TERRORISM By Skyler J. Morgan A THESIS Submitted to OVERSHADOWED: THE MEDIA’S REPRESENTATION OF GENDER AND DOMESTIC Michigan State University in partial fulfillment of the requirements for the degree of Criminal Justice – Master of Science 2019 ABSTRACT OVERSHADOWED: THE MEDIA’S REPRESENTATION OF GENDER AND DOMESTIC TERRORISM By Skyler J. Morgan Women are underrepresented in the mass media and are most frequently portrayed in stereotypical and limited ways. Recent efforts to expand our understanding of the media’s representation of gender have examined the characterization of women in crimes stories as both victim and offenders. One avenue of this research focuses on the portrayal of women involved in terrorism. Although this body of research is limited and has been strictly qualitative, it has provided a useful framework for the current study. This current study sought to expand our knowledge of the media’s representation of gender and crime by systematically testing the six frames identified by Brigitte Nacos (2005) that she argues are used to explain women’s involvement in terrorism. Using the United States Extremist Crime Data Base (ECDB), this study analyzed the news media articles that covered male and female perpetrators between the years of 1990 and 2016 from far-right, animal/environmental rights and jihadist movements. Findings indicated limited support for the original frames identified by Nacos (2005). However, significant differences were found in the terrorist for the sake of love frame, such that female perpetrators were more likely to be presented through this frame. The current study identified four new frames used by the media to discuss perpetrators of terrorism, which included the disregarded frame, the overshadowed frame, the fact frame, and the life frame. Of new frames identified in the current study, the overshadowed frame had large statistically significant differences between male and female perpetrators. Copyright by SKYLER J. MORGAN 2019 Dedicated to my nieces and nephews: Eden Eleanor, Willow Faye, Reagan Remington, Reed Henry, Lakin McKenna, and Drake Morgan. Every smile inspires me to make this world a better place for you. I love you all. iv ACKNOWLEDGEMENTS I would like to start by expressing my profound gratitude to my chair, Dr. Steven Chermak, who guided and supported me through this process. I truly appreciate the time and energy that you have put into me over the past two years. Thank you for believing in me. I would like to thank Dr. Christina DeJong for being a committee member and consistently being one of the most genuine people I have ever met. I appreciate your unique approach to scholarship, and I hope that someday soon I will reflect what I have learned from you when I’m on the other side. I would also like to thank Dr. Karen Holt for being a committee member and providing me with excellent feedback that shaped the final product. I am very grateful for your time and support I would like to thank my supportive, loving, and caring family. A special thanks to my parents, Jon and Kimberly Morgan, for the sacrifices you have made to get me here. Also, to my grandparents, Cleon and Eleanor Morgan, I am most grateful for you. None of this would be possible without your unwavering love, kindness, and support. I love you more. A special thank you to my colleagues who have turned into lifelong friends throughout my master’s degree: Ariel L. Roddy, Youngjae (David) Nam, Kayla M. Hoskins, Jasmine Newton, Roberta Liggett, and Jin R. Lee. Lastly but most importantly, to the two people who I consider my very best friends and soulmates: Hugh John Hales III and Rachel Marie Dazer. Thank you for your consistent support and loyalty. v TABLE OF CONTENTS LIST OF TABLES ........................................................................................................................ vii INTRODUCTION ...........................................................................................................................1 A REVIEW OF EXISTING LITERATURE ...................................................................................7 Representations of Gender in the Media ....................................................................................7 Representations of Gender in Crime Stories ............................................................................14 Reviewing the Literature on Terrorism and Gender ................................................................17 METHODOLOGY ........................................................................................................................24 Overview ..................................................................................................................................24 Method .....................................................................................................................................26 RESULTS ......................................................................................................................................31 DISCUSSION ................................................................................................................................50 APPENDICES ...............................................................................................................................56 APPENDIX A: News Sources Utilized ...................................................................................57 APPENDIX B: Codebook for Thesis ......................................................................................59 REFERENCES ..............................................................................................................................70 vi LIST OF TABLES Table 1 Descriptive Statistics of the Articles (N=270) ..................................................................30 Table 2 Descriptive Statistics of the Frames for Female Perpetrators (N=135) ..........................31 Table 3 Descriptive Statistics of the Frames for Male Perpetrators (N=135) ..............................32 Table 4 Bivariate Analysis of Frames and Representation by Gender (N=270) ...........................37 Table 5 Bivariate Analysis of Frame by Extremist Ideology (N=270) ..........................................39 Table 6 Bivariate Analysis of Frames by Extremist Ideology and Gender (N=270) ....................44 Table 7 Binary Logistic Regression of the Terrorist for the Sake of Love Frame (N=270) ..........46 Table 8 Binary Logistic Regression of the Disregarded Frame (N=270) .....................................47 Table 9 Binary Logistic Regression of the Fact Frame (N=270) ..................................................48 Table 10 Binary Logistic Regression of the Overshadowed Frame (N=270) ...............................49 vii INTRODUCTION There is a common understanding among scholars that sex is biological, and gender is socially and culturally constructed. Specifically, sex signifies chromosomal differences in people, which designates the categorization of male and female. In contrast, gender denotes masculine or feminine cultural norms that are primarily learned and taught (Richardson & Wearing, 2014). These ideas are shaped in a social constructionist view, meaning that behavior is not pre-determined by their biology, but rather shaped by society (Gauntlett, 2002). In the early research, West & Zimmerman (1987) made the critical distinction between sex and gender and argued that “doing gender” means that our culture creates and enforces differences between girls and boys and women and men that are not natural, essential, or biological. For example, culture in the United States shapes and influences the differences in what it means to be a man or a woman. While scholars acknowledge the difference between biological sexes, they are more concerned with the much larger differences in the social construction of gender (Craig, 1992). Since this conception, research has taken different avenues to understand the consequences, outcomes, and significance of living in a society that socially constructs gender. The media’s representation of gender is one of the primary avenues that gender scholars have used to examine the cultural and social construction of gender. Media in many forms (news, television, advertisements, etc.) is entrenched in the fabric of our society and research has indicated that more time is devoted to media than any other “waking” activity (Roberts, 2000). Because it is such an integral part of our lives, the influence that media has on our identity, beliefs, and understanding of the world around us is significant. Media has large impacts specifically on our perceptions of gender and individual’s masculine or feminine identities. Since scholars recognize that gender is socially constructed, many tend to focus on the representations of masculinity and femininity that we see daily in the media (Craig, 1992). Wood (1994) argues 1 that media is the most powerful and pervasive influence on how we as a society view men and women. As a result, there has been a great deal of research on gender and the media, commonly known as feminist media studies. Two subjects are most frequently examined in feminist media studies – the differences in how the media represents women and men, and how the media influences society. In regard to the first subject, research has shown that media is gendered, meaning that there is a large difference in the way media represents men and women and, consequently, masculinity and femininity (Richardson & Wearing, 2014). Historically, women have been relatively invisible in the media when compared with their male counterparts (Tuchman, 1978). However, when they are visible, they are most frequently portrayed as “household functionaries” or as “men’s domestic adjuncts” (Busby, 1975). Depictions of men and masculinity are different such that they construct an image of strength, dominance and aggressive behaviors (Messner, Dunbar, & Hunt, 2000; Wood, 1994). In addition to an image of power and independence, men and masculinity are represented certainly not as feminine or emotional (Wood, 1994). Due to this stark difference in the representation of men and women, media have a large influence at the individual level on gender identity and perceptions of gender roles. This leads to the second concept that is frequently studied in regard to the media’s representation of gender: how the media influences society. The mass media is much more pervasive and influential than many recognize. Research has indicated that the mass media has large influences on public thinking, behavior, and policy (Christen & Gunther, 2003; Entman, 1989; Green-Pedersen & Stubager, 2010; Happer & Philo, 2013). For example, research has shown that the mass media has influence on the agendas of political parties, individual political involvement, perceived public opinion, and the individual’s projections of their personal opinion on others (Christen & Gunther, 2003; Green-Pedersen & Stubager, 2010; Gunther, 1998; 2 Scheufele, Shanahan, & Kim, 2002). In regard to decision making and health-related behaviors, the literature has indicated that the mass media can have large impacts on individuals understanding and behaviors regarding their health (Boles, Adams, Gredler, & Manhas, 2014; Wakefield, Loken, & Hornik, 2010; Young, Norman, & Humphreys, 2008). Furthermore, research has also indicated mass media has also been shown to influence perceptions around sexuality and behavior of LGBTQ youth (Brown, 2002; Craig & McInroy. 2014; Gillig & Murphy, 2016). Taken together, it is clear that research has indicated that the media has a strong influence on public opinion, decision-making, and policy. Through acknowledging the strong influence of the mass media, recent research has also examined the influence of the mass media on gender. Taking into consideration the contrasting representations of gender in the media, this research seeks to understand the influences of gendered media on men and women. Broadly speaking, this body of work indicates that media reinforces stereotypical masculine and feminine roles (Ey, 2014; Goodall, 2012). Furthermore, it has indicated that media has distinct impacts on the body image and identity of men and women (Benowitz-Fredericks, Garcia, Massey, & Borzekowski, 2012; Gauntlett, 2002; Wright & Pritchard, 2009). Although it is difficult to untangle or explain the influence of the gendered media at the individual level, it is clear from the literature that the media reinforces these stereotypical gendered roles and influences identity. Recent efforts to deepen our understanding of the media’s representation of gender have looked at the representation of female criminals in the media (Benson & Gottschalk, 2015; Collins, 2016; Easteal, Holland, & Judd, 2015; Gottschalk & Smith, 2015; Howe, 1997; Lloyd & Ramon, 2017). This literature has primarily sought to understand if and how women are characterized in the media in regard to their criminal activity. Though criminal activity can be considered deviant behavior for both men and women, there tends to be more leniency when men 3 are engaged in crime, as it more closely fits society’s understanding of masculinity traits including dominance and aggressiveness. When women engage in crime, they tend to be perceived as more deviant as they are contradicting the ideals of femininity (Friedman, 2008). Society is more tolerant of men’s deviant and criminal behavior. When women do commit crime, there is a conflicting narrative that the media must construct, especially considering their stereotypical portrayals of women and femininity. Past research has examined the representation of gender in many different forms of crime, including white-collar crime, sexual offending, and intimate partner violence (Benson & Gottschalk, 2015; Christensen, 2018; Gottschalk & Smith, 2015; Landor & Eisenchlas, 2012; Lloyd & Ramon, 2017). Generally speaking, this research has indicated that the media creates a stereotypical narrative of women that is often a deceptive representation of women as victims and offenders. For example, women are often blamed for their circumstances as victims, while female offenders are presented as either sexualized bad girls or malicious black widows (Collins, 2011; Lloyd & Ramon, 2017). Furthermore, depending on the context of the situation and the type of crime committed, the media have ways in which they will commonly depict women as victims and offenders of crime. For example, the media tends to create a dangerous and unsympathetic narrative for female sex offenders (Christensen, 2018), an out of touch with reality narrative for female terrorists (Nacos, 2005), and frequently ignores female white-collar criminals (Benson & Gottschalk, 2015; Gottschalk & Smith, 2015). A more recent avenue of this research has looked at the media’s portrayal of gender and terrorism. This literature is limited, as there are only three studies that have examined the media’s representation of women in terrorism (Friedman, 2008; Nacos, 2005; Plaza, Rivas- Nieto, & Rey-García, 2017). Brigitte Nacos (2005) pioneered this work by arguing that the frames used to represent female terrorists were similar to the frames used when the media 4 represents female politicians. There were six frames identified in her work, which included the physical appearance frame, the family connection frame, terrorist for the sake of love, the women’s liberation/equality frame, the tough-as-males/tougher-than-men frame, and the bored, naïve, out-of-touch-with-reality frame. Nacos (2005) argued that these frames were used by the media to misrepresent female terrorists and discussed the possible negative influences of these frames on future counterterrorism policies. Friedman (2008) followed up this work by focusing on the representation of female suicide bombers. In this study, five themes were identified that were frequently used to explain the motivations of female suicide bombers, including strategic desirability, the influence of men, revenge, desperation, and liberation. Friedman (2008) argued that female suicide bombers challenge universally shared ideas of women and femininity. Plaza et al. (2017) is the most recent study to look at the representation of female terrorist in the media. Their analysis focused primarily on the characterization of women terrorist in Spanish newspapers. This work extended the findings of Nacos (2005) by providing new frames found in Spanish newspapers, including the discourse of exceptionalism, the discourse of expertise, the discourse of sensationalism, and the discourse of bravery. While there are few studies that have looked at the media’s representation of female terrorists (Friedman, 2008; Nacos, 2005; Plaza et al., 2017), to date there has been no systematic study of the media’s representation of gender in terrorist organization in the United States. This study seeks to expand our understanding of the media’s representations of gender and crime. This study will analyze the media’s representation of male and female perpetrators of violent terrorist acts. Using the Extremist Crime Database (ECDB), this study will take a random sample of 90 male and female perpetrators from far-right, far-left, and jihadist organizations between the years of 1990 and 2016 (15 female/15 male perpetrators from each movement). For 5 every perpetrator in the ECDB, there is all open source material pulled from 26 search engines about the respective perpetrator. Most of these materials are from newspaper reports. When the sample was determined, this study analyzed the news media that is associated with each perpetrator within each case file. The primary focus was to systematically test the six frames identified by Nacos (2005) in her work on the media’s representations of female terrorists. To account for the large variation in the number of articles within each case file, three newspaper articles from each case file were randomly selected and analyzed. Once the newspaper articles were determined, each article was then systematically analyzed for the six major frames. More specifically, the text in the newspaper articles were used to classify how each perpetrator was being represented. Each frame was analyzed within each article individually, as there could be multiple frames used to represent a male or female perpetrator. In addition to the primary analysis, an overall count of the number of newspaper articles within each case file will be analyzed to better understand the visibility of gender in the news reporting of male and female perpetrators of terrorism. Furthermore, this study will consider how the media represents masculinity in the male perpetrators and the differences in the representation of male and female extremists in the media. 6 A REVIEW OF EXISTING LITERATURE Representations of Gender in the Media The media has a long history of pushing a narrative of traditional masculinity and femininity through the construction of stereotypical ideas of men and women (Gauntlett, 2002; Gill, 2007; Krijnen & Bauwel, 2015; MacKinnon, 2003; Richardson & Wearing, 2014). Based in social constructionism, feminist media studies have produced a diverse body of research showing the modern western world’s representation of gender in many forms of media. These forms of media include advertisements, news, magazines, television, movies, and music. Regardless of the type of media, scholars have recognized that the representations of men and women in the media are distinct, and that these differences are a larger illustration of the inequality between men and women in society (Gauntlett, 2002). The large differences between the representation of gender can be seen in both the frequency and nature of the representation of men and women (Gunter, 1995). Due to these vast differences in the representations of gender in the media, scholars have concluded that all media is gendered in some way (Ross, 2010; Wood, 1994). Analyses have also indicated that the gendered media has a significant influence at the individual level, primarily with identity and self-image (Gauntlett, 2002). Although there has been a well-documented and consistent pattern in the media’s portrayal of gender over the past five decades, scholars have noticed a recent transformation in these portrayals (Gauntlett, 2002; Gill, 2007). Women’s representation in media has long been characterized by invisibility, dependence, and submissiveness. Tuchman (1978) suggested that the representation of women in media was symbolic annihilation, meaning that women are not equally represented, and when they are represented, they are confined to traditional stereotypical roles. There has been a long trend in the media to underrepresent women, in terms of frequency of representation compared to men (Baumann & de Laat, 2012; Davis, 1982; Humprecht & Esser, 2017; Krijnen & Bauwel, 7 2015; Lont & Bridge, 2004; Potter, 1985; Shor, van de Rijt, Miltsov, Kulkarni, & Skiena, 2015). Regardless of the fact that women and men make up equal proportions of our society, the media often creates a false narrative that men vastly outnumber women by showing men much more than women. Research has documented this underrepresentation in newspapers for the past four decades. For example, in the mid-1980s 2,224 news stories were analyzed from five elite U.S. newspapers and of these new stories 61.6% featured men, 27.8% featured neither men or women, 3.2% featured both men and women, while only 7.3% featured women (Potter, 1985). Similarly, Greenwald (1990) analyzed the representation of gender in newspaper business sections and found similar results, such that women were underrepresented as both subjects and reporters. Furthermore, Luebke (1985) reported that of 497 new stories only 16.5% were news about women. Despite the fact that women create more than half of the population in California, DeLouth, Pirson, Hitchcock, & Rienzi (1995) found that in three California newspapers women were consistently disproportionally underrepresented in photographic images in national and sports sections. More recent research documents the persisting underrepresentation of women in the newspapers and even suggests that having a female publisher or executive editor has no influence on the representation of women (Shor et al., 2015). The Global Media Monitoring Project, the world's longest longitudinal research on gender equality in the news, has consistently reported the underrepresentation of women in the media since 1995. Although there was a gradual increase in representation from 1995 to 2010, they reported a halt in the progress of equal representation in the media from 2010 to 2015 (GMMP, 2015). In a study of Spanish newspapers, Matud, Rodríguez, and Espinosa (2011) found that women only appeared in 6.8% of 3,206 articles and that men were more likely to be cited as sources than women. Taken 8 together, the years of research on gender, representation, and newspapers have consistently documented the incredibly limited representation of women and indicates that newspapers have reinforced the false narrative that men outnumber women in our society. Scholars have found that there is also an intersectionality component to the media’s representation of gender. Specifically, when gender and race or gender and age are considered simultaneously, we see even less representation (Araüna, Dhaenens, & Van Bauwel, 2017). For example, the underrepresentation of older people in advertising and other mass media is much more significant for older women when compared to older men (Baumann & de Laat, 2012; Edström, 2018). Furthermore, research has shown that black women are proportionally less likely to be represented in popular magazines relative to white women (Schug, Alt, Lu, Gosin & Fay, 2017). There is also a large underrepresentation in the portrayal of black women as victims in the media regarding violent crimes and homicide when compared to their white counterparts (Neely, 2015). It’s also important to note that the underrepresentation of women in the media is paralleled with the underrepresentation of women in fields like politics and science (Ceci, Williams, & Thompson, 2011; Teele, Kalla, & Rosenbluth, 2018). The media creating a narrative that women are invisible is a symbolic illustration of the inequality between men and women in society. Not only does the invisibility of women in the media create a deceitful narrative that men outnumber women, but it also indicates that men are the societal standard (Wood, 1994). When women are given representation in the media, there is a consistent stereotypical narrative of women and femininity (Collins, 2011; Gauntlett, 2002; Gill, 2007; Gunter, 1995; Richardson & Wearing, 2014; Wood, 1994). The narrative depicts women as white, young, thin housewives whose existence are defined by their relationships with her family and housework (Wood, 1994). Gendered media constructs a false narrative that women all look the same and 9 have similar priorities. These women are most commonly depicted as passive, objects, weak, and dependent on a man (Busby, 1975; Collins, 2011; Gauntlett, 2002; Gill, 2007; Gunter, 1995; Lauzen, Dozier, & Horan, 2008; Richardson & Wearing, 2014; Wood, 1994). At the same time, women are also frequently sexualized in media, with a strong emphasis on their sexual appearance and appeal (Ward, 2016). Historically speaking, women’s existence in the media has little to do with independence, intelligence, or power, but a strong and distorted emphasis on image, dependency, and incapability. Although feminist media scholars recognize that there are instances when women are not represented by the media in this way, more often than not they are portrayed in a similar stereotypical way (Collins, 2011). There has also been a trend that when women in media break the traditional stereotypical roles and are portrayed as working women that they are often condemned and trivialized (Tuchman,1978). Furthermore, research indicates differences in the ways that different types of media represent women. For example, women in television and movies are often shown as weak, in need of protection, and dependent on the male lead characters, while women in advertising and magazines are portrayed in stereotypical housewife gender roles, with an emphasis on beauty (Gauntlett, 2002). Although different types of media may portray women differently, there is still the overarching issue that the portrayals of women and femininity are constructed in ways that trivialize, disrespect, and generalize women. The representation of men and masculinity in the media can be characterized by almost the exact opposite of women’s representation: visibility, independence, and strength. Regardless of the media type, men’s prominence and representation in the media greatly surpasses that of women’s (Baumann & de Laat, 2012; Humprecht & Esser, 2017; Lont & Bridge, 2004; Potter, 1985; Shor et al., 2015). For example, an analysis of newspapers indicated that men were eight times more likely to be the main character in front-page news stories than women (Davis, 1982). While women are underrepresented, men are overrepresented in the media, especially when 10 considering the almost equal proportion of men and women in society. The media’s consistent portrayal of more men than woman and men as the social standard is a symbolic illustration of the inequality between men and women in society. When men are represented in the media, there is a consistent stereotypical narrative of men and masculinity (Craig, 1992; Gauntlett, 2002; Gunter, 1995; MacKinnon, 2003; Richardson & Wearing, 2014; Wood, 1994). However, this narrative does not limit or diminish the value of men in the same way it does for women. Men are most frequently presented as active, powerful, and confident in the media (Gauntlett, 2002; Wood, 1994). Masculinity in the media embodies strength, independence, and even at times dominance. Although men are arguably represented in more positive ways than women (independent, confident, etc.), they can also be portrayed in undesirable ways (e.g. as violent or aggressive). Their stereotype pushes a narrative that masculinity is defined by the power and control over people. From war films to classic westerns, to the Bond films, men have been long represented as heroic, tough and hypermasculine (MacKinnon, 2003). Media depictions not only favor the narrative of the “manliest men” but the media rarely portrays men doing stereotypical female activities, such as housework or cooking (Wood, 1994). Additionally, when the media has representations of men that do not adhere to the stereotypical narrative of masculinity, there are common distinctions that this behavior is atypical. For example, research has shown that newspapers representations of eating disorders in men, which do not fit the stereotypical narrative of masculinity, use language to ensure that eating disorders are “gender anomalous” for men (MacLean, Sweeting, Walker, Patterson, Räisänen, & Hunt, 2015). The literature has also shown that newspaper’s representations of men’s dieting has a detachment from stereotypical feminized types of dieting (Gough, 2007). In essence, the media both forces a narrative of masculinity and anti-femininity for men. Past research indicates that when the media does represent men or masculinity in a non- 11 traditional way, it makes a distinction that the representation is anomalous, which even further drives the stereotypical narrative of masculinity. Research has also examined the media’s representation of relationships between men and women. In her work on gendered media, Wood (1994) identified four common themes in which relationships between men and women are represented in the media. These include women’s dependence/men’s independence, women’s incompetence/men’s authority, women as primary caregivers/men as breadwinners, and women as victims and sex objects/men as aggressors. These themes and dynamics that the media portrays are often seen in male-female romantic relationships (Galician, 2004; Gill, 2007), and can also be shown in a workplace setting with a male boss and a female subordinate. In male-female relationships, women are generally represented as dependent on men, incompetent, caregivers, and sexual objects while men are independent, authoritarian, breadwinners, and aggressors, which does not deviate from the individual stereotypical representations of men and women. These narratives of men’s and women’s relationships are incredibly limiting and reinforcements of the broader stereotypical narrative of masculinity and femininity. For example, men are rarely portrayed as dependent on the woman, especially in terms of financial income or knowledge (Wood, 1994). The media’s depiction of male-female relationships highlights the polarizing narratives of masculinity and femininity. Recognizing that media plays a significant role in the social construction of femininity and masculinity, scholars have studied the influence of the gendered media on society. This literature indicates that the media’s construction of masculinity and femininity can have large effects of the identity and self-image of both men and women (Benowitz-Fredericks et al., 2012; Gauntlett, 2002; Goodall, 2012; Ey, 2014; Wood, 1994: Wright & Pritchard, 2009). As for the media’s representations of relationships between men and women, Wood (1994) argues that this 12 media normalizes violence toward women. The common structure of the unequal relationships depicted in the media, especially in romantic relationships, normalizes violence towards women because women are most often dependent and objects while the men have power and dominance. In terms of influence, the mass media has made little progress in recognizing or addressing the consequences of gendered media. A reoccurring theme in recent literature is the shift and the changes in the media’s representation of gender. Some scholars have argued that there has been an increase in the representation of women and a decrease in the stereotypical narratives forced by the media regarding gender (Gauntlett, 2002; Gill, 2007). Women in certain media are making progress in terms of the roles they are playing and their overall visibility. However, sources like the Global Media Monitoring Project have found a halt in the progress of gender representation of women in the news (GMMP, 2015). Though they had documented a gradual increase in the overall representation of women in the mass media from 1995 to 2010, in their most recent report they found no increase in the representation of women in the mass media from 2010 to 2015. Although these may seem like conflicting findings, it is likely the representation of women is shifting in specific media like television or movies, while the larger mass media is still seeing the long-term trends of the unequal representation of gender. For example, the visibility of women in television are increasing because of shows like Orange is the New Black, Big Little Lies, and Handmaids Tale, which feature primarily female casts. In these shows, women are also breaking the limiting and stereotypical narratives most frequently portrayed by the media. Though there may be a few isolated examples of the shift in the representation of gender in the media, the consistent trend of underrepresentation and stereotypical narratives of gender in the media remains prevalent. 13 Representations of Gender in Crime Stories Several studies have looked at the media’s representation of gender in crime stories, emphasizing women as both victims and offenders of crime. Paralleled with the media’s general representation of women, there is a trend in crimes stories of both underrepresentation of women and a stereotypical narrative constructed by the media that trivializes women. As victims of crime, women tend to be represented in the media as either “bad victims” or “good victims”, which is accompanied by a great deal of victim blaming. There is also a trend in the media’s underrepresentation of women of color as victims, as they are deemed “deserving victims” compared to white women who are often depicted by the media as “undeserving victims” (Neely, 2015). The media also frequently fails to discuss or frame the underlying social causes of the trend in violence against women (Collins, 2016; Easteal et al., 2015; Howe, 1997; Lloyd & Ramon, 2017). As perpetrators of crime, women are often underrepresented in their criminal activity (Benson & Gottschalk, 2015; Gottschalk & Smith, 2015). Research has looked at the representation of women in different types of crimes, including white-collar, sexual offending, interpersonal violence, murder, and violence against women (Benson & Gottschalk, 2015; Christensen, 2018; Gottschalk & Smith, 2015; Landor & Eisenchlas, 2012; Lloyd & Ramon, 2017; Seal, 2010). When women are represented in the media, they are most frequently depicted as interlopers, more masculine, or sexualized as bad girls, of all which depends on the context of the situation and type of crime (Collins, 2016; Nacos, 2005; Friedman, 2008). Although there can be different representations of women depending on the type of crime, the media still creates a stereotypical narrative of women that is often a deceptive representative of women as victims and offenders. 14 Although there has been some research that examines how women are portrayed for specific types of crimes, there have only been a few studies on the media’s depiction of female terrorists. In her pioneering work on the representations of female terrorists, Brigitte Nacos pointed out similarities between the representation of female politicians and female terrorists in the mass media. Nacos (2005) conducted a content analysis of both print and broadcast news and existing literature on female terrorist. Using LexisNexis and Proquest, Nacos (2005) examined news and articles that resulted from searches that included these phrases, “female terrorist,” “woman (as) terrorist,” “women (as) terrorist,” and “female suicide bomber”. In doing this, Nacos (2005) argued that female terrorists are presented within six frames. These frames include the physical appearance frame, the family connection frame, terrorist for the sake of love, the women’s liberation/equality frame, the tough-as-males/tougher-than-men frame, and the bored, naïve, out-of-touch-with-reality frame. Each of these frames of female terrorists has parallels with the representations of female politicians, as women in politics are frequently represented in a similar manner. Furthermore, these general frames of female terrorists are often the opposite of the frames created in the media’s representation of male terrorists. The physical appearance frame has an emphasis on the appearances, the apparel, the facial expressions of female terrorist as opposed to their criminal activity or extremist ideology. This frame strategically creates a softer image as opposed to a tough image of female terrorists by focusing on appearance as opposed to their actions. The family connection frame has a focus on the female terrorist's connection to the terrorist organization strictly in terms of her family. In doing this, the media downplays women's involvement in terrorists' organizations as simply a family correlation, meaning that without their family they would not be involved. Similarly, the terrorist for the sake of love frame creates an image of women’s involvement with a love connection with a male terrorist. Although both men and women are recruited by romantic 15 partners, the media typically will only discuss this phenomenon with female terrorists. The women’s liberation/equality frame tends to paint female terrorist’s extremist expression of feminism. This frame presents an image of women’s involvement in terrorism as a means of furthering gender equality, which is rarely the case. The tough-as males/tougher-than men frame represents women’s involvement in terrorism as cruel, aggressive, and deadlier. This narrative pushes an image of women who have these violent traits strictly in order to fit into the organization. Finally, the bored, naïve, out-of- touch-with-reality frame presents women in exactly what the title depicts. This frame diminishes the motives or intelligence of female terrorist, while pushes a false narrative of delusional women in these organizations. With these findings, Nacos (2005) argued that these general patterns of the media's representation of female terrorists could have many negative influences on anti- and counterterrorism policies, as many of these frames disregard the seriousness of women’s involvement in terrorist activity. If policies are influenced or shaped by the narrative of women’s involvement in terrorism created by the media, the prevalence and treat of women’s activity will only increase. There have been a few notable studies that have also looked at the media’s representation of gender and terrorism. Friedman (2008) looked at the portrayal of female suicide bombers in four US news sources from 2002 to 2004. This research primarily looked at the media’s depiction of the motives of female suicide bombers. In her analysis of 120 articles, Friedman (2008) reported five themes that were most frequently used to explain the motivations of female suicide bombers. These included strategic desirability, the influence of men, revenge, desperation, and liberation. These findings support a few (influence of men, desperation, liberation) of the frames identified by Nacos (2005) while adding new elements of the media's representation of female terrorists. Friedman (2008) concludes that while female suicide bombers 16 challenge commonly shared ideas of women and femininity, the media frequently reinforces the ideals of women by responding with disbelief of female suicide bombers. The most recent study to investigate the characterization of women terrorists was conducted on Spanish newspapers (Plaza et al., 2017). Using 50 newspapers, this qualitative study looked at the similarities in the type of frames used in American and Spanish news. Their findings suggested new interpretive frames from the Spanish newspapers, including women as terrorists (the discourse of exceptionalism), women as recruiters (the discourse of expertise), women as victims (the discourse of sensationalism), and finally women as warriors (the discourse of bravery) (Plaza et al., 2017). They also reported some similarities in the original frames identified by Nacos (2005) in the Spanish newspaper. These included the physical appearance frame, the family connection frame, and the tough-as-males/tougher-than-men frame but failed to find the terrorist for the sake of love frame, the women’s liberation/equality frame, and the bored, naïve, out-of-touch-with-reality frame in Spanish newspapers. These findings provide insight into the cross-cultural representation of women and terrorism and provide support for half of Nacos (2005) original arguments in regard to the frames that media creates for female terrorists. While there have been three qualitative studies that have looked at the media’s representation of woman terrorists (Friedman, 2008; Nacos, 2005; Plaza et al., 2017), to date no study has quantitatively tested the six frames identified by Nacos (2005) in American newspapers. Reviewing the Literature on Terrorism and Gender There have been several studies that have examined the connection between gender and terrorism (Braudy, 2003; Eager, 2008; Gonzalez-Perez, 2008; Sjoberg & Gentry, 2011). Consistent with the overall gender gap in crime, scholars see a gender gap in terrorism, such that men commit more acts of terrorism than women (Makin & Hoard, 2014). Although women are 17 generally less involved in terrorism, women’s involvement in terrorist organizations is both lethal and growing. Despite this, many people disregard the role that women play in terrorist organizations (Cunningham, 2003; Weinberg & Eubank, 2011). Recent research on gender and terrorism has looked at women as perpetrators of terrorism and indicates women's involvement in terrorism is greatly shaped by the contextual and situation aspects of the given organizations (Eager, 2008; Makin & Hoard, 2014; Raghavan & Balasubramaniyan, 2014; Weinberg & Eubank, 2011). Women's experiences, roles, and involvement within terrorist organizations have been anything but similar across extremists’ groups. This research has investigated the nature of the different roles of women in terrorist organizations, including women as enforcers, recruiters, suicide bombers, and leaders within terrorist organizations (Alexander, 2017; Bloom, 2011a, 2011b; O’Rourke, 2009; Rajan, 2011; Skaine, 2006; Speckhard, 2008). Research has also looked at the current status of women in different types of terrorist organizations compared to their historical and traditional roles (Berko & Erez, 2007; Raghavan & Balasubramaniyan, 2014). Women’s roles in terrorist organizations are greatly dependent on the type and ideological beliefs of the given organization. Depending on the organization, women have been in simple supporting roles, as wives and mothers of the men, or they have been recruiters, propagandist, and even suicide bombers. For example, women in the Popular Front for Liberation of Palestine have traditionally been serving on the frontline, while women in Harakat al-Muqaqama al-Islamiyya have had main roles in the recruitment for the terrorist group (Raghavan & Balasubramaniyan, 2014). Alexander (2017) has identified three categories that women’s roles in terrorist organizations most frequently look like; these roles include influencers in their communities and families, enforcers and informants, and leaders and recruiters. First, women in terrorist groups who take on the influencer role support the cause of the group from the sidelines. They are generally not involved in the planning or action, as they are the wives and 18 mothers of the men in the terrorist’s organization. They support the group through their roles as mothers and educators, as they have an influence on their children to carry on the extremist ideology. These women are similar to other moms in the sense that they teach and pass down beliefs, morals, and ideas. However, the beliefs, morals, and ideas that they pass down are radical. Second, when women are enforcers or informants they are most frequently in charge of the social cohesion within the group and watching others, primarily to ensure they are not breaking the established rules. Some of these women work alongside men, while others will work in female-only units (Alexander, 2017). An example of this includes the Al-Khansaa Brigade, an all-female police force, who are connected to ISIS and whose primary role is monitoring other women’s attire and enforcing the law (Boghani, 2015; Moaveni, 2015). Their role is primarily concerned with the maintenance of the rules of the organization, as opposed to leadership roles, or planning of acts of terrorism. Alexander (2017) notes that when women are in these roles, the terrorist organization often follows ultraconservative and fundamentalist ideals. Women’s roles are central to the organization but are less about carrying out the mission of the organization and much more focused on the enforcement of the established standards. Finally, in some terrorist organizations, women’s roles are more similar to men’s roles such that they are leaders and recruiters. When women are in these roles, they are essential to the operation of advancing the agenda of the organization, which is most frequently done through organizing and inciting violence (Alexander, 2017). These women are not on the sidelines, they are active within the organization and involved in the planning and implementation of the organization's political ideology. Women have operated in these roles cross-culturally, including terrorist organizations in Colombia, Italy, Iran, and the United States (Cunningham, 2003). 19 Research on women’s active involvement has primarily focused on female suicide bombers and trends in women’s leadership in terrorist organizations. The role of women as suicide bombers has recently gained significant attention (Bloom, 2011a, 2011b; Cunningham, 2003, 2007; O’Rourke, 2009; Rajan, 2011; Skaine, 2006; Speckhard, 2008). From 1985 to 2010, there were 257 executed suicide bombing attacks by women in various terrorist organizations and many more failed attempts (Bloom, 2011b). This increase in suicide bombers has primarily occurred in middle eastern countries, including Iraq, Israel, and Palestinian territories, and Iran (Bloom 2011b; Cunningham, 2007; Raghavan & Balasubramaniyan, 2014; Victor, 2004). However, Sri Lanka has the highest rate of female suicide bombers at 75, coming from the Liberation Tigers for Tamil Eelam (Cunningham, 2003; Raghavan & Balasubramaniyan, 2014; Speckhard, 2008). Irrespective of the patriarchal structure of many terrorist groups, women suicide bombers are lethal and prominent in modern terrorism. Research has indicated that on average women kill more per suicide bombing than male suicide bombers (O’Rourke, 2009). In the larger picture of efforts to prevent terrorism, this increase in female suicide bombers poses a novel and complex threat. Scholars agree that women have recently taken on these roles because of ways in which their gender allows for the element of surprise and the ability to cover up suspicious activity (Bloom, 2011b; Speckhard, 2008). Although women in these roles may have more responsibility, they do not seem to have equal respect or rights as the organization is simply using their gender to their advantage. Raghavan & Balasubramaniyan (2014) argue four factors that have contributed to the rise of women as suicide bombers, including operational advantage, greater publicity, increased recruitment, and competition among terrorists’ groups in the same space. When terrorist organizations have women in these roles, they are often less searched, better able to pass security measures, not as suspicious, and underestimated. This allows for surprise attacks 20 and decreases the morale of the opposing soldiers (Bloom, 2011b; Speckhard, 2008). In fact, female suicide bombers in Iraq will often disguise the explosives under their robes as a late-term pregnancy, which creates the illusion of a woman of a lesser threat to the opponent (Bloom, 2011b). Furthermore, Speckhard (2008) adds that women are being put in these suicide bombing roles because terrorist organizations often see women as more dispensable than men. This further indicates that although women are researching higher and more active positions within terrorist organizations, they generally are still not considered equal with men. Although research has provided many insights into the emerging threat of female suiciding bombings, there is still a great deal that we do not know. There is a common understanding that women are more likely to be active and, in a leadership, or recruitment role in a left-wing terrorist organization compared to a right-wing terrorist organization (Cunningham, 2003, 2007; Eager, 2008; Makin & Hoard, 2014; Raghavan & Balasubramaniyan, 2014; Weinberg & Eubank, 2011). Women in the far-left are generally more present and in a higher leadership role while women on the far-right are less present and in more supportive roles. Scholars link this gap in women’s representation between right-wing and left-wing terrorism to the common political and ideological beliefs of the respective groups. Men in far-right terrorist organizations are more reluctant to accept or create the space for women in leadership or recruitment roles because their ideological understandings of women do not allow for women to reach these higher positions. The extreme patriarchal structure of far-right terrorists’ organizations is a representation of their beliefs and keeps women in supporting roles. On the other hand, far-left organizations typically do not have the same beliefs on women’s roles, which allows for more women to be leaders, recruiters, or suicide bombers in these organizations. 21 In her work on women’s engagement in political violence, Eager (2008) examines three left-wing terrorist groups where women were highly involved and took leadership roles, including the Red Brigades, Weather Underground, and Red Army Faction. Weinberg & Eubank (2011) reported that throughout modern terrorism history, women have been active and leaders of left-wing terrorist groups, while women have largely been missing in right-wing terrorist groups, including neo-Nazi and racial supremacist groups. There are exceptions in which women have been leaders in far-right terrorist organizations. For example, Eager (2008) highlights Kathy Ainsworth and Rachel Pendergraft, women who were both active and in leadership roles within the KKK and the Knights’ Party. Recent research has also indicated a surge in women’s activity and inclusion in far-right terrorist groups (Makin & Hoard, 2014). This shift is representative of the overall increase in women's involvement and engagement in terrorism. Research has also examined the current status of women in terrorist organizations (Berko & Erez, 2007; Raghavan & Balasubramaniyan, 2014). Historically speaking, there has been a shift in the roles that women have played in terrorists’ organizations. Women are now more prominent in higher up roles within the terrorist organization, including enforcers, recruiters, and leaders, as opposed to the simpler supportive roles that women have traditionally adopted. In their study on the evolving role of women in terror groups, Raghavan & Balasubramaniyan (2014) show the shifting roles of women within many terrorist groups from 1960 to 2010. For example, women in Harakat al Tahrir al-Watsni al Filastini and Harakat al Muqawama al- Islamiyya, both religious-fundamentalists groups in Palestine, had original roles as recruiters, while their primary roles are now suicide bombers. On the other hand, in organizations such as Fuerzas Armadas Revolucionarias de Colombia and Maoists, both leftwing extremism groups, women’s roles started on the frontline and have advanced to the leadership roles over the years. 22 Their findings also indicate that there has been a large shift from women’s original role (frontline, propaganda, recruiters, etc.) into suicide bombers. While these shifts may indicate steps towards gender equivalence or women’s liberation through terrorist activity, Berko & Erez (2007) find that women in Palestine terrorist organization are instead put in “no-win” situations and are ultimately disempowered from their involvement. Though women’s roles are shifting within terrorist organizations, they are often still secondary to the men, especially in far-right terrorist groups. 23 Overview METHODOLOGY Past research has found that there are significant differences in the media’s representation of gender, particularly concerning the underrepresentation of women and the stereotypical narratives of masculinity and femininity. Recent research has extended these ideas to examine the media’s depictions of gender in crime stories. This research has indicated that there are varying representations of women depending on the type of crime and the context of the situation, including depictions as interlopers, more masculine, or sexualized as bad girls. Furthermore, research has indicated that the media still creates a stereotypical narrative of women as victims and offenders. One avenue of this research has looked at the media’s representation of women involved in terrorists’ organizations. While there have been qualitative studies that have explored the media’s representation of female terrorists (Friedman, 2008; Nacos, 2005; Plaza et al., 2017), there has been no quantitative study of this phenomena. Taken together, this research provides the foundations for the current study. The current study seeks to fill the gaps in the literature of both the media’s representation of gender and gender and terrorism. This study will analyze the media’s representation of male and female perpetrators of violent terrorist acts by using the United States Extremist Crime Data Base (ECDB). Specifically, this study will analyze the news media that covered the perpetrators of illegal violent incidents from far-right extremists, jihadist extremists, and far-left extremists between the years of 1990 and 2016. The current study will test how the media represents the six frames identified by Brigitte Nacos (2005) outlined above. It will also consider the media’s representation of masculinity for men in these organizations and the differences in the representation of male and female extremists in the media. The primary research questions include the following: 24 Research Question 1(RQ1): Which of the six frames identified by Nacos (2005) are most prevalent in the news coverage of illegal violence incidents by women and men in extremists’ organizations? Research Question 2 (RQ2): Using the frames identified by Nacos (2005), is there any difference in the types of frames used by the media to represent violent incidents by women in far-right, far-left, or jihadist organizations? Research Question 3 (RQ3): When looking at the interaction of gender and extremist ideology what are the trends in representation? Research Question 4 (RQ4): What are the major patterns in the representation of men compared to women extremists in the media? Based on the previous findings in the literature, these hypotheses have been formulated in response to the research questions: Hypothesis 1 (H1): The physical appearance frame will be most frequently used by the media for jihadist women compared to women in far-right and far-left terrorist organizations and all of the male perpetrators. Hypothesis 2 (H2): The women’s liberation/equality frame will be most frequently used by the media for far-left women compared to women in far-right and jihadist organizations and all of the male perpetrators. Hypothesis 3 (H3): The family connection frame will be utilized most frequently by the media to explain far-right women compared to women in far-left and jihadist organizations and all of the male perpetrators. Hypothesis 4 (H4): The terrorist for the sake of love frame will be most frequently used by the media to explain far-right women compared to women in far-left and jihadist organizations and all of the male perpetrators. 25 Hypothesis 5 (H5): The tough-as-males/tougher-than-men frame will be most frequently used by the media to explain far-left women compared to women in far-right and jihadist organizations and all of the male perpetrators. Hypothesis 6 (H6): The bored, naïve, out-of-touch-with-reality frame will be most frequently used to explain far-right women compared to women in far-left and jihadist organizations and all of the male perpetrators. Hypothesis 7 (H7): The media will fail to discuss any of the six frames when explaining men’s involvement in terrorist organizations. Hypothesis 8 (H8): Overall, women will be significantly underrepresented by the media in terms of news articles associated with their crimes when compared to the representation of men. Method In order to study the media’s representation of gender in terrorist organizations and answer the research questions, media documents related to extremist perpetrators were examined. The sample of perpetrator studied and media documents about them were pulled from the United States Extremist Crime Database (ECDB). The ECDB was constructed by first identifying crimes from various sources in an attempt to create a list of illegal financial schemes and illegal violent incidents by extremists in the United States (Freilich, Chermak, Belli, Gruenewald, & Parkin, 2014). These sources included existing terrorism databases, official sources (FBI, DOJ), scholarly and journalistic accounts, material published by private watch groups and media publications. While there are many conflicting definitions of terrorism, this database defines acts of terrorism through a two-pronged test: the crime had to be either an illegal violent incident or illegal financial scheme from 1990 to 2016 (behavioral prong) and at least one of the suspects had to be subscribed to an extremist belief system (attitudinal prong) 26 (Freilich et al., 2014). Opposed to other databases and definitions of terrorism, the ECDB is unique such that it specifically includes illegal financial schemes, as they recognize the financial loss of over $700,000,000 that has been perpetrated by far-right and jihadist perpetrators. Once these crimes were identified, each case, suspect, and victim were searched to uncover relevant open source information. Each incident included in the ECDB was searched through 26 search engines, and all available open source material was compiled about it. The materials compiled for each incident provides an excellent source of materials and news articles written about an incident and the perpetrators involved in the incident. Both inter-rater reliability and selectivity bias were critically considered in the creation of the ECDB. To increase inter- rater reliability, there was a strict procedure set into place for the multiple coders. Also, a measurement of the inter-rater reliability was created which indicated coder agreement between 89% and 98% of the time. To deal with selectivity bias, the creators examined the sources used to identify the crimes and “normalized their criteria to accurately assess variations in the events they included” (Freilich et al., 2014). In doing this, the creators of the ECDB report that selectivity bias was less of a problem than originally perceived. Taken together, the ECDB was constructed carefully indicating both a database with both reliability and validity. First, a list of all males and females from the ECDB was pulled. From this list, there was a random selection procedure to obtain a random sample of males and females from the database. The random sample was strategically comprised of 45 males and 45 females so that the analysis can include comparisons between gender and within gender. The sample included individuals from far-right groups, far-left groups, and jihadist terrorist organizations between the years of 1990 and 2016. As shown in Table 1, the sample was stratified by ideological difference and included 15 far-right, 15 far-left, and 15 jihadist men and women to ensure equal coverage and comparison. 27 Media articles about the male and female were then collected from the ECDB case files. Each case file was prepared by removing all of the information (court documents, watch groups, blogs, etc.) within the case file except for the newspaper documents. Each news article within the case file was then counted and documented. There is considerable variation in the number of articles written about incidents as some of the cases in the ECDB are celebrated (e.g. the September 11th attacks) and others have a small number of articles. In order to account for the variation, this study randomly selected three newspaper articles from the case file. In randomly sampling three articles for each perpetrator the study was able to get a more complete understanding of how the media was representing the respective perpetrator instead of just analyzing a single article. This also allowed for variability in the sampling procedure as some articles went into much more depths than others. While analyzing more articles would have provided a more comprehensive understanding of the way the media was representing each perpetrator, some of the case files only had three articles which limited the number of articles that could be randomly sample in the other case files to three articles. All types of news articles were available for selection, including local and global news sources. The news articles associated with each offender within each file were used to systematically analyze how the media represents gender for men and women. Each newspaper article was systematically analyzed for the six major frames identified by Nacos (2005). The text in the articles was used to identify how each perpetrator is being represented by the media and then coded into a database. For example, if a news article discussed the clothing of a female perpetrator the article was coded into the physical appearance frame and if the article discussed the woman’s involvement in connection with her romantic partner the article was coded into the terrorist for the sake of love frame. Each frame was considered individually, as there could be multiple frames used to discuss a male or female perpetrator in 28 each article. For example, within an article, both the family connection frame and the bored, naïve, out of touch with reality frame could be used to explain a sole female perpetrator. The sample of articles covered primarily white perpetrators (61%), followed by Arab perpetrator (42%), and black/African American perpetrators (24%). For twelve of the perpetrators, race was not able to be identified or confirmed in the ECDB, the case files, or through searches on google. The mean age of the perpetrators was 31.79. The articles that were sampled covered a 25-year span and ranged in date from 11/28/1993 to 07/24/2018. The current study utilized Vanessa Otero’s (2018) Media Bias Chart: Version 4.0 for the classification of right-wing, left-wing, and neutral/unbiased news. The vast majority of articles that were randomly sampled came from unbiased news sources (89%), including global and local news sources. Only around 10% of the articles were classified as either right- or left-wing news source. While there was a lack of diversity in the political leaning of news sources, the articles came from 109 different news sources (see Appendix A for a list of the news sources used). The Associated Press (67 articles), the New York Times (17 articles), and the Seattle Times (10 articles) were the most common news sources. Roughly 44% of the articles were written by a male lead, while only 15% were written by a female lead. Additionally, 40% of the articles had no author reported. The articles that were analyzed had a mean of 604 words. Along with the creation of the database, a codebook was created that provided examples of the frames identified by Nacos (2005), explanations of the variables used in the study, and the new frames found in the current study (see Appendix B for Codebook). 29 n 135 135 50.0 50.0 165 42 24 3 61.1 15.6 8.9 1.1 Table 1 Descriptive Statistics of the Articles (N=270) Gender of Perpetrator Male Female Race of Perpetrator* White Arab Black/African American Bi-Racial Age of Perpetrator Extremists Ideology Far-Right Far-Left Jihadist News Type Right-Wing Left-Wing Unbiased Gender of Reporter Male Female Not Reported Word Count *Percentages may not add up to 100% because of missing cases 13 15 242 119 41 110 33.3 33.3 33.3 4.8 5.6 89.6 44.1 15.2 40.7 90 90 90 % Mean Std. Deviation 11.37 516.34 31.79 604.42 30 After analyzing 135 newspaper articles for female perpetrators, there was little support RESULTS found for the frames identified by Nacos (2005). As shown in Table 2, only four of the six identified frames (physical appearance frame, family connection frame, terrorist for the sake of love frame, and bored, naïve, out-of-touch-with-reality frame) were found in the newspaper articles that were analyzed for women. There was no evidence in the articles that were reviewed for the women’s liberation/equality frame or the tough-as-males/tougher-than-men frame. The most prevalent of the frames for female perpetrators, the terrorist for the sake of love frame, was found in only 26 of the 270 newspaper articles analyzed, which was only 19.3% of the articles. This was followed by the family connection frame (6.7%), the physical appearance frame (3.7%), and the bored, naïve, out-of-touch-with-reality frame (3%). The underwhelming evidence of these frames indicates that the media may not be portraying female perpetrators of terrorism through these more obvious and lurid gendered frames. Table 2 Descriptive Statistics of the Frames for Female Perpetrators (N=135) Frames Identified by Nacos (2005) Physical Appearance Frame Family Connection Frame Terrorist for the Sake of Love Frame Women’s Liberation/Equality Frame Bored, Naïve, Out-of-Touch-with-Reality Frame Identified Frames Disregarded Frame Overshadowed Frame Tough-as-Males/Tougher-than-Men Frame Fact Frame Life Frame % 3.7 6.7 19.3 0.0 0.0 3.0 30.4 34.8 69.6 13.3 n 5 9 26 0 0 4 41 47 94 18 31 Unsurprisingly, the vast majority of the articles that covered male perpetrators had no gendered frames identified by Nacos (2005), which is illustrated below in Table 3. The most frequent frame for male perpetrators was the family connection frame, which was in only 5.9 percent of news articles. This was closely followed by the physical appearance frame (4.4%), and then the terrorist for the sake of love frame and the bored, naïve, out-of-touch-with-reality frame, which both were in only 2.2% of news articles reviewed. These findings actually contradicted H7 such there was evidence that male perpetrators can be presented in a similar nature as female perpetrators. Expectedly, there were no articles that had the women’s liberation/equality frame or the tough-as-males/tougher-than-men frame for male perpetrators. Table 3 Descriptive Statistics of the Frames for Male Perpetrators (N=135) Frames Identified by Nacos (2005) Physical Appearance Frame Family Connection Frame Terrorist for the Sake of Love Frame Women’s Liberation/Equality Frame Bored, Naïve, Out-of-Touch-with-Reality Frame Identified Frames Disregarded Frame Overshadowed Frame Tough-as-Males/Tougher-than-Men Frame Fact Frame Life Frame n 6 8 3 0 0 3 47 22 87 21 % 4.4 5.9 2.2 0.0 0.0 2.2 34.8 16.3 64.4 15.6 The primary contribution of the current study is new frames that have been identified in reviewing the newspapers 270 articles. These frames were in response to finding little support for the frames identified by Nacos (2005). After recognizing that many of the articles that were being analyzed failed to show any of the frames identified by Nacos (2005), the current study sought to understand what themes were most frequently being utilized to explain male and 32 female perpetrators of terrorism. After reading through many of the articles that were randomly sampled four primary frames emerged, which included the disregarded frame, the overshadowed frame, the fact frame, and the life frame. These frames emerged after seeing multiple articles that had similar trends, such as leaving out the name of the perpetrator or overshadowing the actions of the perpetrators by focusing on other aspects of the crime. Each of the frames that were identified was found more frequently overall when compared to the original frames identified by Nacos (2005). The first frame identified was the “Disregarded Frame”, which is defined in the codebook as, “An article that discusses an act of terrorism, however, fails to mention the name of the perpetrator” (Appendix B). This frame emerged after recognizing that many of the articles that were randomly sampled for a specific perpetrator failed to directly reference or mention the name of that perpetrator as the article focused on other aspects of the crime. It’s important to note that this frame was not used when the article failed to mention the perpetrator’s name because it was unknown, but because the article highlighted another perpetrator, victim, or aspect of the crime while failing to mention the perpetrator’s name and in doing so disregards the perpetrator. This frame was found in a total of 88 of the newspaper articles that were randomly sampled, 41 news articles of female perpetrators and 47 news articles of male perpetrators. A prime example of this was in an article that was randomly sampled from the case file for Lacey Phillabaum, a far-left female terrorist who was involved in the arson at the University of Washington Center for Urban Horticulture in Seattle Washington in 2001. The news article (Case 016) which was published by the Seattle Post-Intelligencer, a local Seattle newspaper, almost eight years after the crime was committed failed to even mention the name of Lacey Phillabaum. The 789-word news article focused on two other perpetrators involved in the crime, including Brianna Waters and Justin Solondz, however, failed to discuss Phillabaum and therefore disregarded her activity in the terrorist incident (Shukovsky, 2008). Unexpectedly this 33 frame was found in close to 1/3 of the news articles and was a very prevalent theme in which articles would discuss the crimes committed and give hyper-attention to a single perpetrator but fail to mention the perpetrator that had been randomly sampled. An extension of the disregarded frame was the “Overshadowed Frame”, which was defined in the codebook as, “An article that does reference the perpetrator however overshadows the perpetrator of a crime by giving more attention to another perpetrator or other logistical matters of the crime (basic information of the crime, court hearings, victims, etc.)” (Appendix B). This frame emerged when many of the articles that were randomly sampled would only briefly mention the perpetrator, and not provide additional information besides the name and age of the perpetrator. An important coding note was that if a news article had the disregarded frame, the article could not also be categorized under the overshadowed frame. These frames were mutually exclusive. The primary difference between the frames lies that the overshadowed frame had to directly reference the perpetrator, while still overshadowing the perpetrator. The overshadowed frame was identified in a total of 69 of the newspaper articles that were randomly sampled, 47 for female perpetrators and 22 for male perpetrators. An example of the overshadowed frame is a news article randomly sampled from the case file of Tina Mae Stone, a far-right terrorist who was connected to the Hutaree Christina Militia and plotted to kill members of the judicial system. The article (Case 106) was published by the Windsor Star, a daily newspaper published in Windsor Ontario. In the 609-word article Tina’s name was only mentioned once, while her husband David Stone was directly mentioned six times (Windsor Star, 2010). In this news article and 68 others, the perpetrator was overshadowed by another perpetrator or other logistical matters relating to the crime. The third newly identified frame was the “Fact Frame”. This frame was defined as, “An article that directly referenced a perpetrator and provided simple facts about the perpetrator’s 34 case or crime committed” (Appendix B). This frame emerged directly in response to failing to find the frames identified by Nacos (2005), as many of the news articles did not present any of the information with a slanted viewpoint but as simply as facts. While it may be obvious to show that most of news reporting is factually based, it was important to include this within the current study as it shows the stark contrast between the frames identified by Nacos (2005) and the vast majority of news coverage of terrorism. Unsurprisingly, this frame was incredibly common throughout and over 65% of articles had this frame. Ninety-four of the articles that covered the female perpetrators had this frame, while 87 of the articles that covered male perpetrators had this frame. Similar to the overshadowed frame, this frame and the disregarded frame were mutually exclusive. However, this frame and the overshadowed frame could appear within the same article, especially when the only information provided about a perpetrator was only one sentence of information on the perpetrator’s court case. The only articles that did not have the fact frame were news articles that failed to mention the perpetrator, which were categorized under the disregarded frame. Examples of this frame include, “U.S. District Judge Richard Story announced the sentence Wednesday after a hearing in which 73-year-old Frederick Thomas and 67-year-old Dan Roberts addressed the court and apologized. It was the maximum sentence allowed under the plea deal (Brumback, 2012) (Case 142)”, and “The Anti-Defamation League plans an award ceremony Tuesday to honor the FBI agents who investigated Farooque Ahmed, who was charged last year and was sentenced to 23 years in prison after pleading guilty in April (Tucker, 2011) (Case 184)”. While this frame may seem obvious, it was important to show how much of the news is just covering simple facts about the perpetrates case and crimes committed. Being the most frequent frame found in the analysis (181 news articles in total) indicates that the vast majority of news is reporting facts opposed to prevailing gendered explanations of terrorism. 35 The final identified frame was the “Life Frame”, which is defined as, “An article that discusses other aspects of the perpetrator’s life outside of their terrorist activity” (Appendix B). This emerged after reading multiple articles that would discuss the lives of perpetrators outside of their terrorist activity and thus softened their crimes by providing reference points for the reader to normal life activities that are a stark contrast from terrorism. This frame was mutually exclusive with the disregarded frame but considered individually with the overshadowed frame and life frame. This frame took the perpetrator out of the context of their criminal activity and essentially humanized them by discussing other aspects of their lives. This was the least prevalent of the new frames identified with only 39 articles having this frame, however, was still more common than any of the original frames identified by Nacos (2005). This was identified in 18 articles that were sampled for female perpetrators and 21 for male. An example of this frame includes a news article that was randomly sampled for Chelsea Dawn Gerlach (far-left) taken from the Associated Press. Within the 714-word news article (Case 005), the following was stated, “Gerlach, 28, who grew up in the Eugene area and worked as a disc jockey in Portland, faces trial on indictments she and Meyerhoff helped topple a high tension power line outside Bend in 1999 and that she served as a lookout while others were setting fire to the Childers Meat Co. plant in Eugene in 2001 (Barnard, 2005)”. Discussing the city where an individual grew up and providing background occupational information humanizes the perpetrator in a way that many articles fail to do. In the initial analysis, t-tests were used to compare the means of male and female perpetrators for the dependent variables. The dependent variables included four of the original frames identified by Nacos and four of the newly identified frames. Two of the frames, Women’s Liberation/Equality Frame and Tough-as-Males/Tougher-than-Men Frame, were not used in any of the analysis because no support was found for them in the 270 articles analyzed. All of the 36 dependent variables in the current study were coded as a binary, with “0” as no and “1” as yes. If the article included the frame the article was coded as yes, and then if the article did not have the frame the article was coded as no. The results of the comparison of means for the different frames broken down by gender are presented in Table 4. Female Male Table 4 Bivariate Analysis of Frames and Representation by Gender (N=270) Dependent Variables Frames Identified by Nacos (2005) Physical Appearance Frame Family connection Frame Terrorist for the Sake of Love Frame Bored, Naïve, Out-of-Touch-with-Reality Frame Identified Frames Disregarded Frame Overshadowed Frame Fact Frame Life Frame Newspaper Articles in Case Files *p<.05. **p < .01. ***p < .001. .04 .07 .19 .03 .30 .35 .70 .13 .04 .06 .02 .02 .35 .16 .64 .16 54.29 57.22 T-Test .307 -.250 -4.685*** -.382 .777 -3.556*** -.904 -.518 -.156 The only significant differences in means between female and male perpetrators were found in the terrorist for the sake of love frame (p < .001) and the overshadowed frame (p < .001). Expectedly, there were significant differences between female and male perpetrators regarding the terrorist for the sake of love frame, such that female perpetrators were significantly more likely to have been represented in this frame when compared to male perpetrators. This finding gives support to the argument made by Nacos (2005), that female terrorists are being presented in the media in a way different than men. However, it’s important to note that the other frames, physical appearance frame (p = .759), family connection frame (p = .803), and the bored, naïve, out-of-touch-with-reality frame (p = .703) were not significant. These findings indicate that were no differences in the way in which the media was representing male and female 37 perpetrators regarding these frames. While there was support found for gendered differences for the terrorist for the sake of love frame, there seems to be an overall lack of support for the frames identified by Nacos (2005) regarding gendered differences. Similar to the terrorist for the sake of love frame, there were significant differences between the overshadowed frame for males and females, such that female perpetrators were significantly more likely to be presented by the media in an overshadowed frame when compared to male perpetrators. This shows that when women were mentioned in the article, they were significantly more likely to be overshadowed by another perpetrator, victims, or other logistical matters of the crime. This finding further indicates that there are gendered frames in which the media uses to represent perpetrators of terrorism. While there was little support regarding the original frames identified by Nacos (2005), this new frame, the overshadowed frame, may provide insights into how the media has gendered frames in more subtle ways. It is also important to note that while there were gendered differences for the overshadowed frame, there were no differences found in the disregarded frame, fact, frame, and life frame. To test H8, each newspaper article was counted within each perpetrator’s case file, which ranged from 3 to 431 news articles within a case file. A t-test was used to compare the means of news coverage. Since each case file was created using the same open source searching methods, we compared the means of news article coverage for male and female perpetrates. As shown in Table 4, while the male perpetrators (57.22) had a higher mean number of articles when compared to female perpetrators (54.22), there were no significant differences found. However, analysis on the entire sample of the ECDB indicated significant differences between male and female perpetrators, such that males have significantly higher coverage in newspaper articles when compared to female perpetrators (Morgan, 2018). Both of these findings give 38 support for H8, such that female perpetrators are overall significantly underrepresented by the media in terms of news articles associated with their crimes when compared to male perpetrators. To examine the ways in which the media represents extremist ideology through different frames, multiple ANOVAs were conducted on the dependent variables. Based on the classification in the ECDB, extremist ideology was broken up into far-right, far-left, and jihadist motivated perpetrators. Identical to the previous analysis on gender differences, the dependent variables in this analysis were all coded as a binary. The results of the comparison of means of the eight different frames across extremist ideology are presented in Table 5, which included all 270 articles covering both male and female perpetrators. The analysis indicated no significant differences across extremist ideology for the original four frames identified by Nacos (2005). Unsurprisingly, there were not large differences in the gendered frames identified by Nacos (2005) when broken down by extremist ideology. Far-Right Far-Left Table 5 Bivariate Analysis of Frames by Extremist Ideology (N=270) Dependent Variables Frames Identified by Nacos (2005) Physical Appearance Frame Family connection Frame Terrorist for the Sake of Love Frame Out-of-Touch-with-Reality Frame Identified Frames Disregarded Frame Overshadowed Frame Fact Frame Life Frame *p<.05. **p < .01. ***p < .001. .26 .31 .73 .19 .44 .30 .56 .14 .04 .07 .17 .03 .02 .03 .06 .01 Jihadist ANOVA (F Value) .06 .09 .10 .03 .28 .16 .72 .10 .659 1.190 2.968 .583 4.462* 3.619* 4.118* 1.438 In contrast, three of the four newly identified frames were significant when looking at differences in means across extremist ideology. The significant frames included the disregarded frame (p = .012), the overshadowed frame (p = .028), and the fact frame (p=.017). The only 39 frame that was not significant was the life frame (p = .239). For the disregarded frame, we found large differences between the far-left when compared to far-right and jihadist. Close to half of the articles for far-left perpetrators fit the disregarded frame, while the far-right and jihadist had only around a quarter. This indicates that far-left perpetrators are significantly more likely to be disregarded and not even mentioned in news coverage. The current study suspects this finding to be connected to the situational factors regarding the crimes committed by the far-left. Many of the crimes committed by the far-left are committed in groups of people, thus the media can only focus on some of the perpetrators. Closely linked to the disregarded frame, the means across extremist ideology were significantly different for the fact frame. Just over half of the articles that covered the far-left had the fact frame, while the articles for the far-right and jihadist had the fact frame close to ¾ of the time. Since these two frames were mutually exclusive, it is unsurprising that the fact frame is significant knowing that the disregarded frame was significant. If an article disregarded a perpetrator then it could not present a fact regarding that perpetrator. Since 44% of the articles covering the far-left were disregarded, it was anticipated that they would have a significantly lower frequency of the fact frame when compared to the far-right and jihadis perpetrators who were only disregarded a quarter of the time. Furthermore, the overshadowed frame was significantly different across the means of extremist ideology. The analysis indicated that far-right, and far-left perpetrators had significantly higher means than jihadist perpetrators. This indicates that perpetrators classified as far-right and far-left tend to be more overshadowed when compared to jihadist perpetrators. Not only does this indicate differences in the overshadowed frame, but it indicates that when a jihadist perpetrator commits a crime, they are more likely to receive attention within a news 40 article. This shows us that jihadist perpetrators may receive more attention in the media, and their actions are less overshadowed when compared to far-right and far-left perpetrators. To test hypotheses 1 through 6, ANOVAs were conducted on four of the six frames identified by Nacos (2005). The results of the comparison of means across the interaction of extremist ideology and gender on the original four frames and newly identified four frames are presented in Table 6. The interaction of gender and extremist ideology included six categories: far-left women, far-left men, far-right women, far-left men, jihadist women, and jihadist men. Since there were no articles analyzed that fit the women’s liberation/equality frame or the tough- as-males/tougher-than-men frame, there was no support found for H2 and H5. Of the four frames identified by Nacos (2005), only two of these were significantly different across gender and extremist ideology. The analysis indicated little support for H1 and H6. Regarding H1, not only did the physical appearance not significantly differ across gender and extremist ideology, but this frame was most prevalent in jihadist men. The current study had originally predicted that this frame would be used most frequently for jihadist women, however, their mean matched the means of far-left men, far-right men, and women, and was less than jihadist men. This finding contradicts Nacos (2005) argument that the media focuses more on the physical appearance of female perpetrators of terrorism when compared to male perpetrators. In fact, the male perpetrators in all three different types of extremist ideology had higher or equal means when compared to the respective female perpetrators. There was more support for H6 because far-right women were represented through the bored, naïve, out-of-touch-with-reality frame most frequently, however, the analysis indicated no significant differences between the groups. While this frame was used to explain far-right females most frequently, the lack of significance contradicts the findings in Nacos (2005) original work because we did not find any evidence that this frame is gendered. 41 Not only were significant differences found across gender and extremist ideology for the family connection frame and the terrorist for the sake of love frame, but there was also support found for H3 and H4. The data confirmed H3, such that the family connection frame was utilized most frequently by the media to explain far-right females compared to every other group. However, unexpectedly jihadist men were represented by the media through the family connection frame second most frequently followed by far-right men and jihadist women. While there does seem to be some gendered trends and support for Nacos (2005) original argument with this frame, the findings are not as clear cut as the terrorist for the sake of love frame. As previously discussed, this frame was significant by gender and when considering the interaction of extremist ideology and gender we found higher means for women regardless of extremist ideology. There was support for H4 such that the terrorist for the sake of love frame was most frequently used by the media to explain far-right women compared to every other category. The analysis has consistently shown that this frame identified by Nacos (2005) to be the most significant regarding a gendered way in which the media represents male and female perpetrators of terrorism. Analysis was also conducted on the new frames identified in the current study regarding the interaction of extremist ideology and gender. Of the four frames, three were statistically significant: the disregarded frame, the overshadowed frame, and the fact frame. For the disregarded frame, far-left females were most frequently overshadowed in the media followed by far-left males. Considering the findings in Table 4, it was unsurprising to see that male and female far-left perpetrators had higher means when compared to the other groups. While the means were higher for far-left female than far-left males, we did not see this trend for the female far-right and jihadist perpetrators. This indicates that while there may be a gendered media aspect to this frame for far-left perpetrators, there was no such finding for far-right and jihadist 42 motived incidents. Contrary to the disregarded frame, the far-left men and women were least frequently discussed through the fact frame when compared to far-right and jihadist men and women. While the findings were opposite to that of the disregarded frame, there was also no indication that this frame has a gendered aspect as the means by gender did not follow any consistent trends across extremist ideology. In stark contrast to the disregarded frame and the fact frame, there were consistent gender differences across extremist ideology for the overshadowed frame. Findings indicated that the overshadowed frame was utilized most frequently by the media for far-right women when compared to all other groups. Similarly, the means of far-left and jihadist females were larger when compared to far-left and jihadist male perpetrators. The data indicated that female perpetrators are overshadowed in news articles more frequently than men, regardless of the type of extremist ideology. This finding gives support to the idea that the media represents female terrorists through gendered frames, however in possible more subtle ways than Nacos (2005) originally proposed. 43 Far-Left F M Table 6 Bivariate Analysis of Frames by Extremist Ideology and Gender (N=270) Frames Identified by Nacos (2005) Physical Appearance Frame Family Connection Frame Terrorist for the Sake of Love Frame Out-of-Touch-with-Reality Frame Identified Frames Disregarded Frame Overshadowed Frame Fact Frame Life Frame *p<.05. **p < .01. ***p < .001. .49 .35 .51 .13 .22 .44 .78 .16 .04 .07 .04 .02 .40 .24 .60 .16 Far-Right F M Jihadist F M ANOVA (F Value) .00 .00 .07 .00 .04 .13 .31 .07 .04 .00 .02 .00 .29 .18 .69 .22 .04 .07 .20 .02 .20 .24 .80 .11 .07 .11 .00 .04 .36 .07 .64 .09 .543 2.373* 7.854*** 1.203 2.558* 4.392** 2.488* .767 44 Furthermore, because the dependent variables were coded as a binary (yes, no), the current study utilized multivariate binary logistic regressions for four of the dependent variables. The primary independent variable in all of the regression models was gender, which was coded as a binary (male = 0, female =1). The age of the perpetrator, the year the crime was committed, and word count were all included in the models primarily as control variables. Multicollinearity was assessed on all of the variables included in the model and the date of the article was taken out of the model because it was too closely associated with the year the crime was committed. Similarly, race was not included in the models because there was such a strong overlap between race and extremist ideology, such that the non-white perpetrators (Arab, Black/African American, Bi-Racial) were almost all categorized as jihadist. Extremist ideology, news type, and gender of the reporter were also included in the binary logistic regression models and dummy coded. For extremist ideology, we included a far-right variable and far-left variable, each with jihadist as the reference category. For news type, we created a dummy binary variable with biased news (right-wing and left-wing) coded as 0, and unbiased news (independent, local, global, etc.) coded as 1. Finally, for the gender of the reporter, we created a dummy binary variable with male reporters and no gender reporters coded as 0, and female reporters coded as 1. Analysis was conducted on only one of the frames identified by Nacos (2005), the terrorist for the sake of love frame. Unfortunately, because most of the frames identified by Nacos (2005) had little variance, the analysis was severely limited. It is important to note, that while analysis was conducted on the terrorist for the sake of love frame the findings need to be met with caution because we only found 29 articles out of 270 with this frame. Analysis was however conducted on three of the newly identified frames, including the disregarded frame, the overshadowed frame, the fact frame. There was no further analysis was conducted on the life frame, because there were no significant differences found in the bivariate analysis. 45 The results of the binary logistic regression on the terrorist for the sake of love frame are displayed in Table 7. Of the eight variables included in the model, only two were significant. As shown, when controlling for all of the other variables in the model female perpetrators have 10.43 higher odds of being framed through the terrorist for the sake of love frame than male perpetrators. Similarly, articles written by female reporters have 4.46 higher odds for the terrorist for the sake of love frame than articles written by male reporters and articles that failed to report the gender of the author when controlling for all other variables. Not only are articles that are written by women more likely to have this frame, but articles for female perpetrators are also more likely to have this frame. These findings give support for Nacos (2005) original argument solely regarding this frame and indicate gendered differences within the terrorist for the sake of love frame for both the perpetrators and for the reporters. Nagelkerke R2 was used as an indicator of the strength of relationships between predictors and the dependent variable. For this model, there was a moderate relationship found at .33. Table 7 Binary Logistic Regression of the Terrorist for the Sake of Love Frame (N=270) Independent Variables Age Year of Crime Word Count Unbiased News Source Nagelkerke R2 *p<.05. **p < .01. ***p < .001. B 2.34 -.01 .06 .00 .87 -.08 .80 1.50 .33 SE .65 .02 .04 .00 .56 .76 .85 .48 Far-Right Far-Left Exp(B) 10.43*** .98 1.06 1.00 2.40 .92 2.22 4.46** Female Female Reporter There was a great deal of overlap between the analysis on the disregarded frame and the fact frame. Both logistic regressions indicated no significant differences between male and female perpetrators. Since gender was not significant in these models, these frames do not 46 provide significant contributions to our understanding of how the media represents male and female terrorists. However, as shown in Table 8 and Table 9, of the remaining seven independent variables included in the models only three were statistically significant, which included the age of the perpetrator, the year of the crime, and female reporters. Table 8 Binary Logistic Regression of the Disregarded Frame (N=270) Independent Variables Age Year of Crime Word Count Unbiased News Source Nagelkerke R2 *p<.05. **p < .01. ***p < .001. B .07 -.03 -.10 .00 -.71 -.17 -.72 -1.05 .17 SE .30 .02 .03 .00 .41 .41 .46 .49 Female Far-Right Far-Left Female Reporter Exp(B) 1.07 .97* .91** 1.00 .50 .84 .49 .35* Results indicated that the older the perpetrator was, and the more recent the article was published lowered the odds that the disregarded frame would be present when all of the variables were held constant. On the contrary, the older the perpetrator was, and the more recent the article was published actually increased the odds that the fact frame was present. Findings also indicated that articles written by female reporters have .35 lowered odds for the disregarded frames than articles written by male reporters and articles that failed to report the gender of the author when controlling for all other variables. Contrasting this finding, female reporters have 2.89 higher odds for the fact frame than articles written by male reporters and articles that failed to report the gender of the author. The Nagelkerke R2 for the disregarded frame was .17 and for the fact frame was .16. 47 Table 9 Binary Logistic Regression of the Fact Frame (N=270) Independent Variables Age Year of Crime Word Count Unbiased News Source Nagelkerke R2 *p<.05. **p < .01. ***p < .001. B -.02 .03 .10 .00 .62 .15 .69 1.06 .16 SE .30 .02 .03 .00 .41 .41 .46 .49 Female Far-Right Far-Left Female Reporter Exp(B) .98 1.03* 1.10** 1.00 1.86 1.16 2.00 2.89* Half of the variables in the binary logistic regression on the overshadowed frame were significant, including gender. The model, displayed in Table 10, indicated that when controlling for all of the other variables in the model female perpetrators have 2.50 higher odds of being overshadowed than male perpetrators. This gendered finding indicates that the media does represent female perpetrators of terrorism in a different way when compared to male perpetrators. Furthermore, far right perpetrators have 3.26 higher odds of being overshadowed in news articles than jihadist perpetrators, while far-left perpetrators have 3.14 higher odds of being overshadowed than jihadist perpetrators. Word count was also a variable that was significant within the model. The Nagelkerke R2 for this model was at .18. This model provides the greatest insights into the media’s representation of terrorism and the differences across gender and extremist ideology, such that female, far-right and far-left perpetrators are most likely to be overshadowed. 48 Table 10 Female Binary Logistic Regression of the Overshadowed Frame (N=270) Independent Variables Age Year of Crime Word Count Unbiased News Source Nagelkerke R2 *p<.05. **p < .01. ***p < .001. B .92 .01 .00 .00 1.18 1.14 1.06 .51 .18 SE .32 .01 .03 .00 .44 .49 .67 .40 Far-Right Far-Left Female Reporter Exp(B) 2.50** 1.01 1.00 1.00* 3.26** 3.14* 2.90 1.67 49 DISCUSSION The findings in the current study provided the first quantitative approach in analyzing the media’s representation of gender and terrorism. The analysis indicated little support for the six frames identified by Nacos (2005). While there was evidence of four of these frames within the articles analyzed, the current study found no evidence of the women’s liberation/equality frame or the tough-as-males frame. Nacos (2005) original work provided insights into the ways in which the media represents female terrorists similar to female politicians, however, there does not seem to be support for the frames she identified when analyzing a large number of news articles on terrorism. In addition to testing the six frames identified by Nacos (2005), this study identified four news frames that were common in the media’s representation of terrorism. These frames include the disregarded frame, the overshadowed frame, the fact frame, and the life frame. As shown in the analysis, each of these frames was found more frequently when compared to the frames identified by Nacos (2005). This primary purpose of the current study was to compare the frames by gender to better understand if female perpetrators of terrorism were more likely to be represented through these frames than male perpetrators. The two frames that emerged with the most significant gendered differences included the terrorist for the sake of love frame and the overshadowed frame. Although it is common for both male and female terrorists to have romantic partners within a terrorist organization, the findings in the current study show that the news media presents this stereotype for female perpetrators significantly more frequently than their male counterpart. The findings indicate that this frame is most common in the media’s representations of far-right females. As we know from previous research, the media often represents women through their dependence on men and submissiveness, while presenting men through their independence (Wood, 1994). The findings of the terrorist for the sake of love frame seem to be no different 50 from this research, as female perpetrators of terrorism are portrayed in the media through their dependence on their romantic relationship more frequently, while male perpetrators are not connected to their significant others and thus portrayed through their dependence. When looking at gender differences among the newly identified frames, there were clear differences for the overshadowed frame, such that female perpetrators were significantly more likely to be overshadowed in the news coverage of their crimes. This may indicate that while the media does not necessarily represent female perpetrators in the more lurid frames identified by Nacos (2005), that the media has more subtle gendered frames such as the overshadowed frame when representing female terrorists. In overshadowing female perpetrators, the media does not have to force a narrative of a violent female terrorist which would contradict the traditional feminine image that they portray of women. Because the media tends to portray men with strength and independence, they likely have a much easier time representing acts of terrorism committed by men because it fits within their representation of men. However, with women, we suspect they tend to overshadow their actions because the media may not know how else to represent a violent female terrorist. Furthermore, the results indicated no significant differences between male and female perpetrators for the physical appearance frame, the family connection frame, and the bored, naïve, out-of-touch-with-reality frame. These findings contradict Nacos (2005) argument that female perpetrators are presented in the media through these stereotyping and gendered frames. The current study suspects that the frames identified by Nacos (2005) would likely be more present in letters to the editor and opinion editorial news coverage, however, the majority of the news articles that were randomly sampled in the current study were straightforward editorials. While gendered clichés and frames can be present in some news coverage of terrorism, when considering the large number of articles that actually cover terrorism these frames may be 51 significantly less influential on anti- and counterterrorism policies than Nacos (2005) suggests. In addition to no significant differences found between male and female perpetrators for the majority of the frames identified by Nacos (2005), there were no gender differences found in the disregarded frame, the fact frame, and the life frame. The current study also found that the female perpetrators were more underrepresented in terms of frequency of newspaper coverage when compared to male perpetrators. These findings are consistent with a long body of literature that had documented the underrepresentation of women in the media (Baumann & de Laat, 2012; Davis, 1982; Humprecht & Esser, 2017; Krijnen & Bauwel, 2015; Lont & Bridge, 2004; Potter, 1985; Shor et al., 2015). In underrepresenting women, the media often facilitates a false narrative that men vastly outnumber women in society While a common critique of the underrepresentation of women in media is that men and women make up equal proportions in our society, it is important to recognize that women do tend to be less involved in crime and terrorism than men. Thus, the differences found in the underrepresentation of women may be more natural compared to some of the unnatural disparities research has found in the underrepresentation of women in the news and films. These differences may also be linked to the celebration of certain terrorists’ acts, where the media provides significantly more coverage. Research has indicated that the vast majority of terrorist incidents receive little to no news coverage, while specific cases, such as 9/11 and the Oklahoma Bombing, are sensationalized by the media due to certain incident-level factors (Chermak & Gruenwald, 2006; Kearns, Betus, & Lemieux, 2019; Mitnik, Freilich, & Chermak, 2018). Gender of the perpetrator has not been considered in this previous research likely because the vast majority (if not all) of sensationalized terrorist attacks are committed by male perpetrators. If the sensationalized cases are all committed by male perpetrators and women are less involved in terrorist activities, it is only natural that their representation in the news media is 52 significantly less. The underrepresentation of female perpetrators found in the current study is consistent with our understanding of sensationalized cases. The current study suspects that terrorist acts committed by female perpetrators tend to not be sensationalized by the media for two primary reasons. First, female terrorists do not fit within the stereotypical narratives that the media constructs regarding gender. Secondly, the types of offenses committed by female perpetrators may have fewer casualties, and thus are less likely to be sensationalized. Future research should examine how gender plays a role in determining if an act of terrorism is sensationalized. While gender was the primary focus of the current study, this study also considered which frames were most common across extremist ideology, including far-right, far-left, and jihadist motivated crimes. The analysis indicated no significant on the frames identified by Nacos (2005), which was expected. However, there were significant differences found across extremist ideology for the disregarded frame, the overshadowed frame, and the fact frame. The results suggested that articles covering far-left motivated incidents had the highest frequencies of the disregarded frame and the lowest frequencies of the fact frame. The current study suspects that this is connected to the nature in which crimes are committed, such that many of the crimes committed by the far-left are committed in groups of people and thus an individual perpetrator is more likely to be disregarded in the news coverage. This would be in contrast to far-right and jihadist motivated incidents that are committed by fewer people or a sole perpetrator in which the news coverage would not disregard anyone and present factual information on the given perpetrator. Furthermore, the results suggest that the articles covering jihadist motivated incidents had the lowest frequencies of the overshadowed frame, indicating that jihadist perpetrators tend to not be least overshadowed in the media. The current study believes that this finding is possibly 53 connected to the United States cultural understandings of terrorism as many people fail to recognize far-right and far-left motivated incidents as terrorism. When a jihadist perpetrator commits an act of terrorism, journalist likely have an easier time explaining and focusing on each perpetrator and thus these perpetrators are not overshadowed in the media coverage. This is also likely because society is more receptive and interested in reading about a jihadist terrorist because it fits their understanding of terrorism. The strengths of the current study are primarily connected to the systematic and quantitative methodical approach that was used to study the media’s representation of gender and terrorism. Employing both random sampling of the perpetrators and the newspaper articles greatly reduced bias that was possibly present in the previous quantitative work done. Furthermore, using case files from the ECDB and the random sampling technique allowed for news coverage of terrorism from over 100 different news sources, including small local newspapers and even the New York Times. Since media is so pervasive, including both local and national news was important as it allowed the current study to consider how all news media represents male and female perpetrators of terrorism. Finally, having news coverage of terrorism that dated back to 1993 allowed the current study to account for a 25-year range of news coverage on the media’s representation of terrorism. This study had several limitations, including the limited statistical analysis, the type of new sources primarily sampled, and the limited number of jihadist female offenders in the ECDB. As discussed, because there were so few cases of Nacos (2005) frames found in the 270 articles reviewed, the statistical analysis that was conducted was limited and should be met with caution, especially the binary logistic regression on the terrorist for the sake of love frame. Another limitation was that the vast majority (89%) of the news articles we reviewed came from unbiased new sources. Unfortunately, this could not have been controlled because it was simply 54 a result of random sampling news articles from the respective case files. However, having a more equal distribution of unbiased new sources, right-wing, and left-wing new sources would likely have provided a more accurate representation of the media’s representation of gender and terrorism. Finally, as discussed, the current study ran into a sampling issue because there were only eight identified female jihadist perpetrators who had committed an illegal violent incident in the ECDB and fifteen were needed. To solve this issue, seven more female jihadist perpetrators were sampled from the illegal financial schemes data. Although there did not seem to be any evident differences between the media’s representations between violent and financial schemes jihadist perpetrator, the current study recognizes this a sampling limitation. More research is needed to better understand the media’s representation of gender and terrorism. While the current study focused only on domestic terrorism, more work is needed on the media’s representation of global terrorism. Future research can also extend these findings by looking at different types of media (television, movies, political advertisements, etc.) to see if there are differences in the way in which the mass media represents male and female terrorists and different types of extremist ideology. The overshadowed frame should also be further studied within the context of the media’s representation of other types of crimes and/or professional occupations. Research should examine if female politicians are more frequently overshadowed when compared to male politicians. Additionally, looking at the frames identified by Nacos (2005) and the frames identified in the current study within the context of female and male preparators of non-terrorist related crimes would provide a better picture of how the mass media represents male and female offenders. 55 APPENDICES 56 APPENDIX A: News Sources Utilized News Sources Utilized in Alphabetical Order: 1 ABC News 2 ADL 3 Agence France-Presse 4 Alaksa Dispatch 5 Albuquerque Journal 6 Anchorage Daily News 7 Anchorage Press 8 Associated Press 9 Atlanta Journal and Constitution 10 Augusta Chronicle 11 Billings Gazette 12 Boston Globe 13 Burlington County Times 14 Canadian Press 15 Carleton Place 16 CBS 17 Christian Science Monitor 18 Cleveland Plain Dealer 19 CNN 20 Courthouse News Service 21 Daily Beast 22 Daily Mail 23 Daily News (Los Angeles) 24 Denver Post 25 Denver Rocky Mountain News 26 Deseret News 27 Detroit Free Press 28 Deutsche Presse-Agentur 29 Fairbanks Daily News-Miner 30 Fox 31 Fox 5 Vegas 32 Global Report 33 Go UpState 34 Greenwire 35 Herald Bulletin 36 Hinterland Gazette 37 Houston Chronicle 38 Info Shop 39 Insurance News 40 KCRG 41 Las Vegas Sun 42 Legal Monitor Worldwide 43 Los Angeles Times 44 Media Mouse 45 Metrolic 46 Metropolitan News Enterprise 47 MetroWest Jewish News 48 Milwaukee Journal Sentinel 49 Missoula Independent 50 MLive 51 National Post 52 New Jersey Record 53 New York Daily News 54 New York Times 55 News10/KXTV 56 Newsday 57 NewsOK 58 NewWest 59 NY Dailey News 60 Orange County Register 61 Orlando Sentinel 62 Philadelphia Daily News 63 Pittsburgh Tribune Review 64 Post Bulletin 65 PR Newswire 66 Prince Rupert Daily News 67 Reuters 68 Right Vision News 69 Sacramento Bee 70 Salt Lake Tribune 57 71 San Bernardino Sun 72 San Jose Mercury News 73 Seattle Post-Intelligencer 74 Seattle Times 75 St. Louis Post-Dispatch 76 St. Petersburg Times 77 Star-News 78 The Arab American News 79 The Bismarck Tribune 80 The Commercial Appeal 81 The Courier 82 The Guardian 83 The Independent 84 The New York Post 85 The Oregonian 86 The Ottawa Citizen 87 The Patriot-News 88 The Philadelphia Inquirer 89 The Press of Atlantic City 90 The Press-Enterprise 91 The Register-Guard 92 The Telegraph India 93 The Times of India 94 Times of Oman 95 Twin Cities Dailey Planet 96 U.S. Fed News 97 U.S. Newswire 98 Union Daily Times 99 United Press International 100 US Official News 101 USA Today 102 Vail Daily 103 Voice of America 104 Washington Post 105 Washington Times 106 WBTV 107 WFMZ-TV 69 News 108 Willamette Week 109 Windsor Star 58 APPENDIX B: Codebook for Thesis 1. Article Number: The identification number for the article. 2. Perpetrator ID: The identification number from the list of perpetrators. 3. Gender of Perpetrator: The classification of the gender of the perpetrator from the ECDB. a. Male (0) b. Female (1) 4. Extremist Ideology: The classification of extremist ideology of the perpetrator from the ECDB. a. Far-Right Extremist (0) b. Far-Left Extremist (1) c. Jihadist (2) 5. Crime Type: The classification of the most serious crime of the perpetrator from the ECDB. a. Homicide (0) b. Plot Offenders (1) c. Bombing/Arsons (2) 6. Number of crimes: The total number of crimes the perpetrator was charged with. 7. Date of Crime: The date the crime was committed. 8. Date of Article: The date the article was written. 9. News Source: The source of the newspaper article. 10. News Type: The political leaning of the news source. The categorization will be used from Vanessa Otero’s Media Bias Chart 4.0. Link: https://www.adfontesmedia.com. a. Left-wing news source (0) 59 b. Neutral news source (1) c. Right-wing news source (2) 11. Gender of Reporter: The gender of the lead author of the article. a. Male (0) b. Female (1) 12. Word Count: The overall number of words within the newspaper article. 13. Picture Count: The overall number of picture/images within the newspaper article. 14. Frames: Each frame will be analyzed within each article individually, as there could be multiple frames used to represent a male or female perpetrator. a. The physical appearance frame: No (0), Yes (1). i. This frame has an emphasis on the appearances, the apparel, the facial expressions of female terrorist as opposed to their criminal activity or extremist ideology. ii. If an article discusses or describes the physical appearance, apparel, attractiveness, and/or facial expression of the perpetrator it would be categorized under the physical appearance frame. iii. “The news dwells on the looks, the ready smiles, or the carefully chosen apparel of female terrorists that seem in sharp contrast to the image of a tough terrorist” – Nacos (2005) iv. Examples include: Taken from Nacos (2005) 1. A newspaper article about the first female Palestinian suicide bomber Wafra Idris began with the sentence, “She was an attractive, auburn haired graduate who had a loving family and likes to wear sleeveless dresses and make-up.” 60 2. In another report Idris was described as a woman with “long, dark hair tied back with a black-and-white keffiyeh”. 3. A report about the wave of “Palestinian women strapping explosives to their bodies and becoming martyrs” on the website of the Christian Broadcasting Network was headlined “Lipstick Martyrs: A New Breed of Palestinian Terrorists.” 4. An article in the New York Times that emphasized the similarities between a Palestinian suicide bomber and her Israeli victim, both girls in their teens, began with the following words: “The suicide bomber and her victim look strikingly similar. Two high school seniors in jeans with flowing black hair.” b. The family connection frame: No (0), Yes (1). i. The family connection frame has an emphasis on the female terrorist's connection to the terrorist organization in terms of her family. ii. If an article discusses or describes the perpetrator in regard to their family connection or family background to the terrorist organization, it would be categorized under the family connection frame. This would only be categorized under the family connection frame, if the family member(s) are somehow involved in the terrorist organization as well. If an article simply discusses the family of the perpetrator and the family isn’t involved, we would not count this. iii. “In view of this kind of media focus, it is hardly surprising that the news pays a great deal of attention to the family backgrounds of female terrorists.” – Nacos (2005) 61 iv. Examples include: Taken from Nacos (2005) 1. One instructive example is the catchy sound bite “Black Widows” that the news media coined and repeated over and over again, when reporting on female Chechen terrorists. By invoking the image of the widow, clad from head to toe in black, the news perpetuated the image of the vengeance-seeking widow who becomes a terrorist because her husband was killed by Russian troops—a woman with a strong personality rather than political motive. c. Terrorist for the sake of love: No (0), Yes (1). i. The terrorist for the sake of love frame creates an image of women’s involvement to a love connection with a male terrorist. Although both men and women are recruited by romantic partners, the media typically will only discuss this phenomenon with female terrorists. ii. If an article discusses or describes the perpetrator in regard to a love interest, it would be categorized under the terrorist for the sake of love frame. This would only be categorized under the terrorist for the sake of love frame, if the love interest/connection are somehow involved in the terrorist organization as well. If an article simply discusses the romantic partner of the perpetrator and the romantic partner isn’t involved, we would not count this. iii. “The popular image of the women terrorist for the sake of love—not for deeply held political reasons. The “love connection” has been a frequent theme in the media’s coverage of gender terrorism.” – Nacos (2005) 62 iv. Examples include: Taken from Nacos (2005) 1. To explain the large number of female members of the German Red Army Fraction in the 1970s, the news cited male criminologists who said that “a few male terrorists and extremist lawyers in West Germany have had the fanatical devotion of female gang members” and that women join because they “admire someone in the terrorist movement.” 2. According to one long-time female KKK member, a woman who uses the pseudonym Klaliff, “My introduction into the White Pride Movement (WP Movement) was in college where I fell in love with another college student, a man who had been an activist in the WP Movement.” She reveals that many women got involved because they had a boyfriend in the Movement. “I cannot speak for all women in the WP Movement,” she wrote, “but I see the men in the WP Movement as manly men with strong ideals and courage.” d. The women’s liberation/equality frame: No (0), Yes (1). i. This frame represents female terrorist’s extremist expression of feminism. This frame presents an image of women’s involvement in terrorism as a means of furthering gender equality, which is rarely the case. ii. If an article discusses or describes the perpetrator in regard to a feminist, liberation or equality motive, the article would be categorized under the women’s liberation/equality frame. iii. “Seemingly not at all concerned about their public image, female terrorists have been often described as women’s lib extremists. Although this was 63 very common in the past, the contemporary news still explains the motives of female terrorists as the expression of gender equality or the struggle to achieve gender equality quite frequently.” – Nacos (2005) iv. Example include: Taken from Nacos (2005) 1. According to the Times, Dr. Adler said that the publicity surrounding terrorism gives female terrorists “a platform to say, ‘I am liberated from past stereotypes, I am accepted in the ultimate masculine roles. 2. Earlier, in her book Sisters in Crime, Adler wrote, “Despite their broad political pronouncements, what the new revolutionaries [such as the Weather Underground] wanted was not simply urban social gains, but sexual equality.” 3. In Europe, experts provided similar explanations for the large number of female members in terrorist organizations, such as the Red Brigades in Italy and the Red Army Faction in West Germany. According to one news account in 1977, “Italian and German sociologists and news commentators, all of them men, have suggested over the last few weeks that the significant female membership in radical and terrorist groups was an unwelcome consequence of the women’s liberation movement.” e. The tough-as-males/tougher-than-men frame: No (0), Yes (1). i. The tough-as males/tougher-than men frame represents women’s involvement in terrorism as cruel, aggressive and lethal. This narrative 64 pushes an image of women who have these violent traits strictly in order to fit into the organization. ii. If an article discusses or describes women as more aggressive/colder blooded/lethal than men, the article would be categorized under the tough- as males/tougher-than men frame. iii. “Similarly, there is the mass-mediated notion of the female terrorist who, in order to prove that she belongs, tends to be more fanatical, more cruel, more deadly.” – Nacos (2005) iv. Examples include: Taken from Nacos (2005) 1. Matias Antolin, the author of a of a book about female members of the Basque ETA organizations, told a correspondent, “Once in an active service unit they tend to be more cold-blooded and more lethal than the men because they have to prove their worth.” 2. Reports about the violent take-over of a Moscow theater by heavily armed Chechen men and women emphasized that the females were “the most determined and aggressive of the hostage takers” and that they were especially “cruel and threatening and eager to die…” 3. Another image creeps into the tough-as-male frame—that “of the terrorist as lesbian, because everyone knows no ‘real woman’ would hijack planes or cripple middle- age men by shooting them in the kneecaps.” f. The bored, naïve, out-of-touch-with-reality frame: No (0), Yes (1). 65 i. This frame presents women as out of touch with reality and delusional. This frame diminishes the motives or intelligence of female terrorist, while pushes a false narrative of delusional women in these organizations. ii. If an article discusses or describes women as bored, naive, delusional or out of touch with reality, the article would be categorized under the bored, naïve, out-of-touch-with-reality frame. iii. “With respect to female terrorists the notion of the naïve, bored, non- political, out-of-touch-with-reality woman who turns to terrorism, too, is perpetuated in the media—mostly by news sources, sometimes by reporters themselves.” – Nacos (2005) iv. Examples include: Taken from Nacos (2005) 1. In the late 1970s a female criminologist said in an interview: Sometimes a woman turns to terrorism out of simple boredom. It sounds strange, I know, but boredom is one of the pathetic rights and privileges of the middle-class woman. What does a middle- class woman do who doesn’t happen to be interested in a career or college? What does she do in 1978? 2. More than 20 years later, after interviewing two young women who had been recruited by a male cousin to plant incendiary devices in a store in Bahrain, the interviewer wrote a story that was headlined, “From Boredom to Bombs: Two Female Terrorists.” The report described the women as non-political and clueless about the motives of the young man who had recruited them. 15. Extension of the frames: 66 a. The physical appearance frame: i. Did the article directly discuss the clothing of the perpetrator? No (0), Yes (1) ii. Did the article directly discuss facial expressions of the perpetrator? No (0), Yes (1) iii. Did the article directly discuss the attractiveness of the perpetrator? No (0), Yes (1) iv. Did the article directly discuss a physical feature (hair, eyes, etc) of the perpetrator? No (0), Yes (1) b. The family connection frame: i. Did the article link the motive of the perpetrator to their family connection? No (0), Yes (1) ii. Did the article discuss the role of the perpetrator in the family? No (0), Yes (1) iii. If Yes, what was the role of the perpetrator in the family? Leader (1), Supporter (2), Follower (3) c. Terrorist for the sake of love: i. Did the article link the motive of the perpetrator to their love connection? No (0), Yes (1) ii. Did the article describe the history of the relationships between the couple? No (0), Yes (1) iii. Did the article give more attention (word count) to the man or the woman? No (0), Yes (1) d. The women’s liberation/equality frame: 67 i. Did the article directly reference feminism/gender equality? No (0), Yes (1) ii. Did the article directly reference women’s liberation? No (0), Yes (1) iii. Did the article use derogatory terminology (such as feminazi, radical feminist, man-hater, etc) to discuss the perpetrator? No (0), Yes (1) e. The tough-as-males/tougher-than-men frame: i. Did the article discuss the perpetrator as aggressive or hostile? No (0), Yes (1) ii. Did the article discuss the perpetrator cruel or savage? No (0), Yes (1) iii. Did the article discuss the perpetrator as lethal or deadly? No (0), Yes (1) f. The bored, naïve, out-of-touch-with-reality frame: Since this frame is a broad category, I thought it would be best to just break it down. i. Did the article discuss the perpetrator as bored? No (0), Yes (1) ii. Did the article discuss the perpetrator as naïve? No (0), Yes (1) iii. Did the article discuss the perpetrator as out of touch with reality? No (0), Yes (1) 16. New frames: a. Disregarded Frame: An article that discusses an act of terrorism however fails to mention the name of the perpetrator. i. Does the article discuss the crime committed but fail to mention the perpetrator? No (0), Yes (1) b. Overshadowed Frame: An article that does reference the perpetrator however overshadows the perpetrator of a crime by giving more attention to another 68 perpetrator or other logistical matters of the crime (basic information of the crime, court hearings, victims, etc.). i. Does the article overshadow the perpetrator of a crime by giving more attention to another perpetrator or other logistical matters of the crime (basic information of the crime, court hearings, victims, etc.)? No (0), Yes (1) c. Fact Frame: An article that directly referenced a perpetrator and provided simple facts about the perpetrator’s case or crime committed. i. Does the article present simple facts about the perpetrator’s case or crime committed? No (0), Yes (1) d. Life Frame: An article that discusses other aspects of the perpetrator’s life outside of their terrorist activity. i. Doe the article discuss other aspects of the perpetrator’s life besides terrorism? No (0), Yes (1) ii. Example: 1. Briana Waters, 32, was an Evergreen State College student in 2001 when others set fire to the Center for Urban Horticulture in Seattle. 2. Gerlach, 28, who grew up in the Eugene area and worked as a disc jockey in Portland, faces trial on indictments she and Meyerhoff helped topple a high tension power line outside Bend in 1999 and that she served as a lookout while others were setting fire to the Childers Meat Co. plant in Eugene in 2001. 69 REFERENCES 70 REFERENCES Alexander, A. (2017). Women & violent extremism. Future Without Violence, 1-8. Retrieved from https://www.futureswithoutviolence.org. Araüna, N., Dhaenens, F., & Van Bauwel, S. (2017). Historical, temporal and contemporary trends on gender and media. Catalan Journal of Communication & Cultural Studies, 9(2), 177-184. doi:10.1386/cjcs.9.2.177_7. Barnard, J. (2005, December 21). Motion says Virginia man charged in case to testify against others. The Associated Press. Retrieved from https://www.ap.org. Baumann, S., & de Laat, K. (2012). Socially defunct: A comparative analysis of the underrepresentation of older women in advertising. Poetics, 40(6), 514-541. doi: 10.1016/j.poetic.2012.08.002. Benowitz-Fredericks, C. A., Garcia, K., Massey, M., Vasagar, B., & Borzekowski, D. L. G. (2012). Body image, eating disorders, and the relationship to adolescent media use. Pediatric Clinics of North America, 59(3), 693-704. doi: 10.1016/j.pcl.2012.03.017. Benson, M. L., & Gottschalk, P. (2015). Gender and white-collar crime in Norway: An empirical study of media reports. International Journal of Law, Crime, and Justice, 43(4), 535-552. doi: 10.1016/j.ijlcj.2015.01.001. Berko, A., & Erez, E. (2007). Gender, Palestinian women, and terrorism: Women's liberation or oppression? Studies in Conflict & Terrorism, 30(6), 493-519. doi:10.1080/10576100701329550. Bloom, M. (2011a). Bombshell: Women and terrorism. Philadelphia: University of Pennsylvania Press. Bloom, M. (2011b). Bombshells: Women and terror. Gender Issues, 28(1), 1-21. doi:10.1007/s12147-011-9098-z. Boghani, P. (2015, July 13). How ISIS is using women to police other women. PBS. Retrieved from http://www.pbs.org. Boles, M., Adams, A., Gredler, A., & Manhas, S. (2014). Ability of a mass media campaign to influence knowledge, attitudes, and behaviors about sugary drinks and obesity. Preventive Medicine, 67, S40-S45. doi: 10.1016/j.ypmed.2014.07.023. Braudy, L. (2003). From chivalry to terrorism: War and the changing nature of masculinity. New York: Alfred A. Knopf. Brown, J. D. (2002). Mass media influences on sexuality. Journal of Sex Research, 39(1), 42-45. doi:10.1080/00224490209552118. 71 Brumback, K. (2012, August 22). Men sentenced to 5 years each in Ga. militia plot. The Associated Press. Retrieved from https://www.ap.org. Busby, L. J. (1975). Sex-role research on the mass media. The Journal of Communication, 25(4), 107-131. doi:10.1111/j.1460-2466.1975.tb00646.x. Ceci, S. J., Williams, W. M., & Thompson, R. F. (2011). Understanding current causes of women's underrepresentation in science. Proceedings of the National Academy of Sciences of the United States of America, 108(8), 3157-3162. doi:10.1073/pnas.1014871108. Chermak, S. M., & Gruenewald, J. (2006). The media's coverage of domestic terrorism. Justice Quarterly, 23(4), 428. doi:10.1080/07418820600985305. Christen, C. T., & Gunther, A. C. (2003). The influence of mass media and other culprits on the projection of personal opinion. Communication Research, 30(4), 414-431. doi:10.1177/0093650203253366. Christensen, L. S. (2018). The new portrayal of female child sexual offenders in the print media: A qualitative content analysis. Sexuality & Culture, 22(1), 176-189. doi:10.1007/s12119- 017-9459-1. Collins, R. L. (2011). Content analysis of gender roles in media: Where are we now and where should we go? Sex Roles, 64(3), 290-298. doi:10.1007/s11199-010-9929-5. Collins, R. E. (2016). ‘Beauty and bullets’: A content analysis of female offenders and victims in four Canadian newspapers. Journal of Sociology, 52(2), 296-310. doi:10.1177/1440783314528594. Craig, S. (Eds.). (1992). Men, masculinity, and the media. Newbury Park: Sage. Craig, S. L., & McInroy, L. (2014). You can form a part of yourself online: The influence of new media on identity development and coming out for LGBTQ youth. Journal of Gay & Lesbian Mental Health, 18(1), 95-109. doi:10.1080/19359705.2013.77700. Cunningham, K. J. (2003). Cross-regional trends in female terrorism. Studies in Conflict & Terrorism, 26(3), 171-195. doi:10.1080/10576100390211419. Cunningham, K. J. (2007). Countering female terrorism. Studies in Conflict & Terrorism, 30(2), 113-129. doi:10.1080/10576100601101067. Davis, J. (1982). Sexist bias in eight newspapers. Journalism & Mass Communication Quarterly, 59(3), 456-460. doi:10.1177/107769908205900316. DeLouth, T. B., Pirson, B., Hitchcock, D., & Rienzi, B. M. (1995). Gender and ethnic role portrayals: Photographic images in three California newspapers. Psychological Reports, 76, 493–494. 72 Eager, P. W. (2008). From freedom fighters to terrorists: Women and political violence. Burlington, VT: Ashgate Publishing. Easteal, P., Holland, K., & Judd, K. (2015). Enduring themes and silences in media portrayals of violence against women. Women's Studies International Forum, 48, 103-113. doi:10.1016/j.wsif.2014.10.015. Edström, M. (2018). Visibility patterns of gendered ageism in the media buzz: A study of the representation of gender and age over three decades. Feminist Media Studies, 18(1), 77- 93. doi:10.1080/14680777.2018.1409989. Entman, R. M. (1989). How the media affect what people think: An information processing approach. The Journal of Politics, 51(2), 347-370. doi:10.2307/2131346. Ey, L. (2014). The influence of music media on gender role and self-identity: Perceptions of children aged 6 and 10 years. Children Australia, 39(3), 147-160. doi:10.1017/cha.2014.17. Freilich, J. D., Chermak, S. M., Belli, R., Gruenewald, J., & Parkin, W. S. (2014). Introducing the United States extremis crime database (ECDB). Terrorism and Political Violence, 26(2), 372-384. doi:10.1080/09546553.2012.713229. Friedman, B. (2008). Unlikely warriors: How four U.S. news sources explained female suicide bombers. Journalism & Mass Communication Quarterly, 85(4), 841-859. doi:10.1177/107769900808500408. Galician, M. (2004). Sex, love & romance in the mass media: Analysis & criticism of unrealistic portrayals & their influence. Mahwah, N.J: Lawrence Erlbaum Associates. Gauntlett, D. (2002). Media, gender, and identity: An introduction. London, New York: Routledge. Gill, R. (2007). Gender and the media. Cambridge, UK; Malden, MA, USA: Polity. Gillig, T. K., & Murphy, S. T. (2016). Fostering support for LGBTQ youth? The effects of a gay adolescent media portrayal on young viewers. International Journal of Communication, 10, 3828-3850. Global Media Monitoring Project (2015). Gender inequality in the news 1995-2015. Retrieved from http://whomakesthenews.org Gonzalez-Perez, M. (2008). Women and terrorism: Female activity in domestic and international terror groups. New York, New York: Routledge. Goodall, H. (2012). Media's influence on gender stereotypes. Media Asia, 39(3), 160-163. doi:10.1080/01296612.2012.11689932. 73 Gottschalk, P., & Smith, R. (2015). Gender and white-collar crime: Examining representations of women in media? Journal of Gender Studies, 24(3), 310-325. doi:10.1080/09589236.2013.841571. Gough, B. (2007). 'Real men don't diet': An analysis of contemporary newspaper representations of men, food and health. Social Science & Medicine, 64(2), 326-337. doi: 10.1016/j.socscimed.2006.09.011. Green-Pedersen, C., & Stubager, R. (2010). The political conditionality of mass media influence: When do parties follow mass media attention? British Journal of Political Science, 40(3), 663-677. doi:10.1017/S0007123410000037. Greenwald, M. S. (1990). Gender representation in newspaper business sections. Newspaper Research Journal, 11(1), 68-74. doi:10.1177/073953298901100106. Gunter, B. (1995). Television and gender representation. London: John Libbey. Gunther, A. C. (1998). The persuasive press inference: Effects of mass media on perceived public opinion. Communication Research, 25(5), 486-504. doi:10.1177/009365098025005002. Happer, C., & Philo, G. (2013). The role of the media in the construction of public belief and social change. Journal of Social and Political Psychology, 1(1), 321-336. doi:10.5964/jspp.v1i1.96. Howe, A. (1997). "The war against women". Media representations of men's violence against women in Australia. Violence Against Women, 3(1), 59-75. Humprecht, E., & Esser, F. (2017). A glass ceiling in the online age? Explaining the underrepresentation of women in online political news. European Journal of Communication, 32(5), 439-456. doi:10.1177/0267323117720343. Kearns, E. M., Betus, A. E., & Lemieux, A. F. (2019). Why do some terrorist attacks receive more media attention than others? Justice Quarterly,1-24. doi:10.1080/07418825.2018.1524507. Krijnen, T., & Bauwel, S. V. (2015). Gender and media: Representing, producing, consuming. New York; London: Routledge, Taylor & Francis Group. Landor, R. V., & Eisenchlas, S. A. (2012). "Coming clean" on duty of care: Australian print media's representation of male versus female sex offenders in institutional contexts. Sexuality and Culture, 16(4), 486. doi:10.1007/s12119-012-9134-5. Lauzen, M. M., Dozier, D. M., & Horan, N. (2008). Constructing gender stereotypes through social roles in prime-time television. Journal of Broadcasting & Electronic Media, 52(2), 200-214. doi:10.1080/08838150801991971. 74 Lloyd, M., & Ramon, S. (2017). Smoke and mirrors: U.K. newspaper representations of intimate partner domestic violence. Violence Against Women, 23(1), 114-139. doi:10.1177/1077801216634468. Lont, C., & Bridge, M. (2004). The face of the front page: A content analysis of U.S. newspapers. In R. Lind (Ed.), Race/Gender/Media (pp. 140-146). Boston: Pearson/Allyn & Bacon. Luebke, B. F. (1985). News about women on the wire. Journalism & Mass Communication Quarterly, 62(2), 329-333. doi:10.1177/107769908506200215. MacKinnon, K. (2003). Representing men: Maleness and masculinity in the media. London: Arnold. MacLean, A., Sweeting, H., Walker, L., Patterson, C., Räisänen, U., & Hunt, K. (2015). "It's not healthy and it's decidedly not masculine": A media analysis of UK newspaper representations of eating disorders in males. BMJ Open, 5(5), e007468. doi:10.1136/bmjopen-2014-007468. Makin, D. A., & Hoard, S. (2014). Understanding the gender gap in domestic terrorism through criminal participation. Criminal Justice Policy Review, 25(5), 531-552. doi:10.1177/0887403413478016. Matud, M. P., Rodríguez, C., & Espinosa, I. (2011). Gender in Spanish daily newspapers. Sex Roles, 64(3), 253-264. doi:10.1007/s11199-010-9874-3. Messner, M. A., Dunbar, M., & Hunt, D. (2000). The televised sports manhood formula. Journal of Sport & Social Issues, 24(4), 380-394. doi:10.1177/0193723500244006. Mitnik, Z. S., Freilich, J. D., & Chermak, S. M. (2018). Post-9/11 coverage of terrorism in the new york times. Justice Quarterly, 1-25. doi:10.1080/07418825.2018.1488985. Moaveni, A. (2015). ISIS women and enforcers in Syria recount collaboration, anguish and escape. New York Times. Retrieved from http://www.nytimes.com. Morgan, Skyler J. (2018, November). An exploration of the media’s representation of gender in far right and far left terrorist organizations. Paper presentation at the 74th Annual Meeting of the American Society of Criminology, Atlanta, GA. Nacos, B. L. (2005). The portrayal of female terrorists in the media: Similar framing patterns in the news coverage of women in politics and in terrorism. Studies in Conflict & Terrorism, 28(5), 435-451. doi:10.1080/10576100500180352. Neely, C. L. (2015). You're dead? so what? Media, police, and the invisibility of black women as victims of homicide. East Lansing, Michigan: Michigan State University Press. O'Rourke, L. A. (2009). What's special about female suicide terrorism? Security Studies, 18(4), 681-718. doi:10.1080/09636410903369084. 75 Otero, V. (2018). Media bias chart: Version 4.0. Retrieved from: https://www.adfontesmedia.com. Plaza, J. F., Rivas-Nieto, P., & Rey-García, P. (2017). Representation of women as terrorists and victims of terrorism in the Spanish press. the case of el país newspaper. Revista Latina De Comunicación Social, (72), 129-144. doi:10.4185/RLCS-2017-1157. Potter, W. J. (1985). Gender representation in elite newspapers. Journalism & Mass Communication Quarterly, 62(3), 636-640. doi:10.1177/107769908506200327. Raghavan, S. V., & Balasubramaniyan, V. (2014). Evolving role of women in terror groups: Progression or regression? Journal of International Women's Studies, 15(2), 197-211. Rajan, V. G. J. (2011). Women suicide bombers: Narratives of violence. Milton Park, Abingdon, Oxon; New York: Routledge. Richardson, N., & Wearing, S. (2014). Gender in the media. Basingstoke: Palgrave Macmillan. Roberts, D. F. (2000). Media and youth: Access, exposure, and privatization. Journal of Adolescent Health, 27(2), 8-14. doi:10.1016/S1054-139X(00)00128-2. Ross, K. (2010). Gendered media: Women, men, and identity politics. Lanham: Rowman & Littlefield Publishers. Scheufele, D. A., Shanahan, J., & Kim, S. (2002). Who cares about local politics? media influences on local political involvement, issue awareness, and attitude strength. Journalism & Mass Communication Quarterly, 79(2), 427-444. doi:10.1177/107769900207900211. Schug, J., Alt, N. P., Lu, P. S., Gosin, M., & Fay, J. L. (2017). Gendered race in mass media: Invisibility of Asian men and black women in popular magazines. Psychology of Popular Media Culture, 6(3), 222-236. doi:10.1037/ppm0000096. Seal, L. (2010). Women, murder, and femininity: Gender representations of women who kill. New York: Palgrave Macmillan. Shor, E.,van de Rijt, A., Miltsov, A., Kulkarni, V., & Skiena, S. (2015). A paper ceiling: Explaining the persistent underrepresentation of women in printed news. American Sociological Review, 80(5), 960-984. doi:10.1177/0003122415596999. Shukovsky, P. (2008, June 20). Teacher gets 6 years for 2001 UW firebombing. Seattle Post- Intelligencer. Retrieved from https://www.seattlepi.com. Sjoberg, L., & Gentry, C. E. (2011). Women, gender, and terrorism. Athens: University of Georgia Press. Skaine, R. (2006). Female suicide bombers. Jefferson, N.C: McFarland. 76 Speckhard, A. (2008). The emergence of female suicide terrorists. Studies in Conflict & Terrorism, 31(11), 995-1023. doi:10.1080/10576100802408121. Teele, D. L., Kalla, J., & Rosenbluth, F. (2018). The ties that double bind: Social roles and women's underrepresentation in politics. American Political Science Review, 112(3), 525- 541. doi:10.1017/S0003055418000217. Tuchman, G. (1978). The symbolic annihilation of women by the mass media. In G. Tuchman, A. K. Daniels, & J. Benet (Eds.), Hearth and home: Images of women in the mass media (pp. 3–38). New York: Oxford University Press. Tucker, E. (2011, October 4). FBI to be honored for work in DC subway bomb plot. The Associated Press. Retrieved from https://www.ap.org. Victor, B. (2004). Army of roses: Inside the world of Palestinian women suicide bombers. London, England: Constable and Robinson. Wakefield, M. A., Loken, B., & Hornik, R. C. (2010). Use of mass media campaigns to change health behaviour. The Lancet, 376(9748), 1261-1271. doi:10.1016/S0140- 6736(10)60809-4. Ward, L. M. (2016). Media and sexualization: State of empirical research, 1995-2015. The Journal of Sex Research, 53(4-5), 560-577. doi:10.1080/00224499.2016.1142496. Weinberg, L., & Eubank, W. (2011). Women’s involvement in terrorism. Gender Issues, 28(1), 22-49. doi:10.1007/s12147-011-9101-8. doi:10.1177/0891243287001002002. West, C., & Zimmerman, D. H. (1987). Doing gender. Gender and Society, 1(2), 125-151. Windsor Star. (2010, March 30). Michigan says sect plotted to kill cops. Windsor Star. Retrieved from https://windsorstar.com. Wood, J.T. (1994) Gendered media: the influence of media on views of gender. In J. T. Wood, Gendered Lives: Communication, Gender, and Culture (pp. 231-244). Scarborough, ON, Wadsworth. Wright, A., & Pritchard, M. E. (2009). An examination of the relation of gender, mass media influence, and loneliness to disordered eating among college students. Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity, 14(2), e144-e147. doi:10.1007/BF03327813. Young, M. E., Norman, G. R., & Humphreys, K. R. (2008). Medicine in the popular press: The influence of the media on perceptions of disease. PloS One, 3(10), e3552. doi: 10.1371/journal.pone.0003552. 77