyone, especially for male survivors, female survivors of same-sex rape, and transgender survivors. Studying from a less genderassumptive angle may allow us to think about “concrete possibilities for another world” (Muñoz, 2009), and also to deliberately bring silenced voices to the fore and challenge cultural construction of victims as primarily disempowered females. This project represents the first steps toward investigating RRA, and will utilize both feminist and queer theories to develop an instrument that can be administered to all genders and will consider not only male perpetration and female victimization but also male victims, female perpetrators, and same-sex sexual violence. Because the methods used to develop the kind of psychometric assessment tool recommended by Hockett and Saucier provide a sound methodological path to creating any research instrument, this project will utilize best practices for scale development as laid out by DeVellis (2017), with some modifications as explained in Chapter 2. I will begin with a search of relevant literature and interviews with experts in the field of sexual violence prevention, education, and treatment of survivors and perpetrators, then develop an instrument in the format of a scale to be administered in two stages; a pilot study in which feedback to refine the instrument will be solicited from a 5 development sample, followed by administration of the refined instrument to a larger sample with sufficient sampling adequacy. After the administration of the final research instrument, exploratory and confirmatory factor analysis models will be estimated, and the psychometric properties of the instrument will be analyzed to ascertain whether or not a reliable, valid attitudinal scale emerges. Statistical analysis of demographic associations will also be estimated. The research questions that guide the substantive portion of this project are as follows: • • • What attitudes are not captured through the current lens of RSA that might be resistant to dominant rape-supportive narratives? How is studying RRA different from studying RSA? What are the possibilities and potentialities of shifting our thinking from RSA to RRA? Cultural Context At the time of this writing, the political and cultural climate of the United States provides a particularly good illustration of the possibilities inherent in studying attitudes toward sexual violence, and perpetrators and survivors of that violence. In October 2017, sexual violence came to the fore of public discourse in the wake of a string of allegations of sexual misconduct committed by high profile individuals, such as acclaimed film executive and producer Harvey Weinstein, and the subsequent decisions made by their places of employment. Weinstein was fired from his own company after many women came forward with accusations of sexual misconduct ranging from harassment to rape. During the ensuing media activity actress Alyssa Milano tweeted this; Me too. Suggested by a friend: ‘If all the women who have been sexually harassed or assaulted wrote ‘Me too.’ as a status, we might give people a sense of the magnitude of the problem.” (Alyssa Milano, October 15, 2017 at 2:21 p.m.) 6 The accompanying hashtag, #MeToo, immediately went viral, inspiring 55,000 replies in the next twelve hours, trending No 1 on Twitter, and ultimately spreading across 85 countries and inspiring more than 85 million mentions on Twitter, Facebook, and Instagram over the next month and a half. The potential of expressing “me too” when others disclosed sexual violence had been seen two decades earlier by activist Tarana Burke, whose organization Just Be Inc. is geared toward promoting the wellness of young female minorities. She created the slogan in 1997 in response to a young woman who disclosed the abuse she suffered at the hands of her mother’s boyfriend. Burke felt that the simple solidarity in saying, “Me too” could help countless women and girls understand they were not alone, and she encouraged its use in the now largely defunct MySpace social media platform. However, in the current era of viral hashtags on Twitter, Facebook, and Instagram, with their enormous ability to engage in broad social dialogue, this phrase spawned a movement in 2017. Solidarity among survivors, holding perpetrators accountable, and unprecedented awareness of, not to mention belief in, the widespread nature of sexual violence in American society have been daily topics of discussion on social media since Milano’s Tweet on October 15th. In July 2018, President Donald Trump nominated Brett Kavanaugh for the Supreme Court of the United States (SCOTUS). Two months later, Dr. Christine Blasey Ford publicly alleged that nominee Kavanaugh had sexually assaulted her while in high school. Reminiscent of Anita Hill’s 1991 allegations of sexual harassment by then SCOTUS nominee Clarence Thomas, a public Senate Judiciary Committee hearing was held to hear Blasey Ford’s allegations. Like Thomas, Kavanaugh was ultimately sworn in as a Supreme Court justice despite the allegations. Like Hill, Blasey Ford experienced significant disbelief and violent backlash, including so many death threats she was forced to leave her home and take a leave of absence from her job (Martin, 2018). At the same time, support was expressed for 7 Blasey Ford and her family, both monetarily through a GoFundMe account that facilitated her several moves and private security, and in words across social media and the vast number of articles written about the allegations and hearing. The prevalence of online comments sections provided innumerable opportunities for public dialogue and either the reifying or dismantling of rape culture. Both enculturated rape-supportive narratives and the resistance to them were writ large across social media, news articles and reports, and political radio, cable, and network shows. Specific RSA identified in the 1980s were present in any comments section regarding the #MeToo movement or the Kavanaugh Supreme Court hearings, such as the double standard of alcohol consumption used to condemn and disbelieve Dr. Blasey Ford’s testimony before the Senate Judiciary Committee, while simultaneously used to excuse Brett Kavanaugh’s behavior. However, although every comments section reeked of specific RSA and RMA identified nearly four decades ago, there were also more egalitarian attitudes expressed, some particularly aligned with attitudes explored in this project. A few individuals in each comments section were usually able to identify RSA and RMA specifically and point out the logical fallacies behind them, often with precise statistics and solid sources. This phenomenon lends encouragement to the possibility of identifying and measuring a dimension of RRA that could conceivably function in similar ways on societal, interpersonal, and intrapsychic levels. Decades of research help contextualize this unprecedented moment in history, situating scholars and the general public for these vital considerations and public conversations. A general review of relevant work would be beneficial in understanding how the literature informed this project, as well as appreciating the need for the instrument created herein. Specifically, this chapter will consider the interdisciplinary study of sexual 8 violence, a few relevant theories and necessary considerations, and the broad functions of RSA and RMA, which may inform the functions of RRA. Interdisciplinary Study Research intended to understand, and thus eliminate, sexual violence takes places within most social science disciplines. Each adopts their own study of factors salient to individuals’ choices to be sexually violent within their social contexts, and the collective body of knowledge generated by diverse interdisciplinary study, much of it within the frameworks of RSA and RMA, is phenomenal. Anthropologists look cross-culturally across time and space at how sexual violence functions in different cultural contexts (Mead, 1935; Sanday, 2003; Watson-Franke, 2002). Anthropological work was the first to identify associations between women’s status and rape prevalence. In general, the more deeply held the social belief in natural differences between men and women and the more male work is privileged, the greater the incidences of rape and the less likely they are to be punished (Watson-Franke, 2002). By contrast, in societies where there is environmental, and thus economic, stability, and where women have high status and their work and community contributions are valued more equally, rape rates are low (Sanday, 2003). Peggy Reeves Sanday terms these “rape-free societies” and found 45 such societies in the 95 she studied (Sanday, 1981). The existence of these societies informs this project, as RRA are assumed to be cultivated in these societies to some degree. Sociologists generally study macrolevel influences on sexual violence. Interrogation of social contributors to sexual violence is the crux of sociological study of the topic. Sociological study focuses on cultural ideals of masculinity and femininity (Kimmel, 2017), equality and generalized misogyny (Koss, Dinero, Seibel, & Cox, 1988; Malamuth, 1981), 9 rape culture (Brownmiller, 1976), rape myth acceptance (Burt, 1980), and public health factors (National Sexual Violence Resource Center, 2004). Sociologists also study the unequal distribution of risks of perpetration and victimization across various demographic groups in society (Olive, 2012), as well as chosen group affiliations, including hypermasculine occupational groups like the military (O'Brien, Keith, & Shoemaker, 2015) and social groups like fraternities (Calzo, 2014). In general, sociology focuses on group outcomes related to social dominance orientations, finding broadly that privilege, especially gender privilege, is associated with risks of perpetration, and marginalization with risks of victimization Psychologists focus more on the intrapsychic and interpersonal causes and effects of sexual violence. Psychologists study the effects of sexual violence on survivors and their families (Ahrens, 2006), personal risk factors for offending (Maniglio, 2011), revictimization and secondary trauma (Campbell, Wasco, Ahrens, Sefl, & Barnes, 2001), and vicarious traumatization in survivor support systems (Baird & Jenkins, 2003). Evolutionary psychologists also ask the question of whether sexual violence is about sex or control (McKibbin, Shackelford, Goetz, & Starratt, 2008) and generally arrive at the same conclusion anthropologists do, that regardless of the roots of sexual violence, it is a crime that should be curbed through social sanctions and social change. Psychological study concerns itself with perpetrator motivations and impact of sexual violence on individuals. Social work scholars study best practices for counseling survivors of sexual violence, and contributors to both survivor trauma and healing (Campbell et al., 2001). Secondary trauma for individuals who counsel survivors (Choi, 2011) and study sexual violence (Coles, Astbury, Dartnall, & Limjerwala, 2014) is also included in social work research. Social work is primarily concerned with encouraging healing through counseling of survivors and their 10 families, as well as protecting counselors and researchers from vicarious trauma. Criminology has extended our understanding of sexual violence through the pursuit of criminal justice. In particular, criminologists have studied the concept of consent at great length. As consent is the crux of both criminal and social justice, this contribution is significant. Some criminological studies view consent as “any yes,” or any agreement to have sex, even an agreement made after coercion (Panichas, 2001) or force (Dripps, 1992). Some criminologists view consent as a moralizing legal act, one that transforms the crime of rape into consensual sex (Beaudrow, 2014). Some delve into whether the consent is behavioral (Muehlenhard, 1996) or a state of mind (Hurd, 1996). While the strengths and weaknesses of each argument can and should be debated, in all cases each is related to whether or not consent was given according to the rule of law, and how the legal system might handle the offender. Criminologists also look at the history of laws/legal reform (Anderson, 2005), incidences and prevalence of sexual violence (McGlynn, 2011), punishment of perpetrators (Hildebrand & Najdowski 2015), risk assessment (McGlynn, 2011), the handling of cases by the criminal justice system (Grubb & Turner, 2012; Kerstetter & van Winkle, 1990), false allegations (Rennison, 2002), and the legal framework surrounding consent (Beres, 2007). Theories of Sexual Violence Interdisciplinary work of all kinds has offered many theories related to sexual violence. The scope of these theories is far too broad to be fully encompassed here. However, a few vital theories have particularly informed this project and should be discussed. Gender. Because sexual violence is such a gendered phenomenon, all sexual violence research addresses gender in some way. Women are at much greater risk of sexual violence 11 than straight men and alter their lives far more significantly to protect themselves. Transgender women, particularly transwomen of color, are at the highest risk of sexual violence, with half of transwomen reporting sexual assault or rape because they were transgender (James et al., 2016). Cisgender men are at the highest risk of perpetration, committing upwards of 90% of all sexual violence with the result that being male is the primary predictor of perpetration (Beaudrow, 2014). Male survivors of sexual violence are most likely to have been assaulted by a man (Capers, 2011), but women also perpetuate sexual violence. Female perpetration of sexual violence against adult males is both oversexualized1 and understudied (Fisher & Pina, 2013). Same-sex rape between cisgender women is often not seen as “real” rape, either by survivors, perpetrators, society, and/or the criminal justice system (Langenderfer-Magruder, Whitfield, & Walls, 2016). Gender is a constant consideration in the study –and indeed, the experience– of sexual violence. Stemming from the function of rape myth acceptance and gender essentialist beliefs, traditional feminist theories of sexual violence are rooted not only in the prevalence of incidences that involve male perpetrators and female survivors, but also in the daily lived experience of women in cisheteropatriarchal cultural systems rife with RMA. Rates of sexual violence are not evenly distributed across groups of women, particularly with regard to race/ethnicity and transgender experience (Olive, 2012), but all women engage daily with the risk of being victimized (McGlynn, 2011). Feminist interrogation of sexual violence during and after the Women’s Liberation Movement of the 1970s revealed myriad interlocking perceptions, practices, and policies that contributed to not only the risk of sexual violence 1 See Urban Dictionary for the definition of “Mrs. Robinson,” or various porn sites for the trope of a hypersexualized aggressive female who “won’t take no for an answer” and their partners who respond enthusiastically to that aggression. 12 but also the lived experience of women (Burt, 1980). Scholars began to understand how rape myths normalized sexual violence, blamed victims, and excused perpetrators and perpetration in a distinctly gender-essentialist way (Schwendinger & Schwendinger, 1974). Rape myths were seen by feminist activist Susan Brownmiller as an active tool of patriarchy, narratives purposefully utilized by a gendered hierarchical status quo to oppress women and ensure that power remained in the hands of men (Brownmiller, 1976). Widespread acceptance of rape myths have been observed in individuals of all genders, even women against whom sexual violence was justified and excused by RMA (Briere, Malamuth, & Check, 1985). Regardless of the gender of perpetrator and victim, all incidences of sexual violence are inevitably viewed through societal myths about gender, primarily regarding violence and sexuality. Rape myths relate to broad constructions of masculinity and femininity along with gendered sexual norms and mores. In consideration of all of these, Connell’s theory of hegemonic masculinity, emphasized femininity, and subjugated masculinities is often employed in the study of gendered sexual violence (Connell, 1987). Hegemonic masculinity, or the form of masculinity held out as ideal in any particular culture, is conceptualized in U.S. culture as generally “toxic” in that associated outcomes are often harmful (Kaufman, 1993). Hegemonic masculinity is theorized as a way in which many men navigate an androcentric gendered power hierarchy as they jockey for position at the top. In this conceptualization, rape myths construct male aggression as normative, and women are not socialized to be assertive about their own needs or to resist male wishes. In a rape culture like the U.S., men are encouraged to be sexually aggressive and promiscuous (Debowska, Boduszek, Dhingra, & DeLisi, 2016), to “not take no for an answer” (Struckman-Johnson & Struckman-Johnson, 2003), and to dominate females and men whose masculinity cannot be 13 considered hegemonic (Kupers, 2005). For women, emphasized femininity is structured around compliance and is “oriented to accommodating the interests and desires of men” (Connell, 1987). In keeping with the oversexualization of women in general, the social construction of the virgin/whore dichotomy for women applies to emphasized femininity in ways that relate to sexual violence (Lebowitz & Roth, 1994). Rape myths send conflicting messages about “proper” emphasized femininity: how women should dress, what they can drink, where they can go, and how they should experience their sexuality. These messages are further complicated by race, social class, religion, and sexual orientation (Hamilton & Armstrong, 2009). With this combination of RMA and emphasized femininity, women can’t do dating or sex “right” regardless of what they do. Incompatible messages about sexiness, purity, sexual availability, and the public consumption of female bodies interact with rape myths to create a culture rife with victim blaming, slut shaming, and deliberate obfuscation of the social problem of sexual violence (Carr, 2013). Attempting to navigate, in real time, emphasized femininity with its accommodation of the interests and desires of men, as well as the possibilities of slut shaming and victim blaming, may contribute to a common response reported by female victims of sexual violence; tonic immobility, or “freeze” (Bracha, 2004). Given the insistence of some that it was not “real rape” if the victim did not fight back, this construction is highly problematic for female survivors. Power versus sex. Extending from feminist theories is the thought-provoking question of whether sexual violence is motivated by power or sex. Wartime sexual violence proves to be a fruitful analytic for this question. While not ubiquitous in all militarized conflicts, centuries of wartime sexual violence certainly suggest large-scale patterns of using sexual violence to clearly establish dominance over a group of people (Sjoberg & Via, 2010). 14 Combined with hegemonic norms of masculinity that condone violence during wartime and the prevalence of males in military service around the world, wartime rape serves to reify the predominantly heterosexual nature of sexual violence, its connection to hegemonic masculinity, and its use as a societal-level tool of power (Eriksson Baaz & Stern, 2009). Feminist scholar Susan Brownmiller argued that this usage of wartime rape as thorough establishment of dominance and social control was mimicked in women’s lived experience in the United States, establishing the dominance of heteropatriarchy and controlling women’s access to public spheres in a way she described as “gendered terrorism” (Brownmiller, 1976). Since that time, feminist scholarship has largely conceived of sexual violence in terms of power and dominance, rather than uncontrollable male sex drive and female responsibility as sexual gatekeepers, as had been previously conceived before that subtle shift. As women have consistently, albeit erratically, made gains in most public spheres since the Women’s Liberation Movement and social dialogue takes place not only interpersonally but publicly, particularly online, the threat of sexual violence is a common response to public feminist discourse (Megarry, 2014). A search of responses to Twitter hashtags such as #MeToo, #WhyIDidntReport, #MenCallMeThings, and #YesAllWomen quickly show how quickly there is violent backlash when women speak openly and critically about their lived experiences in cisheteropatriarchy. This backlash has again been likened to gendered terrorism, with renewed calls to attend to the public silencing of all women through threats of sexual violence, particularly women of color and queer women (Theriault, 2015). Social dominance orientations. Broad theories related to many different dominance and power structures have emerged from social science research. This research has shown that social dominance orientations (SDO), or an individual’s comfort with and belief in group-based social hierarchies, often function together (Ho et al., 2015). For 15 example, an individual’s high tolerance for RMA, as measured by an attitudinal scale, is predictive of their high tolerance for sexism, racism, homophobia, ageism, classism, and religious intolerance (Aosved & Long, 2006). Associations between SDO can be challenging for survivors of sexual assault, particularly when they occupy more than one minoritized status (Sloan & Wahab, 2000). Individuals who occupy more privileged positions in social hierarchies often hold greater levels of SDO (Ho et al., 2015), a social configuration that is particularly salient within the gendered realities of sexual violence and the criminal justice system. In particular, a negative association between RMA and attitudes toward rape victims has been observed in which as RMA increases, attitudes toward rape victims become more negative and greater blame is ascribed to them (Hockett, Saucier, Hoffman, Smith, & Craig, 2009) Social learning theories. Alongside these broad power machination, social learning theories explore more specific theories to address explanations or predictors for perpetration. Social learning theories are of principal importance, as the sociological imagination demands consideration of individuals within their social context. These theories posit that partnered sexual behaviors, both prosocial and deviant, are learned behaviors (Ferguson & Hartley, 2009). Theorists point to various social learning contexts associated with problematic observed outcomes, including background-situational models (Luthra & Gidycz, 2006) , social groups such as fraternities (Calzo, 2014), hypermasculine occupational groups such as the military (O'Brien et al., 2015), and intergenerational transmission of violence that includes sexual violence (Kwong, 2003). Criminologists extend social learning theory with situational considerations such as routine activity theory, which states that crime requires a motivated offender, suitable target, and lack of capable guardian, particularly with regard to perceptions of “suitable targets” of sexual violence (Schwartz & Pitts, 1995). 16 Sociosexual scripts. Within the context of social learning, script theory is an insightful way to consider sexual initiation and consent-seeking behaviors (or lack thereof). Generalized understandings of norms, values, expectations, and beliefs about sexual activity and sexuality are learned in social contexts, and inform both the envisioning and the performance of sexual interactions. Schemas anticipating sexual interactions, as informed by social context, are known as sociosexual scripts. These scripts specify generally accepted and anticipated features and behaviors of sexual interactions, as well as the general order in which events unfold and expected signals indicating a move to the “next” event in a typically prescribed series (Edgar & Fitzpatrick, 1993). Sociosexual scripts in the United States (and elsewhere) are problematic in terms of obtaining consent before the escalation of sexual activity and in the assumptions they make about gendered sexuality (Simon & Gagnon, 1984). The culturally agreed-upon script for sexual consent is for a partner, usually male, to proceed with sexual escalation until another partner, usually female, stops them (Allen, 2003). Heterosexual sociosexual scripts are at odds with the promotion of affirmative consent, in which anyone escalating sexual activity seeks consent before that escalation (Tuerkheimer, 2016). Uninterrogated expectations combined with norms of noncommunication are pervasive problems in all levels of sociosexual scripting. Sexual scripts are usually examined through cultural scenarios, interpersonal scripts, and intrapsychic scripts (Simon & Gagnon, 1984). All three levels of sociosexual scripting are informed by gender essentialist beliefs about the close relationship between hegemonic masculinity and sexual conquest, principally constructed as the heterosexual conquest of women, and female sexual passivity, particularly the obligation placed on women to be sexual “gatekeepers” in the face of men’s “natural” sexual appetites (Allen, 2003). Cultural scenarios are constructed and reified through these 17 gendered beliefs as well as puritanical assumptions about sexual desire, contributing to sociosexual scripts of dominant male sexuality and an oft-romanticized “overcoming” of female resistance (Struckman-Johnson & Struckman-Johnson, 2003). Interpersonally, neither taking the time nor committing to a practice to obtain consent are normalized, and even seen as an obstacle to pleasure by some (Humphreys & Brousseau, 2010). Intrapsychically, cultural or religious expectations, as well as dominant narratives about gender and sexuality often mediate individual sexual desire. Women engage deeply with the Madonna/whore dichotomy and the sexual double standard during sexual decision making (Jackson & Cram, 2010). Men engage internally with masculine expectations of sexual conquest, often in ways that reproduce gendered power hierarchies (Sanday, 1996). Sexual scripts from all levels become entangled and emerge problematically in various socially encouraged incongruent sexual behaviors, such as token compliance and resistance, that do not necessarily represent individuals’ authentic desires (Krahe, 1998). Biological desires and imperatives are certainly part of the equation, and function in ways that are difficult to disentangle from norms, values, expectations, and beliefs that inform noncommunicative, assumptive sociosexual scripts (Thakker & Ward, 2012). Queer sexual scripts are constructed in conversation with dominant heterosexual scripts and gender essentialist beliefs, but do not always serve the same functions (Power, McNair, & Carr, 2009). For instance, in gay male relationships there is no clear tasking of the “gatekeeper” role (Klinkenberg & Rose, 1994), and no clear culturally prescribed “initiator” in lesbian relationships (McCormick, 1987). Despite this, consent in queer sexual interactions is primarily communicated and ascertained nonverbally just as it is in heterosexual sexual interactions (Beres, Herold, & Maitland, 2004). Gender essentialist beliefs remain in queer cultural scenarios and interpersonal experiences. For example, lesbians typically wait longer 18 to have sex than gay male partnerships (Beres et al., 2004) and men often overperceive the sexual interest of their partners, perhaps especially their male partners (Kohl & Robertson, 2014). In all gender configurations, coercion and consent seeking are entwined with sociosexual scripts on all three levels. Problematic assumptions within norms of noncommunication regarding consent could conceivably impact coercion in any gender configuration of sexual partners. Biology. Biological theories often augment social theories. Some theories consider the biological result of attachment style, deprivation during development, and the processing of emotional stimuli (Mitchell & Beech, 2011). Others consider the biological result of positive feedback loops that reinforce the variables that lead to offending (Thakker & Ward, 2012). Most do not intend to offer a purely biological explanation for sexual violence (Turchik, Hebenstreit, & Judson, 2016). However, the dramatically gendered nature of perpetration has prompted some evolutionary biologists to promote the idea that rape is a biologically based evolutionary strategy to increase male fecundity and encourage aggression in males, seen as an adaptive reproductive trait (Thornhill & Palmer, 2000). This controversial position has caused many anthropologists and social scientists to weigh in, most taking a position opposite Thornhill and Palmer (Travis, 2003). In general, it is believed that if rape is a biologically influenced event, it can and should be restricted through social mores and laws. The existence of societies in which biologically similar men live and do not rape at the same rates as their counterparts in more rape-prone societies indicate that culture does indeed hold sway. 19 The Function of Rape-Myth Acceptance and Rape-Supportive Attitudes Broadly, RMA and RSA are the conceptual frameworks that build theories, generate specific research questions, and measure outcomes when those theories and questions are tested. These attitudinal frameworks were identified, conceptualized, and operationalized during the 1980s, formalizing with Martha Burt’s publication of Cultural Myths and Support for Rape, which offered her original Rape Myth Acceptance Scale (Burt, 1980). Several years later, a few researchers extended Burt’s concepts of RMA to more precisely identify the broad rape-supportive attitudinal dimension they associated with RMA (Briere & Malamuth, 1983; Briere et al., 1985). While Burt assumed but did not test a loose causal relationship beginning with RSA and leading to RMA, most scholars since then consider RMA as evidence of RSA and vice versa, and do not become overly concerned about the time order of the two. By way of example, a common RSA is that if individuals behave “properly” they can keep themselves safe from sexual violence (Burgess, 2007). Multiple specific rape myths are related to this belief, such as that sexually promiscuous women or women who are dressed provocatively are “asking for it” and thus partially responsible if sexual violence is perpetrated against them (Beaudrow, 2014; Hockett et al., 2009; Lebowitz & Roth, 1994). However, it is not known, nor is it particularly important to know in many instances, whether an individual accepts a rape myth because they hold a rape-supportive attitude, or whether they hold a rape-supportive attitude because they accept a rape myth. Regardless of time order, understanding the ways RMA and RSA function could prove illustrative when considering how RRA may function. Rape myths and RSA are tightly interwoven and decades of work in this area has helped inform policy, programs, educational endeavors, and the purposeful disruption of rape-supportive cultural narratives. Rape- 20 resistant attitudes may function similarly and have similar applications. Because of this possibility, I will first consider RMA here, followed by several specific aspects of RSA. Rape Myth Acceptance In general, it can be said that rape myths blame victims, deny harm, and excuse perpetration. These false beliefs are deeply heteronormative, lack any kind of nuance, and work within a rape culture to create a climate that is hostile and dismissive toward victims of sexual violence. Female perpetration and male victimization directly challenge our highly gendered assumptions about sexual violence and its contributors, and RMA offers no explanation. Rape myths associated with female victims are primarily rooted in the belief that women lie; they exaggerate sexual violence or want sex but must make a show of resisting to avoid slut shaming, disingenuosly “asking for it” in various ways and crying wolf after the fact. Rape myths associated with male victims are primarily rooted in beliefs that men have uncontrollable libidos, always want sex no matter what, and are physically capable enough that force is impossible. Myths about rape in LGBTQ+ communities deny sexual violence at all, or treat it as an inevitable result of the “lifestyle.” All rape myths deny the complexity of coercion, the reality of arousal nonconcordance and physical response during violence, and the influence of gendered sociosexual scripting. The following sections will discuss heteronormative constructs that imbue rape myths, and then consider research that explores RMA outside of the assumed heteronormative context and the attitudinal scales that have been developed RMA. Heteronormativity in rape myths. Because a cisgender man rapes a woman every two minutes in the United States (Bureau of Justice Statistics, 2015), gender must be the first consideration in the study of sexual violence, and this cultural conceptualization informs 21 sexual violence research, understandings, policies, and language. However, the near-exclusive image of male perpetrator/female survivor (mp/fs) presents an inadequate understanding of sexual violence in general, and its functions across society in particular (Capers, 2011). Erasure of male victims, female perpetration, and same-sex sexual violence presents a twofold problem; not only does it ignore potential contributions to research and policy, but it also fails to adequately serve survivors of sexual violence who fall outside of this conceptualization (Turchik et al., 2016). These understudied areas are of vital importance in understanding sexual violence and caring for survivors. Even within the gendered reality of overwhelming male perpetration and female victimization, the heteronormativity of societal rape myths serves neither social nor criminal justice. Rather, rape myths often rely on deeply essentialist beliefs about gender, and construct male perpetration against women as natural and inevitable, female perpetration against men as impossible, and same-sex sexual violence as invisible. They function to excuse and exonerate perpetrators, blame victims, and task potential victims (women in particular) with keeping themselves safe rather than tasking all individuals with keeping themselves from becoming perpetrators. For example, the belief that women’s provocative clothing cause men to lose control and rape them is an example of several different rape myths functioning together; men are “visual” creatures who can’t control their sexual appetites, women are responsible to anticipate predatory male gaze and alter their dress and behavior to combat it, and at least some blame lies with her if she doesn’t (Moffett, 2006). Acceptance of these myths and their relationship to sexual violence, victim blaming, problematic justice, and understanding of the social problem of sexual violence has been studied extensively (Coles et al., 2014; Klement, Sagarin, & Lee, 2016; Payne, Lonsway, & Fitzgerald, 1999; Scott & Tetreault, 1987; Wegner, Abbey, Pierce, Pegram, & Woerner, 2015). 22 Male rape myth acceptance. The heteronormativity of RMA means relatively little is understood about male survivors (Capers, 2011). Understanding male rape myths and male rape myth acceptance (MRMA) is vitally important in caring for male survivors who seek services, often in spaces understandably created for female survivors, and justice within a system rife with myths that excuse perpetration against men as well as women. Male rape myths generally revolve around essentialist ideals of masculinity, assuming strength and sexual prowess that make sexual violence impossible. When a male survivor is heterosexual and the perpetrator is female, socially constructed gender conceptualizations come into direct conversation with heteronormative MRMA in deeply problematic ways. Many male survivors recount a feeling they described as “broken” masculinity when they experienced denial, shame, and continued feelings of sexual vulnerability after violence (Javaid, 2018). Compounding this internal stigma, male survivors are often disbelieved due to the rape myth that “all men want sex all the time” (Fisher & Pina, 2013). Disbelief is particularly strong if a male-bodied individual had an erection during the assault (Sleath & Bull, 2009). Men’s generally superior size and strength relative to females leads to the myth that men are too strong to be forced into sex by women (Kassing, Beesley, & Frey, 2005) and that this, combined with gendered expectations about male sexual aggression and female sexual passivity, means that male rape by females is extremely rare, which it is not (StruckmanJohnson & Struckman-Johnson, 1992). Stemming from this heteronormative construction, sexual orientation becomes associated with male same-sex sexual violence, and homosexuality assumed in both perpetration and victimization (Coxell & King, 1996). Because it is true that males are more likely to be victimized by other males, homosexuality and the study of male perpetrators dominates research regarding male rape (Kassing et al., 2005). Male rape myth acceptance 23 encompasses the belief that gay men are raped by other gay men because there are two strong male libidos, two promiscuous individuals involved, and no individual tasked with sexual “gatekeeping” (Rumney, 2010). Relatedly, socially ascribed assumptions regarding the relatively higher number of sexual partners in homosexual men result in the myth that gay men “ask for it” through their sexual behavior (Kassing et al., 2005). A somewhat contradictory assumption is that male sexual violence happens primarily in prison and is an establishment of dominance outside of the norm (Kimmel, 2017). Erection and/or ejaculation in male victims are often assumed to be consent (Sleath & Bull, 2009), indicating an attendant lack of understanding that an individual can desire sexual activity and still not consent to sexual activity for any number of reasons, or that bodies can respond sexually even during trauma through desire nonconcordance (Nagoski, 2015). The impact of sexual violence on men is also misunderstood, with emotional stoicism assumed to prevent or resolve the trauma often seen in female victims of sexual violence (Snyder, 2008). The significance of these harmful realities for male survivors, combined with the lack of research and resources allocated specifically to men, has led some to call for a completely degendered approach to sexual violence research, prevention, and conceptualization (Capers, 2011). Female perpetration. Rape myths generally do not address female perpetration at all. When they do, they perpetuate the idea that female perpetration is so rare it does not warrant significant attention. Consequently, the paucity of work studying female perpetration is notable (Fisher & Pina, 2013). For example, at the time of this writing, a search for “perpetrators of sexual violence” in Google Scholar returns nearly 3,100 articles whereas a search for “female perpetrators of sexual violence” returns only 62. Primarily, these 62 articles focus on female perpetration of sexual abuse against a minor, as this is more common than female perpetration against adults (Watford, 2010). However, female 24 perpetration against adults, both male and female, does happen, and some general observations have been made in the literature. Similar to the erasure of male victimization, female perpetration is often erased because it is antithetical to gendered stereotypes about female docility, empathy, and sexuality (Fisher & Pina, 2013). Female perpetrators rely more on coercion and verbal aggression than physical force/intimidation, or take advantage of a man’s incapacitated state (Krahe, Waizenhofer, & Moller, 2003). When aggressing against other female-bodied individuals, female perpetration is often ignored completely or categorized as a “cat fight” by those who should intervene (Langenderfer-Magruder et al., 2016). Female perpetration and social responses to it operate within female socialization and broad gendered assumptions. Understanding female perpetrators provides vital depth to discussions about sexual violence in general, usually through disconfirming observations to the norm. Same-sex sexual violence. Heteronormative RMA leads to the erasure of same-sex sexual violence, rendering its victims invisible. However, same-sex sexual violence does happen, and should be understood for better prevention approaches and survivor care. There are many similarities between other-gender and same-gender violence, and research has offered other context-specific points to consider. For example, sexual minority stress has been identified as a contributor to intimate partner violence, including sexual violence, in same-sex relationships (Carvalho, Lewis, Derlega, Winstead, & Viggiano, 2011). However, seeking services is often complicated both by the inability of the criminal justice system to recognize same-sex violence for a variety of gender-essentialist reasons (Rumney, 2010), as well as the gendered nature of survivor support (Richards & Marcum, 2015). Resources are often gender- or sex-segregated; thus a female survivor who seeks services at a women’s center or shelter runs the risk of encountering her perpetrator there, and transwomen have 25 difficulty accessing services at all because of their biological sex (Langenderfer-Magruder et al., 2016). An added layer of complexity in reporting for queer survivors of same-sex sexual violence is the risk to the survivor of losing or disrupting their queer community support (Carvalho et al., 2011). When women are raped by other women, it is sometimes not recognized as rape in the law, which still specifically defines rape in terms of penile penetration in some jurisdictions (Federal Bureau of Investigation, 2012), and societal-level beliefs that closely associate rape with penises (Franklin, 2004). Transgender individuals are at extremely high risk of sexual violence due to their queer identity, but this is rarely understood and legal protections are sporadic at best (James et al., 2016). Both the criminal justice system and general social attitudes are imbued with misinformation about same-gender sexual violence. Policies and protocols intended to prevent and respond to sexual violence are often deeply cisheteronormative, leaving queer individuals and those in a position to help them unsure of how to proceed with prevention programs and survivor services (Potter, Fountain, & Stapleton, 2012). This cisheteronormative construction of prevention, education, and survivor response can lead to retraumatization and/or erasure for queer individuals, significantly impacting their healing trajectories (Richards & Marcum, 2015). Rape myth acceptance scales. Research about RMA is often accomplished through the use of various RMA scales. Burt developed the Rape Myth Acceptance Scale as the first psychometric tool to test theories surrounding RMA and various other attitudinal dimensions such as sexual conservatism and acceptance of interpersonal violence (Burt, 1980). Stemming from Dr. Burt’s work, the Illinois Rape Myth Acceptance Scale (IRMA) was developed in 1999 by Payne, Lonsway, and Fitzgerald in an attempt to update the language and concepts surrounding RMA since Burt’s conceptualization in 1980 (Payne et 26 al., 1999). Since that time, assessment of RMA has been an important part of the study of sexual violence, with the IRMA rising to prominence as the most reliable, most widely used psychometric tool available to scholars. Because cultural norms surrounding sexuality continually evolve as social attitudes evolve, the IRMA was recently updated to reflect more current language (McMahon & Farmer, 2011). The updated IRMA was also purposefully written to capture subtle, persistent rape myths that are often unexposed due to social desirability bias in research participants. Despite the updated language of the IRMA, its utility is only in measuring myths centered primarily in heteronormative assumptions of male perpetration and female victimization, as it does not deviate from the mp/fs concept. Thus, it fails to adequately assess myths surrounding male victims, female perpetration, or same-gender sexual violence. A set of rape myths specific to male victimization that function similarly to those specific to female victimization was identified in the Male Rape Myth Acceptance Scale (Kerr Melanson, 1998). Heteronormative assumptions persisted in the MRMA, thus a scale specific to female perpetration against males was developed to overcome biases regarding sexual orientation measured in the MRMA (Snyder, 2008). No scale could be located that specifically assessed rape myths regarding same-sex sexual violence. Rape-Supportive Attitudes The prevalence of RSA in the rape culture of the United States is such that all individuals engage with them on some level, and most individuals hold RSA to some degree. Because of this, scholars have studied how RSA specifically relate to various related outcomes, and how they function in many different groups and contexts. For examples, a vast range of specific associations with RSA have been studied, including but not limited to 27 gender role stereotypes, ideals of masculinity and femininity, pornography and alcohol consumption, enjoyment of sexist humor, media consumption including music, film, and books, male peer association, sexual fantasies, intimacy and loneliness, hetero/homosociality, willingness to blame victims, video game participation, and various social domination orientations including sexism, racism, right-wing authoritarianism, and religious dogmatism. Researchers have studied RSA in many groups, including gender, age cohort, religion, race, ethnicity, sexual orientation, Greek affiliation in college, and every role possible in the criminal justice system, from police officers and sexual assault nurse examiners to judges and jury members. This breadth of work was considered in a relatively recent literature review, which identified eight primary RSA domains that are correlated with higher rape proclivity, or self-reported willingness to engage in sexually predatory behavior, and aggressive sexual history. Broadly, RSA can be grouped together in these categories; Not Rape, Woman Caused, Alcohol, Sex Role, Misogyny, Gender Role, Coercion, and Misinterpretation (Burgess, 2007). The scope of associations and contexts related to RSA is far too broad to adequately conceptualize here. Four specific group contexts prove illustrative to consider within this project and will be discussed in this section: 1) the general public, all of whom are socialized and enculturated within a rape culture; 2) perpetrators of sexual violence; 3) survivors of sexual violence; and 4) the criminal justice system, particularly those who investigate, prosecute, and adjudicate criminal sexual violence cases. Rape-supportive attitudes in the general public. Susan Brownmiller first identified the United States as a rape culture during the 1970s. In her conceptualization, a rape culture is one in which sexual aggression, which she speaks of primarily in terms of male violence perpetuated against females, is tolerated and even glorified (Brownmiller, 28 1976). Peggy Reeves Sanday notably utilized this conceptualization in a global study of various cultures’ approaches to sexual violence. Sanday observed that around the globe, societal-level rates of rape were associated with a local cultural milieu that emphasized gender differences, and accepted male domination and interpersonal violence as norms (Sanday, 1981). Since that time, other scholars have studied RSA and RMA as indicators of rape culture, noting in particular that high levels of both are present where cultures blame victims, deny harm, and excuse perpetrators (Campbell, 1998). Importantly, just as all cultures do, rape cultures inform norms, values, expectations, and beliefs in ways that are not always obvious but that hold sway on intrapsychic, interpersonal, and societal levels. This broad prevalence of RSA is problematic in many ways. Individuals of all genders with higher levels of RSA tend to assign more blame to victims (Grubb & Turner, 2012) and fail to adequately address perpetration of sexual violence, instead focusing on victimization (Conrad, 2006). Focusing on potential victimization rather than potential perpetration does nothing to dismantle rape culture or stop perpetration, it simply attempts to ensure that rapists rape someone else in a more vulnerable position. Despite the lack of progress in these narratives, U.S. culture is so imbued with RSA it is no surprise that during broad, societal-level conversations about sexual violence, rape-supportive myths and victim blaming narratives emerge in public dialogue. For example, during the 2018 Kavanaugh Supreme Court confirmation hearing, the survivor was blamed (Payne, 2018), harm was denied (Tolentino, 2018), and perpetration excused (Axelrod, 2018). This can serve to reinforce already enculturated RSA, further entrenching these attitudes in the public consciousness (Franiuk, Seefelt, & Vandello, 2008). Similarly to RMA, the heteronormative nature of RSA, which presupposes a male perpetrator and a female survivor, is reiterated over and over again, effectively erasing male victims (Capers, 2011), female perpetrators 29 (Gilroy & Carroll, 2009), and same-sex violence (Turchik et al., 2016). This is deeply concerning, as individuals who experience sexual violence outside assumed gender/sex narratives often experience difficulty obtaining appropriate support, both institutionally and personally (Capers, 2011). Rape-supportive attitudes in those who perpetrate sexual violence. Primarily, RSA function to encourage and possibly create rape-prone individuals (Beaudrow, 2014; Brownmiller, 1976). They also function to excuse and normalize perpetration of sexual violence as a norm of masculinity (Murnen, Wright, & Kaluzny, 2002), deny the harm of sexual violence (Hermann, Babchishin, Nunes, Leth-Steensen, & Cortoni, 2012), and shift blame from perpetrators to victims (Wegner et al., 2015). Unsurprisingly, higher levels of both RSA and RMA are often observed in individuals convicted of perpetrating sexual violence (Scott & Tetreault, 1987). In a chicken-and-egg relationship whose causality has not been established, men with higher levels of RSA also tend to overperceive women’s sexual interest and interpret behaviors as consent that were not consent (Farris, Treat, Viken, & McFall, 2008). Similarly, men with sexually aggressive histories and higher levels of rape proclivity often also have high levels of RSA (Kupers, 2005), and tend to identify nonnormative individuals (e.g., sex workers or gay men) as complicit in their own violence due to their deviance from societal norms (Grubb & Turner, 2012). When individuals are surrounded by others with high levels of RSA, (e.g., convicted perpetrators) their levels of RSA tend to increase (Debowska et al., 2016). While holding RSA alone does not always predict rape proclivity, high RSA combined with feelings of entitlement are often observed together in convicted perpetrators (Wegner et al., 2015). In short, enculturated RSA teach would-be perpetrators that overcoming their partner’s resistance is a normal part of masculinity, helps them identify potential targets the system is predisposed to discount and 30 disbelieve, deny harm to those victimized, and shift blame away from themselves. Rape-supportive attitudes in survivors of sexual violence. It is not only perpetrators and individuals with high levels of rape proclivity who hold high levels of RSA. Of particular interest in this project are the ways RSA function in nonperpetrating groups, especially survivors of sexual violence. Individuals who are themselves victims of sexual violence often fall back on enculturated victim blaming narratives when they describe their own experiences. Principally, many survivors of sexual violence do not recognize what happened to them as rape (Peterson & Muehlenhard, 2004). The persistent myth that “real rape” is a violent attack on a woman by a strange man causes many individuals to deny sexual violence, toward both themselves and others, in instances that do not match the culturally acknowledged scenario (DuMont, Miller, & Myhr, 2003). Survivors with higher levels of RSA often ascribe negative cultural assumptions about rape victims onto themselves after a rape (e.g., slut), even if they had not identified with those labels before (Lebowitz & Roth, 1994). Rape-supportive attitudes can interfere with survivors’ healing trajectory, and those who have higher levels of RSA have worse outcomes in recovery than survivors who reject RSA (Katz & Burt, 1988). Self-blame rooted in RSA has been cited by survivors as among the reasons they did not pursue criminal charges, or even disclose their experience to anyone else (DuMont et al., 2003), contributing to the significant underreporting that masks the pervasity of sexual violence and contributes to the RSA that survivors, particularly women, lie or falsely amplify the problem (Lonsway & Fitzgerald, 1994). If survivors do disclose their experience to another person and their response is imbued with RSA, self-blame can be reinforced (Ahrens, 2006). Effects of self-blame are described as “crippling” in the literature, and survivor care has focused on diffusing selfblame since the 1970s (Katz & Burt, 1988). 31 Rape-supportive attitudes in the criminal justice system. Criminal justice is also influenced by RSA and RMA, as dozens of studies have explored. This system is not an isolated entity but rather a series of laws and policies made and interpreted by individuals imbued with the narratives of their culture. Laws are not value-free, and RSA are found in laws and policies on every level of government (Schwendinger & Schwendinger, 1974). Rape laws, originally intended to protect the property rights of men as “owners” of women, recognized only an exclusively heteronormative construction of male perpetrator and female victim (Sheffield, 1998). Legal definitions of rape did not account for the possibility of male victims, female perpetrators, same-sex rape, or object, oral, or anal rape (Tracy, Fromson, Gentile-Long, & Whitman, 2011). Forceful resistance was often required as the benchmark of nonconsent, disregarding the reality of tonic immobility during traumatic events (Fuse, Forsyth, Marx, Gallup, & Weaver, 2007). Spousal rape was not recognized in the law until 1977, and while all states had withdrawn exemptions for spousal rape by 1993, assumed continual consent in marriage remains as a measurable RSA (Ferro, Carmele, & Saltzman, 2008). Traditional rape laws focused more on the behavior of the victim than the offender (Spohn, 1999). Rape law reform and the implementation of rape shield laws began in earnest in the 1990s, but vestiges of the RSA that permeated traditional laws remain in current laws, policies, and practices (Tracy et al., 2011). Also within the criminal justice system are individuals who investigate, prosecute, and adjudicate criminal cases of sexual violence. Within a rape culture, all of these individuals have been enculturated with RSA. At the time of the crime, police stations have been identified as sites of retraumatization, often due to RSA and RMA in police officers (Campbell, 1998; Campbell et al., 2001). Problematic interactions with police officers, all related to various RSA, have been observed and studied in many ways, including disbelief 32 when the crime did not fit the culturally acknowledged scenario of stranger rape (Campbell et al., 2001), blame when a female victim was intoxicated or dressed provocatively (Goodman-Delahunty & Graham, 2010), differential response to “good victims” (e.g., visible distress and injuries) and “bad victims” (e.g., victims without those visible signs) in the officer’s perception (Madigan & Gamble, 1991), transgender victims and offenders (Buist & Stone, 2014), male victims (Capers, 2011), queer victims (Rumney, 2010), and sex workers (Page, 2007). Once a criminal case is opened, investigatory decisions are often influenced by RSA. False reports of sexual violence, which fall between 2% - 8%, are often perceived by detectives to be a much greater rate (Page, 2008). Detectives have been observed assessing the credibility of the victim before taking a report seriously in vignette research, with low arrest rates projected when a victim’s credibility was deemed low (Sleath & Bull, 2017). Survivors who had experienced sexual violence more than once were viewed with suspicion, and subsequent investigations lukewarm (Page, 2008). When the perpetrator and victim had been or were in a relationship, officers assigned more blame to victims and less authenticity to the rape (Hine & Murphy, 2017). Unskilled detective interviews have been shown to influence survivor decisions, sometimes discouraging further pursuit of criminal justice when the survivor did not feel supported and believed (Rich K. , 2014) or impacting how much a survivor was willing to disclose during an investigation (Patterson, 2011). Detective skepticism relating to RSA and RMA have been identified as a barrier to case progression in the criminal justice system (Jordan, 2011). Courtrooms too are significantly affected by RSA and RMA. Relatively few sexual violence cases reach trial, and those that do bear significant marks of societal- and personallevel RSA. The specter of false accusation has been present in courtrooms since Matthew 33 Hale so infamously and influentially stated in 1609 that rape is “an accusation easily to be made and hard to be proved, and harder to be defended by the party accused” (Geis, 1978). These words have been quoted in courtrooms even into the 21st century; indeed, they may still be (Sanday, 2003). In combination with widespread rape-supportive narratives, this specter of false accusation is particularly problematic during court proceedings and jury deliberations. Rape myths have also been codified into law in some places, often focusing on victim behavior (e.g., resisting “to the utmost” to establish that the event was indeed “real” rape) rather than perpetrator behavior (Earnshaw, Pitpitan, & Chaudoir, 2011). Before rape law reform in the 1990s and the implementation of rape shield laws, the victim’s sexual history, occupation (particularly that of sex worker), dress, and other behaviors have been interrogated in trials as evidence of false accusation (Oakley, 2007). Even after the implementation of rape shield laws, judges occasionally allow these questions if they feel they are pertinent (Spohn, 1999). Victim credibility has been linked to trial outcomes in terms of gender and the belief that only women are raped (Capers, 2011), “proper” behavior by the witness on the stand as informed by societal-level RSA (Hildebrand & Najdowski, 2015), alcohol consumption (Goodman-Delahunty & Graham, 2010) and dress at the time of the assault (Hildebrand & Najdowski, 2015), religion (Edwards, Turchik, Dardis, Reynolds, & Gidycz, 2011), race/ethnicity (Campbell, 1998), and queerness in both sexual orientation and gender identity (Buist & Stone, 2014). In a jury of peers, the outcome of a trial can be affected by even one person who believes the victim-blaming narratives readily supplied by U.S. culture (Hildebrand & Najdowski, 2015). Despite the likelihood of such a circumstance, mistrials and pronouncements of “not guilty” are often lumped into statistics of false accusation and general rape-supportive beliefs that victims, usually women, lie or overreact to normal sexual behavior (Lonsway, 2010). 34 Rape-supportive attitudes scales. Many scales exist to measure RSA. Martha Burt, who developed the first Rape Myth Acceptance Scale, also developed the Adversarial Sexual Beliefs scale, a close precursor to the conceptualization and measurement of RSA (Burt, 1980). Briere, Malamuth, and Check (1980) extended the analyses of Burt’s own work as, “Burt’s analysis appears to have markedly underestimated the variety of rapesupportive beliefs tapped by her scales” (Briere, Malamuth, & Check, 1985). This led to a more formal conceptualization of RSA, and scales were developed to measure them. While many scales tap various RSA with specificity, such as Measuring a Macho Personality Constellation (Mosher & Sirkin, 1984) and Attraction to Sexual Aggression (Malamuth, 1989), other scales have measured RSA more broadly. The Rape-supportive Attitudes Scale developed by Ilsa Lottes in 1991 drew from multiples scales developed in previous years, and presented an aggregate scale that measured the breadth of RSA (Lottes, 1991). This instrument has been used when a broad assessment of RSA is appropriate to the research project. Other RSA scales, such at the Rape Attitudes and Beliefs Scale (Burgess, 2007), have updated language and attempted to tap into more nuance as research about sexual violence continually becomes more finegrained. In sum, the pervasiveness of heteronormative RSA and RMA in a rape culture influence laws, policies, and practices that govern all individuals within that culture. Individuals interact with those systemic realities in various ways, some leading to rape proclivity, victim blaming, denial of harm to victims, and erasure of nonnormative incidences of sexual violence. Some lead to self-blame if an individual experiences sexual violence, or even the inability to recognize oneself as a survivor of sexual violence. The pursuit of social and criminal justice brings victims and perpetrators in direct contact with institutions and individuals that have been enculturated with RSA and RMA, confusing 35 assignation of blame and making criminal justice elusive. The subtle machinations of pervasive RSA and RMA influence every facet of engagement with sexual violence, from individual experiences of perpetration and victimization to the laws intended to provide justice. A great deal of research has been done in the attempt to understand, and thus eliminate, sexual violence. Psychometric tests in the form of attitudinal scales are often used in this research. Literature-Informed Assumptions After this search of literature surrounding sexual violence, I have chosen to move forward in the construction of a research instrument that attends carefully to several considerations and perspectives. First and foremost, I have chosen to create a research instrument that does not assume the mp/fs model. While this gendered assumption is understandable given that this is the most common gender configuration of perpetrator and victim, studying same-sex violence, male victimization, and female perpetration is vital to understanding and eliminating sexual violence in its entirety. Research instruments must be equipped for use in these important areas of study. Unless an attitude to be explored was particularly gendered (e.g., the sexual double standard that stigmatizes female sexuality while glorifying male sexuality), the instrument was purposefully written to be gender neutral. It also did not assume heterosexuality, monogamy, or assume any sexual behavior is “proper” except the clear establishment of enthusiastic and ongoing consent. Critical, intersectional feminism demands that all matrices of domination be challenged and dismantled, and so this choice to consider future interrogation of all contributors to sexual violence was purposeful. This stems directly from my feminist criticism of an overarching system of power and control that oppresses many. While cisheteropatriarchy certainly oppresses principally 36 women and femmes, it also does not serve men whose hegemonic masculinities contribute to risks of perpetration, men whose subordinate or queer masculinities contribute to risks of victimization within this system, and men whose participation in hegemonically masculine activities and spaces increases risks of both perpetration and victimization (e.g., male peer support for sexual conquest or even group rape in fraternity and athletic settings, as well as hazing through sexual violence in these same settings). Socially influenced risks of perpetration in particular are considered within this project. In particular, I recognize the existence of a “rape culture,” or, in one circumferential definition, “a social system that has slowly normalised rape and sexual assault through the bombardment of images, language, laws and social attitudes” (Ford, 2013). Rape culture, which is informed and exacerbated by cultural sex negativity, lack of healthy sexual scripts, problematic gender ideals, and systemic discrimination, is believed to contribute to the uneven experience of sexual violence across gender, race, class, sexual orientation, ability, occupation et cetera. Social positionality is key to both risks of victimization and risks of perpetration in a rape culture. However, I did appreciate and was informed by an oppositional perspective that must be considered with regard to the existence of and emphasis on rape culture in prevention strategies and the application of justice in instances of sexual violence. Recently, Scott Berkowitz, the president of Rape, Abuse & Incest National Network (RAINN), the nation’s largest anti sexual violence organization and prominent leader in conversations regarding sexual violence and prevention, explicated this stance on rape culture in a letter to a White House Task Force appointed by President Obama to protect college students from sexual violence on campuses; “In the last few years, there has been an unfortunate trend towards blaming ‘rape culture’ for the extensive problem of sexual violence... While it is helpful to point out the systemic barriers to 37 addressing the problem, it is important to not lose sight of a simple fact: Rape is caused not by cultural factors but by the conscious decisions, of a small percentage of the community, to commit a violent crime” (Berkowitz, 2014). While this perspective certainly has merit and was carefully considered, this project was undertaken with a firm belief in the existence of rape culture and recognition of social contributors to sexual violence. It was not assumed that culture was more salient to sexual violence than personal choices made by violent individuals. Rather, the assumption was that the recognition of rape culture does not mitigate personal responsibility on any level, and we stand to gain a great deal by understanding the social contexts that influence individuals’ behaviors. The relationship between personal choice and systemic contexts were specifically recognized and explored at length throughout this project. Indeed, the research instrument created was purposefully developed to continue to do just that in the future. Finally, when considering the lessons learned through moving beyond the assumption of male power exercised over female victims and recognizing the social contributors to rape culture, it became clear that this instrument needed to capture far more than simply attitudes that would discourage someone from perpetrating violent sexual assault. Rather, this instrument was written with an attempt to assess nuanced understanding of the necessity of recognizing and seeking proper consent during interpersonal sexual interactions rather than following problematic sociosexual scripts. With that in mind, the project is guided by a synthesized theory that recognizes both sex drive and the drive to exert dominance as contributors to sexual violence (Ellis, 1991). It attends to the mediating influence of entitlement in individuals who may stop short of committing the kind of violent sexual assault we collectively recognize as “real” rape, but who commit sexual assault nonetheless through failing to seek consent, instead relying on cultural scripts and gendered 38 assumptions about sexuality to exert pressure on reluctant partners, or on tonic immobility to circumvent resistance. It is not the intent of synthesized theory (or indeed, this project) to excuse perpetrators who did not view their actions as sexually violent. Rather, it is an attempt to understand the many cases of sexual violence that do not fit our culturally recognized model but still contribute significantly to survivor trauma. This is delicate work, as perpetrators “not understanding” sexual violence is a defense that has been used successfully to acquit many rapists, particularly before rape law reform when laws focused on victim behavior rather than perpetrator behavior. However, studying the nuance in various kinds of perpetrators and motivations for perpetration, which this instrument allows when combined with existing attitudinal scales, could be illustrative and valuable work. The next chapter describes my methods, informed by this body of literature and my broad paradigmatic frameworks, in detail. CHAPTER 2 METHODS In this chapter, I will describe the methods I followed to explore the theoretical dimension of rape-resistant attitudes (RRA) and develop an instrument to test specific RRA. Interrogation of the elusive, previously unexplored concept of RRA provides a unique challenge in that no literature exists that directly examines the concept, and no research instruments currently exist to measure it. Because of this, I selected scale development as a research method for this largely unstudied attitudinal dimension (DeVellis, 2017; Faul & van Zyl, 2004) because the method provides systematic opportunities for understanding. Scale development begins by generating theories pertinent to the topic at hand through qualitative work and ends by statistically testing an instrument developed to assess those theories. The interim process requires various forms of development, oversight by experts, checks, and tests, all of which encourage knowledge production about the topic itself. A great deal of attitudinal research, including research concerning sexual violence, is performed through the administration of multiple attitudinal scales (see Hockett et al., 2009 for an excellent example of the use of scales in the study of sexual violence). The existence of a scale measuring RRA could be a valuable addition to the plethora of sexuality-related measures that currently exist, offering the potential to study sexual violence in a new way and through a new lens. The method of scale building is illustrative of the topic in and of itself, and the possible contribution if I am successful made scale building a solid method for this exploratory 40 project. This chapter begins with theories that inform the methods, then proceeds to describe the specific steps in the method I followed to generate the instrument tested in this project. Theory While this exploratory project is purposefully theory generating, several theoretical perspectives are salient to my perspective and positionality as a researcher. First and foremost, my work is informed by queer and feminist critical theories that acknowledge and attend to power imbalances on both systemic and interpersonal levels (Abbinnett, 2003). Relatedly, theories of masculinities are important to consider in the study of sexual violence, as the variable most predictive of perpetration is being male-identified (Beaudrow, 2014). As a symbolic interactionist, I am interested in the influence of cultural context on sexual violence, particularly through the interactions of individuals within socially constructed meanings and contexts (Blumer, 1973). Stemming from that, I understand that my data are gathered in a rape culture rife with RSA (Watson-Franke, 2002), and that individuals who hold what may be considered RRA have forged those attitudes within a largely rapesupportive context. Similarly to Sanday’s theories regarding rape-free societies (Sanday, 2003), I hope these individuals can provide insight into what may “work” in approaching the goal of eliminating sexual violence from a rape-prone society. Extending from Sanday’s anthropological work, I have considered the theory of rape as an evolutionary strategy for procreation in unsuccessful males (Thornhill & Palmer, 2000). While stopping far short of accepting this theory, I accept that biological influences are likely relevant to some degree, as the genes of violent individuals may have been propagated through violent procreation. Regardless, the solution to the social problem of sexual violence is sociocultural, as 41 demonstrated when societies make laws against, police, and punish these unacceptable behaviors (Belknap, 2007). Beyond the substantive theories that provide the framework for this project, my approach to scale building itself is driven by classical measurement theory (Faul & van Zyl, 2004). Classical measurement theory is often used in the development and initial validation of unidimensional scales, as it provides a basic assessment of the observed score (DeVellis, 2017). The observed score is equal to the true score, or the actual experience, plus the error score, or the gap between actual experience and participants’ perception of actual experience (Nunnally & Bernstein, 1994). The assumptions underlying classical measurement theory drove methodological choices, such as writing items with similarly strong wording (DeVellis, 2017). Importantly, classical measurement theory is also related to the psychometric properties of any instrument developed with it. Domain sampling, or the ability to employ randomly selected items that yield the same result as the total of the items, is part of classical measurement theory. Sampling from an adequate domain requires the researcher to seek the “theoretical totality” of explanations for the latent variable (DeVellis, 2017). This demand of classical measurement theory also satisfies the actions required of a scale developer that are associated with content validity during initial construction of the instrument. Any psychometric properties assessed through tests related to error should be in line with the assumptions of classical measurement theory. Because attitudes are unobservable, psychometric instruments are necessary to study them. Consequently, the psychometric properties of any instrument intended to measure unobservable attitudes are vital indicators of the appropriateness and utility of the instrument. Psychometric properties are related to deliberate methodological choices made by the researcher to attend reliability, validity, and consistency. Throughout this chapter and 42 indeed this paper, I will indicate these deliberate methodological choices and their relationship to psychometric properties. Nine Steps From Conceptualization to Statistical Analyses In general, I followed the steps of scale development outlined by DeVellis (2017) with some modifications. Specifically, DeVellis recommends that the researcher generate an initial pool of items following a literature search, and then ask experts on the topic to review this initial pool and provide feedback (DeVellis, 2017). However, because at present RRA is a theoretical concept with no literature directly studying it, I chose to include experts both in the generation of a broad exploration of attitudes as well as to provide feedback after the initial pool of items had been written. A slightly different method of scale development recommended by Faul and van Zyl (2004) informed this choice. Ultimately, my method of scale development followed these steps: Step 1 – Clearly Conceptualize Rape-Resistant as an Attitudinal Dimension Step 2 – Literature Search Step 3 – Expert Interviews Step 4 – Determine Format Step 5 – Generate Initial Pool of Items Step 6 – Pool Reviewed by Experts Step 7 – Administer Items to Development Sample Step 8 – Administer Items to Full Sample Step 9 – Statistical Analyses Step 1: clearly conceptualize rape resistance as an attitudinal dimension. Scale development requires clear conceptualization of the latent variable being studied, as having a precise definition of the phenomenon the scale intends to measure helps the content of the scale stay inside specific parameters, increasing content validity. This is challenging, as attitudinal dimensions are elusive and difficult to measure or observe directly. In addition to this, I knew that specific RRA would develop as I spoke to experts and performed data 43 analyses. For my method of conceptualization, I chose to clearly define what I felt a person who held RRA might be like and how the set of attitudes may function within the context of a rape culture. Ultimately, the conceptualization narrowed to three vital considerations; 1) a person who could be considered rape-resistant would not have high rape proclivity, 2) they would recognize, understand, and value consent (both when considering narratives and in their own lives), and 3) they would not blame victims for the sexual violence perpetrated against them. Considerable effort informed the clear conceptualization of RRA, both in the interest of content validity and in having a concept to discuss with experts. In this, I was supremely conscious that nearly everyone holds some level of RSA, particularly when considering rape myth acceptance (RMA). Merely holding RSA or having high rates of RMA do not in themselves indicate rape proclivity or a sexually aggressive history (Schwendinger & Schwendinger, 1974). Rather, there are mediating factors in individual decisions to commit sexual violence, such as entitlement (Wegner et al., 2015). Rape-supportive attitudes are problematic, regardless of whether those who hold them perpetrate sexual violence or not. False beliefs about sexual violence contribute to rape culture in ways that do not specifically relate to perpetration. High societal rates of RSA contribute to the societal and individual practices of victim blaming, self blame in those who experience sexual violence, reticence of victims to report, willingness of detectives to close cases in which the victim isn’t the “right” kind of victim, jurors not recognizing perpetrator responsibility or assigning blame and responsibility to victims, and the deep shame survivors grapple with due to enculturated beliefs about sexual violence. The functions of rape culture informed my conceptualization of the underlying construct of rape resistance in this process. Being mindful of this context was also important in ensuring content validity, as the instrument was intentionally developed to test RRA in various contexts and still tap into 44 deeply held attitudes and beliefs related to sexual violence. Step 2: literature search. I followed the methodological path for scale development laid out by DeVellis (2017) and Faul and van Zyl (2004). As a largely theoretical dimension at present, exploration of the topic first required a strong understanding of sexual violence itself, as well as the functions of RSA. Thus, my literature search began by reading literature related to RSA (Google Scholar keyword search, “rape-supportive attitudes”), articles cited by this body of literature, and articles that cited this body of literature. This project is based on the consideration that the absence of RSA may not necessarily assume the presence of RRA, as the concepts may be distinct from one another. Since RRA has not previously been conceptualized, the first priority of investigation was to consider whether RRA could be a measurable dimension, just as RSA has been identified as a measurable dimension. Ultimately, this method led to a broad search of sexual violence literature, deliberately read within this new paradigm, to find evidence of individuals who lived in a rape-prone society yet resisted common rape-supportive narratives. In performing a literature search, I did not assume that RRA were the opposite of RSA. Rather, upon encountering a specific RSA in the literature, I generated ideas about how that particular attitude might be engaged with by individuals who would not blame victims when they learned of sexual violence, not blame themselves if sexual violence was perpetrated against them, and who would not commit sexual violence themselves. I also considered the extent to which the absence of a particular RSA might be indicative of RRA, if at all. Through this process, I attempted to think beyond the paradigm of our current focus on RSA and trust to further research and refinement to expose my own biases and correct my course. As an individual who was socialized in a rape culture myself, unconscious bias was continually at the forefront of my mind. My thinking was purposefully critical and 45 my interpretations against the grain as I considered possible RRA. As an example of my thought process, many individuals do not recognize acquaintance rape as rape in example vignettes, or believe acquaintance rape is less harmful to the victim than stranger rape (Viki, Abrams, & Masser, 2004). This denial of sexual violence and harm to victims has been clustered under the domain of Not Rape in an assessment of RSA in college men (Burgess, 2007). When considering this rape-supportive belief, it seemed to follow that related rape-resistant beliefs might be not only the ability to recognize rape but also the ability to recognize sexual consent in similar vignettes and willingness to call acquaintance rape “real” rape rather than “a misunderstanding” or any of the other dog whistles used by individuals with high RSA. Literature from many disciplines contributed to my understanding and consideration of RRA. In particular, research in Criminology yielded immediate insight, especially in the conceptualization of consent. Consent is the point upon which sexual activity pivots between criminal and noncriminal in courts of law. Some arguments view consent in terms of “moralizing” the event, and that all sexual behavior is criminal (and thus problematic and illegal) until the moral transformation of consent occurs (Hurd, 1996). Despite the problematics in that conceptualization, legal arguments surrounding how to operationalize and recognize moralizing consent during rape trials offered clarity I may not have found otherwise. For example, it has been argued that consent given after repeated refusals constitutes legal consent regardless of direct or indirect coercion (Walker, 1997), that jurors should consider certain behaviors over the course of a date (e.g., flirting or accepting a drink) as “inadvertent” behavioral consent (Ostler, 2003), and that legal consent is a performative act rather than a personal attitude, thus certain acts communicate consent regardless of intent (Wertheimer, 1996). Engaging with these and other gender essentialist, victim-blaming 46 perspectives prompted me to think with more nuance precisely because they were problematic in my critical feminist and queer paradigms. Sharpening my arguments against the belief that certain flirting behaviors constituted “inadvertent” heteronormative consent led me to consider several facets that may be considered RRA, such as a desire to fully understand a partner’s willingness before escalating sexual activity or how assumed consent based on heteronormative sociosexual scripts is inappropriate in a queer context. Considering these arguments from a criminological perspective was extremely helpful. Step 3: expert interviews. This careful, against-the-grain reading of the current literature generated a series of questions and ideas I wanted to explore with more specificity. As the first step in an exploration of an unstudied, theoretical attitudinal dimension, I sought open-ended input from individuals working in many capacities related to sexual violence. This input was intended to generate a plethora of possible items from multiple areas of expertise in sexual violence work rather than perform a deep analysis of qualitative data. Seeking interdisciplinary expertise informed both content validity related to conceptualization and construct validity related to ensuring the instrument measured what it intended to measure. Interrogating RRA from multiple perspectives would not only increase validity, but also extend both the concept of RRA itself and the applicability of any instrument I developed. These considerations prompted me to pursue semistructured interviews in order to have both the structure of specific questions as well as the ability to explore relevant new topics that emerged from interviews (Wolcott, 2005). With my sample of experts in mind, I developed an interview guide that allowed me to both ask specific questions and also explore new ideas through unscheduled probes into comments made by participants (Berg, 1995). The Institutional Review Board of the University of Utah approved the method, sample, and interview guide during Spring semester 2017 and I began 47 interviews in May, concluding in August of that year. Due to the exploratory nature of this process, the recommended method is to identify and interview people with a range of expertise in the topic who can generate a voluminous list of attitudes to consider as possibly rape-resistant (DeVellis, 2017; Faul & van Zyl, 2004). Snowball sampling for convenience began with two particularly well-connected individuals in different areas of the field of sexual violence prevention/education/survivor support. One was an employee at the University of Utah’s Student Health and Wellness Center, a passionate educator and the cocreator of the Men’s Anti Violence Network, an organization intended to engage men in the fight against both domestic and sexual violence. The other was a Sexual Assault Nurse Examiner (SANE) on the board of Wasatch Forensic Nurses who worked with multiple prevention/education organizations on local, state, and national levels. Upon completion of these interviews, I asked each of them to recommend other individuals who could contribute to the breadth of understanding I sought at this stage. I asked for individuals who were actively involved in prevention/education, survivor support, law enforcement, or individuals who worked with perpetrators postarrest. I specifically sought individuals who worked with communities of color and the queer community. From these two individuals, I scheduled interviews with two Licensed Clinical Social Workers (LCSW); one who counseled juvenile sexual offenders and one who counseled male survivors and queer individuals, the founder of a local survivor support organization, an individual from the Utah Department of Health Sexual Violence Task Force, a caseworker from the Salt Lake Urban Indian Center, and an officer in a nationally affiliated university fraternity whose charter specifically addressed the elevated risk of sexual violence in fraternity settings on college campuses. Upon completion of these interviews, data saturation had not been reached. I then solicited interviews with employees and board 48 members of the Utah Coalition Against Sexual Assault (UCASA) and interviewed three individuals; a gay man with expertise regarding sexual violence in the queer community, the former Executive Director of a women’s shelter who was key in developing Utah’s 40-hour certification training for all individuals who interface with survivors in any official capacity, and a college-aged woman who developed and implemented an oncampus student group for prevention/education about sexual violence at Weber State University. In addition to UCASA, I also attended a meeting of the Utah Sexual Violence Council and solicited the input of any member of that council. From that meeting, I interviewed a pediatrician from Safe and Healthy Families, one individual from the Family Crisis Center, and one from the Utah Office for Victims of Crime. Upon completion of these 14 interviews, data saturation was reached and I concluded this portion of the project. Step 4: determine format. In keeping with the format of other sexuality-related measures (Fisher, Davis, Yarber, & Davis, 2011), this scale was written as a series of unambiguous statements followed by a category partition scale in which response categories are assumed to be broken up into equal intervals. I chose a seven-point Likert scale indicating agreement from Strongly Disagree to Strongly Agree. A seven point scale was selected in an effort to avoid an excess of unbounded “Strongly Dis/Agree” responses and to increase the reliability of the scale (Faul & van Zyl, 2004). Statements were carefully written to be unambiguous, measure only one attitude, and were strongly worded in an effort to encourage respondents to carefully consider the strength of their dis/agreement. Step 5: generate initial pool of items. Upon concluding the interview portion of this project, I transcribed the interviews. Rather than coding for emergent themes from the body of qualitative data at this time, I searched each transcript for the mention of any possible attribute that may be considered rape-resistant in the way I had operationalized it. 49 As per best practices in creating a unidimensional scale, I purposefully considered the specific construct of RRA with each item I wrote (Faul & van Zyl, 2004). I followed a twostep list method of scale item creation; 1) I wrote down one attribute of the defined construct, and 2) wrote an item based on that attribute (Hudson, 1994). For each item on this list, I wrote a statement reflecting what I felt was the gist of the attitude it intended to measure. For example, an attitude mentioned by multiple experts was something I called “personal responsibility to not commit assault.” This attitude seemed to reference the problematic emphasis in rape cultures on rape avoidance and resistance strategies in potential victims, rather than education and prevention in potential perpetrators (Ullman, 2007). From this attitude identified by experts, I wrote the statement, “It is up to me to be sure I do not commit sexual violence.” At this point, I was not concerned about overlapping items or statements that may measure similar attitudes but was rather concerned with being as expansive as possible in my thinking. In generating possible RRA, I was purposefully broad and creative in my thinking, and was not concerned about limiting the number of items at this point. DeVellis recommended generating a large initial pool of items, as this was an attempt to identify the “theoretical totality” of contributors to rape resistance (DeVellis, 2017). Experts identified 201 potentially rape-resistant attitudes, a list of which can be found in Appendix A. With the goal of theoretical totality in mind, I wrote a list of 326 potential items to measure these attitudes. Multiple wordings of a few important attitudes (e.g., valuing consent) were included at this stage, as this allowed the possibility of assessing which wording generated the most statistically sound responses. Each statement was strongly written in order to reduce the possibility of primarily “strongly dis/agree” responses. Knowing I would use a seven-point Likert-type category partition scale, I wrote items that were strong enough to 50 encourage participants to pause and reflect before they strongly dis/agreed too readily. Each statement measured only one concept and was purposefully nonassumptive about gender or sexual orientation, with the exception of a few concepts that are specifically gendered and lose their vitality when presented as gender neutral, such as the Madonna/whore dichotomy and the sexual double standard (Conrad, 2006). A few items were negatively worded to test the utility of both negative and positive wording, as well as to explore their potential as participant attention checks. Once I had a statement written to correspond with each possible RRA, I coded for emergent themes. Many research scales contain a series of shorter subscales, which makes analysis and interpretation both easier and more fruitful, and also provide some information about broader characteristics of the construct. For example, one theme that emerged from coding was “Consent.” Unsurprisingly given the importance of consent in the context of sexual violence, “Consent” was the theme that had the most items written for it, indicating that for those with expertise in sexual violence prevention/education/survivor support, consent was an important part of what could be considered rape-resistant. Ten such themes emerged, and I treated them as subscales within the broader context of RRA. These will be described in detail in Chapter 3. Step 6: pool reviewed by experts. After organizing the items in this manner, I arranged for a focus group with most of the interviewees to give them time to read over the items and provide feedback of any kind, as well as to discuss with one another which items were likely to be salient to the measurement of RRA and which were not. This step has been identified as crucial to maximizing content validity in the final instrument (DeVellis, 2017). Eight experts were available to attend the focus group. Each participant in the focus group indicated the three statements in each themed category that seemed most salient to the 51 concept of RRA, and gave feedback about the wording of each item and whether or not it was likely to measure the attitude it was intended to measure. There was a great deal of discussion around the table as experts compared responses, debated about the efficacy of item wording, and discussed how well or not an item would interrogate the attitude it was intended to interrogate. The focus group narrowed the list down to an administrable 108, refined the wording of almost every item, and gave me many things to consider as they talked with one another about sexual violence in general. The focus group was not recorded or transcribed. Rather, I took notes when important conversations that would enhance construct validity took place between interviewees. Step 7: administer items to a development sample. As recommended by DeVellis (2017), upon completion of the focus group I refined the items for administration to a development sample. A vast majority of the original items were eliminated, and most of the remaining items were reworded based on feedback from interviewees. The length of the scale was a primary concern at this stage of development. In general, scale reliability increases with length (Nunnally & Bernstein, 1994). However, scale length is subject to diminishing returns; reliability increases dramatically from one item to 10, but not as significantly from 11 items to 20 (Faul & van Zyl, 2004). Sexuality-related measures vary greatly in length (Fisher et al., 2011), and the utility of different measures often depends on the ultimate context of implementation and likelihood of research participants to fully engage with the number of items they encounter in survey research. A balance must be struck between desired solicitation of information and participant motivation to complete a survey. At this preliminary stage, I made the decision to include more items rather than fewer, hoping that the topic of the survey would hold participants’ interest and encourage completion of a longer instrument. 52 Related to the length of the scale I developed was the decision about which other scales to administer along with my own instrument. Attitudinal and behavioral scales are often administered together, and I considered many scales related to sexual violence that could be illustrative when considering a participant’s level of RRA in the ways I had conceptualized it. This method also helps establish external construct validity, as correlations between instruments should indicate valid results in line with theory and previous empirical work. Primarily, I chose to administer the Illinois Rape Myth Acceptance (IRMA) scale (Payne et al., 1999) along with my own. The IRMA has been tested continually since its development, and is closest measurement of my conceptualization of RRA that currently exists. In addition to tapping RRA closely, the IRMA is often the scale used in the manner I envision using a RRA scale; to assess the efficacy of sexual violence prevention/education programs, for example. Administering the IRMA in full alongside my own scale allowed me to compare that score along with the score generated by the RRA scale, providing the opportunity to consider with greater precision the similarities and/or differences in low levels of RMA and high levels of RRA. Alongside the IRMA, I chose to administer the Sexual Consent Scale – Revised (SCSR; Humphreys & Brousseau, 2010) due to the emphasis placed on consent by my interviewees and in sexual violence literature. Similarly to the IRMA, the SCSR has been tested and validated, and the resultant comparison of scores would be potentially illustrative in many ways. Despite the “Consent” subscale being the longest subscale in the preliminary scale I developed, I chose to include the SCSR because the statements were different enough to tap into slightly disparate things (e.g., obtaining consent as a way to avoid accusal versus obtaining consent as a way to ensure a partner’s pleasure). This heavy emphasis on consent in the administration of these scales was warranted, given the pivotal role consent plays in 53 the realm of sexual violence and the potential for more finegrained analysis with reliable tools that allow consideration for deeply nuanced attitudes. Before settling on the IRMA and SCSR, I considered multiple other scales rooted in current scholarship surrounding sexual violence, such as the Hostility Toward Women Scale (Lonsway & Fitzgerald, 1995), the Sexually Aggressive History Questionnaire (Burgess, 2007), and the Situational Rape Proclivity Scale (Malamuth, 1981). In particular, I considered the Bystander Attitude Scale (Banyard, Moynihan, & Plante, 2007)and the Bystander Attitude Scale – Revised (McMahon, 2010) because the literature on bystander intervention identifies as vital the ability to recognize problematic behavior. This ability struck me as an important feature of rape resistance in the ways I had conceptualized it, particularly because not recognizing rape has been identified as part of RSA (Burt, 1980). Ultimately, I chose to include several items about recognizing problematic behavior rather than admininster the full Bystander Attitude Scale. However, philosophical engagement with bystander literature was particularly helpful during my formulation of ideas about RRA and is thus worth noting, since careful consideration of various sexuality-related measures became important as a method. Upon making final decisions about which items and scales to include in the development sample, I utilized Qualtrics Survey Software to create an administrable survey that could be accessed anonymously. Surveys contained a consent page indicating that participation was completely voluntary, and that any participant could refrain from answering any question for any reason, and could quit the survey at any time. This consent page also provided my contact information and the contact information for the University of Utah Institutional Review Board in the case of an adverse event. Various digital platforms of access to the survey were tested (e.g., different electronic tablets with internet capabilities, 54 both Apple and Android smartphones, PC and MAC computers and laptops), and as many different internet service providers as possible. The survey included an openended text box along with every statement, and feedback on wording et cetera. was solicited from the development sample. Participants were sought from University of Utah students enrolled in Sociology classes, largely through one supportive professor with several large classes. At the close of the pilot study, all qualitative feedback on scale items was reviewed. These comments were extremely helpful, as they revealed several items that were misunderstood by many participants. Other feedback indicated that some items contained too much jargon or were misleading. Most items elicited a range of responses, indicating that they were written strongly enough to cause participants to consider the nuance in responses of the category partition scale. Several items were completely rewritten, and a few were eliminated altogether. In addition to this, some demographic catagories were altered based on participant feedback. Due to the high rate of survey completion in the development sample, I made the decision to administer almost all the scale items to the full sample since the length of the survey seemed not to be a deterrent to completion. Step 8: administer items to full sample. After refining items and updating demographic categories based on the pilot study, I created another survey in Qualtrics to be administered to a full sample. Participants were recruited in three waves during spring, summer, and fall semesters at the University of Utah. Several instructors sent the link via email to their students, others made announcements in person and through Canvas, and others distributed a paper flier containing both an easy to remember bit.ly link and a QR code that took participants directly to the survey on their mobile device. Students were largely recruited from Sociology classes and through word of mouth. Lassonde Entrepreneurial Institute at the University of Utah also sent the information to their email 55 listserv, reaching a broader variety of majors. The required sample size was reached in September, but as responses continually increased the survey was left open until early November. Step 9: statistical analysis. There are multiple methodological options using factor analysis to guide the process of identifying the most important indicators of the latent variable (rape-resistant attitudes in this case) and narrowing a broad pool of items down to a smaller set that does a good job of capturing desired information. These methods rely on both statistical criteria and conceptual fit as determined by the researcher. Because RRA is previously unstudied in this way, I used a high number of items in an attempt to consider the theoretical totality of contributors to the latent, unobservable variable of rape resistance. This allowed me to test many options but required a significant reduction in number of items; from theoretical totality to an administrable number of high-performing items that functioned together and made conceptual sense. For this, principal component factor analysis (PCF) was ideal, as it accounts for all the variance in a correlation matrix and thus is a good tool for this reduction process (Acock, 2014). Structural Equation Modeling (SEM) was then used to confirm the findings from PCF. I used Stata statistical software to perform all statistical analyses. Preliminary statistical analysis. For ease of interpretation, I began by reverse coding any variables necessary so a higher score on the scale was indicative of more rape resistance. Before beginning PCF, it was important to assess whether the data collected was sufficient to perform factor analysis. This was done by using a command (syntax: factortest varlist) to ascertain the Kaiser-Meyer-Olkin (KMO) value and ensure that my sample was adequate, as well as Bartlett’s test of sphericity to ensure that each model was significant at the p ≥ 0.001 level (Child, 2006). A KMO value of ≥ 0.8 was desired, although I would 56 accept a KMO that would round up to 0.8. Values ≥ 0.7 were considered by Kaiser to be “middling” but not unacceptable (Kaiser, 1974). I also ran a bivariate correlation measure to ensure concurrent validity (Acock, 2014). Once I was certain my models met these criteria, I proceeded with preliminary analyses. For each model, I estimated descriptive statistics and bivariate correlations. I used Stata (syntax: sum) to calculate the mean, standard deviation, and to ascertain the range of responses for each item across the seven available categories of dis/agreement in the Likert response options. I then used Stata to measure skew and kurtosis (syntax: sktest) in an effort to understand the shape of each distribution. Next, I used Stata to calculate bivariate correlations between each of the items in the model (syntax: pwcorr). According to best practices, I removed any item that did not correlate at ≥ 0.30 with at least one other item in the model (Acock, 2014). Exploratory factor analysis. The Stata output for PCF (syntax: factor varlist, pcf) yields a correlation matrix that contains a list of factors, which are an indication of how many dimensions explain the latent variable. These factors have an associated eigenvalue, which suggests the factor(s) to retain for their statistical performance. The accepted rule of thumb is that any eigenvalue < 1 underperforms and the factor should not be retained (Gie Yong & Pearce, 2013). If a correlation matrix yields multiple factors with eigenvalues ≥ 1, these factors should be considered conceptually to ascertain their fit. Conceptual fit can be suggested by which items load the highest on those factors. In general, a factor loading of ≥ 0.4 is considered a good indicator, and factor loadings of ≤ 0.3 should be considered for elimination. Once items were eliminated through PCF and models reduced to good statistical performers that made conceptual sense, it was important to ensure reliability through a 57 Cronbach’s alpha score (syntax: alpha varlist, item asis). This test is an indicator of internal validity, and according to best practices I rejected any model whose score was not ≥ 0.7 (Acock, 2013). Even above this benchmark, if removing an item increased the alpha score, I carefully considered whether the item should remain in the model. I began with PCF analysis on the items within each subscale, making note of the number of factors onto which the items loaded. I looked at factors beyond Factor1 to ascertain whether any made conceptual sense as a separate but related factor, thus indicating that more than one latent variable explained the subscale. This did not occur for any of the 10 subscales, so I proceeded with interpretation. I ascertained the eigenvalue of Factor1, ensuring it was higher than 1. I removed any items that did not load at ≥ 0.4 on the factor (Acock, 2014), then reran the analysis. Most of the 108 items loaded ≥ 0.4 in their subscale, but I was interested in drastically reducing the number of items. For some models, I set the benchmark for removal at ≥ 0.5 or ≥ 0.6 in an effort to use only the top statistical performers in the final set. This process of exploratory factor analysis yielded a statistically sound index of 65 items, informed by each of the 10 subscales, with good conceptual fit. I felt this index could be further refined into an administrable scale, so I used a different approach and went through the processes of exploratory factor analysis again. Relying on domain sampling, I selected the top three statistical performers in each of the 10 subscales, which yielded a 30-item scale that met or exceeded all statistical benchmarks. However, considering the items as a whole indicated that the conceptual fit was not ideal. I then performed PCF analysis on all 108 items tested in this instrument, removing any that did not load ≥ .4 on Factor1 or that loaded too similarly onto a subsequent factor. This yielded 60 high-performing items. I then selected the items that seemed most consistent with the way I conceptualized RRA. This approach yielded a 25-item scale with good reliability, 58 internal consistency, and validity, and a conceptual fit I was completely comfortable with. Each of the 10 subscales contributed to the final scale. The psychometric properties of all PCF analyses are described in Chapter 4. Confirmatory factor analysis. At this point, I had two possible instruments; one a set of 65 items arranged in 10 subscales and one a scale of 25 items. Both were statistically and conceptually sound. To ensure that this was indeed the case, I proceeded with confirmatory factor analysis through structural equation modeling (SEM), a causal model with the ability to assess latent attitudinal variables through observed responses to survey items (Acock, 2013). I again used Stata for all statistical analyses, the results of which will be described in Chapter 4. Structural equation modeling allows each item to have its own unique variance, which contributes to the shared meaning of a set of items. This shared meaning is represented in a scale score of a single dimension. Exploratory factor analysis revealed the dimensions and allowed me to ensure they measured a single concept. Structural equation modeling computed variances and fit statistics, which allowed the generation of a meaningful score from each unidimensional scale (Acock, 2013). Because each item in SEM has unique variance, it is possible to correlate error between variables to reduce the noise in the model, improving goodness of fit. In Stata, SEM modification indices suggest which items in the model could be allowed to covary for better fit when it made conceptual sense to do so. Appropriately correlating error terms increases the reliability of the instrument. Standardized factor loadings in SEM should be ≥ 0.4 and have significant p values (Acock, 2013). Once the adequacy of factor loadings is ascertained, the next step is to assess goodness of fit. Goodness-of-fit tests used here were Chi-square, the root mean square error of approximation (RMSEA), and the comparative fit index (CFI). Scores for the RMSEA 59 range from 0 to 1, and indicate how well the model fits the population covariance matrix. Smaller scores indicate better fit. Scores for the CFI range from 0 to 1 and examine discrepancies between the data and the model. For the CFI, higher scores indicate better fit (Child, 2006). To do this, I used a Stata command (syntax: sem) to test the model created through PCF. I asked Stata for standardized scores (syntax: sem, standardized) and goodness-of-fit statistics (syntax: estat gof, stats (all)) to analyze how well the data fit the model. Statistical benchmarks for goodness-of-fit were set in keeping with best practices; ≥ 0.4 for each standardized factor loading, a RMSEA score of ≤ 0.05, a comparative fit index (CFI) score of ≥ 0.95, and an insignificant chi-square (Acock, 2013). Upon ensuring that the standardized factor loadings were sufficient, I executed a command (syntax: estat mindices) to return modification indices that could suggest changes to the structure of the model through allowing covariance. Any additional parameter estimated in the model must be approached with caution (Acock, 2013), and so no changes were made unless they fit both of these criteria; 1) the changes significantly improved the model, and 2) the modification suggested could be justified theoretically. To avoid making substantive decisions based on chance, I added only one additional parameter at a time, and then only when that parameter had a substantial modification index and there was a conceptually sound reason to correlate the error between two variables (Acock, 2013). Following this method ultimately led to models with comfortable theoretical fit and better goodness-of-fit. I did not proceed until the models were statistically perfect, which could have been achieved in each case. Rather, I made careful decisions that made good conceptual sense, trusting that focusing purely on statistical fit without the benefit of researcher insight rarely yields a good instrument (DeVellis, 2017). The psychometric properties of these CFA tests are described in Chapter 4. 60 Regression analysis. After testing the reliability and validity of the index and scale, I went on to explore the associations between RRA and demographic variables. Criterion validity could be established if these outcomes were in keeping with more broad empirical observations (e.g., that men tend to hold more RMA than women). I began by comparing means between groups after creating dichotomous variables from demographic categories of interest. This was done carefully so as not to inadvertently include any participant into a demographic category they specifically did not select into (e.g., in the model comparing male- and female-identified participants, I did not include those who preferred not to state their gender or whose gender identity was not in the male/female binary). I conducted t tests (syntax: ttest) for each model in the categories of gender identity, sexual orientation, racial/ethnic identity, religious affiliation, and age. Area of study data were also gathered, but were of less interest to me than the principal identity demographic categories. Next, I conducted multiple linear regression to ascertain which demographic variables in particular were associated with RRA. I conducted regression models on each subscale in the 25-item scale. Using RRA as the dependent variable, I ran these models (syntax: reg) considering gender identity, sexual orientation, racial/ethnic identity, religious affiliation, and age. I also correlated the results with the IRMA and the SCSR as a check of external validity. The results of regression analyses are described in Chapter 5. Summary In this chapter, I described the methods I used to explore the theoretical dimension of RRA, develop an instrument to test this dimension, and analyze the results of those tests. I began with a literature search, then used in-depth interviews and a focus group with experts in the field of sexual violence to further cement the conceptualization of RRA and 61 develop a solid instrument. This instrument was tested in a pilot study, refined according to their feedback, and administered to a full sample. The results were analyzed statistically using Principal Component factor analysis and Structural Equation Modeling, which allowed for the narrowing of the total number of items into administrable scales. Regression models were estimated to assess the associations between RRA and various demographic variables of interest, and correlations between RRA, RMA, and attitudes toward consent were tested through comparison of correlations to the IRMA and SCSR. In the next chapter, I will describe the results from the qualitative methods of this project. CHAPTER 3 QUALITATIVE RESULTS In this chapter, I will describe the results from the various qualitative methods I used during this project. I began by searching relevant literature with the same methods I would use during a content analysis. This yielded two primary outcomes; one, I knew that indeed, no work relating to rape-resistant attitudes was available, and two, attitudes had been studied that while not considered specifically in terms of being rape-resistant were still able to inform the conceptualization of the term. These attitudes and the work that identified them also informed the development of a semi structured interview guide to use with experts in the field of sexual violence. Once these interviews had identified the “theoretical totality” of attitudes that could be considered rape-resistant, my interviewees came together in a focus group to help refine the instrument I would ultimately administer during the quantitative portion of this project. And finally, qualitative data was gathered during a pilot study of the research instrument. This qualitative data was used to further refine the final instrument before it was administered to the full sample. Content Analysis A content analysis of literature surrounding sexual violence yielded a preliminary list of 24 possible RRA. These attitudes, along with the knowledge gleaned from reading the literature with this topic in mind, aided in the development of an interview guide for indepth 63 semi structured interviews with experts. The search for and articulation of these preliminary attitudes provided language to describe the theoretical dimension of rape resistance to experts in the way I conceptualized it, and to help them verbally explore the way raperesistant attitudes could conceivably function. The 24 preliminary attitudes were: • • • • • • • • • • • • • • • • • • • • • • • • Low rape myth acceptance Recognizing rape Egalitarian attitudes toward women Willingness to intervene when witnessing predatory behavior Willingness to disrupt gendered sexual scripts Willingness to remain engaged in sexual violence prevention education Desire for mutual sexual satisfaction Willingness to engage with survivors’ experience Understand that prevention education needs to be perpetrator-centered Valuing consent as part of pleasure Knowledge of consent Unwillingness to bypass consent Obligation to be sexually frustrated rather than coerce/force sex Acceptance of non-toxic masculinities Disruption of sexual double standard Careful alcohol use Understanding of incapacitation Discernment in porn consumption Commitment to humanizing others Unwillingness to blame victims Empathy Heterosociality Taking responsibility for own actions Lack of entitlement Each of these attitudes emerged from a cross-grain reading of at least one article, and some seemed to be suggested by more than one. The following section shows the thought processes that led to the preliminary list of possible attitudes, and the ways my purposive literature review informed the conceptualization of RRA. Low rape myth acceptance. The importance and impact of rape myth acceptance cannot be understated. Many problematic contributors to rape culture and sexual violence 64 have been linked to rape myth acceptance. Relatively lower rape myth acceptance (RMA) is the measure most closely resembling rape-resistant attitudes and beliefs currently. Higher rates of RMA have been found to correlate positively with many problematic attitudes, beliefs, and behaviors (Wegner et al., 2015). Thus, reduction in RMA has been used as evidence of successful rape-prevention education programs (Hines & Palm-Reed, 2015) and lower RMA has been associated with lower self-reported rape proclivity (Malamuth 1981). This is the item I will investigate most closely, as it is possible that the established and oftenrefined Illinois Rape Myth Acceptance Scale (Payne et al., 1999) will function in the same way as the scale I intend to create. Because of this, I will take the opportunity to carefully interrogate the nuanced differences, if any, between relatively lower RMA and espousal of rape-resistant attitudes and beliefs. Recognizing rape. Many survivors of sexual violence do not recognize the experience as such, either because they do not understand what constitutes sexual violence or they blame themselves (Peterson & Muehlenhard, 2004). Similarly, many perpetrators do not recognize what they themselves did as sexual violence because they also do not understand what constitutes sexual violence. Perpetrators also often blame their victims for sending “mixed signals” causing arousal, and thus having a responsibility to provide sex (Wegner et al., 2015), for exhibiting “inappropriate” personal strength and autonomy (Moffett, 2006), and for withholding what perpetrators feel is “owed” to them for various reasons (Beech, Ward, & Fisher, 2006). Additionally, some perpetrators simply do not consider their partner’s consent but rather go forward without discussion, otherwise known as the “no engagement” strategy (Conte, 1989). Each of these facilitators of sexual violence entails the lack of recognition of sexual violence. However, despite the assumption that sexual violence is always an event perpetrated willfully with full understanding, ignorance is 65 not a legitimate defense. For these reasons, I believe the ability to recognize sexual violence, particularly rape, can be considered a rape-resistant belief. Egalitarian attitudes toward women. Egalitarian attitudes toward women are perhaps the broadest systemic contributor to the reduction of sexual violence. This assumption is based in the work of anthropologists who observe rape-free societies and have made these associations (Sanday 1981, Watson-Franke 2002) as well as current observations about relatively less risk of rape in more egalitarian cultures (Watson-Franke, 2002). This macrolevel orientation can be measured in individuals, and an item on the scale will attempt to do that. Willingness to intervene. Bystander intervention is perhaps the vein of research that most closely mirrors what I am attempting here (McMahon & Farmer 2011). Individuals who are willing to disrupt sexually predatory events are likely to hold what I hope to identify as rape-resistant attitudes and beliefs. However, most bystander intervention work focuses on an individual’s willingness to act, not necessarily to recognize a potentially predatory event and believe it to be wrong, or any of a host of other attitudes I hope to interrogate. Nevertheless, combing this literature inspired a great deal of critical thinking and provided many possible attitudes and beliefs to consider for this scale. Willingness to disrupt gendered sexual scripts. Individuals follow sexual scripts culturally, interpersonally, and intrapsychically in order to direct behavior and construct meaning (Simon & Gagnon, 1984). Sexual scripts are extremely heteronormative and gendered (Wiederman, 2005), rooted in assumptions about aggressive masculine sexuality and submissive female sexuality, and problematic beliefs about entitlement. Because of this, I believe a personal willingness to disrupt current heteronormative gender-essentialist sexual scripts may prove to be a rape-resistant attitude. With regard to queer respondents, it is 66 recognized that the formation of queer identity and the coming out process usually entails a disruption of heteronormative sexual scripts (Magruder, 1993) but this in itself is not necessarily indicative of a rape-resistant attitude. This will be considered during the development of this item, and an attempt will be made to capture queerness in this attitude. Willingness to engage in sexual violence-related education. College campuses are common sites for community disruption of rape culture and attempts to educate and empower individuals in order to reduce sexual violence. In general, these kinds of events are attended primarily by women, discussions are problematically focused on women keeping themselves safe rather than on individuals keeping themselves from becoming perpetrators, and attempts to engage men are sometimes met with hostility (Rich, Utley, Janke, & Moldoveanu, 2010). However, within this small body of literature, some men emerge who recognize the gendered nature of sexual violence, are willing to engage wholeheartedly with sexual violence prevention education, and understand their role as extending beyond simply committing to not be a perpetrator (Rich et al., 2010). While a large portion of their sample believed they were not responsible for rape culture and therefore not responsible to dismantle it, 11% expressed willingness to engage with male-directed prevention education. Many of them did so emphatically, explicitly stating their recognition of the gendered nature of sexual violence and affirming their role as male peers to disrupt it. This seems to suggest not only an understanding of the topic but a continual engagement with it, and willingness to see themselves as part of the solution without becoming defensive about the gendered nature of sexual violence. This item (or items) will attempt to capture that ongoing, willing engagement. Desire for mutual sexual satisfaction. Counseling psychologists and social workers have identified the desire in many couples to achieve mutual sexual satisfaction 67 through the development of an exchange model of sexual satisfaction in interpersonal relationships (Byers, 1999). Navigating mutual sexual satisfaction hinges on reciprocal consideration of the costs and rewards each person experiences in the sexual relationship. Within this model, sexual partners carefully consider the needs and desires of the other(s). Willingness, expectation, and desire to negotiate these highly relational aspects of sexual intimacy could conceivably be considered rape-resistant attitudes. Willingness to engage with survivor’s experience. People are often reticent to discuss their own experiences of sexual trauma for a variety of reasons including victimblaming, social shaming, and the extremely personal nature of the trauma. When survivors do come forward, they are often told to put it behind them or stop seeking attention. However, listening to rather than avoiding survivors’ voices not only empowers survivors but also encourages understanding about rape culture and victim-blaming narrative. Thus, a willingness to engage with survivors’ experiences of sexual trauma may function as a raperesistant attitude (Rich et al., 2010). Understanding necessity of rape prevention as perpetrator-centered. Many rape-prevention strategies focus on encouraging potential victims to avoid the crime rather than on potential perpetrators to not commit it. A review of 21 rape-prevention programs across the country revealed that only two were developed to influence potential perpetrators rather than potential victims (O'Donohue, Yeater, & Fanetti, 2003). However, teaching potential victims to avoid rape by following a set of protective practices not only does not work, but also does nothing to reduce rates of rape. Rather, these practices attempt to ensure that a perpetrator rapes somebody else, rather than does not rape at all. Focusing sexual violence prevention efforts on potential victims feeds victim-blaming narratives, promotes the idea that rape is an easily preventable crime, and that survivors can eliminate sexual 68 violence by never being targetable (Rich et al., 2010). A nuanced understanding that sexual violence can only be reduced by reducing the number of perpetrators could be considered meaningfully rape-resistant. Valuing consent as integral to pleasure. Obtaining consent is the key component in addressing sexual violence. Some scholars view consent as “the critical element that separates healthy sexual encounters from assault” (Klement et al., 2016). Individuals who practice Bondage-Discipline-Sado-Masochism (BDSM) often follow an extremely thorough and purposeful affirmative consent model. Because some activities during a BDSM “scene” would be considered violent if not undertaken for mutual pleasure, BDSM practitioners navigate expectations, desire, and consensual activities during prescene negotiation. It has been found that individuals who consistently participate in this subculture that values consent as integral to pleasure have lower rates of benevolent sexism, rape myth acceptance, and victim blaming beliefs than other groups (Klement et al., 2016), all of which may contribute to rape-resistant attitudes and beliefs. Knowledge about consent. Similarly to the hallmark of practicing consent in BDSM scenes, understanding what does and does not convey consent has also been identified as necessary to ethical BDSM (Klement et al., 2016). When placed alongside other work that identifies that some perpetrators did not understand when consent was not present (Gillen & Muncer, 1995), knowledge about consent could be considered a rape-resistant attitude. Unwillingness to bypass consent. A companion belief to understanding consent and viewing it as necessary for mutual sexual pleasure may be a personal unwillingness to bypass consent (Klement et al., 2016). This item will attempt to assess individual unwillingness to move forward with a sexual act when consent is ambiguous or absent. Obligation to be sexually frustrated rather than force sex. An interesting 69 justification made by many perpetrators is that of sexual arousal. Many perpetrators reported feeling that it was their victims’ obligation to satisfy them sexually once they were aroused (Wegner et al., 2015). An individual’s recognition of their obligation to be sexually frustrated rather than coerce or force sex is likely to be a rape-resistant belief. Acceptance of nontoxic masculinities. The work regarding sexual violence and toxic masculinities is extensive (Kupers, 2005). Hegemonic ideals of masculinity are imbued with problematic, toxic beliefs about men, women, sex, and aggression, and all men in any given culture are in conversation with hegemonic masculinity (Connell, 1987). However, a recent study showed that men’s willingness to engage in a nontoxic way with sexual violence prevention training increased their perceived masculinity among their peers (Carr & VanDeusen, 2004). Thus, willingness to accept, value, and perhaps espouse masculinities that are not toxic may be related to rape resistance. Disruption of sexual double standard. Essentialist beliefs about men, women and sex have led to a sexual double standard wherein men and women are viewed differently when engaging in the same behavior (Milhausen & Herold, 1999). These beliefs stem from assumptions about aggressive male sexuality and submissive female sexuality, defining what is “proper” for men and women as oppositional during sex (Hamilton & Armstrong, 2009). Willingness and ability to disrupt the sexual double standard may indicate an individual’s willingness and ability to disrupt socially conditioned narratives that couch resistance as part of normal heterosexual interaction. Careful alcohol use. Alcohol is often either blamed for vulnerability to sexual violence, particularly when a woman consumes it, or used as an excuse for sexual violence, particularly when a man consumes it (Koss & Gaines, 1993). Women are told to watch their drinks in public places, not to accept drinks they did not see made, to use the “buddy 70 system” when drinking in public, to wear special date-rape drug detecting nail polish, or not to drink at all (Testa & Livingston, 2009). However, interviews done with sexual predators reveals that alcohol is utilized specifically by perpetrators to compromise their victims, whether through lowering their inhibitions, rendering them unable to resist, or ultimately discrediting them if accusations result (McCauley, Ruggiero, Resnick, & Kilpatrick, 2010). This item will be specifically written to capture the use of alcohol with intent to lower the inhibition of another, not necessarily to govern ones own use of alcohol. Understanding of incapacitation. Incapacitation as a coercive strategy to bypass consent has long been recognized (Tyler, Hoyt, & Whitbeck, 1998). Inability to give consent as a result of incapacitation related to alcohol or drugs has long been a site of contestation in criminal cases. Establishing a “legal limit” for consent (similar to a legal blood alcohol limit for drunk driving) is so problematic it has not been done. However, awareness of incapacitation due to substance impairment could be considered a rape-resistant attitude. Careful porn consumption. Porn consumption in general is understood to be associated with sexually predatory behavior (Malamuth, 1981). Newer work has attempted to identify specific genres of porn (Ferguson & Hartley, 2009) and/or specific porn viewing habits, such as those that are intended to increase male peer support through group viewing (Sanday, 1996), that are particularly problematic with regard to espousal of rape-supportive attitudes and self-reported rape proclivity. Porn consumption is extremely common, and rates sharply increased during the 1990s and early 2000s with the availability of internet pornography. However, there was no corresponding sharp increase of sexual violence that should have followed if porn consumption alone were the critical factor, and some research points to protective effects of porn (Castleman, 2016; Kendall, 2006). Taken together, these things suggest the need to consider porn consumption specifically as it relates to sexual 71 violence. Commitment to humanizing others. Sexually objectifying others as a matter of course has been identified as a risk factor for rape proclivity (Malamuth, 1981) and a contributor to sexual violence (Scott & Tetreault, 1987). By contrast, humanizing practices are often employed as a tactic to reduce violence and stereotypes against sex workers (Oakley, 2007) and transgender individuals (Tompkins, Shields, Hillman, & White, 2015), both groups who suffer human rights violations due to objectification. Thus, a personal commitment to humanizing others may function as a rape-resistant attitude. Unwillingness to blame victims. Sexual violence occurs in particular social contexts and in the midst of enculturated attitudes of victim blaming. Victim blaming is the cultural norm, affecting not only how survivors are treated upon disclosure and during a criminal trial, but also how survivors see themselves and attribute self-blame (Grubb & Turner, 2012). However, a recognition of this common practice and an unwillingness to fall to victim blaming is likely to be found in people who are rape-resistant. Empathy. Theories of aggression often involve analysis of empathy and perspective-taking in perpetrators (Miller & Eisenberg, 1988). Because of its effectiveness, empathy training has been identified as one of the three most studied mediators of prejudice (Pettigrew & Tropp, 2008). Additionally, empathy for victims of sexual harassment has shown to decrease likelihood to harass in men (Diehl, Glaser, & Bohner, 2014). Empathy seems likely to be a rape-resistant attitude. Heterosociality. Hypersexualization of male-female relationships in a heteronormative culture relates to the assumed inability of men and women to form friendships without it becoming a sexual relationship. This has in turn led to the establishment of a heterosocial taboo in which men and women are discouraged from 72 entering into friendships with individuals of another gender, particularly if all parties are heterosexual (Calzo, 2014). Homosociality in the form of male peer support has been identified as a contributor to rape proclivity (Koss et al., 1988). In addition to this, hypersexualization of women to the point of objectification has also been found to be related to rape proclivity (Malamuth, 1981). Heterosexual women’s sexual objectification of men may function similarly. Thus, it may be possible that an individual’s willingness and ability to eschew hypersexuality and homosociality by establishing nonsexual relationships with individuals of other genders is indicative of a rape-resistant attitude. Taking responsibility for ones own actions. Interviews with sexual violence perpetrators have revealed that victim blaming among this group is extremely high (Wegner et. al. 2015). It is quite common for perpetrators to blame their victims for being “seductresses,” drinking alcohol, saying no when they really meant yes, not “relaxing and enjoying it,” wearing revealing clothing, having been sexually active in the past (either with the perpetrator or someone else) or going somewhere alone with the perpetrator. Victim blaming like this has been seen in both incarcerated (Beech et al., 2006) and nonincarcerated perpetrators (Abbey & McAuslan, 2004). Rarely did perpetrators, particularly incarcerated perpetrators, indicate self-reflection and engagement with their personal responsibility in the incidences. For these reasons, the ability to take responsibility for ones own actions and recognizing the obligation to do so may be considered a rape-resistant attitude. Low entitlement. Many individuals who hold rape-supportive attitudes do not commit sexual violence. Research has attempted to investigate the mediating factor between holding rape-supportive attitudes and having increased rape proclivity. In general, entitlement is seen as the mediating factor between these attitudes and criminal behavior (Hill & Fischer, 2001). For this reason, exploring entitlement in the context of rape-resistant 73 attitudes and beliefs is worthwhile. While likely to function differently, it will be interesting to examine the relationship between entitlement and rape-resistant attitudes and beliefs. Ultimately, only about half of the attitudes originally identified as potentially raperesistant performed well enough both conceptually and statistically to remain in the research instrument. Low rape myth acceptance was proved through the administration and statistical proving of the Illinois Rape Myth Acceptance Scale (McMahon & Farmer, 2011), and twelve of the preliminary attitudes remained in one form or another: • • • • • • • • • • • • Recognizing rape Willingness to intervene when witnessing predatory behavior Desire for mutual sexual satisfaction Understand that prevention education needs to be perpetrator-centered Valuing consent as part of pleasure Knowledge of consent Unwillingness to bypass consent Obligation to be sexually frustrated rather than coerce/force sex Discernment in porn consumption Commitment to humanizing others Unwillingness to blame victims Taking responsibility for own actions Notably, many items that are conceptually related to “knowledge of consent” performed well. The proliferation of consent-related items from both the literature and experts seems to reinforce the importance of consent within sexual violence work. Interestingly, no expert connected the idea of consent to violent stranger rapists, only to acquaintance/date rapists. Among experts, they generally spoke of consent seeking as a concept related to entitlement, objectification, egregious dating behavior, and inexcusable ignorance on the part of perpetrators. 74 Semistructured Interviews Interviewees tried to maintain their focus on rape resistance as a broad dimension that could be related to many kinds of attitudes, not simply those that predicted higher rape proclivity. This proved to be challenging, as rape proclivity is a principal concern to individuals involved with prevention, survivor support, criminal justice, and perpetrator rehabilitation. It thus became a common follow-up question to ask experts to specifically consider individuals who did not have high rape proclivity but would still have various attitudes toward sexual violence and survivors of sexual violence. When thinking beyond rape proclivity, experts generally spoke in terms of reduced likelihood to blame victims, as well as identifying concepts that could potentially be fruitful for inclusion in prevention education. They also spoke in terms of survivors who espoused RRA, and the potential that this could reduce self-blame. In the words of a former Executive Director of a women’s shelter, when asked how it might impact survivors to understand and hold RRA themselves; It would change everything. You would see changes in victims themselves, being empowered…. But if victims came into the experience and didn’t question anything that happened to them and they knew that what happened to them was rape, whether by a stranger or boyfriend or partner or spouse, it would change everything…. We’d have victims that don’t doubt themselves. First of all, they would report because they would know what happened wasn’t consensual… more people reporting means laws and law enforcement would have to change. Detectives would have to believe them, and they would not fall into the myths. They’d be strong, they’d be better…. My work experience is with victims, and when they come in and have been so traumatized 90% will blame themselves. I’m usually with them within two hours of them reporting, and they always say, ‘[Name redacted], I shouldn’t have been there. I shouldn’t have been drinking. I got in trouble. I should have known better. I shouldn’t have gone to that party. I shouldn’t have trusted. Do you think it really was rape? Because I didn’t say anything, and I wanted some of that but not all of that and he wouldn’t listen.’ And it’s not just women. I had a gay man come in who was assaulted by his partner and three other men and even he said, ‘I should have known.’ You just listen and validate… but if people who were victimized had a rape-resistant mindset already, I see it would change the whole thing. And then hopefully we would have rape-resistant people move into law enforcement, or be advocates and therapists and even judges. You wouldn’t have legislators say stupid stuff because they’d get it. We wouldn’t have to go up to Capitol Hill and spend two years advocating for people to understand you can’t have sex with your wife while 75 she’s asleep. I think this could change the whole thing. Most experts also drifted into gendered language and assumptions about male perpetrators and female survivors (mp/fs). Most informants who did this eventually caught themselves and verbally acknowledged that the emphasis on mp/fs was a problematic, albeit understandable, habit they were attempting to break. This seemed to reiterate two things; 1) the assumption of mp/fs is a challenging thing to overcome due to the huge proportion of instances in which this gender configuration is the case, and 2) even so, those who work in fields relating to sexual violence recognize the need to expand language to include instances that do not fit the normative (and indeed, normal) gender configuration. An interesting exception, one that seemed to further support both of these suppositions, was a expert who worked with adolescent male sex offenders. His language was markedly more genderinclusive than other experts. When asked why, he expressed that most of his male-identified clientele occupied the status of both survivor and perpetrator, and that many of them had been harmed by another male and had acted out against another male. The specialized, daily work of this expert did not fit the broad mp/fs emphasis, which was consequently not present in his language to the same degree. Observing the language used by my experts further cemented for me the importance of developing a measure that better fit the actual range of sexual violence, not simply the most normative scenario. DeVellis (2017) recommended as much proliferation as possible at this stage of scale development in an effort to identify the “theoretical totality” of the dimension a scale was intended to measure. Following this methodological path yielded over 300 attitudes that could conceivably considered rape-resistant. This list included attitudes verbally identified by experts during interviews as fitting the dimension in the way they understood it, as well as attitudes that emerged during coding. Because I did not want to limit the theoretical totality 76 of RRA in any way, I did not code transcripts for themes before writing statements. Rather, I wrote a statement to correspond with every conceivable RRA that emerged during my literature search or mentioned by any of my experts, then coded those. In this, related statements revealed themes that functioned as subscales underlying the concept of RRA. Ten such subscales emerged from the qualitative data in this project. Communication. For my experts, communication as an integral concept within RRA extended far beyond communication between sexual partners during sex, although that was certainly considered important. Communication encompassed things like willingness to speak about both sex and sexual violence with others, disrupting rape culture and RMA, using proper terminology for body parts and sex acts, and the importance of thoroughly understanding a partner’s nonverbal communication before assuming consent. A forensic nurse observed, A person who is rape-resistant would communicate with their partner in making sure. Whether it’s body language or they are asking questions that their partner is interested and engaging in the activity. This is versus another person who thinks, ‘This is what I want to do and I’m just assuming it’s cool, that it’s what you want to do unless you stop me. Communication between partners about and during sex included the desire for communication about sexual interests, understanding differing ways of communicating those interests, and communication about sex as integral to both healthy relationships and healthy sexuality. In general, they felt that taboos surrounding communication about sex made obtaining consent problematic, and also interfered with adequately assessing ones own perpetration. This lack of normative communication was seen as part of rape culture, as evidenced by this quote from a fraternity man; Sex is everywhere but it’s taboo to talk about sex…. If it’s taboo to talk about sex but you are supposed to be having a lot of sex if you’re a strong, macho man and if there’s no natural conversation that’s supposed to happen before sex, then it’s a lot harder to identify yourself as a rapist if you sexually assault someone. If it’s not 77 perceived that you need to get verbal consent before sex (that’s part of rape culture) then it allows, I hate this term, but ‘gray lines’ to be portrayed as normal. Sex positivity. Sex positivity and the importance of sexual pleasure emerged as one of the strongest themes from my experts. This subscale included valuing ones own healthy sexuality and the sexuality of others, recognizing that all people experience their sexuality differently, eschewing shame, recognizing the limitations of stereotypical gendered expectations during sex, valuing queerness, and emphasizing sexual pleasure for all partners, as indicated by this male advocate. My dad told me that sex was like Chinese food – it wasn’t over until you both eat your cookie. Silly analogy, but [it meant] I should be focusing on her pleasure…. You say sex is like Chinese food. You define sex and healthy relationships as a not sexually violent. This subscale recognized that sexual behaviors, desires, and orientations were broad, often queer, and completely normal. The director of a primary prevention program through the Utah Department of Health put it this way, “It’s normal and human for people to be interested in sex… we need to help people know how to be healthy and safe in those activities if they choose to engage.” My experts recognized sex as a primary drive, and felt that individuals who were rape-resistant would be comfortable with a diversity of sexual practices, in themselves or in others, as long as there was consent. Discussions around sex positivity and pleasure approached sex in a pragmatic way, indicating that individuals should be able to pursue their sexual bliss without shame as long as they were completely ethical, honest, and communicative with their partners. More than one expert indicated that perhaps the antidote to rape culture was not simply consent culture in terms of behavior, but rather pleasure culture; one expert went so far as to say, “It’s more than consent culture. It’s valuing pleasure. It’s orgasm culture.” For all my experts, an orientation toward the pleasurable aspects of sex and away from procreative emphasis, particularly within a 78 heteronormative gendered hierarchy, was envisioned as deeply rape-resistant. When asked specifically about this, a men’s antiviolence activist said; “I think it would be an attitude about their partner’s pleasure, about shared pleasure… it’s the concept of pleasure for your partner…. There is a rape-preventive component to sex positivity.” Understanding patriarchy. Unsurprisingly given the gendered nature of sexual violence, a subscale relating to gender emerged. Gendered power dynamics, assumptions about male and female sexuality, toxic masculinity, and patriarchy were recognized as contributors to rape culture, and alternative attitudes explored for their likelihood of being rape-resistant. The organizer of a men’s antiviolence activism group addressed sociosexual scripts and the normalizing of overcoming resistance as part of hegemonic masculinity; “[Rape-resistant men] would understand the nuance of what is sexually abusive behavior, what is sexually violent behavior.” More than any other subscale, these statements addressed heteronormative, binary assumptions about men and women. Rape-resistant attitudes here were considered as accepting of female sexuality, valuing nontoxic masculinities, rejecting the madonna/whore dichotomy and the sexual double standard, acceptance of queer sexuality and nonnormative behaviors in cishetero individuals, and careful attention to gendered power dynamics within heterosexual relationships. This subscale also addressed the real possibility of sexual violence for different groups, and recognized that women in particular were not hysterical for their fear and vigilance. Understanding privilege and entitlement. The differences between power addressed in the previous subscale and power addressed in this subscale was one of nuance. Statements about power related to gender and patriarchy were included there; power related to personality or status, regardless of gender, were included here. For the ED of a women’s shelter, carefully navigating entitlement applied equally to all genders and was a matter of 79 understanding interpersonal power; They would recognize that power changes the consent part. They would be aware that they do hold power, but they wouldn’t use it for the sexual part. They would know the boundary. Holding the power wouldn’t be a risk factor because they wouldn’t use that power wrong. Closely associated to that were statements about entitlement. Clearly, it could be argued that both power and entitlement are more accessible to men and masculinity in U.S. patriarchy, but the statements in this subscale were placed here due to their possibility of applying to any gender. Rape-resistant attitudes here were envisioned by my experts as not assuming sexual access in any kind of relationship, willingness to consider ones own inadvertent perpetration of sexual violence, personal responsibility to ask for consent before escalating sexual activity, understanding other people’s definitions of sexual aggression, and obligation to be sexually frustrated rather than put pressure on a partner for sex they did not want. Understanding perpetrators and survivors. This subscale related to the understanding of sexual violence in terms of perpetrators and survivors. Willingness to listen to, believe, and care for survivors, individual responsibility to keep oneself from perpetrating sexual violence, understanding that sexual violence prevention should focus on potential perpetrators rather than potential victims, and willingness to consider perpetration outside of “real rape” constructs were all considered rape-resistant. As a Sexual Assault Nurse Examiner (SANE) put it, “It goes back to making it clear with potential perpetrators that unless you get positive affirmative consent, it’s rape. I think we need to make the perpetrator responsible.” Knowledge about sexual violence. Closely related to the subscale regarding perpetrators and survivors, this subscale addressed sexual violence in terms of rape culture, RMA, and other systemic issues that influence rates of and response to sexual violence. 80 “More often than not, we are using cultural norms to commit assault, then using victim blaming to hold victims accountable. It’s classic victim blaming where perpetrators are erased.” Experts viewed RRA in terms of recognizing and rejecting rape myths, understanding the nuances of coercion and force, understanding sexuality-related microaggressions, recognizing that approaches to sexual violence prevention are politically charged, and being eager for ongoing education regarding sexual violence. “I’m on a soapbox with this…. Education is integral to sexual violence prevention.” One expert felt that teaching about sexual violence was like teaching anything related to fairness; “Because eventually you’re going to have to talk about sex but if you have a foundation and framework about equity, you can add sex to that just as easily as you can add alcohol, notions of academic honesty, distribution of work.” For him, RRA were an educational issue related to sexual equity. Self-governance. Related to accepting personal responsibility for all sexual behavior, the subscale of self-governance probed attitudes toward individual choices. These choices included understanding the influence of pornography and alcohol consumption, acknowledging media influence, impulse control, and purposeful self-restriction in terms of being careful not to push ones own or ones sexual partners’ boundaries. In the words of one male expert, I keep wondering when I realized that I could have been a perpetrator. Because if you had asked me that 10 years ago, no. Five years ago, no…. But I’ve arguably done sexually aggressive things to people… I’m still accountable and I’m getting more nuanced in my accountability…. Reflexivity is part of it. Setting and respecting boundaries in general were important in this subscale. “I think of making sure that people treat women and all vulnerable people with respect.” Consent. By far, experts talked most frequently about the valuing of consent as integral to any conception of RRA that could be identified. As consent is the point upon 81 which sexual violence hinges, this was unsurprising. “The only answer at every turn is consent…. The big one would be consent all the time. And you need to know that consent is revocable.” However, the depth and breadth with which experts discussed consent was remarkable. “A change in the activity requires a change in consent. If you’ve initiated a kiss but the other person wants to do something different, the activity has changed… consent to one thing isn’t consent to everything.” Consent was discussed both generally and specifically, from the perspective of those giving it to the perspective of those asking for it, and in conjunction with other social contexts such as relationship type and gendered expectations. In particular, consent in terms of RRA was seen as a purposeful practice that was integral to pleasure and respect. Nuanced understandings (e.g., the withdrawal of consent, understanding tonic immobility, or the “freeze” response, as lack of consent, asking for consent in ways that were ongoing, etc.) were explicated. Experts spoke about the dimension of RRA being thoroughly informed by consent as a freely given, case by case, ongoing, deliberately obtained requirement in every relationship, including long-term sexual partnerships. As an employee of UCASA put it, It’s having personal boundaries, like feeling empowered and equipped to assert those boundaries, but also respect those boundaries in someone else…. It’s beneficial to have conversations about accepting when consent is not given, because that can be a hard thing to navigate. Ambiguity was seen as antithetical to affirmative consent, which my experts viewed as being extremely clear and fully understood. The connection to pleasure was broad, ranging from the attitude that stopping to obtain consent did not “spoil the mood” to requiring affirmative consent before sexual pleasure could be enjoyed. Interestingly, this subscale had the most items in it even after I removed those that were similar to items asked in the Sexual Consent Scale (Humphreys & Brousseau, 2010) that would also be administered to research participants in this project. 82 Personal values regarding sex and sexual violence. Experts also identified a group of attitudes related to personal values about sex and sexual violence. These included personal reflection ones own language as a contributor to rape culture and commitment to disrupting “soft” sexism in ones friend group; “For men, if you were in a locker room or a fraternity and somebody [participated in slut shaming] and someone else turned to them and said, ‘That’s messed up, she’s our friend.’” Acceptance of ones own queerness in both orientation and behaviors, the cultivation of intimacy even with casual sex partners, recognition of sexual vulnerability in both self and partners, were examples of the influence of personal values on sexual behavior. Heterosociality was viewed by my experts as raperesistant, as was the assumption that every potential sexual partner would have different personal values regarding sex that should be discovered and respected. This subscale was marked by deliberate care for oneself and ones sexual partners in terms of honoring diverse value systems that informed sexual activity. Personal values in general. Surprising to me was the emergence of a subscale relating to personal values that had nothing to do with sexual activity on the surface. However, most of my experts spoke to some degree about general attitudes they believed would be rape-resistant. An attitude spoken of by a forensic nurse was body positivity; “You know another thing that… would be rape-resistant? Teaching people positive body image…. Part of it is how we promote healthy body image but also healthy self-esteem and selfworth.” These attitudes included a commitment to living within ones general value system, drawing personal strength from spiritual practices (of any kind), feeling confident, having value and personal empowerment that had nothing to do with sexual activity, not judging others, valuing diversity in ones personal circle, and respecting others even during disagreement. “That is something else here – lack of judgment. We need good judgement, 83 not judging others.” For my experts, the dimension of RRA would certainly be informed by particular kinds of attitudes and knowledge about sexuality and sexual violence, but it would also extend beyond that into openness to others and steadiness in oneself. Focus Group Once every possible RRA was identified, one or more scale item written about each, and these scale items coded into themes, a focus group consisting of interviewees was held. The intention of the focus group was to consider the themes that had emerged and to narrow the list of 326 possible statements down to an administrable list. Members of the focus group got together and went through a hard copy of all statements with instructions to talk to one another, debate the utility of the themes, the identified RRA, and the wording of the statements I had written. I asked them to star the statements they felt were vital to the dimension of RRA. Most of these individuals knew each other from community events, and the conversation was lively. In general, they all felt the statements were worded using too much jargon and specific language. They helped me rephrase many of the statements they felt I should keep. All of my experts had participated in education and awareness projects where they interfaced with the general public, and consequently had a better idea than I did about how casual the language should be to ensure accessibility. Ultimately, the focus group agreed that the themes were relevant and contributed to the understanding of RRA. The list of 326 items measuring possible RRA was narrowed down to the 108 they considered vital. Focus group participants also gave their opinions regarding the two previously developed scales I intended to administer along with mine as validity checks, the IRMA (McMahon & Farmer, 2011; Payne et al., 1999), and the Sexual Consent Scale - Revised (Humphreys & Brousseau, 2010). The result of the focus group was the final instrument, ready to be 84 submitted to the University of Utah Institutional Review Board for approval before its administration to a student sample for a pilot study. Pilot Study The purpose of the pilot study was primarily to check the wording used in writing the items, to clarify meaning, and to solicit any feedback participants wanted to offer on the items. To facilitate qualitative communication with student pilot participants, each item ended with an open-ended text box so participants could speak to anything they liked. This qualitative feedback was incredibly helpful in allowing me to eliminate confusing language, clarify concepts, and see what items elicited strong responses from participants. For example, I eliminated use of the term “trope” after multiple participants expressed they didn’t fully know what it meant. Other items were rewritten completely for clarity. Other qualitative feedback at this stage was also fascinating, as some participants chose to express the strength with which they responded to particular statements. One participant wrote, regarding an item intended to measure obligation to experience sexual frustration rather than pressure a sexual partner into sexual behavior they didn’t want, “Put this on a billboard in Times Square!!!” Another, relating to the same item, asked, “Why isn’t masturbation an option?” Feedback like this helped me refine the scale I ultimately offered to the full sample at the University of Utah and Salt Lake Community College. Summary This chapter described the qualitative results obtained through content analysis, indepth interviews with experts in the field of sexual violence, a focus group of those same experts, and the qualitative feedback obtained from text boxes in the pilot study. Twenty- 85 four RRA were identified during the content analysis and used to conceptualize rape resistance as an attitudinal dimension, and to inform the semi structured interview guide used during interviews with experts. These interviews generated 201 RRA, and I wrote 326 potential items to measure them. These items were presented to a focus group of those same experts, and they gave feedback on which items they felt must be included in the pilot study instrument, and on the wording of the items. In total, 108 items made up the portion of the research instrument testing RRA. These items were administered to a pilot study who gave qualitative feedback on wording and content. That feedback was used to refine the final instrument ultimately administered at the University of Utah and Salt Lake Community College. The quantitative results obtained from administering that survey will be discussed in the next chapter. CHAPTER 4 QUANTITATIVE RESULTS This chapter will describe the sample from which these data were drawn, then offer the results of the exploratory and confirmatory factor analysis models used to test the items administered in the quantitative portion of this project. Ten dimensions related to RRA as identified through qualitative analysis were tested, and a stand-alone research instrument with two subscales was developed. The statistical models estimated here refined the items a great deal, further clarifying the attitudinal dimension of rape resistance. Description of Participants Of the 808 individuals who opened the survey, 100% consented to the survey and began responding to items. However, I was only interested in participants who completed the entire survey, which took around 25 minutes for most participants. The final sample size for this project included 524 participants from the University of Utah. An additional 169 were recruited from Salt Lake Community College (SLCC), but unfortunately this sample was too small to make any meaningful comparisons. Consequently, the SLCC data were not included in this project. As described in Table 4.1, the sample from which these data were drawn was predominantly White (72%), female (69%), nonreligious (59%), straight (53%), under the age of 24 (71%), and heavily pursuing social science degrees (59%). In these data, there are a few 87 observations of note. First, this overwhelmingly White sample is unfortunate, but unsurprising. Utah is a predominately White state, and the University of Utah has a predominantly White student body. Statistical analyses were thus performed on a dichotomous variable of White/Persons of Color. It is unfortunate that all People of Color were combined into one category, but insufficient numbers of any racial/ethnic category other than White prohibited more detailed analysis. Second, disproportionate female participation was expected, as women generally participate in social science research at higher rates than men (Curtin, Presser, & Singer, 2000), but a sample of 363 female-identified respondents compared to 141 male-identified respondents was surprising. In addition to female- and male-identified respondents, there were 18 respondents (3%) who identified themselves as a better fit outside a gender binary. This number was too small to make meaningful inferences. Third, religious affiliation was also interesting given the unique religious character of the state of Utah. Salt Lake City is the worldwide headquarters of the Church of Jesus Christ of Latter-Day Saints, known as Mormons, and 62% of the population of the state are purported to be Mormon (Canham, 2018). And so, while neither the University of Utah nor the Mormon Church offer statistics regarding the percentage of Mormons in the student body at the University of Utah, it was assumed that many Mormons had the opportunity to participate in this research. However, the sample includes only 83 self-identified Mormons, or 16% of the sample. While this may be attributable to a number of possibilities (e.g., perhaps Mormons largely self-select into majors other than social science or are reticent to participate in research regarding human sexuality), the number was far lower than expected. Finally, a higher percentage of students than expected (43%) expressed some kind of queer sexual identity. This could be a matter of semantics, as 121 participants (23%) 88 identified as “Mostly Heterosexual” and may have been considered straight by another researcher. Notably, only 54% selected “Exclusively Heterosexual” as their sexual identity. Even when combining these two categories, only 77% of the sample expressed heterosexuality as their primary identity. The sample includes 52 (10%) participants who identified as bisexual, 22 (4%) who identified primarily as gay or lesbian, eight (1.5%) who identified as asexual, and 17 (3%) who identified as pansexual. All of these categories are higher than self-reported national averages found in the General Social Survey (Smith, Davern, Freese, & Hout, 2016) or any other nationally representative data set (e.g., Gallup, Pew Research, or the Williams Institute). Because any assumptions regarding this high percentage of queer-identified participants would be inappropriate within the scope of this project, I will make no speculations. However, this intriguing finding merits notice. Ten Dimensions Underlying Rape Resistance I first performed exploratory factor analysis (EFA) through Principal Component Factor Analysis (PCF) as described in Chapter 2 on each of the 10 subscales identified by experts. Each model was strengthened by removing a few items, as described below. Once each of the 10 models had items that loaded ≥ 0.40 onto one factor, with a sufficient KMO of ≥ 0.80 (or would round up to 0.80), and Cronbach’s alpha of ≥ 0.70, I proceeded through confirmatory factor analysis through Structural Equation Modeling (SEM) as described in Chapter 2. In SEM, ideal goodness-of-fit is achieved with Root Square Mean Error of Approximation (RMSEA) of ≤ 0.50 and Comparative Fit Index (CFI) of ≥ 0.95. A description of the psychometric properties of each model is included. Communication. Eight items related to communication as part of RRA were tested in this sample. Bivariate correlation indicated that each item correlated ≥ 0.30 with at least 89 one other item in the group, indicating concurrent validity. Any item that did not meet this benchmark was removed from the model. As shown in Table 4.2, upon running PCF five items were retained for their good statistical performance. Three were removed because they loaded poorly onto the factor (below the benchmark of ≥ 0.40) and/or negatively affected the alpha coefficient indicating internal validity. Table 4.3 shows the factor loadings of the Communication dimension which indicated that in this model, n = 520, retained items loaded onto Factor1 with an eigenvalue of 2.53 and a factor range of 0.57-0.83. These were well above statistical benchmarks for item retention. Cronbach’s alpha for this model, 0.745, indicated good internal consistency, Bartlett’s test indicated significance at p < 0.001, and KMO measure of sampling adequacy was acceptable at 0.764. Exploratory analysis indicated that this could be an appropriate model for Communication, so I proceeded to further test the model through confirmatory factor analysis with SEM. As shown in Figure 4.1, results of SEM modification indices showed that better model fit could be achieved by allowing two items to covary. Because these items were conceptually related, as shown in Table 4.4, and improvement of model fit was substantial, I allowed the covariance. When the error terms were correlated, goodnessof-fit estimations returned RMSEA of 0.000 and CFI of 1.00. Ultimately, PCF and SEM indicated that this model was statistically sound. The psychometric properties of this model indicated that the items did indeed measure the dimension of Communication as it had been conceptualized in this project, and these items could be administered in other research projects to assess participants’ attitudes related to communication as part of rape resistance. Understanding patriarchy. Ten items related to understanding patriarchy as part of RRA were tested in this sample. Bivariate correlation indicated that each item correlated ≥ 90 0.30 with at least one other item in the group, indicating concurrent validity. Any item that did not meet this benchmark was removed from the model. As shown in Table 4.5, upon running PCF, six items were retained for their good statistical performance. Four were removed because they loaded poorly onto the factor (below the benchmark of ≥ 0.40) and/or negatively affected the alpha coefficient indicating internal validity. Table 4.6 shows the factor loadings of the Understanding Patriarchy dimension, which indicated that in this model, n = 519, retained items loaded onto Factor1 with an eigenvalue of 3.31 and a factor range of 0.59-0.86. These were well above statistical benchmarks for item retention. Cronbach’s alpha for this model, 0.809, indicated good internal consistency, Bartlett’s test indicated significance at p < 0.001, and KMO measure of sampling adequacy was meritorious 0.818. Exploratory analysis indicated that this could be an appropriate model for Understanding Patriarchy, so I proceeded to further test the model through confirmatory factor analysis with SEM. Upon running SEM on all six variables in Understanding Patriarchy and estimating covariances that would improve model fit, it became clear that q6_3, “There is something distasteful about people who enjoy sex and have a lot of partners” was problematic in the model. For this reason, I have chosen to remove it from the dimension. Without q6_3, the Eigenvalue, factor loadings, p value, KMO, and alpha coefficient of the model remained sufficient. As shown in Figure 4.2, once it was removed results of SEM modification indices showed that better model fit could be achieved by allowing two items to covary. Because these items were conceptually related, as shown in Table 4.7, and improvement of model fit was substantial, I allowed the covariance. When the error terms were correlated, goodnessof-fit estimations returned RMSEA of 0.059 and CFI of 0.993. Ultimately, PCF and SEM indicated that this model was statistically sound. The 91 psychometric properties of this model indicate that the items did indeed measure the dimension of Understanding Patriarchy as it had been conceptualized in this project, and these items could be administered in other research projects to assess participants’ attitudes related to understanding patriarchy as part of rape resistance. Understanding privilege and entitlement. Eleven items related to understanding privilege and entitlement as part of RRA were tested in this sample. Bivariate correlation indicated that each item correlated ≥ 0.30 with at least one other item in the group, indicating concurrent validity. Any item that did not meet this benchmark was removed from the model. As shown in Table 4.8, upon running PCF, nine items were retained for their good statistical performance. Two were removed because they loaded poorly onto the factor (below the benchmark of ≥ 0.40) and/or negatively affected the alpha coefficient indicating internal validity. Table 4.9 shows the factor loadings of the Understanding Privilege and Entitlement dimension, which indicated that in this model, n = 512, retained items loaded onto Factor1 with an eigenvalue of 3.62 and a factor range of 0.51-0.73. These were well above statistical benchmarks for item retention. Cronbach’s alpha for this model, 0.785, indicated good internal consistency, Bartlett’s test indicated significance at p < 0.001, and KMO measure of sampling adequacy was a meritorious 0.863. Exploratory analysis indicated that this could be an appropriate model for Understanding Privilege and Entitlement, so I proceeded to further test the model through confirmatory factor analysis with SEM. As shown in Figure 4.3, results of SEM modification indices showed that better model fit could be achieved by allowing four sets of items to covary. Because these items were conceptually related, as shown in Table 4.10, and improvement of model fit was substantial, I allowed the covariance. When the error terms were correlated, goodness-of-fit estimations returned 92 RMSEA of 0.040 and CFI of 0.982. Ultimately, PCF and SEM indicated that this model was statistically sound. The psychometric properties of this model indicated that the items did indeed measure the dimension of Understanding Privilege and Entitlement as it had been conceptualized in this project, and these items could be administered in other research projects to assess participants’ attitudes related to understanding the functions of privilege and entitlement as part of rape resistance. Sex positivity. Twelve items related to sex positivity as part of RRA were tested in this sample. Bivariate correlation indicated that each item correlated ≥ 0.30 with at least one other item in the group, indicating concurrent validity. Any item that did not meet this benchmark was removed from the model. As shown in Table 4.11, upon running PCF, five items were retained for their good statistical performance. Seven were removed because they loaded poorly onto the factor (below the benchmark of ≥ 0.40) and/or negatively affected the alpha coefficient indicating internal validity. Table 4.12 shows the factor structure of the Sex Positivity dimension, which indicated that in this model, n = 517, retained items loaded onto Factor1 with an eigenvalue of 2.64 and a factor range of 0.65-0.79. These were well above statistical benchmarks for item retention. Cronbach’s alpha for this model, 0.766, indicated good internal consistency, Bartlett’s test indicated significance at p < 0.001, and KMO measure of sampling adequacy was an acceptable 0.734. Exploratory analysis indicated that this could be an appropriate model for Sex Positivity, so I proceeded to further test the model through confirmatory factor analysis with SEM. Upon running SEM on all five variables in Sex Positivity and estimating covariances that would improve model fit, it became clear that q5_1, “I understand my own sexual 93 needs,” was problematic in the model. Modification indices recommended that it covary with all but one other variable, and goodness-of-fit statistics were still unacceptable when those covariances were allowed. For this reason, I have chosen to remove it from the dimension. Without q5_1, the Eigenvalue, factor loadings, p value, KMO, and alpha coefficient of the model remained sufficient. As shown in Figure 4.4, results of SEM modification indices showed that better model fit could be achieved by allowing two items to covary. Because these are conceptually related, as shown in Table 4.13, and improvement of mode fit was substantial, I allowed the covariance. When the error terms were correlated, goodness-of-fit estimations returned RMSEA of 0.000 and CFI of 1.00. Ultimately, once q5_1 was removed, PCF and SEM indicated that this model was statistically sound. The psychometric properties of this model indicate that the items did indeed measure the dimension of Sex Positivity as it had been conceptualized in this project. The remaining four items could be administered in other research projects to assess participants’ attitudes related to sex positivity as part of rape resistance. Understanding perpetrators and survivors. Six items related to understanding perpetrators and survivors as part of RRA were tested in this sample. Bivariate correlation indicated that each item correlated ≥ 0.30 with at least one other item in the group, indicating concurrent validity. Any item that did not meet this benchmark was removed from the model. As shown in Table 4.14, upon running PCF, all items were retained for their good statistical performance. Table 4.15 shows the factor structure of the Understanding Perpetrators and Survivors dimension, which indicated that in this model, n = 517, retained items loaded onto Factor1 with an eigenvalue of 2.57 and a factor range of 0.55-0.73. These were well above 94 statistical benchmarks for item retention. Cronbach’s alpha for this model, 0.716, indicated good internal consistency, Bartlett’s test indicated significance at p < 0.001, and KMO measure of sampling adequacy was a meritorious 0.803. Exploratory analysis indicated that this could be an appropriate model for Understanding Perpetrators and Survivors, so I proceeded to further test the model through confirmatory factor analysis with SEM. As shown in Figure 4.5, results of SEM modification indices showed that better model fit could be achieved by allowing two items to covary. Because these are conceptually related, as shown in Table 4.16, and improvement of model fit was substantial, I allowed the covariance. When the error terms were correlated, goodness-of-fit estimations returned RMSEA of 0.26 and CFI of 0.993. Ultimately, PCF and SEM indicated that this model was statistically sound. The psychometric properties of this model indicate that the items did indeed measure the dimension of Understanding Perpetrators and Survivors as it had been conceptualized in this project, and these items could be administered in other research projects to assess participants’ attitudes related to understanding perpetrators and survivors as part of rape resistance. Self-governance. Eight items related to self-governance as part of RRA were tested in this sample. Bivariate correlation indicated that each item correlated ≥ 0.30 with at least one other item in the group, indicating concurrent validity. Any item that did not meet this benchmark was removed from the model. As shown in Table 4.17, upon running PCF, six items were retained for their good statistical performance. Two were removed because they loaded poorly onto the factor (below the benchmark of ≥ 0.40) and/or negatively affected the alpha coefficient indicating internal validity. Table 4.18 shows the factor structure of the Self-Governance dimension, which 95 indicated that in this model, n = 517, retained items loaded onto Factor1 with an eigenvalue of 3.49 and a factor range of 0.60-0.83. These were well above statistical benchmarks for item retention. Cronbach’s alpha for this model, 0.823, indicated good internal consistency, Bartlett’s test indicated significance at p < 0.001, and KMO measure of sampling adequacy was meritorious at 0.867. Exploratory analysis indicated that this could be an appropriate model for Self-Governance, so I proceeded to further test the model through confirmatory factor analysis with SEM. As shown in Figure 4.6, results of SEM modification indices showed that better model fit could be achieved by allowing two sets of items to covary. Because these are conceptually related, as shown in Table 4.19, and improvement of mode fit was substantial, I allowed the covariance. When the error terms were correlated, goodness-of-fit statistics returned RMSEA of 0.042 and CFI of 0.995. Ultimately, PCF and SEM indicated that this model was statistically sound. The psychometric properties of this model indicate that the items did indeed measure the dimension of Self-Governance as it had been conceptualized in this project, and these items could be administered in other research projects to assess participants’ attitudes related to self-governance as part of rape resistance. Consent. Eleven items related to consent as part of RRA were tested in this sample. Bivariate correlation indicated that each item correlated ≥ 0.30 with at least one other item in the group, indicating concurrent validity. Any item that did not meet this benchmark was removed from the model. As shown in Table 4.20, upon running PCF eight items were retained for their good statistical performance. Three were removed for poor statistical performance because they loaded poorly onto the factor below the benchmark of ≥ 0.40) and/or negatively affected the alpha coefficient indicating internal validity. Table 4.21 shows the factor structure of the Consent dimension, which indicated 96 that in this model, n = 517, retained items loaded onto Factor1 with an eigenvalue of 3.38 and a factor range of 0.53-0.73. These were well above statistical benchmarks for item retention. Cronbach’s alpha for this model, 0.778, indicated good internal consistency, Bartlett’s test indicated significance at p < 0.001, and KMO measure of sampling adequacy was meritorious at 0.837. Exploratory analysis indicated that this could be an appropriate model for Consent, so I proceeded to further test the model through confirmatory factor analysis with SEM. As shown in Figure 4.7, results of SEM modification indices showed that better model fit could be achieved by allowing three sets of items to covary. Because these are conceptually related, as shown in Table 4.22, I allowed the covariance. When the error terms were correlated, goodness-of-fit estimations returned RMSEA of 0.059 and CFI of 0.97-. Ultimately, PCF and SEM indicated that this model was statistically sound. The psychometric properties of this model indicate that the items did indeed measure the dimension of Consent as it had been conceptualized in this project, and these items could be administered in other research projects to assess participants’ attitudes related to consent as part of rape resistance. Knowledge about sexual violence. Six items related to knowledge about sexual violence as part of RRA were tested in this sample. Bivariate correlation indicated that each item correlated ≥ 0.30 with at least one other item in the group, indicating concurrent validity. Any item that did not meet this benchmark was removed from the model. As shown in Table 4.23, upon running PCF, all items were retained for their good statistical performance. Table 4.24 shows the factor structure of the dimension of Knowledge about Sexual Violence, which indicated that in this model, n = 518, retained items loaded onto Factor1 97 with an eigenvalue of 2.73 and a factor range of 0.52-0.74. These were well above statistical benchmarks for item retention. Cronbach’s alpha for this model, 0.742, indicated good internal consistency, Bartlett’s test indicated significance, at p < 0.001, and KMO measure of sampling adequacy was meritorious at 0.806. Exploratory analysis indicated that this could be an appropriate model for Knowledge about Sexual Violence, so I proceeded to further test the model through confirmatory factor analysis with SEM. As shown in Figure 4.8, results of SEM indicated that better model fit could be achieved by allowing three sets of items to covary. Because these are conceptually related, as shown in Table 4.25, I allowed the covariance. When the error terms were correlated, goodness-of-fit estimations returned RMSEA of 0.055 and CFI of 0.985. Ultimately, PCF and SEM indicated that this model was statistically sound. The psychometric properties of this model indicate that the items did indeed measure the dimension of Knowledge about Sexual Violence as it had been conceptualized in this project, and these items could be administered in other research projects to assess participants’ attitudes related to knowledge about sexual violence as part of rape resistance. Personal values – sex. Nineteen items related to personal values surrounding sex as part of RRA were tested in this sample. Bivariate correlation indicated that each item correlated ≥ 0.30 with at least one other item in the group, indicating concurrent validity. Any item that did not meet this benchmark was removed from the model. As shown in Table 4.26, upon running PCF, seven items were retained for their good statistical performance. Twelve were removed because they loaded poorly onto the factor (below the benchmark of ≥ 0.40) and/or negatively affected the alpha coefficient indicating internal validity. Table 4.27 shows the factor structure of the Personal Values – Sex dimension, which 98 indicated that in this model, n = 509, retained items loaded onto Factor1 with an eigenvalue of 3.47 and a factor range of 0.52-0.83. These were well above statistical benchmarks for item retention. Cronbach’s alpha for this model, 0.812, indicated good internal consistency, Bartlett’s test indicated significance at p < 0.001, and KMO measure of sampling adequacy was meritorious at 0.831. Exploratory analysis indicated that this could be an appropriate model for Personal Values – Sex, so I proceeded to further test the model through confirmatory factor analysis with SEM. Upon running SEM on all seven variables in Sex Positivity and estimating covariances that would improve model fit, it became clear that q13_6 “I communicate with my partner if I am not ready for sex in the relationship” was problematic in the model. Modification indices recommended that it covary with three other variables, and not all were conceptually related. For this reason, I have chosen to remove it from the dimension. Without q13_6, the Eigenvalue, factor loadings, p value, KMO, and alpha coefficient of the model remained sufficient. As shown in Figure 4.9, results of SEM indicated that better model fit could be achieved by allowing two sets of items to covary. Because these are conceptually related, as shown in Table 4.28, and improvement of model fit was significant, I allowed the covariance. When error terms were correlated, goodness-of-fit estimations returned RMSEA of 0.062 and CFI of 0.985. Ultimately, PCF and SEM indicated that this model was not statistically ideal; RMSEA should ideally be ≥ .05. However, it was statistically acceptable and the covariances made good conceptual sense. The psychometric properties of this model indicate that the items did indeed measure the dimension of Personal Values – Sex as it had been conceptualized in this project, and these items could be administered in other research projects to assess participants’ attitudes related to personal values surrounding sex as part of 99 rape resistance. Personal values – general. Fifteen items related to communication as part of RRA were tested in this sample. Bivariate correlation indicated that each item correlated ≥ 0.30 with at least one other item in the group, indicating concurrent validity. Any item that did not meet this benchmark was removed from the model. As shown in Table 4.29, upon running PCF, six items were retained for their good statistical performance. Nine were removed because they loaded poorly onto the factor (below the benchmark of ≥ 0.40) and/or negatively affected the alpha coefficient indicating internal validity. Table 4.30 shows the factor structure of the Personal Values – General dimension, which indicated that in this model, n = 514, retained items loaded onto Factor1 with an eigenvalue of 3.11 and a factor range of 0.51-0.82. These were well above statistical benchmarks for item retention. Cronbach’s alpha for this model, 0.786, indicated good internal consistency, Bartlett’s test indicated significance at - < 0.001, and KMO measure of sampling adequacy was meritorious at 0.845. Exploratory analysis indicated that this could be an appropriate model for Personal Values - General, so I proceeded to further test the model through confirmatory factor analysis with SEM. As shown in Figure 4.10, no covariance was allowed in this model. Results of SEM indicated that better model fit could be achieved by allowing a few sets of items to covary, but they were not clearly conceptually related so I was not comfortable allowing the covariance. Without covariance, the model is not ideal but still within acceptable ranges with RMSEA of 0.069 and CFI of 0.977. Ultimately, PCF and SEM indicated that this model was statistically acceptable. The psychometric properties of this model indicate that the items did indeed measure the dimension of Personal Values – General as it had been conceptualized in this project, and these items could be administered in other research projects to assess participants’ attitudes 100 related to communication as part of rape resistance. Creation of the Abbreviated Rape-Resistant Attitudes (ARRA) Scale After performing PCF and SEM using the methods described in Chapter 2, the reduction of items to those that were statistically sound yielded an index of 65 items, informed by each of the 10 dimensions identified by experts during the interviews described in Chapter 3. This index represents the first significant finding of this project. It is my contention that RRA do exist, and the concept appears to be informed by various underlying categories that could be considered alone or together. In this project, 10 dimensions within the context of RRA were identified and tested. Each dimension could be used as its own RRA-related scale or subscale, with the particular connotations of the dimension it represents. Identifying 10 dimensions with associated sets of items to test various facets of RRA was a valuable and worthwhile finding, but I still wondered whether a more easily administrable research instrument representing the range of attitudes in the dimension of rape resistance could be drawn from the data. Having used classical measurement theory with its ability to offer domain sampling, such an instrument was theoretically possible. Within domain sampling, a model meeting the needs of any theoretical application can be comprised of any combination of items from the theoretical totality of items that measure the latent variable. To test precisely that possibility, I ran principal component factor analysis on all 108 items in the dataset in an effort to ascertain their relationships to each other with no conceptual organization imposed at the outset by me as a researcher. The results were startling. Outputs indicated that 60 items loaded ≥ 0.40 onto the first factor. The few items that loaded ≥ .40 on subsequent factors were sufficiently different between factors (a 101 difference of ≥ .50 between factor loadings indicates the item is functioning on its own in both factors) and factor loadings ≥ .40 became sporadic and unrelated after Factor2.2 A list of these items and their factor loadings can be found in Appendix B. In running test models with between six to 10 items pulled at random from the list of 60, all combinations of items I tested loaded well (predominantly ≥ 0.40) onto one factor with no subsequent factor loadings that would indicate an additional theme. Exploratory factor analysis indicates that this index functions well as a domain and may be sampled at random, although the ability to think in terms of the dimensions identified in this project might largely be lost. However, researchers could generate their own dimensions based on any conceptual or theoretical criteria they wished, and use these items to test those dimensions. Encouraged by the results of EFA, I used domain sampling and my own conceptualization of RRA to choose the items I felt were most vital in include in an instrument measuring the breadth of RRA. This process yielded a set of 25 items, an ideal length for an administrable scale. Interestingly, these items were informed by all 10 dimensions in the RRA index. This scale, known hereafter as the Abbreviated Rape-resistant Attitudes (ARRA) scale, provides a snapshot of the breadth of RRA. In looking at the items selected for this scale, two subscales emerged; one with items relating to a reluctance to blame victims of sexual violence (RBV) and one with items related to self-monitoring one’s own behavior for perpetration (SMP). I performed exploratory and confirmatory factor 2 T he five items that loaded onto Factor2 ≥ .40 and differently enough from Factor1 to be considered as a separate factor are an interesting group. Taken together, they are a set of remarkably “woke” attitudes reflecting a strong understanding of gendered sexual violence and acceptance of queerness in sexual orientation and behavior. While too specific for the function I seek with the ARRA, these items are an interesting and relevant example of the possibilities of domain sampling. 102 analysis for each of these as well as the full ARRA. Results will be described below. Reluctance to blame victims. The items in this scale relate to individuals’ appropriate understanding of sexual violence itself, and thus their reluctance to blame victims. As shown in Table 4.31, twelve items were tested in this model. Bivariate correlations indicated that each item correlated ≥ 0.30 with at least one other item in the group, indicating concurrent validity. All items performed well statistically, loading onto the first factor ≥ 0.40 with one exception; the item, “Consent is determined by the person giving it, not the person asking for it” loaded onto the factor at 0.37. I chose to leave the item in the scale for two reasons; 1) this factor loading was close to the benchmark of ≥ 0.40 and often loaded above the benchmark in other models, and 2) it is integral to RRA to understand that the person who is being asked to escalate sexual activity is the one whose definition of consent matters. Perpetrators of sexual violence have justified their coercion by their overperception of consent indicators, particularly certain perpetrators (men who endorse stereotypical sex roles) with certain victims (women whom they perceive as sexually available, conceptualized by some perpetrators as women who drink alcohol, make eye contact, or initiate the date; Farris et al., 2008). Understanding this nuanced perspective would help individuals accurately assess interactions and recognize sexual violence when consent was overperceived by the perpetrator. As shown in Table 4.32, the factor structure of the Reticence to Blame Victims subscale indicated that in this model, n = 507, items loaded onto Factor1 with an eigenvalue of 3.89 and a factor range of 0.37-0.66. Cronbach’s alpha for this model, 0.782, indicated good internal consistency, Bartlett test indicated significance at p < 0.001, and KMO measure of sampling adequacy was meritorious at 0.860. Exploratory analysis indicated that this could be an appropriate subscale in the ARRA, so I proceeded to test the model through Confirmatory Factor Analysis (CFA) through Structural 103 Equation Modeling (SEM). As shown in Figure 4.11, results of SEM modification indices showed that better model fit could be achieved by allowing three sets of items to covary. Because these are theoretically related, as shown in Table 4.33, and SEM indicated that covariance would significantly improve model fit, I allowed the covariance. When error terms were correlated, goodness-of-fit estimations returned RMSEA of 0.50 and CFI of 0.944. Many other covariances were suggested through modification indices, and it would have been possible to attain a statistically ideal model. However, this model has comfortable fit and relying on statistical benchmarks over researcher intuition is not a sound method. Ultimately, PCF and SEM indicated that this model was statistically sound, and the psychometric properties were sufficient to indicate the model did indeed measure Reluctance to Blame Victims as it had been conceptualized in this project. This subscale of the ARRA is sufficient to measure individuals’ reluctance to blame victims as part of rape resistance. Self-monitoring for perpetration. The items in this scale relate to individuals selfmonitoring their behavior to ensure they do not perpetrate sexual violence. As shown in Table 4.34, thirteen items were tested in this model. Bivariate correlations indicated that each item correlated ≥ 0.30 with at least one other item in the group, indicating concurrent validity. All items performed well statistically, loading onto the first factor ≥ 0.40 As shown in Table 4.35, the factor structure of the Reticence to Blame Victims subscale indicated that in this model, n = 504, the items loaded onto Factor1 with an eigenvalue of 4.54 and a factor range of 0.47-0.67. Cronbach’s alpha for this model, 0.825, indicated good internal consistency, Bartlett test indicated significance at p < 0.001, and KMO measure of sampling adequacy was meritorious at 0.888. Exploratory analysis indicated that this could be an appropriate subscale in the ARRA, so I proceeded to test the model through Confirmatory Factor Analysis (CFA) through Structural Equation Modeling (SEM). As shown in Figure 104 4.12, results of SEM modification indices showed that better model fit could be achieved by allowing four sets of items to covary. Because these are theoretically related, as shown in Table 4.36, and SEM indicated that covariance would significantly improve model fit, I allowed the covariance. When error terms were correlated, goodness-of-fit estimations returned RMSEA of 0.054 and CFI of 0.942. Many other covariances were suggested through modification indices, and it would have been possible to attain a statistically ideal model. However, this model has comfortable fit and relying on statistical benchmarks over researcher intuition is not a sound method. Ultimately, CFA and goodness-of-fit tests indicated that this model was statistically sound, and the psychometric properties were sufficient to indicate the model did indeed measure Self-Monitoring for Perpetration as it had been conceptualized in this project. This subscale of the ARRA is sufficient to measure individuals’ commitment to monitoring their own behavior to ensure they did not commit sexual violence as part of rape resistance. The Abbreviated Rape-Resistant Attitudes scale. The preceding subscales can be administered alone, or combined into the index I have dubbed the ARRA. As shown in Table 4.37, twenty-five items were tested together in this scale. Bivariate correlations indicated that each item correlated ≥ 0.30 with at least one other item in the group, indicating concurrent validity. All items performed well statistically, loading onto the first factor ≥ 0.40 with two exceptions. 1) “Consent is determined by the person giving it, not the person asking for it” loaded onto the factor at 0.39. I chose to retain the item for the same reasons it remained in the subscale. 2) I left, “I am solely responsible for my own actions when I am sexually aroused, regardless of what anyone else has done” because its factor loading was 0.38, and because it directly challenges the rape-supportive belief that men in particular could become sexually aroused to the point they were no longer responsible for 105 their behavior. Ultimately, as shown in Table 4.38, the factor structure of the ARRA indicated that in this model, n = 494, the items loaded onto Factor1 with an eigenvalue of 7.44 and a factor range of 0.38-0.71. Cronbach’s alpha for this model, 0.884, indicated good internal consistency, Bartlett test indicated significance at p = 0.001, and KMO measure of sampling adequacy was exceptional at 0.913. Exploratory analysis indicated that this could be an appropriate subscale in the ARRA, so I proceeded to test the model through SEM. Due to the size of the model, dozens of covariances were suggested by modification indices to improve model fit. There is no theoretical limit to how many covariances are allowed, provided they meet both the criteria of a) significantly improving model fit and b) having a clear conceptual or theoretical reason to correlate the error terms. The opportunity to capitalize on chance when correlating error terms is high, so I proceeded with extreme caution in estimating additional parameters for this model. I disregarded many suggested covariances that were projected to significantly improve the model but whose conceptual connection was the slightest bit unclear. I estimated one parameter at a time, only when that parameter was clearly appropriate. As shown in Figure 4.13, better model fit could be achieved by allowing thirteen sets of items to covary. As shown in Table 4.39, these items are clearly related and correlating their error terms makes good conceptual sense. When error terms were correlated, goodness-of-fit estimations returned RMSEA of 0.50 and CFI of 0.904. Ultimately, PCF and SEM indicated that this model was statistically sound, and the psychometric properties were sufficient to indicate the model did indeed measure RapeResistant Attitudes broadly as they had been conceptualized in this project. The ARRA is sufficient to measure individuals’ RRA in various research contexts and in conjunction with other scales and methods. 106 Summary This chapter described the processes of Exploratory Factor Analysis through Principal Component Factor analysis and Confirmatory Factor Analysis through Structural Equation Modeling, estimated on 10 dimensions of RRA and a concise instrument, with its two subscales, that can be used in any research project to test various conceptualizations of RRA. Descriptions of items removed and item covariance was included for each model. Indicators of psychometric properties were included where applicable. The next chapter will describe additional statistical tests of the ARRA, including means comparisons and multiple regression models considering various demographic categories. Table 4.1 Descriptive Statistics of Demographic Variables of Survey Participants Gender Identity Race/Ethnicity Agender 4 Asian 27 Female 364 Black or African American 11 Male 140 Hispanic 33 Non-Binary 6 Latinx 5 Transgender 4 Native American 4 Not listed 1 Pacific Islander 9 Don't know 3 White or Caucasian 375 Prefer not to answer 2 Not listed 9 Don't know 4 Prefer not to answer 8 Area of Study Humanities Social science Natural science Formal science Applied science 66 310 39 29 67 Sexuality Exclusively heterosexual Mostly heterosexual Bisexual Mostly homosexual Exclusively homosexual Not sexually attracted Pansexual Queer Not listed Don't know Prefer not to answer Not listed Don't know Prefer not to answer 282 121 52 11 11 8 17 2 4 1 12 21 15 13 Age 18-24 25-34 35-44 45-54 55-64 65+ 375 98 33 14 3 1 Religious Affiliation Not religious 285 Buddhist 7 Catholic 26 Confucian 0 Hindu 0 Jewish 3 Mormon (LDS) 82 Muslim 9 Small branch Christian 0 Protestant 4 Sikh 1 107 108 Table 4.2 Items Tested for Communication I want my partner to communicate their desires to me during sex. I always talk to my partner directly about their likes, dislikes, and desires before we have sex. I make sure I understand how my partner expresses their sexual desires, intentions, and willingness before I make assumptions. I feel comfortable talking about my sexual desires with my partner. I make sure I understand my partner’s body language before I assume they are giving nonverbal consent. People who experience sexual violence do not need to feel ashamed of that violence.* When I talk about rape, I call it rape and not some other term.* It is important that my partner and I communicate about our desires during sex.* *These items were removed for poor statistical performance. Table 4.3 Factor Loadings for Communication n = 520 Eigenvalue 2.53 q4_1 I want my partner to communicate their desires to me during sex. 0.57 q4_2 I always talk to my partner directly about their likes, dislikes, and desires before we have sex. 0.77 q4_3 I make sure I understand how my partner expresses their sexual desires, intentions and willingness before I make assumptions. 0.83 q4_4 I feel comfortable talking about my sexual desires with my partner. 0.71 q4_5 I make sure I understand my partner's body language before I assume they are giving nonverbal consent. 0.64 Alpha = 0.745 KMO = 0.764 p < 0.001 109 Figure 4.1 Communication SEM Table 4.4 Communication Covariances I always talk to my partner directly about their likes, dislikes, and desires before we have sex. I make sure I understand my partner’s body language before I assume they are giving non-verbal consent. 110 Table 4.5 Items Tested for Understanding Patriarchy Women should experience consensual sexual pleasure in any way they want to. Some groups (e.g., women, gay men, trans women of color) are under real threat of sexual violence in our society. There is something distasteful about people who enjoy sex and have a lot of partners. (reverse coded) Our society privileges men and maleness. General societal messages teach men they are entitled to sex. There are more than two genders. I am more comfortable around men who have traditionally masculine roles than those who don’t. (reverse coded)* People of all genders can be equally sexual.* Women should reduce their risk of sexual violence by not having casual sex. (reverse coded)* It bothers me more when some people (e.g., women, old people, fat people) have casual sex than when other people have casual sex. (reverse coded)* *These items were removed for poor statistical performance. Table 4.6 Factor Loadings for Understanding Patriarchy n = 519 Eigenvalue 3.31 q6_1 Women should experience consensual sexual pleasure in any way they want to. 0.63 q6_2 Some groups (e.g., women, gay men, trans women of color) are under real threat of sexual violence in our society. 0.82 q6_3 There is something distasteful about people who enjoy sex and have a lot of partners. 0.59 q6_4 Our society privileges men and maleness. 0.86 q6_5 General societal messages teach men they are entitled to sex. 0.79 q6_6 There are more than two genders. 0.74 Alpha = 0.809 KMO = 0.818 p < 0.001 111 Figure 4.2 Understanding Patriarchy SEM Table 4.7 Patriarchy Covariances Our society privileges men and maleness. General societal messages teach men they are entitled to sex. 112 Table 4.8 Items Tested for Understanding Privilege and Entitlement Even if I have more power in a relationship, I still obtain consent for sex. I am solely responsible for my own actions when I am sexually aroused, regardless of what anyone else has done. I am obligated to be sexually frustrated rather than put pressure on my partner to engage in behavior they don’t want. All people have the right to experience their bodies in consensual ways that bring them pleasure. People have a right to tell me if they feel I am being sexually aggressive. It is up to me to be sure I do not commit sexual violence. I am still responsible for sexual harm I commit unknowingly. If my sexual partner expresses uncertainty, I always allow them to determine our level of sexual activity. My sexual partner does not owe me any sexual activity at any time. People with more privilege are entitled to more power in sexual relationships (reverse coded).* I feel personally criticized when people are angry about sexual violence and rape culture (reverse coded).* *These items were removed for poor statistical performance. 113 Table 4.9 Factor Loadings for Understanding Privilege and Entitlement n = 512 Eigenvalue 3.62 q7_1 Even if I have more power in a relationship, I still obtain consent for sex. 0.55 q7_2 I am solely responsible for my own actions when I am sexually aroused, regardless of what anyone else has done. 0.51 q7_3 I am obligated to be sexually frustrated rather than put pressure on my partner to engage in behavior they don’t want. 0.63 q7_4 All people have the right to experience their bodies in consensual ways that bring them pleasure. 0.63 q7_5 People have a right to tell me if they feel I am being sexually aggressive. 0.71 q7_6 It is up to me to be sure I do not commit sexual violence. 0.73 q7_7 I am still responsible for sexual harm I commit unknowingly. 0.55 q7_8 If my sexual partner expresses uncertainty, I always allow them to determine our level of sexual activity. 0.72 q7_9 My sexual partner does not owe me any sexual activity at any time. 0.65 Alpha = 0.785 KMO = 0.863 p < 0.001 114 Figure 4.3 Understanding Privilege and Entitlement SEM Table 4.10 Understanding Privilege and Entitlement Covariances If my partner expresses uncertainty, I My sexual partner does not owe me any always allow them to determine our level of sexual activity at any time. sexual activity. I am solely responsible for my own actions I am obligated to be sexually frustrated when I am sexually aroused, regardless of rather than put pressure on my partner to what anyone else has done. engage in behavior they don’t want. Even if I have more power in a If my partner expresses uncertainty, I relationship, I still obtain consent for sex. always allow them to determine our level of sexual activity. I am obligated to be sexually frustrated I am solely responsible for my own actions rather than put pressure on my partner to when I am sexually aroused, regardless of engage in behavior they don’t want. what anyone else has done. 115 Table 4.11 Items Tested for Sex Positivity I understand my own sexual needs. I am not ashamed of my sexual desires. I enjoy exploration in my sexual relationship(s). I know how my body experiences sexual pleasure. I enjoy fantasizing about what I want to experience sexually. Adults have the right to be as sexually active/inactive as they want to be.* Orgasm is the most important form of sexual pleasure.* It is vital that I understand my partner’s sexual needs.* If my partner does not experience pleasure during sex, I do not experience pleasure.* My partner’s sexual pleasure is more important to me than my own sexual pleasure.* Solo masturbation is a normal part of healthy sexuality, even if someone has a sexual partner.* A person’s sexuality is an important part of who they are.* *These items were removed for poor statistical performance. Table 4.12 Factor Loadings for Sex Positivity n = 517 Eigenvalue 2.64 q5_1 I understand my own sexual needs. q5_2 I am not ashamed of my sexual desires. 0.72 0.72 q5_3 I enjoy exploration in my sexual relationship(s). 0.75 q5_4 I know how my body experiences sexual pleasure. 0.79 q5_5 I enjoy fantasizing about what I want to experience sexually. 0.65 Alpha = 0.766 KMO = 0.734 p < 0.001 116 Figure 4.4 Sex Positivity SEM Table 4.13 Sex Positivity Covariances I am not ashamed of my sexual needs. I enjoy fantasizing about what I want to experience sexually. 117 Table 4.14 Items Tested for Understanding Perpetrators and Survivors Victims of nonconsensual sexual violence are never at fault for that sexual violence Everyone needs to learn how to keep themselves from perpetrating sexual violence The best way to eliminate sexual violence is to address perpetration of sexual violence. Sexual violence prevention programs that teach people how to keep themselves safe do not sufficiently address the problem of sexual violence If a sexual partner says I coerced them into sex, I am obligated to consider the possibility that I did. If people do not understand consent, they can inadvertently become perpetrators of sexual violence. Table 4.15 Factor Loadings for Understanding Perpetrators and Survivors n = 517 q9_1 Victims of nonconsensual sexual violence are never at fault for that sexual violence. q9_2 Everyone needs to learn how to keep themselves from perpetrating sexual violence. Eigenvalue 2.57 0.59 0.73 q9_3 The best way to eliminate sexual violence is to address perpetration of sexual violence. 0.65 q9_4 Sexual violence prevention programs that teach people how to keep themselves safe do not sufficiently address the problem of sexual violence 0.55 q9_5 If a sexual partner says I coerced them into sex, I am obligated to consider the possibility that I did. 0.66 q9_6 If people do not understand consent, they can inadvertently become perpetrators of sexual violence. 0.72 Alpha = 0.716 KMO = 0.803 p < 0.001 118 Figure 4.5 Understanding Perpetrators and Survivors SEM Table 4.16 Understanding Perpetrators and Survivors Covariances Everyone needs to learn how to keep Sexual violence prevention programs that themselves from perpetrating sexual teach people how to keep themselves safe violence. do not sufficiently address the problem of sexual violence. 119 Table 4.17 Items Tested for Self-Governance It is wrong to initiate sex with a person, even a long-term partner, if they are too drunk/high to give consent. If my sexual partner shows nonverbal discomfort or dismay during sexual activity, I cease the activity. If a person changes their mind during any sexual activity, their partner should immediately halt that activity. It is wrong to push my partner’s sexual boundaries when they are uncomfortable. Sexual boundaries must be discussed when discussing sexual exploration. It is important that I understand my own sexual boundaries. I am often bothered by sexism I see in the media.* My sexual partner(s) are likely to have different sexual boundaries than I have.* *These items were removed for poor statistical performance. Table 4.18 Factor Loadings for Self-Governance n = 517 Eigenvalue 3.49 q10_1 It is wrong to initiate sex with a person, even a long-term partner, if they are too drunk/high to give consent. 0.6 q10_2 If my sexual partner shows nonverbal discomfort or dismay during sexual activity, I cease the activity. 0.75 q10_3 If a person changes their mind during any sexual activity, their partner should immediately halt that activity. 0.83 q10_4 It is wrong to push my partner’s sexual boundaries when they are uncomfortable. 0.76 q10_5 Sexual boundaries must be discussed when discussing sexual exploration. 0.81 q10_6 It is important that I understand my own sexual boundaries. 0.79 Alpha = 0.828 KMO = 0.867 p < 0.001 120 Figure 4.6 Self-Governance SEM Table 4.19 Self-Governance Covariances If my sexual partner shows nonverbal discomfort or dismay during sexual activity, I cease the activity. Sexual boundaries must be discussed when discussing sexual exploration. It is important that I understand my own sexual boundaries. It is important that I understand my own sexual boundaries. 121 Table 4.20 Items Tested for Consent Any penetration without consent is rape. If a person wants to have sex but decides against it and is then forced by their partner, this is rape. Consent has been withdrawn if my partner agrees to a sexual activity and then changes their mind. I need to obtain consent each time, even from people I’ve had sex with before. A person can want to have sex and still not consent to have sex. Pausing sexual activity to gain consent is not an obstacle to sexual pleasure. I am not willing to intensify sexual activity until I have obtained consent. I am able to enjoy sex more after I have obtained consent. If my sexual partner does not stop me from some sexual activity, I can assume they have given consent.* If someone does not want to have sex with me, they have rejected me as a person.* Consent is determined by the person giving it, not the person asking for it.* *These items were removed for poor statistical performance. Table 4.21 Factor Loadings for Consent n = 517 q11_1 Any penetration without consent is rape. Eigenvalue 3.38 0.68 q11_2 If a person wants to have sex but decides against it and is then forced by their partner, this is rape. 0.73 q11_3 Consent has been withdrawn if my partner agrees to a sexual activity and then changes their mind. 0.53 q11_4 I need to obtain consent each time, even from people I’ve had sex with before. 0.73 q11_5 A person can want to have sex and still not consent to have sex. 0.56 q11_6 Pausing sexual activity to gain consent is not an obstacle to sexual pleasure. 0.53 q11_7 I am not willing to intensify sexual activity until I have obtained consent. 0.69 q11_8 I am able to enjoy sex more after I have obtained consent. 0.71 Alpha = 0.778 KMO = 0.837 p < 0.001 122 Figure 4.7 Consent SEM Table 4.22 Consent Covariances Any penetration without consent is rape. If a person wants to have sex but decides against it and is then forced by their partner, this is rape. I am not willing to intensify sexual activity until I have obtained consent. If a person wants to have sex but decides against it and is then forced by their partner, this is rape. Consent has been withdrawn if my partner agrees to a sexual activity and then changes their mind. I am able to enjoy sex more after I have obtained consent. 123 Table 4.23 Items Tested for Knowledge About Sexual Violence Society has well-established beliefs about sexual violence that are often not true. It is everyone’s responsibility to eliminate rape culture. People can commit sexual violence without knowing it. Sexual violence is committed if someone does not obtain consent for sex, even if there is no force. There is no situation that justifies nonconsensual sexual violence. If someone disclosed sexual violence to me, I would believe they were telling the truth. Table 4.24 Factor Loadings for Knowledge About Sexual Violence n = 518 q8_1 Society has well-established beliefs about sexual violence that are often not true. Eigenvalue 2.73 0.64 q8_2 It is everyone’s responsibility to eliminate rape culture. 0.67 q8_3 People can commit sexual violence without knowing it. 0.52 q8_4 Sexual violence is committed if someone does not obtain consent for sex, even if there is no force. 0.70 q8_5 There is no situation that justifies nonconsensual sexual violence. 0.74 q8_6 If someone disclosed sexual violence to me, I would believe they were telling the truth. 0.74 Alpha = 0.742 KMO = 0.806 p < 0.001 124 Figure 4.8 Knowledge About Sexual Violence SEM Table 4.25 Knowledge About Sexual Violence Covariances There is no situation that justifies If someone disclosed sexual violence to me, nonconsensual sexual violence. I would believe they were telling the truth. Sexual violence is committed if someone People can commit sexual violence without does not obtain consent for sex, even if knowing it. there is no force. Society has well-established beliefs about There is no situation that justifies sexual violence that are often not true. nonconsensual sexual violence. 125 Table 4.26 Items Tested for Personal Values – Sex Consenting adults have a right to have sex in any kinds of partnerships they choose (e.g., polyamorous, dominant/subordinate). I enjoy the emotional intimacy I experience in sexual relationships. I enjoy the emotional intimacy I experience in nonsexual relationships. I will not have sex in a new relationship until I’m ready. All partners must be in agreement before sex is introduced into the relationship. I communicate with my partner if I am not ready for sex in the relationship. I am careful with the vulnerabilities my sexual partner shows during sex (e.g., body image concerns, sexual performance, ability or desire to orgasm). I have used alcohol to help overcome my sexual partner’s resistance.* I have persisted to ask about a sexual activity after my partner has declined multiple times.* I have threatened to end a relationship if my partner did not have sex the way I wanted to.* I think about how I contribute to rape culture through my language and behavior.* I do not allow people around me to use language that makes light of sexual violence.* I would not feel ashamed if I felt attracted to someone of my same gender.* How other people choose to have consensual sex is none of my business.* My value system, religious or not, influences my sexual activity.* People often have different needs for intimacy than their sexual partner.* I enjoy connecting with my partner, even if the sex is casual.* It is not a good idea for people in committed relationships to be friends with people of the gender they are sexually attracted to.* I am aware of my own sexual vulnerabilities.* *These items were removed for poor statistical performance. 126 Table 4.27 Factor Loadings for Personal Values - Sex n = 509 Eigenvalue 3.47 q13_1 Consenting adults have a right to have sex in any kinds of partnerships they choose (e.g., polyamorous, dominant/subordinate). 0.52 q13_2 I enjoy the emotional intimacy I experience in sexual relationships. 0.74 q13_3 I enjoy the emotional intimacy I experience in nonsexual relationships. 0.70 q13_4 I will not have sex in a new relationship until I’m ready. 0.71 q13_5 All partners must be in agreement before sex is introduced into the relationship. 0.83 q13_6 I communicate with my partner if I am not ready for sex in the relationship. 0.67 q13_7 I am careful with the vulnerabilities my sexual partner shows during sex (e.g., body image concerns, sexual performance, ability or desire to orgasm). 0.73 Alpha = 0.812 KMO = 0.831 p < 0.001 127 Figure 4.9 Personal Values – Sex SEM Table 4.28 Personal Values – Sex Covariances I enjoy the emotional intimacy I experience in sexual relationships. I will not have sex in a new relationship until I’m ready. I enjoy the emotional intimacy I experience in nonsexual relationships. I am careful with the vulnerabilities my sexual partner shows during sex (e.g., body image concerns, sexual performance, ability or desire to orgasm). 128 Table 4.29 Items Tested for Personal Values – General I am a valuable person whether I am sexually active or not. I can learn a lot from people who are different from me. It doesn’t bother me if adults experience consensual sexual pleasure in things I find distasteful. It is up to me to learn how to nurture my relationships, even nonsexual relationships, in ways my partners need. We all have an obligation to interrupt predatory behavior when we see it. Bodies of all shapes, sizes, and abilities have a right to experience sexual pleasure. I feel strength when I live within my value system.* My spirituality is important to me.* I feel empowered when I am having the kind of sex I want to have.* All people have the same amount of worth.* All people deserve the same amount of respect.* I consciously resist stereotypes when I meet someone different from me.* People generally know what is best for them.* I generally feel a sense of my own intrinsic worth.* I am more interested in how bodies experience pleasure than in how they look.* *These items were removed for poor statistical performance. Table 4.30 Factor Loadings for Personal Values - General n = 514 q14_3 I am a valuable person whether I am sexually active or not. Eigenvalue 3.11 0.61 q14_8 I can learn a lot from people who are different from me. 0.80 q14_10 It doesn’t bother me if adults experience consensual sexual pleasure in things I find distasteful. 0.51 q14_12 It is up to me to learn how to nurture my relationships, even nonsexual relationships, in ways my partners need. 0.76 q14_13 We all have an obligation to interrupt predatory behavior when we see it. 0.77 q14_14 Bodies of all shapes, sizes, and abilities have a right to experience sexual pleasure. 0.82 Alpha = 0.786 KMO = 0.845 p < 0.001 129 Figure 4.10 Personal Values – General SEM 130 Table 4.31 Items Tested for Reluctance to Blame Victims People can commit sexual violence without knowing it. Any penetration without consent is rape. Sexual violence is committed if someone does not obtain consent for sex, even if there is no force. A person can want to have sex and still not consent to have sex. Some groups (e.g., women, gay men, trans women of color) are under real threat of sexual violence in our society. Society has well-established beliefs about sexual violence that are often not true. The best way to eliminate sexual violence is to address perpetration of sexual violence. If a person wants to have sex but decides against it and is then forced by their partner, this is rape. People of all genders can be equally sexual. Consent is determined by the person giving it, not the person asking for it. We all have an obligation to interrupt predatory behavior when we see it. Adults have the right to be as sexually active/inactive as they want to be. 131 Table 4.32 Factor Loadings for Reluctance to Blame Victims n = 507 Eigenvalue 3.89 rbv_1 People can commit sexual violence without knowing it. 0.48 rbv_2 Any penetration without consent is rape. 0.64 rbv_3 Sexual violence is committed if someone does not obtain consent for sex, even if there is no force. 0.66 rbv_4 A person can want to have sex and still not consent to have sex. 0.55 rbv_5 Some groups (e.g., women, gay men, trans women of color) are under real threat of sexual violence in our society. 0.61 rbv_6 Society has well-established beliefs about sexual violence that are often not true. 0.57 rbv_7 The best way to eliminate sexual violence is to address perpetration of sexual violence. 0.52 rbv_8 If a person wants to have sex but decides against it and is then forced by their partner, this is rape. 0.70 rbv_ 9 People of all genders can be equally sexual. 0.61 rbv_10 Consent is determined by the person giving it, not the person asking for it. 0.37 rbv_11 We all have an obligation to interrupt predatory behavior when we see it. 0.53 rbv_12 Adults have the right to be as sexually active/inactive as they want to be 0.51 Alpha = 0.782 KMO = 0.860 p < 0.001 132 Figure 4.11 Reluctance to Blame Victims SEM Table 4.33 Reluctance to Blame Victims Covariances Any penetration without consent is rape. Any penetration without consent is rape. People of all genders can be equally sexual. If a person wants to have sex but decides against it and is then forced by their partner, this is rape. Sexual violence is committed if someone does not obtain consent for sex, even if there is no force. Adults have the right to be as sexually active/inactive as they want to be. 133 Table 4.34 Items Tested for Self-Monitoring for Perpetration I am not owed any sexual activity by any partner. Consent has been withdrawn if my partner agrees to a sexual activity and then changes their mind. If a sexual partner says I coerced them into sex, I am obligated to consider the possibility that I did. It is wrong to push my partner’s sexual boundaries when they are uncomfortable. It is important that my partner and I communicate about our desires during sex. I am able to enjoy sex more after I have obtained consent. It is up to me to be sure I do not commit sexual violence. I am responsible for sexual harm I commit unknowingly. I am obligated to be sexually frustrated rather than put pressure on my partner to engage in behavior they don’t want. I am solely responsible for my own actions when I am sexually aroused, regardless of what anyone else has done. Everyone needs to learn how to keep themselves from perpetrating sexual violence. Sexual boundaries must be discussed when discussing sexual exploration. People have a right to tell me if they feel I am being sexually aggressive. 134 Table 4.35 Factor Loadings for Self-Monitoring for Perpetration n = 504 Eigenvalue 4.54 smp_1 I am not owed any sexual activity by any partner. 0.63 smp_2 Consent has been withdrawn if my partner agrees to a sexual activity and then changes their mind. 0.54 smp_3 If a sexual partner says I coerced them into sex, I am obligated to consider the possibility that I did. 0.58 smp_4 It is wrong to push my partner’s sexual boundaries when they are uncomfortable. 0.63 smp_5 It is important that my partner and I communicate about our desires during sex. 0.48 smp_6 I am able to enjoy sex more after I have obtained consent. 0.57 smp_7 It is up to me to be sure I do not commit sexual violence. 0.67 smp_8 I am responsible for sexual harm I commit unknowingly. 0.54 smp_ 9 I am obligated to be sexually frustrated rather than put pressure on my partner to engage in behavior they don’t want. 0.58 smp_10 I am solely responsible for my own actions when I am sexually aroused, regardless of what anyone else has done. 0.47 smp_11 Everyone needs to learn how to keep themselves from perpetrating sexual violence. 0.60 smp_12 Sexual boundaries must be discussed when discussing sexual exploration. 0.7 smp_13 People have a right to tell me if they feel I am being sexually aggressive. 0.63 Alpha = 0.825 KMO = 0.888 p < 0.001 135 Figure 4.12 Self-Monitoring for Perpetration SEM Table 4.36 Self-Monitoring for Perpetration Covariances It is wrong to push my partner’s sexual boundaries when they are uncomfortable. It is up to me to be sure I do not commit sexual violence. I am obligated to be sexually frustrated rather than put pressure on my partner to engage in behavior they don’t want. It is up to me to be sure I do not commit sexual violence. Sexual boundaries must be discussed when discussing sexual exploration. People have a right to tell me if they feel I am being sexually aggressive. I am solely responsible for my own actions when I am sexually aroused, regardless of what anyone else has done. I am responsible for sexual harm I commit unknowingly. 136 Table 4.37 Items in the Abbreviated Rape-Resistant Attitudes Scale People can commit sexual violence without knowing it. Any penetration without consent is rape. Sexual violence is committed if someone does not obtain consent for sex, even if there is no force. A person can want to have sex and still not consent to have sex. Some groups (e.g., women, gay men, trans women of color) are under real threat of sexual violence in our society. Society has well-established beliefs about sexual violence that are often not true. The best way to eliminate sexual violence is to address perpetration of sexual violence. If a person wants to have sex but decides against it and is then forced by their partner, this is rape. People of all genders can be equally sexual. Consent is determined by the person giving it, not the person asking for it. We all have an obligation to interrupt predatory behavior when we see it. Adults have the right to be as sexually active/inactive as they want to be. I am not owed any sexual activity by any partner. Consent has been withdrawn if my partner agrees to a sexual activity and then changes their mind. If a sexual partner says I coerced them into sex, I am obligated to consider the possibility that I did. It is wrong to push my partner’s sexual boundaries when they are uncomfortable. It is important that my partner and I communicate about our desires during sex. I am able to enjoy sex more after I have obtained consent. It is up to me to be sure I do not commit sexual violence. I am responsible for sexual harm I commit unknowingly. I am obligated to be sexually frustrated rather than put pressure on my partner to engage in behavior they don’t want. I am solely responsible for my own actions when I am sexually aroused, regardless of what anyone else has done. Everyone needs to learn how to keep themselves from perpetrating sexual violence. Sexual boundaries must be discussed when discussing sexual exploration. People have a right to tell me if they feel I am being sexually aggressive. 137 Table 4.38 Factor Loadings for Abbreviated Rape-Resistant Attitudes n = 494 Eigenvalue 7.44 rbv_1 People can commit sexual violence without knowing it. 0.40 rbv_2 Any penetration without consent is rape. 1 rbv_3 Sexual violence is committed if someone does not obtain consent for sex, even if there is no force. 0.60 0.56 rbv_4 A person can want to have sex and still not consent to have sex. 0.50 rbv_5 Some groups (e.g., women, gay men, trans women of color) are under real threat of sexual violence in our society. 0.52 rbv_6 Society has well-established beliefs about sexual violence that are often not true. 0.47 rbv_7 The best way to eliminate sexual violence is to address perpetration of sexual violence. 0.49 rbv_8 If a person wants to have sex but decides against it and is then forced by their partner, this is rape. 0.71 rbv_ 9 People of all genders can be equally sexual. 0.60 rbv_10 Consent is determined by the person giving it, not the person asking for it. 0.39 rbv_11 We all have an obligation to interrupt predatory behavior when we see it. 0.53 rbv_12 Adults have the right to be as sexually active/inactive as they want to be. 0.48 smp_1 I am not owed any sexual activity by any partner. 0.62 smp_2 Consent has been withdrawn if my partner agrees to a sexual activity and then changes their mind. 0.53 smp_3 If a sexual partner says I coerced them into sex, I am obligated to consider the possibility that I did. 0.56 138 Table 4.38 continued Factor Loadings for Abbreviated Rape-Resistant Attitudes n = 494 Eigenvalue 7.44 smp_4 It is wrong to push my partner’s sexual boundaries when they are uncomfortable. 0.64 smp_5 It is important that my partner and I communicate about our desires during sex. 0.45 smp_6 I am able to enjoy sex more after I have obtained consent. 0.58 smp_7 It is up to me to be sure I do not commit sexual violence. 0.60 smp_8 I am responsible for sexual harm I commit unknowingly. 0.54 smp_ 9 I am obligated to be sexually frustrated rather than put pressure on my partner to engage in behavior they don’t want. 0.52 smp_10 I am solely responsible for my own actions when I am sexually aroused, regardless of what anyone else has done. 0.38 smp_11 Everyone needs to learn how to keep themselves from perpetrating sexual violence. 0.60 smp_12 Sexual boundaries must be discussed when discussing sexual exploration. 0.68 smp_13 People have a right to tell me if they feel I am being sexually aggressive. Alpha = 0.884 KMO = 0.913 p < 0.001 0.54 139 Figure 4.13 Abbreviated Rape-Resistant Attitudes SEM 140 Table 4.39 Abbreviated Rape-Resistant Attitudes Covariances It is up to me to be sure I do not commit People have a right to tell me if they feel I sexual violence. am being sexually aggressive. It is wrong to push my partner’s sexual Sexual boundaries must be discussed when boundaries when they are uncomfortable. discussing sexual exploration. Any penetration without consent is rape. If a person wants to have sex but decides against it and is then forced by their partner, this is rape. It is up to me to be sure I do not commit I am responsible for sexual harm I commit sexual violence. unknowingly. The best way to eliminate sexual violence is Everyone needs to learn how to keep to address perpetration of sexual violence. themselves from perpetrating sexual violence. I am obligated to be sexually frustrated I am solely responsible for my own actions rather than put pressure on my partner to when I am sexually aroused, regardless of engage in behavior they don’t want. what anyone else has done. Consent is determined by the person giving I am able to enjoy sex more after I have it, not the person asking for it. obtained consent. People can commit sexual violence without Sexual violence is committed if someone knowing it. does not obtain consent for sex, even if there is no force. Any penetration without consent is rape. Sexual violence is committed if someone does not obtain consent for sex, even if there is no force. If a person wants to have sex but decides Consent has been withdrawn if my partner against it and is then forced by their agrees to a sexual activity and then changes partner, this is rape. their mind. It is important that my partner and I Sexual boundaries must be discussed when communicate about our desires during sex. discussing sexual exploration. People of all genders can be equally sexual. Adults have the right to be as sexually active/inactive as they want to be I am not owed any sexual activity by any If a person wants to have sex but decides partner. against it and is then forced by their partner, this is rape. CHAPTER 5 REGRESSION ANALYSES In this chapter, I will 1) explore how demographic characteristics like gender identity, sexual identity, race/ethnicity, religious affiliation, and age help us understand variation in the ARRA and 2) compare the ARRA and its subscales to two established measures of sexual violence to assess external construct validity. I chose the Illinois Rape Myth Acceptance Scale (Payne et al., 1999) and the Sexual Consent Scale – Revised (Humphreys & Brousseau, 2010) for the reasons described in Chapter 2. The creation of the ARRA and its subscales is discussed in detail in Chapter 4. This chapter will describe the demographic predictors of RRA found in this sample, which also serves to demonstrate the implementation of the ARRA in other research projects. I began with means comparisons through t tests to look for significant differences between demographic groups of interest, then estimated multivariate regression models for all demographic variables. Demographic characteristics of interest were participants’ gender identity, sexual identity, racial/ethnic identity, religious affiliation, and age. Because I was interested in the outcomes for each of these categories with regard to both reluctance to blame victims (RBV) and self-monitoring for perpetration (SMP), I chose to estimate models for each of the two subscales in the ARRA separately, with the scaled variable of each subscale treated as the dependent variable. I also estimated the full ARRA. 142 Mean Differences To estimate mean differences, I estimate t tests on the full sample using binary variables purposefully created to appropriately reflect research participants’ gender identities, as described in Chapter 2. Similarly, I created dichotomous variables for participants who were either exclusively heterosexual or expressed some kind of queer identity, participants who were either religiously affiliated or had no religious affiliation, those who were aged 34 and under and those 35 and older, and participants who were either White or nonWhite. In each case, if a participant did not know or chose not to answer they were excluded from the t tests, with the exception that participants who answered “I do not know” regarding their sexual identity were considered queer, as questioning ones sexuality is a queer identity. Table 5.1 displays the mean differences in the SMP and RBV subscales along with those in the ARRA, categorized by gender identity, sexual identity, religiosity, and race/ethnicity. A higher mean on the SMP indicates more willingness to self-monitor for perpetration. A higher mean on the RBV indicates more reluctance to blame victims of sexual violence for their experiences. A higher mean on the ARRA indicates more RRA in general, in all the ways it has been conceptualized in this project. For both subscales and the full ARRA, the mean for female-identified participants were significantly higher than the mean for male-identified participants. Similarly, participants who identified with a queer identity had significantly higher means on both subscales and the ARRA than those who indicated they were “exclusively heterosexual.” The same is true for participants whose religious identity indicated they were “not religious;” these participants had significantly higher means on all scales than those participants who identified as having a particular religious affiliation. For White-identified participants, only the mean indicating SMP was significantly higher. The means indicating RBV had no significant difference. 143 Multivariate Regression After estimating these mean differences, I estimated regression models controlling for these demographic variables in both the RBV and the SMP, as well as the full ARRA. For the regression models, I used the dichotomous variables I created to measure mean differences, with the exception of gender identity; for the regression models, I used all categories of gender including agender, transgender, genderqueer, and nonbinary. The remaining variables captured the full sample in the dichotomous variable and fulfilled the necessary function. I was not interested in looking at religions individually, and there were so few people of color in each nonWhite category that comparisons were insignificant. Here again, if a participant selected “I do not know” or “Prefer not to answer” on any demographic variable, they were excluded from that variable with the exception of sexual identity where only those who preferred not to answer were excluded. Table 5.2 displays the results of regression models using the SMP, RBV, and the ARRA as dependent variables. For the SMP model, participants who identified as female were significantly more associated with willingness and belief in self-monitoring for perpetration than participants who identified as male. There was no significant difference between transgender/nonbinary individuals and females in this model, likely because of the small size of the nonbinary group (n = 20). There was a significant negative relationship between being straight and SMP. Participants who indicated they were not religious had a significant positive relationship with SMP, as did participants who were White. Age was not statistically significant. All indications in bivariate tests of mean differences held in the multivariate regression context. In general, it can be said that net of all demographic factors considered here, being female, queer, White, and/or of no religion was associated with more willingness and belief in the to self-monitor ones own behavior for perpetration. 144 The regression model using RBV as the dependent variable. In general, this model follows both the bivariate tests of mean differences between demographic groups and predictions from the regression model estimating self-monitoring. Being female, queer, and not religious was significantly associated with more RBV. However, age was significant in this model, with participants 34 and younger significantly more reluctant to blame victims. Race/ethnicity was insignificant in this model. In the regression model using the full ARRA as the dependent variable, all variables that had been insignificant in one of the bivariate tests of mean differences was insignificant here. Net of all demographic factors considered here, being female, queer, and not religious were significantly associated with more RRA. Age and race/ethnicity were insignificant in this model. Likely due to its small sample (n = 18) the nonbinary gender category was insignificant in all regression models, indicating there was no significant difference from the reference category of female. Correlations With Established Scales In addition to estimating mean differences and predictive statistics of various demographic variables, I was also interested in comparing results of the ARRA with the results of both the IRMA and SCSR as a measure of external validity. If everything were theoretically sound, capturing attitudes toward sexual violence and consent in ways that made sense and functioned as expected, scores on each subscale and the full ARRA would be high when scores on the IRMA were low or vice versa, and scores on each subscale and the full ARRA would be similarly high or low to the SCSR should be similarly high or low. Acceptance of rape myths, measured by the IRMA, would be antithetical to RRA. Nuanced beliefs about consent, measured by the SCSR, would be in keeping with RRA as it has been 145 conceptualized. As shown in Table 5.3, this indeed held true with this sample. Correlations between each scale and the IRMA were indeed negative, and positive between each scale and the SCSR. In general, directions and associations in these statistical tests were unsurprising, and in keeping with generally accepted understandings regarding engagement with sexual violence prevention and victim blaming. As discussed in Chapter 1, men are more likely to hold RSA and RMA than women, and queer people tend to hold fewer. There are no generally accepted, widespread observations regarding RSA and RMA and their associations with race/ethnicity, so the irregularity of significance here is in keeping with the various observations in specific studies that have not culminated in consensus. The insignificance of age was surprising, however, as older individuals tend to hold more RMA in particular than younger individuals. This held true in only one model tested here, and not in the full ARRA. This may lend support to generational observations regarding RSA and RMA in the decades after the Women’ Liberation movement. As the first generation raised alongside cultural feminism, Gen X may be bringing their more enlightened attitudes toward sexual violence with them as they age. In the next chapter, I will discuss the findings of this dissertation in more depth, including usage notes, general observations, limitations, and future directions of research. Table 5.1 Mean Differences: ARRA and Subscales Gender Identity (n = 504) SMP (t = -4.01**) n M SD Female 364 6.53 0.50 Male 140 6.31 0.68 RBV (t = -6.25**) n M SD 364 6.39 0.504 140 6.03 0.747 ARRA (t = -5.55**) n M SD 364 6.46 0.46 140 6.18 0.67 Sexual Identity (n = 508) SMP (t = 5.02**) n M SD Straight 282 6.37 0.63 Queer 226 6.61 0.41 RBV (t = 6.67**) n M SD 282 6.13 0.67 226 6.48 0.46 ARRA (t = 6.33**) n M SD 282 6.26 0.602 226 6.55 0.602 Religiosity (n = 495) SMP (t = -4.85**) n M SD Not Religious 309 6.57 0.50 Religious 186 6.33 0.62 RBV (t = -4.92**) n M SD 309 6.40 0.58 186 6.13 0.61 ARRA (t = -5.31**) n M SD 309 6.50 0.50 186 6.23 0.56 Race/Ethnicity (n = 512) SMP (t = -2.89*) n M SD White 376 6.53 0.50 POC 136 6.38 0.65 RBV (t = -0.95) n M 376 6.33 136 6.27 ARRA (t = -2.08*) n M SD 376 6.44 0.50 136 6.33 0.58 SD 0.60 0.60 Age (n = 524) SMP (t = -0.48) RBV (t = 0.73) ARRA (t = 0.14) n M SD n M SD n M SD 18-34 473 6.49 0.54 473 6.40 0.60 473 6.40 0.53 35+ 51 6.44 0.71 51 6.35 0.61 51 6.40 0.58 _____________________________________________________________________________________________ Note * p < 0.05, ** p < 0.001 146 Table 5.2 Regression Analyses for ARRA and Subscales SMP RBV ARRA n = 468 B t B t B t Heterosexual -0.162*** (-2.942) -0.251*** (-4.368) -0.203*** (-3.022) Male -0.137** (-2.395) -0.294*** (-4.928) -0.213*** (-3.957) Nonbinary 0.0977 (-0.724) 0.0251 (-0.178) 0.0635 (-0.501) Not Religious 0.161*** (-2.92) 0.175*** (-3.041) 0.169*** (-3.267) White 0.128*** (-2.284) 0.0187 (-0.319) 0.0765 (-1.445) 18-34 -0.012 (-0.147) -0.144* (-1.680) -0.0755 (-0.977) Constant 6.421*** (-61.66) 6.524*** (-59.99) 6.469*** (-66.02) Adj R-squared 0.093 Female (reference) 0.151 0.134 ***p < 0.01, ** p < 0.05, * p < 0.1 Table 5.3 Correlations With Established Scales n = 518 SMP RBV ARRA SCSR IRMA SMP 1.00 0.73 0.92 0.52 -0.49 RBV ARRA SCSR IRMA 1.00 0.93 0.55 -0.55 1.00 0.58 -0.56 1.00 -0.55 1.00 147 CHAPTER 6 CONCLUSIONS In this chapter I will describe the usage notes for the ARRA as it is, including scoring for four-, five-, or seven-category Likert-type response options. I will also describe how to use domain sampling to select any set of items that fits the research questions, as well as the potential of the attitudes themselves, not only the items written to measure them. I then discuss rape resistance as an ideal type, or a conceptual tool with which to theorize the normalcy (or lack thereof) of resisting rape myths, having a deliberate consent process, and placing blame appropriately with perpetrators rather than victims. I discuss limitation, strengths, future research, and applied possibilities before wrapping up with a reflexive assessment of this project. Usage Notes To use the ARRA as it is, simply administer both subscales for a 25-item scale assessing rape-resistant attitudes. This will offer insight into participants’ propensity to blame victims of sexual violence as well as their willingness to monitor their own behavior to ensure they do not perpetrate sexual violence. These items are intended to be used with a seven-item Likert-type response option; Strongly Disagree (1), Disagree (2), Somewhat Disagree (3), Neither Agree nor Disagree (4), Somewhat Agree (5), Agree (6), Strongly Agree (7). Ascension from Strongly Disagree to Strongly Agree is intentional. When used this way, 149 levels of RRA can be thought to increase as scores increase. To use negatively worded items for attention checks or other methodological ends, simply reword the item negatively and reverse-code the variable in your statistical software package before data analysis. For example, the negatively worded item, “It bothers me more when some people (e.g., women, old people, fat people) have casual sex than when other people have casual sex” is reversecoded in this dataset. Mean scores can be computed for confirmatory factor analysis through structural equation modeling, and for regression analyses. This scale is also appropriate to generate individual scores. Scoring. To score the ARRA with seven response categories, assign points as follows; 1 = 0 points, 2-3 = 1 point, 4 = 2 points, 5-6 = 3 points, and 7 = 4points. Reverse code any negatively worded items and assign the score using the same pattern. Sum the scores of each of the subscales and add them together. The total score, or the sum of responses to each item, has a possible range of 0 – 100. Higher scores indicate more strongly held RRA. It is also possible to use five- or four-item Likert-type response options. Attitudinal scales are often offered with five options rather than seven, omitting the “strongly dis/agree” on each end of the scale. For ease of comparison, a five-item Likerttype response option may be selected and scored as follows; 1 = 0 point, 2 = 1 point, 3 = 2 points, 4 = 3 points, and 5 = 4 points. A four-item response option has the benefit of eliminating the neutral “neither agree nor disagree” option, forcing participants to make a decision to land in one category or the other. In that case, points for each item are equal to its response category number. Domain sampling. Part of classical measurement theory, which informed the development of this instrument, is the domain sampling model. The purpose of domain sampling is to offer researchers the option to select any combination of items for any 150 theoretical or substantive reason, and all will function to estimate measurement similarly to the estimates that would be possible if all items were administered. Consequently, researchers may select any items they feel will best measure their construct of interest. A list of the items whose statistical performance indicates their utility as a domain can be found in Appendix B. To use the scoring method offered here, selecting 25 items is recommended to keep the possible range between 0 – 100 for ease of interpretation. In any case, if these items are used as is or reverse coded when negatively worded, higher scores will indicate more strongly held RRA regardless of the possible range of items administered. The attitudes themselves. The attitudes identified by experts in this project offer their own possibilities for future research. If the purpose is to administer a psychometric instrument, items that better serve the purpose of any particular project may be written differently from how I have written them. Additionally, the theoretical utility of the attitudinal dimension itself can and should be explored and tested. Researchers have many options in using the work in this project to further study RRA as a broad attitudinal dimension, and reliable and valid psychometric tools to use if they fit the methods selected by the researcher. Rape Resistance as an Ideal Type The attitudes and dimensions related to RRA identified from literature and by experts in this project are illustrative in and of themselves. They provide a conceptual tool with which to think about rape resistance broadly in terms of rejecting rape myths, having low rape proclivity, having purposeful consent practices, and being unlikely to blame victims of sexual violence. The conceptual tool offered by these attitudes and dimensions functions similarly to a Weberian ideal type. It is not intended to construct a typology that is perfectly 151 concrete but rather one that is adequate in helping to conceptualize rape resistance as an ideal pattern. While it is certainly my hope that RRA and attendant behaviors would become normative, for now they could function as tools to encourage both interdisciplinary study and self-reflection. In the interest of pursuing philosophical exploration of all kinds, considering rape resistance as an ideal type would be a helpful construct. Rape resistance begins with a thorough understanding of sexual violence itself, as well as the cultural structures that support it. Knowledge of sexual violence goes beyond surface-level, “common-sense” assumptions about the causes of sexual violence, with the recognition that these assumptions are informed by deeply rooted mythos that serve dominant power structures and capitalize on the unequal distribution of various social resources. This knowledge includes firm acknowledgment of gendered and personal power hierarchies, emphasis on perpetrator motivations and responsibility, thoughtful engagement with survivors’ experiences, and recognition that power, privilege, and entitlement influence sexual relationships of all kinds. Education regarding sexual violence is welcome, and consent is thoroughly understood as not only necessary for staying within legal bounds but also conducive to the pleasure and satisfaction of the encounter. With regard to healthy sexuality, rape resistance is sex positive in the belief that sex is important, sexual desires are not shameful, and all people will experience their sexuality in ways specific to them that should be understood and respected to ensure the best interaction possible. While acknowledging the breadth and diversity of appropriate sexual behavior, even outside ones own preferences, rape-resistant sex positivity does not mean that “anything goes;” boundaries are still present and honored, and empathy is a guiding principle. Rape-resistant sex positivity is deeply concerned with affirmative consent; honest, freely given, enthusiastic, “Yes!” given in a context free from coercion or fear of retribution. 152 Consent requires clear communication between partners, both verbal and nonverbal, and a desire to thoroughly understand partners’ sexual boundaries so no lines are inadvertently crossed. Self-monitoring and self-governance are part of rape resistance, and responsibility for not crossing someone else’s lines is never shifted away from oneself. Dominance or submission are never part of a sexual encounter unless they have been agreed upon and are desirable to everyone involved. Within healthy, rape-resistant sexuality, privilege does not extend to entitlement, and unequal power would never be leveraged within an interpersonal sexual experience. Self-governance extends beyond interpersonal sexual experiences to general interaction with others. Rape resistance includes being willing to learn from others, being humble and courteous to others regardless of their social position, recognizing the value of and offering respect to all people, and resisting cultural stereotypes when interacting with people who are different. Having a sense of ones own intrinsic worth and drawing strength from living within a purposeful value system, religious or otherwise, were seen as contributing to rape resistance in a very broad, nonsexual sense. Much like the validity of instruments used to measure it, rape resistance is not conceived of in terms of having it or not. Rather, it is a matter of degree. Clearly, in the interest of eliminating a traumatic social problem and achieving social and criminal justice, higher levels of rape resistance are indeed ideal. However, no person is likely to fully embody the ideal type of rape resistance as it is conceptualized here. Self-identifying as rape-resistant is not the goal; indeed, it would be problematic to do so. Rather, rape resistance as an ideal type must be continually attended to in various kinds of relationships, from casual hookups to long-term monogamy. Personal preference and the contexts of sexual relationships may mediate the emphasis on pleasure as a benchmark, possibly emphasizing intimacy or respect 153 or some other rape-resistant concept. Within the conceptualization of rape resistance as thoroughly understanding rape culture, avoiding victim blaming and RMA, and purposefully pursuing coercion-free consent, mutual satisfaction can be defined in any rape-resistant terms. Individuals and partnerships must regularly engage honestly with rape resistance as an ideal type, and complacency would always indicate the need to dig deeper. Similarly to how overperception of sexual interest can cause individuals to bypass the need for careful consent, overperception of ones own level of rape resistance could cause individuals to bypass the need to examine their own predatory behavior. Carefully considering rape resistance as an ideal type offers the opportunity to seriously and truthfully reflect on our own consent-seeking behaviors (or lack thereof) and feelings of entitlement. It would help us assess the privileges we have, interrogate the power dynamics in our sexual relationships, and understand how enculturated sociosexual scripts fail to normalize affirmative consent. It would help us be unflinchingly honest about the kind of culture we live in and our own engagement with that culture. Rape resistance as an ideal type is a rational objective to engage with and consider our own behavior through. To invoke another Weberian concept, “…man would not have attained the possible unless time and again he had reached out for the impossible” (Weber, 1919). The beauty of reaching for the impossible in terms of perfect rape resistance is that we would come as close as we possibly could in all our interactions, thus pursuing healthy, positive sexuality focused on mutual respect and satisfaction to everyone’s own definition. 154 Limitations, Strengths, Future Research, and Applied Possibilities Limitations. There are four important limitations in this project that should be noted. First, during the qualitative portion of this project, I was unable to connect with a law enforcement officer to schedule an indepth interview. I was constrained by IRB permissions regarding how many times I was allowed to contact the same person, and without personally knowing any law enforcement officers through which to gain entrée, I simply ran out of time. This omission is regrettable, as a law enforcement officer would have offered a valuable voice regarding not only the criminal justice process but also the culture of policing and interaction with survivors in a place where retraumatization occurs. The qualitative process was also limited in that I was able to interview only one person of color and only two who specifically served racial/ethnic minority populations, who are particularly vulnerable to sexual violence. In addition to this, I was unable to schedule indepth interviews with experts working in rural areas, despite repeated attempts to do so. In the future, I will not limit myself so much when I write IRB applications. Secondly, my inability to gather a sufficient sample of students from Salt Lake Community College (SLCC) was a serious limitation given my goals of pursuing external validity by comparing results to another institution. Here again, I reached the limit of my IRB permissions to collect a too-small dataset of 169 respondents, and did not have time to go through SLCC’s IRB process a second time to attempt another round of data collection. Not only did I underestimate how many students in each class would fully complete the survey, but classes at SLCC typically have fewer than 30 students in each class compared to many more at the University of Utah. The time and effort expended to gather too few participants from SLCC also impacted my timeline of completion. 155 Third, this project is severely limited by a lack of racial and ethnic diversity. I am White, most of my expert interviewees were White, and a vast portion of the sample from which these data were drawn were White. Whiteness is problematic in sexual violence research in particular, as White people in general, even White women, are less likely to experience sexual violence than Women of Color, especially Black and Indigenous Women of Color and Transwomen of Color. In addition to that, both social and criminal justice pursuits are marked by racist policies and practices, which my Whiteness and the Whiteness of those who contributed as both experts and participants masks in many ways. My purposeful intersectional practices are insufficient to overcome the blinders of racial privilege, and this project is lacking a great deal of valuable input from People of Color, especially Black and Indigenous People of Color. I deeply valued the voice of a case worker from the Salt Lake Urban Indian Center, but the problem of sexual violence against Indigenous women is so severe that it deserves far more attention than it received in this project, and indeed, in most research. Finally, I believe the greatest limitation in this project was the writing of items to measure the RRA. There are several specific ways these items could have been written better. As many researchers do during the deep dive of a dissertation, I assumed too-broad an understanding of the concepts I intended to address. In three stages of the process (qualitative feedback from the focus group, qualitative feedback from the pilot study, and conversations with my mentor) it was apparent that I was writing items whose concepts were too finegrained to be widely understood. I reworded each time, but I am not sure I was successful. A check of readability statistics indicated that the items were written on a 9th grade reading level, but they should have been easier to understand with more generalized knowledge. I also believe they were not written strongly enough. Roughly 50 participants, or 156 just under 10%, reported perfect rape resistance, indicating that they strongly agreed with all positively worded items and strongly disagreed with all negatively worded items. Ideally, items should be written strongly enough that almost everyone has difficulty answering to the extremes in every case; indeed, the purpose of having a seven-category Likert response option was to push everyone to the limits of their beliefs such that the strongly dis/agree categories were relatively rarely used. I also should have attended more carefully to passive voice and negatively worded items. On the whole, items written negatively tended to be problematic in the statistical models, and I should have been more conscientious about their distribution through the instrument. Strengths. Having said that, there are two principal strengths of this project that I feel are valuable contributions. First, the rape-resistant attitudes identified in the literature and through the experts interviewed are thought-provoking, deeply powerful conceptual tools that offer a great deal of insight into the social problem of sexual violence. They, and all others that are identified in the future, hold the potential to shift conversations about sexual violence prevention to what works. The research that has been done through the conceptual frameworks of RSA and RMA is astoundingly valuable, and will no doubt continue to break ground. Having a dimension of RRA to lay alongside that work holds the potential to expand our theorizing and understanding in powerful ways. Assuredly, I did not successfully identify the theoretical totality of attitudes within the ideal type of rape resistance, but this work must start somewhere. I believe the attitudes identified here will be valuable concepts in keeping pace with the kinds of unflinching research questions we must demand during this unprecedented moment in history. Secondly, having a reliable, valid instrument with which to begin this kind of research is a significant contribution. The index of items is also a valuable resource and 157 research tool, as are the attitudes themselves. This index offers a place to formally start the process of incorporating this kind of theorizing in our work. It is also important that this instrument was purposefully designed to reach outside rigid assumptions regarding gender and sexuality in sexual violence. Future research. While researchers in many social science disciplines are likely to have interest in a scale such as this, some particularly sociological research would be illustrative and add significantly to the body of sociological knowledge about sexual violence. Principally, the ARRA must be administered in a different population to ensure construct validity. While model estimates from this data were good, it may be that this project tapped into particular characteristics of this specific study population at this precise time. Other populations, such as college students across the country, may return data different from those observed here, and this possibility should be explored shortly. I have received a verbal agreement from Dr. Jericho Hockett, whose thought-provoking ask for this kind of psychometric tool was the impetus for this project, to administer this scale to her students in Kansas as soon as I am ready. After that, I hope to partner with academics in other geocultural areas of the United States to test the limits of the ARRA and RRA themselves. Non college educated populations should also be sought to test the validity and reliability of the ARRA. Higher education has a generally inverse relationship to RMA and RSA, with levels of college education predictive of lower RMA (Kassing et al., 2005; Page, 2007). A sample gathered via M-Turk or similar could offer insightful results to lay alongside those from this project and any other project that uses the ARRA. Once replication has offered further refinement of the ARRA and other items in the RRA index, studies should offer both the ARRA as well as an RSA scale to the same sample. Rape-resistant attitudes are anticipated to function alongside RSA, not necessarily to simply 158 replace them. Having data from the full range of attitudes regarding sexual violence holds the potential to offer much greater nuance in the understanding of how these attitudinal dimensions function in general, and how they predict various outcomes. This may offer significant insight into all social processes related to social and criminal justice for survivors of sexual violence. Scales measuring sexually aggressive histories and rape proclivity should be administered alongside the ARRA where appropriate. When considering perpetration of sexual violence and RRA, it is certainly clear that the experts who contributed to this project, and indeed, me as the researcher, were thinking primarily in terms of perpetrators of acquaintance/date rapists rather than violent stranger rapists. Our collective vision of the potential of RRA was more focused on individuals whose predatory behavior could conceivably be altered through education and honest reflection about the parameters of their sexual relationships. We were in agreement that violent stranger rapists were unlikely to be engaged with the nuanced concepts measured in the ARRA – they know what they are doing is violent and criminal. Despite this, it would still be beneficial to administer the ARRA to an incarcerated population along with other measures of social dominance orientations. A great deal of work has been done with individuals convicted on charges of sexual violence, and it offers valuable insight. Using the ARRA in research involving this population would significantly improve our understanding of predictors of this kind of violent crime, as well as investigate whether, and if so, why, convicted perpetrators hold RRA and still chose to perpetrate sexual violence. Any research similar to the body of work that measures RSA could conceivably benefit from including a measure of RRA. It may be that RRA are protective to many. If that could be demonstrated, we could use the ARRA to understand, and indeed encourage, protective factors just as we warn about risk factors. Too many specific possibilities exist to 159 explicate here, but one that would be particularly interesting is pornography and attitudes toward sexual violence. The correlation between pornography involving adults3 and sexual violence is widely debated but not well understood. There are longstanding assertions that consumption of porn and sexual violence go hand in hand (Malamuth, 1981), but correlations are not straightforward and causation extremely difficult to establish given the challenging nature of porn studies. At present the most honest academic answer regarding the influence of porn consumption on sexual violence is, “We don’t know.” Even after considering issues regarding operationalization, conceptualization, methodology, and bias, results are mixed. This does not stop pornography from being overtly politicized, often with opposing ideological camps cherry-picking the data that support their worldview. However, regardless of personal perspective, the advent of internet pornography in the late 1990s demands engagement with the existence and pervasive consumption of readily available, sexually explicit materials. Indeed, many young people report turning to pornography for sex education (Smith, 2012). Dire outcomes connecting pornography and sexual violence are decried by antiporn organizations including Utah’s own Fight the New Drug.4 However, porn consumption is so widespread that a majority of people consume porn, and no sharp increase in sexual violence occurred with the advent of easily accessible internet pornography. Some researchers even claim to see protective effects when considering pornography and sexual (Castleman, 2016; Kendall, 2006). Clearly, more research is needed to understand pornography and its effects, particularly its effects on sexual violence. The RRA index developed here provides an additional tool to look at all possible outcomes of 3 No ambiguity exists regarding child pornography, as the inability of a child to provide consent renders all child pornography criminal sexual violence. 4 fightthenewdrug.org/media/porn-fuels-sexual-assault/ 160 porn consumption. Finally, with the ARRA, sociosexual scripts in various contexts could be studied to find more healthy, sex positive patterns. A great deal of consensual sex occurs with positive engagement between people of all genders and sexual orientations, and the ARRA could help identify where and how healthy sexualities and scripts are more normal. These would likely extend our concepts of healthy masculinity and femininity, as many ways of being exist beyond the toxicity of hegemonic masculinity and emphasized femininity. Queer conceptualizations of gender and sexual fluidity would likely augment these understanding, and the ARRA is equipped to measure those. Applied possibilities. Beyond academic research, the ARRA holds the potential to inform, assess, and suggest programs and policies in applied ways. For instance, one expert who participated in this project works in a group home setting with adolescent males who have been convicted for offending sexually. With my permission, he took from the focus group the list of over 300 items written to assess RRA. He presented the ideas to the adolescents in the group home during a group therapy session, talked the group participants through the nuances I was intending to tap into, and reported that the ensuing conversation about sexual violence was more purposefully rape-resistant than any the group had had before. This anecdotal report offers a provocative consideration regarding the possibilities of including RRA in rehabilitation programs. The potential of teaching adolescents about RRA specifically is most interesting when thinking of sex education in schools. Sex ed is notoriously inadequate in many, if not most, places across the country, with the inclusion of consent and healthy relationship the exception rather than the rule. Regardless of the appropriateness of their formal sex ed, adolescents are bombarded with messages every day that communicate problematic 161 sociosexual norms and gender essentialist beliefs regarding hypersexual masculinity and Madonna/whore femininity. These messages are problematic in terms of seeking consent and having healthy relationships. The possibilities of developing a sex ed program that specifically includes and encourages RRA would be a significant applied contribution. Perhaps most intriguing about that possibility is that some attitudes identified by experts, primarily those grouped into the Personal Values – General subscale in this project, are not controversial and are unlikely to inspire resistance from conservative groups who oppose comprehensive sex ed. The attitudes identified here could inform the development of a healthy relationship component in sex ed programs, even those that are abstinence-based. The items themselves could be used to assess program efficacy, particularly when given in a pre and posttest. In general, programs and policies in many institutions must be research-based in order to be implemented. This is true of quality sex ed curricula, as well as public health programs and policies adopted by initiative groups such as Primary Children’s Hospital Safe and Healthy Families5 and myriad other organizations who are interested in sexual violence education, prevention, and treatment. Having theories and psychometric tools related to RRA would allow the kinds of research needed to inform better programs and policies that better address protective factors and rehabilitation. Program assessment would also be well served by the ARRA. At present, the IRMA is the most widespread instrument used for program assessment, but the exclusively mp/fs focus in the items, as well as phrases that trigger social desirability bias (e.g., “If a girl acts like a slut, eventually she is going to get into trouble”), limit the utility of the IRMA and reify gendered, problematic views of sexual 5 https://healthcare.utah.edu/pediatrics/programs-services/safe-healthy-families.php 162 violence. Beyond programmatic and policy applications, having a society well versed in RRA, not just RSA and RMA as we are now, holds a great deal of possibility in terms of better achieving criminal justice. For all the reasons described in Chapter 1, RSA and RMA are problematic in the pursuit of criminal justice. However, RRA would offer at least a counterpoint if not an antidote to RSA and RMA in the criminal justice process. If law enforcement officers, lawyers, judges, and juries were fully versed in RRA, investigations and outcomes may be less traumatic for survivors and result in more appropriate discipline for perpetrators. So too would social justice for survivors be impacted by a widespread understanding of RRA. The attitudes as a teaching tool not only address predatory behavior, but also blaming victims of sexual violence, including individuals who blame themselves. Low RMA has been shown to be protective in terms of self-blame, and survivor healing trajectories being more positive if their own RMA is low (Peterson & Muehlenhard, 2004). Raperesistant attitudes could conceivably be related to Posttraumatic Growth (Mason, 2013)The possibility of having widespread social dialogue of the kind seen since the #MeToo movement began in earnest, only rooted in RRA rather than RSA and RMA is deeply exciting. Conversations could move in protective, compassionate directions that appropriately consider survivors’ experiences and humanity and discredit inappropriate victim blaming. As the former executive director of a women’s shelter said regarding RRA, “It would change everything.” Finally, perhaps the most important applied possibility of the ARRA itself, and indeed all RRA identified through the course of this project, is for individuals to score themselves with unflinching honestly about their own behaviors, their own privileges, and 163 their own ideas about healthy sexuality. Ultimately, every improvement in the journey and treatment of survivors and all progress toward more equitable applications of social and criminal justice do not sufficiently address the root cause of this social problem – the perpetration of sexual violence. Systemic changes must happen, as societal-level inequalities are predictive of rates of sexual violence, but in every case, regardless of social context, an individual made a choice to perpetrate. Eliminating the perpetration of sexual violence is the only acceptable goal, and the only possible way to attain that goal is for all individuals to honestly contemplate their beliefs and behavior regarding sexual violence. And so while I would never claim sufficiency in what I have produced thus far, both the concept of RRA and the specific attitudes introduced in this project could conceivably be emancipatory to individuals who wish to assess and overcome their own problematic beliefs and behaviors. Reflexive Assessment As a critical feminist scholar, reflexivity is a continual, purposeful part of my praxis. To satisfy the demands of a sociology dissertation, I have attempted, perhaps unsuccessfully, to keep my reflexivity out of the body of this work as much as possible. However, I feel that my critical, reflexive voice should come into this thesis formally at some point, and so I have reserved it until the end. Beginning a dissertation on sexual violence just before the #MeToo movement began has proven to be fascinating, hopeful, and discouraging all at once. Through the process of conceptualizing rape resistance as an attitudinal dimension, I was already asking myself deep questions about the nuances of sexual violence when suddenly those questions were writ large in public dialogue. The gnarly problem of sexual violence defies easy answers, and a plethora of voices from all over the map created a cacophony that was alarming at 164 times in its ignorance. In engaging with public dialogue as I have for the past two years, I admit to some concern that social desirability and the continual construction of perpetrators of sexual violence as “monsters” rather than people just like us may function as a way to selfidentify as “one of the good ones,” thus disengaging with the need to interrogate our own behavior. Indeed, this is the reason I feel the items should have been worded more bluntly and strongly after seeing such high scores in some participants. Of necessity, scale items are written to assess surface-level understanding, but the concepts beneath them are complicated and nuanced. However, I attempted to create a tool that would dismantle the master’s house; my endless optimism helps me believe that in skilled hands, including my own, it may function as such. In all of this, I was continually aware of my positionality. I am a White, middle class, Gen X, former Mormon, cisgender woman with straight privilege. As I say to my students (with my tongue only partially in my cheek), each of those identities makes me suspect to some degree. I am also a survivor of sexual violence, and the trauma surrounding that reality has pushed me in several ways throughout this project. I knew I would have to manage my time carefully to complete the amount of work I wanted to do in the allotted time. I did not anticipate that I would also have to manage the topic in the ways I did. By way of example, part of what I demand from myself as a scholar is to compassionately articulate arguments against my own claims and positions. I do this primarily through seeking disconfirming observations from the literature and in the voices of people who are steadfastly opposed to everything I believe in and stand for. During this project, I attempted to understand the particularly vitriolic voices of radical, prorape groups of predominantly White men who have organized in the dark corners of the internet (see Ging, 2017 for a review). I was particularly interested in Incels (Involuntary Celibates), Betas 165 (men who blame their celibacy on women’s supposed withholding of sex from all but alpha males), and MGTOWs (Men Going Their Own Way, who wish to live without women entirely except to fulfill sexual needs). These men sincerely believe that women are inferior, and our single, God-given purpose is to provide sexual access and nurturant caregiving through mothering, submissive partnering, and homemaking. Having been raised with (albeit vehemently rejecting) these kinds of limiting and demeaning messages, I thought I was prepared to consider with kindness the ideals of these groups and their members. I was wrong. These men believe in rape as an organized strategy to control women who have, in their estimation, far outstepped our bounds. Once, after having spent the day reading through men’s rights activist’s websites and diving into the attendant comments sections, I looked down at my hands and noticed they were shaking violently. It had taken a physical toll to immerse myself in that violent hatred for people like me – pointedly, women like me. I was unable to work well for almost a week. My scholarly practices of critical thinking, which serve me so well in most cases, were unequal to the horror of reading the vicious rage directed toward women like me, and my daily lived reality of managing sexual trauma superseded my scholarly commitment to articulating opposing perspectives with compassion. I cannot articulate the filth I encountered in attempting to understand. I will not. Not wanting to give up something like this practice entirely, I facilitated this scholarly commitment in a less vitriolic sphere – public comments sections regarding the #MeToo movement. For more than a year, I have read at least one comment section in full every day. While discouraging, this practice has sharpened me as a scholar and shown me the possibilities of understanding both RSA and RRA together. Fascinatingly, individuals often expressed both in a lengthy comment or an exchange. I learned a great deal from reading 166 nonacademic musings in the wild, having the benefit of both the lenses of RSA and RRA. I am more convinced than ever that rape resistance as an ideal type is extremely important to understand. Of course, these comments sections were rife with problematic assumptions. Victim blaming was rampant. The belief that women (it was almost always placed in terms of womanhood) should anticipate and accommodate the inherently predatory male gaze was present in every single one. The presupposed “naturalness” of male predation as an excuse for violence was deeply troubling. As a feminist, the frustration of continually encountering these dead end beliefs made my throw my pen at the wall more than once. However, in reading these comments sections looking specifically for RRA, I always found it. In every comments section where I saw rape-supportive attitudes and rape myth acceptance, I also saw rape-resistant attitudes. People pushed back on essentialist, rapesupportive tropes and assumptions, disrupting the ubiquitous lectures to women about how to keep ourselves safe through “proper” femininity (read chastity, modesty, and meekness), and educating about rape culture with both ferocity and compassion. They often had precise and accurate statistics, backed up by quality sources. They recognized that tasking potential victims with keeping themselves safe only attempts to ensure that a rapist rapes someone else, but does absolutely nothing to eliminate sexual violence. As a survivor, the voice I have hoped to hear for so long is present. Rape-resistant attitudes are there, in the wild, waiting to be formally understood as soon as we have the tools. And so I am more hopeful than I otherwise might have been after writing a dissertation about sexual violence at this unprecedented time in history. 167 Conclusion This project began with a thought-provoking ask for a psychometric tool to study rape-resistant attitudes, made by a scholar who studies sexual violence and saw the potential of addressing this gap in research. It ended with an instrument, an index, and an ideal type. Hundreds of people have contributed in the intervening time, through expertise and research participation. Moving forward, this new approach to the study of sexual violence may offer insights we are keen for as a society. No longer will we collectively diminish the social problem of sexual violence and the systemic hierarchies that support and perpetuate rape culture. The time is ripe, and there is much work to be done. The master’s house awaits dismantling, and we now have one more tool. APPENDIX A COMPLETE LIST OF RAPE-RESISTANT ATTITUDES 169 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. Willingness to consider if you are a perpetrator of sexual violence Recognition that sex without consent is rape (and is a crime) Secure masculinity/femininity (at any level of self-determined performance) Comfort with deviation from gendered sexual scrips, both in self and partner Not allowing libido to be more important than consent Recognition that women are not hysterical – there IS risk Willingness to accommodate that risk and not take it personally; walking on the other side of the street, for example Understanding that unintended things can be sexual violence Understanding that unintentional violence is still a big deal Understanding that penetration without consent is rape, even if there is no force Resistance of the “scarlet letter” – understanding women’s sexual autonomy and right to be as sexually active as they want to be Recognition of male privilege Understanding that power in the relationship matters during sex Knowing not to use privilege or power to gain sexual access Knowing you always need to learn more about sexual violence Being willing to remain continually engaged with sexual violence education Embracing nonviolent/positive masculinities Willingness to interrupt light sexually violent behavior Reflexivity about how you personally think about sexual violence Willingness to call rape, rape Willing to learn and do better when called out on soft sexual aggression Being willing to listen if a partner said you were/are sexually aggressive Sex positivity Recognizing the importance of communication during sexual activity Understanding that sexual violence is surrounded by mythos Respecting sexual vulnerability Comfort with speaking out about sexual violence Having positive, appropriate conversations about sex Being open to new messages about sexual violence Critical media literacy (e.g., how media, especially pornography, misrepresents sex) Recognizing sexism in media and pornography Openness to not following sexual script Understands consent is revocable Understanding that consent should be obtained every time sexual activity escalates Understanding the difference between dehumanization and sex positivity (sex positivity does not mean accepting anything and everything) Self-esteem, self-worth Positive body image – acceptance of bodies (self and others) No shame about sex itself Openness to people who are different (values, religions, ethnicities, etc.) Awareness that all bodies experience pleasure differently (self and other) Acceptance of queer feelings in self and others Respecting others bodily pleasure 170 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. Intimacy needs vary from person to person Intimacy is positive Even casual sex can be connective Learning partner(s) boundaries Not relying on body language cues Relational sensitivity Recognizing and using correct labels for body parts (e.g., using vulva and vagina properly) Recognizing people express sexual things differently and you can’t make assumptions Sexual behavior should be purposeful and not reactive Healthy coping behaviors for stress Understanding rape is not your fault Openness to sexual exploration with consent Understanding your own heart and sex Letting people be experts on their own experience (e.g., people know what’s best for themselves) Recognizing that sexual boundaries move, even with the same partner (e.g., place to place, moment to moment, interaction to interaction) Lack of defensiveness when a partner expresses needs (sexual and otherwise) Preferring sex with individuals who are not reactive (opposite of grooming, targeting) Community includes more people than you like Recognizing the approach to sexual violence needs to change (e.g., current social dialogue is problematic) Not judging others, particularly for how they have sex (e.g., polyamory, fetish play) Not being “grossed out” by things other people enjoy sexually, even if you never want to do it Taking responsibility even if we didn’t mean to harm someone We have a personal responsibility NOT to assault Spirituality (nonspecific) Impulse control We are all more than our mistakes Dominance is not sexy unless that is agreed upon by everyone Drawing strength from living ones values, religious or not Feminism People have intrinsic value Understanding expectations vary and need to be navigated purposefully Feeling more comfortable when you understand someone’s expectations Understanding not everyone wants to have sex with you Lack of arrogance Understanding that we usually believe rape myths Knowing we hold rape-supportive attitudes from our culture, even when we are educated Not deciding you can disengage because you are a “good person” Asking for sex is normal Asking for sex creates a mood that enhances the experience 171 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. Willingness to discuss and disrupt sexual scripts Enjoying partner’s pleasure Specifically addressing gray lines (e.g., alcohol consumption) during consent process Having a consent process Avoiding getting as close to the “line” as possible (wanting an experience that doesn’t come close) Recognizing that gender is a broad concept – not “typical” to men or women (maybe understanding of gender spectrum?) Engaging with an individual not the tropes and stereotypes of their group Relating to someone is integral to sex, even casual sex Understanding microaggression Education about sexual violence is empowering Being eager to learn about and better understand sexual violence. Education about sexual violence is empowering to all individuals, not only survivors. Understanding that the scripts in pornography will not necessarily be present in real sex Approaching porn as fantasy, not sex education Willingness to consider that someone you love is a perpetrator Teaching (or being taught) not to perpetrate doesn’t assume perpetration Everyone needs to learn how to avoid becoming a perpetrator of sexual violence, even if they are not “prone” to violence Teaching about perpetration is an assumption about society and social norms, not about the people you are teaching Even nonnormative sex needs consent Knowing that even if you’ve had sex with someone before, you need to obtain consent to have sex with them again Being careful with partner’s sexual vulnerability (e.g., body image concerns, performance worries, ability or desire to orgasm) Valuing connection during sex (even casual sex) Accepting other peoples’ nonnormative configurations of sexual partnerships as long as there is consent Assuming no until you hear yes Not proceeding with any sexual activity until there has been affirmative consent Recognizing nonverbal boundary setting Respecting nonverbal boundary setting All sexual behavior needs to be negotiated Knowing you are not entitled to any sexual activity with any partner at any time Carefully considering new sexual behavior if you haven’t done it before Recognizing the socialization of sexual partners (what are the social norms, what is feasible for someone who is socialized to defer/dominate, is smaller/bigger than you, other gender socialization) Understanding “freeze” is not consent Confidence regardless of sexual activity Healthy sexuality 172 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. Healthy relationships Separating power and sex (unless power is sex play) Being aware of how healthy the sexual relationships were in your family Calling out rape culture Unwillingness to be passive (unless passive is how they enjoy sex) Taking an active role in sexual relationships Actively considering what you want out of sexual relationships Taking time to know what you like sexually Knowing how your body experiences sexual pleasure Personal empowerment in general No masturbation shame Resisting patriarchy Recognizing patriarchy as a contributor to rape culture Being trustworthy with power in a sexual relationship Understanding coercion, not just force, is sexually violent Understanding misremembering in survivors Normalcy of sex – primary drive People should not be ashamed of their sex drives, but should be aware of consent Consent in all ways, not just sex Taking time to learn partner’s likes and desires before sex Purposeful disruption of rape-supportive narratives (interrupting the social “tape” that plays automatically to excuse perpetrators and blame victims) Communication in sex – needing to say things, needing to hear things No situation merits sexual violence Start by believing Consent is defined by the person giving it, not the person asking for it Acknowledge media influence (including porn) Understanding rejection – not wanting sex is not rejection of you as a person Understanding that a person can want to have sex and still not to consent to sex Understanding that any change in activity requires change in consent Understanding when you aren’t ready for sex in a relationship Being honest about your intention of having sex Caution with partner vulnerability Dismantling rape culture is everyone’s responsibility Understanding that discussions about rape are politically charged Sympathy for the impact of sexual violence on survivors Caution with partners who are survivors Willingness to listen to and have uncomfortable conversations with survivors Rejection of Madonna/whore dichotomy Rejection of hypersexuality as inherent to masculinity Orgasm culture – valuing orgasm Consent does not ruin the mood Low rape myth acceptance Recognizing what is and isn’t rape Egalitarian attitudes toward women Willingness to intervene (related to bystander willingness) 173 161. 162. 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. 181. 182. 183. 184. 185. 186. 187. 188. 189. 190. 191. 192. 193. 194. 195. 196. 197. 198. 199. Recognizing predatory behavior Willingness to disrupt gendered sexual scripts Willingness to engage in sexual violence-related education Desire for mutual sexual satisfaction Willingness to understand partner’s definition of sexual satisfaction Willingness to engage with survivor’s experience Understanding necessity of rape prevention as perpetrator-centered Valuing consent as integral to pleasure Knowledge about consent Unwillingness to bypass consent Obligation to be sexually frustrated rather than coerce/force sex Acceptance of nontoxic masculinities Disruption of sexual double standard Careful alcohol use Understanding of incapacity to provide consent (e.g., drug, alcohol, emotional, or physical impairment) Understanding that someone who participates in sexual activity because they are afraid of what their sexual partner would do if they refused has not given consent Seeking porn that is in line with personal values Consuming porn ethically Recognition of the humanity (is this empathy?) Unwillingness to blame victims Empathy Heterosociality (not “risky” because you decide whom to have sex with through your value system, not simply your base desire) Taking responsibility for ones own actions when aroused Low entitlement Low sexual aggression Low narcissism Rejection of hypersexuality as integral to masculinity Not bothered when women in particular have casual sex (no slut shaming) Respecting how people use their bodies, sexually and otherwise (nobody’s business) Knowing socialization and entitlement influence sexual decision making Avoiding crossing any lines, not pushing partner’s boundaries Thinking honestly about what you learned about sex as a kid Understanding sexual violence as a social problem Understanding that education about sexual violence must be ongoing in a person’s life Thinking honestly about how you and your social groups contribute to rape culture Being intentional in your sexual interactions (not “swept away”) Making sexual decisions within a conscious value system Not withholding assumptions about pleasure from people whose bodies are not “ideal” by cultural standards Knowing the difference between social and criminal justice 174 200. 201. Understanding that lack of conviction did not mean false accusation Recognizes the difference between sexual encouragement and pressure APPENDIX B FULL INDEX OF ITEMS FOR DOMAIN SAMPLING 176 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. Adults have the right to be as sexually active/inactive as they want to be. It is important that my partner and I communicate about our desires during sex. I make sure I understand my partner’s body language before I assume they are giving nonverbal consent. People of all genders can be equally sexual. Women should experience consensual sexual pleasure in any way they want to. Some groups (e.g., women, gay men, trans women of color) are under real threat of sexual violence in our society. Women should reduce their risk of sexual violence by not having casual sex. (reverse coded) There is something distasteful about people who enjoy sex and have a lot of partners. (reverse coded) It bothers me more when some people (e.g., women, old people, fat people) have casual sex than when other people have casual sex. (reverse coded) Our society privileges men and maleness. General societal messages teach men they are entitled to sex. There are more than two genders. Even if I have more power in a relationship, I still obtain consent for sex. I am obligated to be sexually frustrated rather than put pressure on my partner to engage in behavior they don’t want. All people have the right to experience their bodies in consensual ways that bring them pleasure. People have a right to tell me if they feel I am being sexually aggressive. It is up to me to be sure I do not commit sexual violence. I am still responsible for sexual harm I commit unknowingly. If my sexual partner expresses uncertainty, I always allow them to determine our level of sexual activity. My sexual partner does not owe me any sexual activity at any time. If someone does not want to have sex with me, they have rejected me as a person. Victims of nonconsensual sexual violence are never at fault for that sexual violence. Everyone needs to learn how to keep themselves from perpetrating sexual violence. The best way to eliminate sexual violence is to address perpetration of sexual violence. If a sexual partner says I coerced them into sex, I am obligated to consider the possibility that I did. If people do not understand consent, they can inadvertently become perpetrators of sexual violence. I am often bothered by sexism I see in the media. It is wrong to initiate sex with a person, even a long-term partner, if they are too drunk/high to give consent. If my sexual partner shows nonverbal discomfort or dismay during sexual activity, I cease the activity. 177 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. If a person changes their mind during any sexual activity, their partner should immediately halt that activity. It is wrong to push my partner’s sexual boundaries when they are uncomfortable. Sexual boundaries must be discussed when discussing sexual exploration. It is important that I understand my own sexual boundaries. Any penetration without consent is rape. If a person wants to have sex but decides against it and is then forced by their partner, this is rape. I need to obtain consent each time, even from people I’ve had sex with before. A person can want to have sex and still not consent to have sex. Pausing sexual activity to gain consent is not an obstacle to sexual pleasure. I am not willing to intensify sexual activity until I have obtained consent. I am able to enjoy sex more after I have obtained consent. Society has well-established beliefs about sexual violence that are often not true. It is everyone’s responsibility to eliminate rape culture. Sexual violence is committed if someone does not obtain consent for sex, even if there is no force. There is no situation that justifies nonconsensual sexual violence. If someone disclosed sexual violence to me, I would believe they were telling the truth. I do not allow people around me to use language that makes light of sexual violence. I would not feel ashamed if I felt attracted to someone of my same gender. How other people choose to have consensual sex is none of my business. Consenting adults have a right to have sex in any kinds of partnerships they choose (e.g., polyamorous, dominant/subordinate). I enjoy the emotional intimacy I experience in nonsexual relationships. All partners must be in agreement before sex is introduced into the relationship. I am careful with the vulnerabilities my sexual partner shows during sex (e.g., body image concerns, sexual performance, ability or desire to orgasm). I consciously resist stereotypes when I meet someone different from me. I can learn a lot from people who are different from me. People generally know what is best for them. It doesn’t bother me if adults experience consensual sexual pleasure in things I find distasteful. It is up to me to learn how to nurture my relationships, even nonsexual relationships, in ways my partners need. We all have an obligation to interrupt predatory behavior when we see it. Bodies of all shapes, sizes, and abilities have a right to experience sexual pleasure. I am more interested in how bodies experience pleasure than in how they look. APPENDIX C SURVEY QUESTIONNAIRE 179 Thank you for your interest and willingness to participate in the following survey. I am a doctoral candidate at the University of Utah in the Sociology department researching a theoretical dimension of attitudes toward sexual activity, sexual consent, and sexual violence. The purpose of this research study is to develop an attitudinal scale that measures these attitudes, which can be administered with several existing scales in related topics. I am doing this study because the existence of such a scale will allow further investigation into the correlations between attitudes toward sex, sexual consent, sexual violence, and a variety of other measurable attitudinal dimensions. I am asking you to complete the following survey by selecting the best answer for each question, and at the end of the survey click “Submit”. Please be honest, as accurate responses are extremely important. There are no wrong answers, and your honest participation will help me understand if the theoretical dimension of attitudes I am exploring does in fact exist. All survey responses will remain anonymous and confidential. Should you choose to enter the drawing for one of the gift cards, your contact information will be disassociated from your survey answers and deleted upon completion of the drawing. No person involved in data analysis will have access to any identifiable information. You must be age 18 or older to participate. When a question prompts you to consider a sexual interaction, please think of your personal ideal when answering the question, regardless of any current sexual partnership(s). We do not anticipate any risk to participation, but survivors of sexual trauma may be sensitive to questions about sexual violence and aggression. You are under no obligation to participate and are free to decline answering any of the questions that will be asked. If you have any questions, complaints, or if you feel you have been harmed by this research please contact CoCo James, doctoral candidate, Department of Sociology, at 801-477-6341 or by email at [e-mail]. Contact the Institutional Review Board (IRB) if you have questions regarding your rights as a research participant. Also, contact the IRB if you have questions, complaints or concerns which you do not feel you can discuss with the investigator. The University of Utah IRB may be reached by phone at [phone] or by e- mail at [e-mail]. It should take approximately 20 minutes to complete the questionnaire. Participation is voluntary, and both participation and nonparticipation are completely anonymous. You can choose not to finish the questionnaire by exiting out of your browser at any time, and you may skip any question you prefer not to answer without penalty or loss of benefits. By submitting this questionnaire, you are giving your consent to participate. Thank you for completing this survey. Sincerely, CoCo James Yes, I consent and agree to complete this survey. No, I do not consent and do not wish to complete this survey. 180 For each of the following statements, please indicate the strength of your agreement/disagreement.6 Adults have the right to be as sexually active/inactive as they want to be. Orgasm is the most important form of sexual pleasure. I understand my own sexual needs. It is vital that I understand my partner’s sexual needs. If my partner does not experience pleasure during sex, I do not experience pleasure. My partner’s sexual pleasure is more important to me than my own sexual pleasure. I am not ashamed of my sexual desires. I enjoy exploration in my sexual relationship(s). Solo masturbation is a normal part of healthy sexuality, even if someone has a sexual partner. I know how my body experiences sexual pleasure. I enjoy fantasizing about what I want to experience sexually. A person’s sexuality is an important part of who they are. 6 Note- in the online survey platform, each item was followed by seven Likert-type response options ranging from “strongly disagree” to “strongly agree.” Some items were negatively worded and the data reverse coded to reflect stronger rape-resistance with higher scores. 181 For each of the following statements, please indicate the strength of your agreement/disagreement. People who experience sexual violence do not need to feel ashamed of that violence. When I talk about rape, I call it rape and not some other term. It is important that my partner and I communicate about our desires during sex. I want my partner to communicate their desires to me during sex. I always talk to my partner directly about their likes, dislikes, and desires before we have sex. I make sure I understand how my partner expresses their sexual desires, intentions, and willingness before I make assumptions. I feel comfortable talking about my sexual desires with my partner. I make sure I understand my partner’s body language before I assume they are giving nonverbal consent. 182 For each of the following statements, please indicate the strength of your agreement/disagreement. I am more comfortable around men who have traditionally masculine roles than those who don’t. People of all genders can be equally sexual. Women should experience consensual sexual pleasure in any way they want to. Some groups (e.g., women, gay men, trans women of color) are under real threat of sexual violence in our society. Women should reduce their risk of sexual violence by not having casual sex. There is something distasteful about people who enjoy sex and have a lot of partners. It bothers me more when some people (e.g., women, old people, fat people) have casual sex than when other people have casual sex. Our society privileges men and maleness. General societal messages teach men they are entitled to sex. There are more than two genders. 183 For each of the following statements, please indicate the strength of your agreement/disagreement. People with more privilege are entitled to more power in sexual relationships. Even if I have more power in a relationship, I still obtain consent for sex. I am solely responsible for my own actions when I am sexually aroused, regardless of what my anyone else has done. I am obligated to be sexually frustrated rather than put pressure on my partner to engage in behavior they don’t want. All people have the right to experience their bodies in consensual ways that bring them pleasure. I feel personally criticized when people are angry about sexual violence and rape culture. People have a right to tell me if they feel I am being sexually aggressive. It is up to me to be sure I do not commit sexual violence. I am responsible for sexual harm I commit unknowingly. If my sexual partner expresses uncertainty, I always allow them to determine our level of sexual activity. My sexual partner does not owe me any sexual activity at any time. 184 For each of the following statements, please indicate the strength of your agreement/disagreement. Victims of nonconsensual sexual violence are never at fault for that sexual violence. Everyone needs to learn how to keep themselves from perpetrating sexual violence. The best way to eliminate sexual violence is to address perpetration of sexual violence. Sexual violence prevention programs that teach people how to keep themselves safe do not sufficiently address the problem of sexual violence. If a sexual partner says I coerced them into sex, I am obligated to consider the possibility that I did. If people do not understand consent, they can inadvertently become perpetrators of sexual violence. 185 Do you consume pornography? Yes No (Skip Logic) For each of the following statements, please indicate the strength of your agreement/disagreement. The porn I consume influences the ways I want to have sex. It is important that the porn I consume is in line with my personal values. Fantasies in porn are separate from the realities of sexual activity with my partner. 186 For each of the following statements, please indicate the strength of your agreement/disagreement. I am often bothered by sexism I see represented in media. It is wrong to initiate sex with a person, even a long-term partner, if they are too drunk/high to give consent. If my sexual partner shows nonverbal discomfort or dismay during sexual activity, I cease the activity. If a person changes their mind during any sexual activity, their partner should immediately halt that activity. My sexual partners are likely to have different sexual boundaries than I have. It is wrong to push my partner’s sexual boundaries when they are uncomfortable. Sexual boundaries must be discussed when discussing sexual exploration. It is important that I understand my own sexual boundaries. 187 For each of the following statements, please indicate the strength of your agreement/disagreement. Any penetration without consent is rape. If a person wants to have sex but decides against it and is then forced by their partner, this is rape. Consent has been withdrawn if my partner agrees to a sexual activity and then changes their mind. If my sexual partner does not stop me from some sexual activity, I can assume they have given consent. I need to obtain consent each time, even from people I’ve had sex with before. If someone does not want to have sex with me, they have rejected me as a person. A person can want to have sex and still not consent to have sex. Pausing sexual activity to gain consent is not an obstacle to sexual pleasure. I am not willing to intensify sexual activity until I have obtained consent. I am able to enjoy sex more after I have obtained consent. Consent is determined by the person giving it, not the person asking for it. 188 For each of the following statements, please indicate the strength of your agreement/disagreement. Society has well-established beliefs about sexual violence that are often not true. It is everyone’s responsibility to eliminate rape culture. People can commit sexual violence without knowing it. Sexual violence is committed if someone does not obtain consent for sex, even if there is no force. There is no situation that justifies nonconsensual sexual violence. If someone disclosed sexual violence to me, I would believe they were telling the truth. 189 For each of the following statements, please indicate the strength of your agreement/disagreement. I have used alcohol to help overcome my sexual partner’s resistance. I have persisted to ask about a sexual activity after my partner has declined multiple times. I have threatened to end a relationship if my partner did not have sex the way I wanted to. I think about how I contribute to rape culture through my language and behavior. I do not allow people around me to use language that makes light of sexual violence. I would not feel ashamed if I felt attracted to someone of my same gender. How other people choose to have consensual sex is none of my business. Consenting adults have a right to have sex in any kinds of partnerships they choose (e.g., polyamorous, dominant/subordinate). My value system, religious or not, influences my sexual activity. People often have different needs for intimacy than their sexual partner. I enjoy the emotional intimacy I experience in sexual relationships. I enjoy the emotional intimacy I experience in nonsexual relationships. I enjoy connecting with my partner, even if the sex is casual. It is not a good idea for people in committed relationships to be friends with people of the gender they are sexually attracted to. I will not have sex in a new relationship until I’m ready. All partners must be in agreement before sex is introduced into the relationship. I communicate with my partner if I am not ready for sex in the relationship. I am careful with the vulnerabilities my sexual partner shows during sex (e.g., body image concerns, sexual performance, ability or desire to orgasm). I am aware of my own sexual vulnerabilities. 190 For each of the following statements, please indicate the strength of your agreement/disagreement. I feel strength when I live within my value system. My spirituality is important to me. I am a valuable person whether I am sexually active or not. I feel empowered when I am having the kind of sex I want to have. All people have the same amount of worth. All people deserve the same amount of respect. I consciously resist stereotypes when I meet someone different from me. I can learn a lot from people who are different from me. People generally know what is best for them. It doesn’t bother me if adults experience consensual sexual pleasure in things I find distasteful. I generally feel a sense of my own intrinsic worth. It is up to me to learn how to nurture my relationships, even nonsexual relationships, in ways my partners need. We all have an obligation to interrupt predatory behavior when we see it. Bodies of all shapes, sizes, and abilities have a right to experience sexual pleasure. I am more interested in how bodies experience pleasure than in how they look. 191 Gender Identity Please choose the description that best fits how you think of yourself. Agender Female Male Nonbinary Transgender A gender identity not listed here I don't know I prefer not to answer Sexual Identity Please choose the description that best fits how you think of yourself. Exclusively heterosexual (straight) Mostly heterosexual (straight) Bisexual (attracted to both men and women) Mostly homosexual (gay/lesbian) Exclusively homosexual (gay/lesbian) Not sexually attracted to either males or females Pansexual Queer A sexual identity not listed here I don't know I prefer not to answer Relationship Status Please choose the description that best fits your current relationship status. Actively dating, but NOT in a committed relationship Not married, but living together in a committed relationship Married Divorced/separated Single (not dating, not in a relationship) Committed (married or not), nonmonogamous Polyamorous Widowed A relationship status not listed here 192 Religious Identity Please choose the description(s) that best fits how you think about yourself. Check all that apply. Not religious Buddhist Catholic Confucian Hindu Jewish Mormon (LDS) Muslim Small branch Christian (nondenominational, etc.) Protestant (Lutheran, Evangelical, Mainline, etc.) Sikh A religious identity not listed here I don't know I prefer not to answer Age 18-24 25-34 35-44 45-54 55-64 65+ Area of Study Humanities (e.g., Art, History, Philosophy) Social Science (e.g., Sociology, Anthropology, Psychology, Law) Natural Science (e.g., Biology, Chemistry, Earth/Space Science, Physics) Formal Science (e.g., Computer Science, Math) Applied Science (e.g., Engineering, Technology, Medicine/Health) 193 Race/Ethnicity Please choose the description(s) that best fits your racial/ethnic identity. Check all that apply. Asian Black or African American Hispanic Latinx Native North American or American Indian Pacific Islander White or Caucasian A racial/ethnic identity not listed here I don't know I prefer not to answer Thank you for your participation in this survey! If your instructor has offered extra credit for completing this survey, please take a screenshot of this screen and email it to them as proof. SPRNG2018-ex0e3f07 Feel free to close your browser window at this time. Have a nice day! Powered by Qualtrics REFERENCES Abbey, A., & McAuslan, P. (2004). A longitudinal examination of male college students' perpetrations of sexual assault. Journal of Counseling Psychology, 72(5), 747-756. Abbinnett, R. (2003). Culture and identity: Critical theories. London, UK: Sage Publications. Acock, A. C. (2013). Discovering structural equation modeling using stata (revised ed.). College Station, TX: Stata Press. Acock, A. C. (2014). A gentle introduction to stata (4th ed.). College Station, TX: Stata Press. Ahrens, C. (2006). Being silenced: The impact of negative social reactions on the disclosure of rape. Americn Journal of Community Psychology, 38(3-4), 31-34. Allen, L. (2003). Girls want sex, boys want love: Resisting dominant discourses of (hetero)sexuality. Sexualities, 6(2), 215-236. Anderson, M. (2005). Negotiating sex . California Law Review, 78(6), 1401-1438. Aosved, A., & Long, P. (2006). Co-occurance of rape myth acceptance, sexism, racism, homophobia, ageism, classism, and religious intolerance. Sex Roles, 55(7), 481-492. Axelrod, T. (2018, Septempber 22). GOP women on Kavanaugh allegations: 'What boy hasn't done this in high school?'. The Hill. https://thehill.com/homenews/senate/407927-gop-women-issue-strong-defenseof-kavanaugh Baird, S., & Jenkins, S. R. (2003). Vicarious traumatization, secondary traumatic stress, and burnout in sexual assault and domestic violence agency staff. Violence and Victims, 18(1), 71-86. Banyard, V., Moynihan, M., & Plante, E. (2007). Sexual violence prevention through bystander education: An experimental evaluation. Journal of Community Psychology, 35(4), 463-381. Beaudrow, J. (2014). The culture of rape: Examining causes and educating for a rape-free society. (Doctoral Dissertation). Available from GenderWatch; ProQuest Dissertations & Theses Gobal. (1563343). Beech, A., Ward, T., & Fisher, D. (2006). The identification of sexual and violent motivations in men who assault women: Implication for treatment. Journal of 195 Interpersonal Violence, 21(12), 1635-1653. Belknap, J. (2007). The invisible woman. Belmont, CA: Wadsworth, Cengage Learning. Beres, M. (2007). “Spontaneous” sexual consent: An analysis of sexual consent literature. Feminism & Psychology , 17(1 ), 93-108. Beres, M., Herold, E., & Maitland, S. (2004). Sexual consent behaviors in same-sex relationships. Archives of Sexual Behavior, 33(5), 475-486. Berg, B. (1995). Qualitative research methods for the social sciences. Needham Heights, MA: Allyn and Bacon. Berkowitz, S. (2014). Letter to White House task force to protect students from sexual assault. Washington, DC: Rape, Abuse & Incest National Network. Blumer, H. (1973). A note on symbolic interactionism. American Sociological Review, 38(6), 797798. Bracha, H. S. (2004). Freeze, flight, fight, fright, faint: Adaptationist perspectives on the acute stress response spectrum. CNS Spectrums, 9(9), 679-685. Briere, J., & Malamuth, N. (1983). Self-reported likelihood of sexually aggressive behavior: Attitudinal versus sexual explanations. Journal of Research in Personality, 17(3), 315-323. Briere, J., Malamuth, N., & Check, J. (1985). Sexuality and rape-supportive beliefs. International Journal of Women's Studies, 8, 398-403. Brownmiller, S. (1976). Against our will: Men, women, and rape. New York, NY: Bantam Books. Buist, C., & Stone, C. (2014). Transgender victims and offenders: Failures of the United States criminal justice system and the necessity of queer criminology. Critical Criminology, 22(1), 35-47. Bureau of Justice Statistics. (2015). National crime victimization survey, 2010-2014. Office of Justice Programs. Washington, DC: Department of Justice. Burgess, G. (2007). Assessment of rape-supportive attitudes and beliefs in college men. Journal of Interpersonal Violence, 22(8). 973-993. Burt, M. (1980). Cultural myths and supports for rape. Journal of Personality and Social Psychology, 38(2), 217-230. Calzo, J. (2014). Applying the pattern-centered approach to understanding how attachment, gender beliefs, and homosociality shape college men's sociosexuality. Journal of Sex Research, 51(2), 221-233. Campbell, R. (1998). The community response to rape: victims' experiences with the legal, 196 medical, and mental health systems. American Journal of Community Psychology, 26, 355379. Campbell, R., Wasco, S., Ahrens, C., Sefl, T., & Barnes, H. (2001). Preventing the "second rape": Rape survivors' experiences with community service providers. Journal of Interspersonal Violence, 16(12), 1239-1259. Canham, M. (2018, December 9). Salt Lake County is now minority Mormon, and the impacts are far reaching. The Salt Lake Tribune. https://www.sltrib.com/religion/2018/12/09/salt-lake-county-is-now/ Capers, B. (2011). Real rape too. California Law Review, 99, 1259-1308. Carr, J. (2013). The SlutWalk movement: A study in transnational feminist activism. Journal of Feminist Scholarship, 4(4), 24-38. Carr, J., & VanDeusen, K. (2004). Risk factors for male sexual aggression on college campuses. Journal of Family Violence, 19(5), 279-289. Carvalho, A., Lewis, R., Derlega, V., Winstead, B., & Viggiano, C. (2011). Internalized sexual minority stressors and same-sex intimate partner violence. Journal of Family Violence, 26(7), 501-509. Castleman, M. (2016, January 14). Evidence mounts: More porn, LESS sexual assault. Psychology Today. https://www.psychologytoday.com/us/blog/all-aboutsex/201601/evidence-mounts-more-porn-less-sexual-assault Child, D. (2006). The essentials of factor analysis (3rd ed.). New York, NY: Continuum International Publishing Press. Choi, G.-Y. (2011). Secondary traumatic stress of service providers who practice with survivors of family or sexual violence: A national survey of social workers. Smith College Studies in Social Work, 47(4), 101-119. Coles, J., Astbury, J., Dartnall, E., & Limjerwala, S. (2014). A qualitative exploration of researcher trauma and researchers' responses to investigating sexual violence. Violence Against Women, 20(1), 95-117. Connell, R. (1987). Gender and power: Society, the person, and sexual politics. Stanford, CA: Stanford University Press. Conrad, B. K. (2006). Neo-institutionalism, social movements, and the cultural reproduction of a mentalite: Promise keepers reconstruct the madonna/whore complex. The Sociological Quarterly, 47(2), 301-331. Conte, J. (1989). What sexual offenders tell us about prevention strategies. Child Abuse & Neglect, 13(2), 293-301. 197 Coxell, A., & King, M. (1996). Male victims of rape and sexual abuse. Sexual and Marital Therapy, 11(3), 297-308. Curtin, R., Presser, S., & Singer, E. (2000). The effects of response rate changes on the index of consumer sentiment. Public Opinion Quarterly, 64, 413-428. Debowska, A., Boduszek, D., Dhingra, K., & DeLisi, M. (2016). The effect of male incarceration on rape myth acceptance. Deviant Behaviors, 37(6), 634-643. DeVellis, R. (2017). Scale development: Theory and applications. Thousand Oaks, CA: Sage Publications. Diehl, C., Glaser, T., & Bohner, G. (2014). Face the consequences: Learning about victim's suffering reduces sexual harassment myth acceptance and men's likelihood to sexually harass. Aggressive Behavior, 40, 489-503. Dripps, D. (1992). Beyond rape: An essay on the difference between the presence of force and the absence of consent. Colombia Law Review, 92(7), 1780-1809. DuMont, J., Miller, K.-L., & Myhr, T. (2003). The role of "real rape" and "real victim" stereotypes in the police reporting practices of sexually assaulted women. Violence Against Women, 9(4), 466-486. Earnshaw, V., Pitpitan, E., & Chaudoir, S. (2011). Intended responses to rape as functions of attitudes, attributes of fault, and emotions. Sex Roles, 64(5), 382-393. Edgar, T., & Fitzpatrick, M. A. (1993). Expectations for sexual interaction: A cognitive test of the sequencing of sexual communication behaviors. Health Communication, 5(4), 239-261. Edwards, K., Turchik, J., Dardis, C., Reynolds, N., & Gidycz, C. (2011). Rape myths: History, individual and institutional-level presence, and implications for change. Sex Roles, 65(11-12), 761-773. Ellis, L. (1991). A Synthesized (Biosocial) Theory of rape. Journal of Consulting and Clinical Psychology, 59(5), 631-642. Eriksson Baaz, M., & Stern, M. (2009). Why do soldiers rape: Masculinity, violence, and sexuality in the armed forces in the congo (DRC). International Studies Quarterly, 53(2), 495-518. Farris, C., Treat, T., Viken, R., & McFall, R. (2008). Sexual coercion and the misperception of sexual intent. Clinical Psychology Review, 28(1), 48-66. Faul, A. C., & van Zyl, M. A. (2004). Constructing and validating a specific multi-item assessment or evaluation tool. In A. R. Roberts & K. R. Yeager, Evidence-based practice manual (pp. 564-584). New York, NY: Oxford University Press. 198 Federal Bureau of Investigation. (2012). An updated definition of rape. Washington, DC: United States Department of Justice. Ferguson, C., & Hartley, R. (2009). The pleasure is momentary... the expense damnable?: The influence of pornography on rape and sexual assault. Aggression and Violent Behavior, 14(5), 323-329. Ferro, C., Carmele, J., & Saltzman, A. (2008). Current perceptions of marital rape: Some good and not-so-good news. Journal of Interpersonal Violence, 23(6), 764-779. Fisher, N., & Pina, A. (2013). An overview of the literature on female-perpetrated adult male sexual victimization. Aggression and Violent Behavior, 18(1), 54-61. Fisher, T., Davis, C., Yarber, W., & Davis, S. (Eds.). (2011). Handbook of sexuality-related measures (3rd ed.). New York, NY: Routledge. Ford, C. (2013, March 12). The two ways to dismantle rape culture. Daily Life. http://www.dailylife.com.au/news-and-views/dl-opinion/the-two-ways-todismantle-rape-culture-20130311-2fvi0.html Franiuk, R., Seefelt, J., & Vandello, J. (2008). Prevalence of rape myths in headlines and their effects on attitudes toward rape. Sex Roles, 58(11-12), 790-801. Franklin, K. (2004). Enacting masculinity: Antigay violence and group rape as participatory theater. Sexuality Research & Social Policy, 1(2), 25-40. Fuse, T., Forsyth, J., Marx, B., Gallup, G., & Weaver, S. (2007). Factor structure of the tonic immobility scale in female sexual assault survivors: An exploratory and confirmatory factor analysis. Journal of Anxiety Disorders, 21(3), 265-283. Geis, G. (1978). Lord Hale, witches, and rape. British Journal of Law and Society, 5(1), 26-44. Gie Yong, A., & Pearce, S. (2013). A beginner's guide to factor analysis: Focusing on exploratory factor analysis. Tutorials in Quantitative Methods for Psychology, 9(2), 79-94. Gillen, K., & Muncer, S. J. (1995). Sex differences in the perceived causal structure of date rape: A preliminary report. Aggressive Behavior, 21(2), 101-112. Gilroy, P., & Carroll, L. (2009). Woman to woman sexual violence. Women & Therapy, 32, 423-435. Ging, D. (2017). Alphas, betas, and incels: Theorizing the masculinities of the manosphere. Men and Masculinities, 1-20. https://doi.org/10.1177/1097184X17706401 Goodman-Delahunty, J., & Graham, K. (2010). The influence of victim intoxication and victim attire on police responses to sexual assault. Journal of Investigative Psychology, 8(1), 1-19. 199 Grubb, A., & Turner, E. (2012). Attribution of blame in rape cases: A review of the impact of rape myth acceptance, gender role conformity and substance use on victim blaming. Aggression and Violent Behavior, 17(5), 443-452. Hamilton, L., & Armstrong, E. (2009). Gendered sexuality in young adulthood: Double binds and flawed options. Gender & Society, 23(5), 589-616. Hermann, C., Babchishin, K., Nunes, K., Leth-Steensen, C., & Cortoni, F. (2012). Factor structure of the Bumby rape scale: A two-factor model. Criminal Justice and Behavior, 39(7), 869-886. Hildebrand, M., & Najdowski, C. (2015). The potential impact of rape culture on juror decision making: Implications for wrongful acquittals in sexual assault trials. Albany Law Review, 78(3), 1059-1086. Hill, M., & Fischer, A. (2001). Does entitlement mediate the link between masculinity and rape-related variables? Journal of Counseling Psychology, 48(1), 39-50. Hine, B., & Murphy, A. (2017). The impact of victim-perpetrator relationship, reputation, and initial point of resistance on officers' responsibility and authenticity ratings towards hypothetical rape cases. Journal of Criminal Justice, 49, 1-13. Hines, D., & Palm-Reed, K. (2015). Predicting improvement after a bystander program for the prevention of sexual and dating violence. Health Promotion Practice, 16(4), 550-559. Ho, A., Sidanius, J., Kteily, N., Sheehy-Skeffington, J., Pratto, F., Henkel, K., . . . Stewar, A. (2015). The nature of social dominance orientation: theorizing and measuring preferences for intergroup inequality using the new SDO7 scale. Journal of Personality and Social Psychology, 109(6), 1003-1028. Hockett, J., & Saucier, D. (2015). A systematic literature review of "rape victims" versus "rape survivors": Implications for theory, research, and practice. Aggression and Violent Behavior, 25, 1-14. Hockett, J., Saucier, D., Hoffman, B., Smith, S., & Craig, A. (2009). Oppression through acceptance? Predicting rape myth acceptance and attitudes toward rape victims. Violence Against Women, 15(8), 877-897. Hudson, W. W. (1994). Developing short form assessment scales. Annual Meeting of the Council on Social Work Education. Atlanta, GA: Council on Social Work Education. Humphreys, T., & Brousseau, M. (2010). The sexual consent scale - revised: Development, reliability, and preliminary validity. Journal of Sex Research, 47(5), 420-428. Hurd, H. (1996). The moral magic of consent. Legal Theory, 2(2), 121-146. Jackson, S., & Cram, F. (2010). Disrupting the sexual double standard: Young women's talk about heterosexuality. British Journal of Social Psychology, 42(1), 113-127. 200 James, S., Herman, J., Rankin, S., Keisling, M., Mottet, L., & Anafi, M. (2016). The report of the 2015 U.S. transgender survey. Washington, DC: The National Center for Transgender Equality. Javaid, A. (2018). Male rape, masculinities, and sexualities. International Journal of Law, Crime, and Justice, 52, 199-210. Jordan, J. (2011). Here we go round the review-go-round: Rape investigation and prosecution- are things getting worse not better? Journal of Sexual Aggression, 17(2), 234-249. Kaiser, H. (1974). An index of factor simplicity. Psychometrika, 39, 31-36. Kassing, L., Beesley, D., & Frey, L. (2005). Gender role conflict, homophobia, age, and education as predictors of male rape myth acceptance. Journal of Mental Health Counseling, 27(4), 311-328. Katz, B., & Burt, M. (1988). Self-blame: Help or hindrance in recovery from rape? Rape and sexual assault II (pp. 151-168). New York, NY: Garland Publishing. Kaufman, M. (1993). Cracking the armour: Power, pain and the lives of men. Toronto, ON: Penguin Books. Kendall, T. (2006). Pornography, rape, and the internet. Paper presented at the Law and Economics Seminar. Clemson University, Ohio. Kerr Melanson, P. (1998). Belief in male rape myths: A test of two competing theories. (Doctoral Dissertation). Available from ProQuest Dissertations & Theses Global. (304461448). Kerstetter, W., & van Winkle, B. (1990). Who decides?: A study of the complainant's decision to prosecute in rape cases. Criminal Justice and Behavior, 17(3), 268-283. Kimmel, M. (2017). The gendered society (6th ed.). New York, NY: Oxford University Press. Klement, K., Sagarin, B., & Lee, E. (2016). Participating in a culture of consent may be associated with lower rape-supportive beliefs. The Journal of Sex Research, 54(1), 130134. Klinkenberg, W. D., & Rose, S. (1994). Dating scripts of gay men and lesbians. Journal of Homosexuality, 26(4), 23-35. Kohl, C., & Robertson, J. (2014). The sexual overperception bias: An exploration of the relationship between mate value and perception of sexual interest. Evolutionary Behavioral Science , 8(1), 31-43. Koss, M., & Gaines, J. (1993). The prediction of sexual aggression by alcohol use, athletic participation, and fraternity affiliation. Journal of Interpersonal Violence, 8(1), 94-108. 201 Koss, M., Dinero, T., Seibel, C., & Cox, S. (1988). Stranger and acquaintance rape: Are there differences in the victim's experience? Psychology of Women Quarterly, 12(1), 1-24. Krahe, B. (1998). Sexual aggression among adolescents: Prevalence and predictors in a German sample. Psychology of Women Quarterly, 22, 537-554. Krahe, B., Waizenhofer, E., & Moller, I. (2003). Women's sexual aggression against men: Prevalence and predictors. Sex Roles, 5(6), 219-232. Kupers, T. (2005). Toxic masculinity as a barrier to mental health treatment in prison. Journal of Clinical Psychology, 61(6), 713-724. Kwong, M. (2003). The intergenerational transmission of relationship violence. Journal of Family Psychology, 17(3), 288-301. Langenderfer-Magruder, L., Whitfield, D., & Walls, N. E. (2016). Experiences of intimate partner violence and subsequent police reporting among lesbian, gay, bisexual, transgender, and queer adults in Colorado. Journal of Interpersonal Violence, 31(5), 855871. Lebowitz, L., & Roth, S. (1994). "I felt like a slut": The cultural context and women's response to being raped. Journal of Traumatic Stress, 7(3), 363-390. Lonsway, K. (2010). Trying to move the elephant in the room: Responding to the challenge of false rape reports. Violence Against Women, 16(12), 1356-1371. Lonsway, K., & Fitzgerald, L. (1994). Rape myths: In review. Psychology of Women Quarterly, 18(2), 133-164. Lonsway, K., & Fitzgerald, L. (1995). Attitudinal antecedents of rape myth acceptance: A theoretical and empirical reexamination. Journal of Personality and Social Psychology, 68(4), 704-711. Lottes, I. (1991). Belief systems: Sexuality and rape. Journal of Psychology and Human Sexuality, 4(1), 37-59. Lugones, M. (2003). Pilgrimages/peregrinajes: Theorizing coalition against multiple oppressions. New York, NY: Lanham, Maryland: Rowman & Littlefied Publishers, Inc. Luthra, R., & Gidycz, C. (2006). Dating violence among college men and women: Evaluation of a theoretical model. Journal of Interpersonal Violence, 21(6), 717-731. Madigan, L., & Gamble, N. (1991). The second rape: Society's continued betrayal of the victim. New York, NY: Lexington Books. Malamuth, N. (1981). Rape proclivity among males. Journal of Social Issues, 37(4), 138-157. 202 Malamuth, N. (1989). The attraction to sexual aggression scale: Part one. The Journal of Sex Research, 26, 26-49. Maniglio, R. (2011). The role of childhood trauma, psychological problems, and coping in the development of deviant sexual fantasies in sexual offenders. Clinical Psychological Review, 31(5), 748-756. Martin, M. (2018, November 9). Christine Blasey Ford has had to move four times, can't go back to work. Metro. https://www.metro.us/news/christine-blasey-ford-has-had-tomove-four-times-cant-go-back-to-work Mason, J. (2013). Posttraumatic growth in female sexual assault survivors. Tennessee Research and Creative Exchange. https://trace.tennesee.edu/utk_graddis/2458 McCauley, J., Ruggiero, K., Resnick, H., & Kilpatrick, D. (2010). Incapacitated, forcible, and drug/alcohol facilitated rape in relation to binge drinking, marijuana use, and illicit drug use: A national survey. Journal of Traumatic Stress, 23(1), 132-140. McCormick, N. (1987). Sexual scripts: Social and therapeutic implications. Sexual and Marital Therapy, 2, 3-27. McGlynn, C. (2011). Feminism, rape and the search for justice. Oxford Journal of Legal Studies, 31(4), 825-842. McKibbin, W., Shackelford, T., Goetz, A., & Starratt, V. (2008). Why do men rape? An evolutionary psychological perspective. Review of General Psychology, 12(1), 86-97. McMahon, S. (2010). Rape myth beliefs and bystander attitudes among incoming college students. Journal of American College Health, 59(1), 3-11. McMahon, S., & Farmer, G. L. (2011). An updated measure for assessing subtle rape myths. Social Work Research, 35(2), 71-81. Mead, M. (1935). Sex and temperament in three primitive societies. New York, NY: Harper Collins Publishers Inc. Megarry, J. (2014). Online incivility or sexual harassment?: Conceptualising women's experiences in the digital age. Women's Studies International Forum, 47, 46-55. Milhausen, R., & Herold, E. (1999). Does the sexual double standard still exist? Perceptions of university women. The Journal of Sex Research, 36(4), 361-368. Miller, P., & Eisenberg, N. (1988). The relation of empathy of aggressive and externalizing/antisocial behavior. Psychological Bulletin, 103(3), 324-344. Mitchell, I., & Beech, A. (2011). Towards a neurobiological model of offending. Clinical Psychology Review, 31(5), 872-882. 203 Moffett, H. (2006). "These women, they force us to rape them": Rape as narrative of social control in post-apartheid South Africa. Journal of South African Studies, 32(1), 129-144. Mosher, D., & Sirkin, M. (1984). Measuring a macho personality constellation. Journal of Research in Personality, 18(2), 150-163. Muehlenhard, C. (1996). The complexities of sexual consent. SIECUS Report, 24(2), 4-7. Muñoz, J. E. (2009). Cruising utopia: The then and there of queer futurity. New York, NY: New York University Press. Murnen, S., Wright, C., & Kaluzny, G. (2002). If "boys will be boys," then girls will be victims? A meta-analytic review of the research that relates masculine ideology to sexual aggression. Sex Roles, 46(11), 359-375. Nagoski, E. (2015). Come as you are. New York, NY: Simon & Schuster. National Sexual Violence Resource Center. (2004). Sexual violence prevention: Beginning the dialogue. Atlanta, GA: Centers for Disease Control and Prevention. Nunnally, J., & Bernstein, I. (1994). Psychometric theory (3rd ed.). New York, NY: McGrawHill. Oakley, A. (2007). Working sex: Sex workers write about a changing industry. Emeryville, CA: Seal Press. O'Brien, C., Keith, J., & Shoemaker, L. (2015). Don't tell: Military culture and male rape. Psychological Services, 12(4), 357-365. O'Donohue, W., Yeater, E., & Fanetti, M. (2003). Rape prevention with college males: The roles of rape myth acceptance, victim empathy, and outcome experiences. Journal of Interpersonal Violence, 18(5), 514-531. Olive, V. (2012). Sexual assault against women of color. Journal of Student Research, 1, 1-9. Ostler, T. A. (2003). Verbal and nonverbal dating behaviors and sexual consent: Implications for miscommunication between men and women. Reno, NV: University of Nevada Press. Page, A. D. (2007). Behind the blue line: Investigating police officers' attitudes toward rape. Journal of Police and Crminal Psychology, 22(1), 22-32. Page, A. D. (2008). Gateway to reform? Policy implications of police officers' attitudes toward rape. American Journal of Criminal Justice, 33(1), 44-58. Panichas, G. (2001). Rape, autonomy, and consent. Law & Society Review, 35, 231-269. Patterson, D. (2011). The impact of detectives' manner of questioning on rape victims' disclosure. Violence Against Women, 17, 1355-1364. 204 Payne, D. (2018, October 1). 3 big questions hanging after Christine Blasey Ford's testimony on Brett Kavanaugh. USA Today. https://www.msn.com/en-ca/news/world/3-bigquestions-hanging-after-christine-blasey-fords-testimony-on-brett-kavanaugh/arBBNOkXZ Payne, D., Lonsway, K., & Fitzgerald, L. (1999). Rape myth acceptance: Exploration of its structure and its measurement using the illinois rape myth acceptance scale. Journal of Research in Personality, 33, 27-68. Peterson, Z., & Muehlenhard, C. (2004). Was it rape? The function of women's rape myth acceptance and definitions of sex in labeling their own experiences. Sex Roles, 51(3/4), 129-144. Pettigrew, T., & Tropp, L. (2008). How does intergroup contact reduce prejudice? Metaanalytic tests of three mediators. European Journal of Social Psychology, 38(6), 922-934. Potter, S., Fountain, K., & Stapleton, J. (2012). Addressing sexual and relationship violence in the LGBT community using a bystander framework. Harvard Review of Psychiatry, 20(4), 201-108. Power, J., McNair, R., & Carr, S. (2009). Absent sexual scripts: Lesbian and bisexual women's knowledge, attitudes and action regarding safer sex and sexual health information. Culture, Health & Sexuality, 11(1), 67-81. Rennison, C. M. (2002). Rape and sexual assault: Reporting to police and medical attention, 19922000. U.S. Department of Justice. Washington, DC: Office of Justice Programs. Rich, K. (2014). Skills for interviewing rape victims. Interviewing rape victims: Practice and policy issues in an international context (pp. 36-72). Palgrave, UK: Macmillan. Rich, M., Utley, E., Janke, K., & Moldoveanu. (2010). "I'd rather be doing something else:" Male resistance to rape prevention programs. The Journal of Men's Studies, 18(3), 268288. Richards, T., & Marcum, C. (2015). Sexual victimization: Then and now. Los Angeles, CA: Sage Publications. Rumney, P. (2010). Gay male rape victims: Law enforcement, social attitudes, and barriers to recognition. The International Journal of Human Rights, 13(2-3), 233-250. Sanday, P. R. (1981). The socio-cultural context of rape: A cross-cultural study. Journal of Social Issues, 37(4), 5-27. Sanday, P. R. (1996). Rape-prone versus rape-free campus cultures. Violence Against Women, 2(2), 191-208. Sanday, P. R. (2003). Rape-free versus rape-prone: How culture makes a difference. In C. Brown Travis (Ed.), Evolution, gender, and rape (pp. 337-361). Boston, MA: 205 Massachusetts Institute of Technology. Schwartz, M., & Pitts, V. (1995). Exploring a feminist routine activities approach to explaining sexual assault. Justice Quarterly, 12(1), 9-31. Schwendinger, J., & Schwendinger, H. (1974). Rape myths: In legal, theoretical, and everyday practice. Crime and Social Justice, 1, 18-26. Scott, R., & Tetreault, L. (1987). Attitudes of rapists and other violent offenders toward women. Journal of Social Psychology, 127(4), 375-381. Sheffield, C. (1998). Sexual terrorism: The social control of women. In D. Anselmi & A. Law, Questions of gender: Perspectives and paradoxes (pp. 642-654). Boston, MA: McGrawHill. Simon, W., & Gagnon, J. (1984). Sexual scripts. Society, 22(1), 53-60. Sjoberg, L., & Via, S. (2010). Gender, war, and militarism: Feminist perspectives. Santa Barbara, CA: ABC-CLIO, LLC. Sleath, E., & Bull, R. (2009). Male rape victim and perpetrator blaming. Journal of Interpersonal Violence, 6, 969-988. Sleath, E., & Bull, R. (2017). Police perceptions of rape victims and the impact on case decision making: A systematic review. Aggression and Violent Behavior, 34, 102-112. Sloan, L., & Wahab, S. (2000). Feminist voices on sex work: Implications for social work. Affilia, 15(4), 457-479. Smith, T., Davern, M., Freese, J., & Hout, M. (2016). General social surveys, 1972-2016. Chicago, IL: National Opinion Research Center. Snyder, E. (2008). The female-to-male rape myths acceptance scale: Initial scale development. (Doctoral Dissertation). Available from ProQuest Dissertations & Theses Global. (304461448). Spohn, C. (1999). The rape reform movement: The traditional common law and rape reform laws. Jurimetrics, 39(2), 119-130. Struckman-Johnson, C., & Struckman-Johnson, D. (1992). Acceptance of male rape myths among college men and women. Sex Roles, 27, 85-100. Struckman-Johnson, C., & Struckman-Johnson, D. (2003). Tactics of sexual coercion: When men and women won't take no for an answer. The Journal of Sex Research, 40(1), 76-86. Testa, M., & Livingston, J. (2009). Alcohol consumption and women's vulnerability to sexual victimization: Can reducing women's drinking prevent rape? Intimate Partner Violence, 44(9-10), 1349-1376. 206 Thakker, J., & Ward, T. (2012). An integrated theory of sexual reoffending. Psychiatry, Psychology, & Law, 2, 236-248. Theriault, A. (2015, February 13). Let's call female online harassment what it really is: Terrorism. Vice News. https://www.vice.com/en_us/article/nnqxe8/lets-call-femaleonline-harassment-what-it-really-is-gender-terrorism-481 Thornhill, R., & Palmer, C. (2000). A natural history of rape. Cambridge, MA: MIT Press. Tolentino, J. (2018, September 20). After the Kavanaugh allegations, Republicans offer a shocking defense: Sexaul assault isn't a big deal. The New Yorker. https://www.newyorker.com/news/our-columnists/after-the-kavanaughallegations-republicans-offer-a-shocking-defense-sexual-assault-isnt-a-big-deal Tompkins, T., Shields, C., Hillman, K., & White, K. (2015). Reducing stigma toward the transgender community: An evlauation of a humanizing and perspective-taking intervention. Psychology of Sexual Orientation and Gender Diversity, 2(1), 34-42. Tracy, C., Fromson, T., Gentile-Long, J., & Whitman, C. (2011). Rape and sexual assault in the legal system. National Research Council of the National Academies, Women's Law Project. Philadelphia, PA: Bureau of Justice Statistics. Travis, C. B. (Ed.). (2003). Evolution, gender, and rape. Cambridge, MA: Massachusetts Institute of Technology Press. Tuerkheimer, D. (2016). Affirmative consent. Ohio State Journal of Criminal Law, 13(2), 441468. Turchik, J., Hebenstreit, C., & Judson, S. (2016). An examination of the gender inclusiveness of current theories of sexual violence in adulthood: Recognizing male victims, female perpetrators, and same-sex violence. Trauma, Violence, & Abuse, 17(2), 133-148. Tyler, K., Hoyt, D., & Whitbeck, L. (1998). Coercive sexual strategies. Violence and Victims, 13(1), 47-61. Ullman, S. (2007). A 10-year update of "review and critique of empirical studies of rape avoidance. Criminal Justice and Behavior, 34(3), 411-429. Viki, G. T., Abrams, D., & Masser, B. (2004). Evaluating stranger and acquaintance rape: The role of benevolent sexism in perpetrator blame and recommended sentence length. Law and Human Behavior, 28(3), 295-303. Walker, S. J. (1997). When "no" becomes "yes": Why girls and women consent to unwanted sex. Applied and Preventitive Psychology, 6(3), 157-166. Watford, D. (2010). Statistics on perpetrators of child sexual abuse. Washington, DC: The National Center for Victims of Crime. 207 Watson-Franke, M.-B. (2002). A world in which women move freely without fear of men: An anthropological perspective on rape. Women's Studies International Forum, 25(6), 599-606. Weber, M. (1919). Politics as a vocation. In H. Gerth & C. Wright Mills, From Max Weber: Essays in sociology (pp. 77-128). New York, NY: Oxford University Press. Wegner, R., Abbey, A., Pierce, J., Pegram, S., & Woerner, J. (2015). Sexual assault perpetrators' justifications for their actions: Relationships to rape-supportive attitudes, incident characteristics, and future perpetration. Violence Against Women, 1018-1037. Wertheimer, A. (1996). Consent and sexual relations. Legal Theory, 2(2), 89-112. Wolcott, H. (2005). The art of fieldwork (2nd ed.). Lanham, MD: AltaMira Press.