FRAMING CAMPUS CARRY: THE CLASH BETWEEN THE FIRST AND SECOND AMENDMENTS ON AMERICAN COLLEGE AND UNIVERSITY CAMPUSES by KRYSTAL EVE NOGA-STYRON A DISSERTATION Presented to the School of Journalism and Communication and the Graduate School of the University of Oregon in partial fulfillment of the requirements for the degree of Doctor of Philosophy March 2020 DISSERTATION APPROVAL PAGE Student: Krystal Eve Noga-Styron Title: Framing Campus Carry: The Clash Between the First and Second Amendment on American College and University Campuses This dissertation has been accepted and approved in partial fulfillment of the requirements for the Doctor of Philosophy degree in the School of Journalism and Communication by: Timothy Gleason Chairperson Jesse Abdenour Core Member Scott Maier Core Member Nancy Shurtz Institutional Representative and Kate Mondloch Interim Vice Provost and Dean of the Graduate School Original approval signatures are on file with the University of Oregon Graduate School. Degree awarded March 2020 ii © 2020 Krystal Eve Noga-Styron iii DISSERTATION ABSTRACT Krystal Eve Noga-Styron Doctor of Philosophy School of Journalism and Communication March 2020 Title: Framing Campus Carry: The Clash Between the First and Second Amendments on American College and University Campuses This mixed-methods study focused on how individuals and the news media frame the subject of campus carry; it did so with a specific focus on determining whether campus carry is understood as being a free speech or a gun rights issue as defined by the First or Second Amendments to the Constitution of the United States. In an attempt to determine whether the news media set an agenda in the debate on campus carry, a cross-sectional survey (N = 517) of university students, faculty, and staff was administered. The goal of the survey was to collect empirical data to determine to what extent these individual’s demographics and perceptions predict their attitudes toward three dependent variables related to the possession of firearms on campus. The survey results indicated that, on average, the respondents (1) would not feel more secure on campus, (2) they would change their behavior in the classroom, and (3) they would refrain from engaging in controversial classroom discussions if individuals were legally allowed to carry firearms on campus. While level of support for the Second Amendment was predictive of these attitudes, level of support for the First Amendment was not. A descriptive content analysis of text extracted from 143 newspaper articles was also conducted and the results revealed that the media frame newspaper articles on the subject of iv campus carry by centering the issue on the right to bear arms, as endorsed by the Second Amendment, and by taking an anti-campus carry position in the debate. Both the survey and the content analysis indicated that there is a general sentiment against campus carry, and that campus carry is considered more of a Second Amendment issue, associated with the right to bear arms, than a First Amendment issue, associated with the freedom of speech. There was, therefore, some alignment between the survey data and the content analysis data indicating that news frames may have the potential to exert an influence on audience perceptions and attitudes toward the issue of campus carry. v CURRICULUM VITAE NAME OF AUTHOR: Krystal Eve Noga-Styron GRADUATE AND UNDERGRADUATE SCHOOLS ATTENDED: University of Oregon, Eugene University of Illinois, Champaign-Urbana DEGREES AWARDED: Doctor of Philosophy, School of Journalism and Communication, 2020, University of Oregon Master of Arts, School of Journalism and Communication, 2004, University of Oregon Juris Doctorate, School of Law, 2004, University of Oregon Bachelor of Arts, Political Science, 1999, University of Illinois AREAS OF SPECIAL INTEREST: Constitutional Law Justice Studies PROFESSIONAL EXPERIENCE: Professor of Law and Justice, Central Washington University, 14 years Judge Pro Tempore, Marysville Municipal Court, 10 years Judge Pro Tempore, Lewis County District Court, 11 years Deputy Prosecuting Attorney, Lewis County District Court, 1 year PUBLICATIONS: Britto, S. & Noga-Styron, K. E. (2018). Media consumption and support for capital punishment. In S. G. Tibbetts & C. T. Hemmens (Eds.), Criminological Theory: a Text/Reader (3rd Edition). Thousand Oaks: Sage. vi Noga-Styron, K. E. & Olivero, J. M. (2017). The legalization of marijuana in Washington State and its effect on the price of the recreational and the medical product. Crime and Human Behavior, vol. 1, 70-84. Dahmen, N. S., Abdenour, J., McIntyre, K. & Noga-Styron, K. E. (2017). Covering mass shootings: journalists’ perceptions of coverage and factors influencing attitudes. Journalism Practice. DOI: 10.1080/17512786.2017.1326832. Noga-Styron, K. E., Olivero, J. M., & Britto, S. (2016). Predatory Journals in the Criminal Justices Sciences: Getting our Cite on the Target. Journal of Criminal Justice Education, DOI 10.1080/10511253.2016.1195421. Britto, S. & Noga-Styron, K. E. (2015). The belief that guns deter crime and support for capital punishment. Criminal Justice Studies, vol. 28, no. 3, 314-335. Noga-Styron, K. E., & Olivero, J. M. (2014). A content analysis of homosexual related issues and introductory undergraduate law-enforcement texts: becoming common curriculum. Journal of Behavioral and Social Sciences, vol. 1, no. 4, 217-225. Britto, S. & Noga-Styron, K. E. (2013). Media consumption and support for capital punishment. Criminal Justice Review, vol. 39, no. 1, 81-100. Noga-Styron, K. E., Reasons, C., & Peacock, D. (2012). The last acceptable prejudice: an overview of LGBT social and criminal injustice issues within the United States. Contemporary Justice Review, vol. 15, no. 4, 369-398. Noga-Styron, K. E., & Ruble, M. (2012). Crime and the Great Recession. In J. Chin & J. Owens, Contemporary Issues in Criminal Justice: a Research-Based Introduction. Flushing: Looseleaf Law Publication, 59 - 73. Noga-Styron, K. E. (2010). The top-two primary in Washington State. National Civic Review, vol. 98, no. 4, 49-51. Noga-Styron, K. E. (2009). Reformatories. In H.T. Greene & S.L. Gabbidon, Encyclopedia of Race and Crime, (pp. 705 - 708). Los Angeles: Sage. Noga-Styron, K. E. (2008). Reconstructing the Fourth Amendment: a history of search & seizure, 1789-1868. Criminal Justice Review, vol. 33, no. 3, 421-423. Noga-Styron, K. E. (2007). Securitizing copyrights: an answer to the Sonny Bono Copyright Term Extension Act. Tulane Journal of Technology and Intellectual Property, vol. IX, Issue 1, 1-73. Lead Article. vii ACKNOWLEDGMENTS I wish to express sincere appreciation to University of Oregon Professors Jesse Abdenour, Patricia Curtin, Timothy Gleason, Scott Maier, Nancy Shurtz, and Anthony Waters for their assistance in the preparation of this manuscript. In addition, special thanks are due to Megan Johnson for her invaluable help in coding the newspaper articles in this study. I also thank my husband, Paul Styron, and my children, Adam, Luke, and Jude for their love, support, and patience during this undertaking. viii This work is dedicated to the victims of gun violence everywhere, including my family. ix TABLE OF CONTENTS Chapter Page I. INTRODUCTION ............................................................................................... 1 Campus Carry Defined ...................................................................................... 2 The Importance of the Campus Carry Issue ....................................................... 3 Purpose of this Research .................................................................................... 5 II. COMMUNICATION LITERATURE REVIEW ................................................. 7 Agenda-Setting Theories: First-Level and Second-Level ................................... 7 Framing Theories .............................................................................................. 10 III. REVIEW OF LAW: THE HISTORY AND EVOLUTION OF THE FIRST AMENDMENT OF THE UNITED STATES CONSTITUTION ....................... 14 First Amendment ............................................................................................... 14 Introduction ....................................................................................................... 14 Early History of Seditious Libel and the Ratification of the First Amendment ... 15 Seditious Libel During the Early Days of the Republic ...................................... 18 Sedition in the Twentieth Century...................................................................... 19 Academic Freedom ............................................................................................ 25 Symbolic Speech ............................................................................................... 29 IV. REVIEW OF LAW: THE HISTORY AND EVOLUTION OF THE SECOND AMENDMENT OF THE UNITED STATES CONSTITUTION ....................... 34 Second Amendment........................................................................................... 34 Introduction ....................................................................................................... 34 x Chapter Page Legislative Intent and the Ratification of the Second Amendment ..................... 36 Plain Wording of the Text ................................................................................. 39 Second Amendment Precedent and Interpretation .............................................. 41 The Path to Heller.............................................................................................. 46 V. THE CONTENTIOUS INTERSECTION OF GUNS AND EXPRESSION ........ 53 The Undeniable Power and Influence of the NRA ............................................. 53 Distinguishing Whether Campus Carry Falls Within the Scope of First Amendment Protected Expression ..................................................................... 55 VI. CONTEXT OF RESEARCH ............................................................................. 58 History and Research on the Campus Carry Debate ........................................... 58 Prior Research Studies on Campus Carry ........................................................... 59 Gap in the Literature on Campus Carry .............................................................. 60 Populations of Interest in this Study ................................................................... 60 VII. RESEARCH QUESTIONS AND RESEARCH METHODS ............................ 63 Introduction ....................................................................................................... 63 Research Questions and Hypothesis ................................................................... 63 Survey ............................................................................................................... 64 Content Analysis of Survey Responses .............................................................. 67 Justification for Using Multivariate Statistical Methods to Test H1 .................... 69 Analysis of Survey Data –Testing H1 ................................................................ 71 Statistical Significance and Correcting for Type I Errors ................................... 75 xi Chapter Page Sample Size ....................................................................................................... 77 Effect Size ......................................................................................................... 79 Content Analysis of Newspaper Articles ............................................................ 80 Inter-Coder Reliability ....................................................................................... 84 Inter-Coder Agreement Scores ........................................................................... 85 Methodological Triangulation ............................................................................ 86 VIII. RESULTS OF SURVEY ................................................................................ 89 Screening and Cleaning of Survey Data ............................................................. 89 Description of the Survey Respondents .............................................................. 90 Descriptive Statistics for Dependent Variables................................................... 91 Descriptive Statistics for Predictor Variables ..................................................... 92 Testing of Assumptions of Multiple Linear Regression ...................................... 94 Prediction of Feeling Secure on Campus using Multiple Linear Regression ....... 95 Prediction of Changing Behavior in Classroom.................................................. 98 Prediction of Refraining from Engaging in Controversial Classroom Discussions ....................................................................................................... 101 IX. RESULTS OF CONTENT ANALYSIS ............................................................ 104 Sources of Data for Content Analysis ................................................................ 104 Methodological Triangulation ............................................................................ 110 X. DISCUSSION AND CONCLUSION ................................................................. 112 Discussion ......................................................................................................... 112 xii Chapter Page Limitations ........................................................................................................ 117 Conclusion ........................................................................................................ 120 Future Research ................................................................................................. 124 XI. APPENDICES .................................................................................................. 126 Appendix A: Survey Instrument ...................................................................... 126 Appendix B: Coding Instrument ...................................................................... 148 REFERENCES CITED ........................................................................................... 155 xiii LIST OF TABLES Table Page 1. Campus Carry Law on U.S. College and University Campuses .......................... 3 2. Dependent Variables in Multiple Linear Regression .......................................... 72 3. Predictor Variables in Multiple Linear Regression ............................................. 74 4. Inter-Coder Agreement ...................................................................................... 87 5. Screening for Missing Values ............................................................................ 89 6. Characteristics of Survey Respondents .............................................................. 91 7. Descriptive Statistics for Dependent Variables in Survey ................................... 92 8. Descriptive Statistics for Predictor Variables in Survey ..................................... 93 9. Prediction of Feeling Secure on Campus using Survey Data .............................. 96 10. Prediction of Changing Behavior in Classroom using Survey Data .................... 99 11. Prediction of Refraining from Engaging in Controversial Classroom Discussions using Survey Data .......................................................................... 101 12. Frequencies of Newspaper Sources .................................................................... 104 13. Frequencies of Types of Newspaper Articles ..................................................... 105 14. Frequencies of Newspaper Articles Regarding Feeling Secure, Changing Behavior, and Classroom Discussions................................................................ 105 15. Frequencies of Newspaper Items Reporting Different Positions on Campus Carry ................................................................................................................. 107 16. Frequencies of Newspaper Articles Discussing the First Amendment and the Second Amendment........................................................................................... 107 xiv Table Page 17. Frequencies of Themes/Frames Identified by Content Analysis ......................... 108 xv I. INTRODUCTION The debate over permitting firearms on public college and university campuses reignites in the aftermath of each school shooting in the United States (U.S.). Conversations erupt about how the situation might have been different if students, faculty, or staff had been armed, or at least allowed to carry concealed weapons on campus (Pohl, 2016). In the wake of several deadly attacks on campuses across the country, some lawmakers, as well as colleges and universities across the U.S., are reconsidering whether to permit guns at institutions of higher education. Some have already allowed them (National Conference of State Legislatures, 2018). Based on the most recent statistics available, by far, 2018 was the worst year on record for gun violence in U.S. schools. According to a database from the Center for Homeland Defense and Security (2019), researching as far back as 1970, there were more school shooting incidents and more resultant deaths in 2018 than in any other year on record. That information is corroborated by Everytown for Gun Safety (2019) which, in response to a lack of data and research on the subject, began tracking shootings on school grounds in 2013, shortly after the mass shooting at Sandy Hook Elementary in Newtown, Connecticut. The data are collected to gain a better understanding of the issue of gun violence on school, college, and university campuses. Those data are factored into campus-carry law and policy decisions. Proponents of campus-carry laws and policies argue that allowing people to possess firearms on public university and college campuses will create safer environments by deterring would-be shooters and by enabling students, faculty, and staff to defend themselves and others in the event of a shooting (Fennell, 2009; Kopel, 2014; 1 Klukowski, 2014; Moody, 2014; Vasek, 2014). Opponents of campus-carry laws argue that permitting guns on college and university campuses will create unsafe environments (Barnes, 2017; Cavanaugh, et al., 2012; DeBrabander, 2016; Horwitz, 2016; Lewis, 2017; Patten, et al., 2012; Smith, 2012; Wolcott, 2017). They worry that people will draw weapons in response to merely perceived or minor threats, or when angry, or that guns might accidentally discharge, causing injury (DeBrabander, 2016). Yet, one of the greatest potential harms of campus-carry laws may not be physical in nature. Rather, it may be the loss of open and critical discourse – two of the basic tenets of both higher education and democracy. Campus Carry Defined The carrying of firearms by students, staff, and faculty on U.S. college and university campuses is known as campus carry. Each individual state has the discretion to mandate its own laws on campus carry and, at present, three different types of campus carry laws have been enacted across the country. These laws include mandatory laws (guns allowed), non-permissive laws (guns banned), and institutional laws (campuses can choose). As of February 2020, eleven states allow for the carrying of firearms on public college and university campuses, seventeen states ban concealed carry on public college and university campuses, and the remaining twenty-two states have some combination of institutional and/or non-permissive laws on campus carry. Table 1 lists which states fall under which type of campus carry law. 2 Table 1 Campus Carry Law on U.S. College and University Campuses Mandatory Laws (guns allowed) Arkansas, Colorado, Georgia, Idaho, Kansas, Mississippi, Oregon, Tennessee*, Texas, Utah, and Wisconsin Non-Permissive Laws (guns banned) California, Florida, Illinois, Louisiana, Massachusetts, Michigan, Missouri, Nebraska, Nevada, New Jersey, New Mexico, New York, North Carolina, North Dakota, South Carolina and Wyoming. Institutional Laws (campuses can choose) Alabama, Alaska, Arizona, Connecticut, Delaware, Hawaii, Indiana, Iowa, Kentucky, Maine, Maryland, Minnesota, Montana, New Hampshire, Ohio, Oklahoma, Pennsylvania, Rhode Island, South Dakota, Vermont, Virginia, Washington and West Virginia. * Tennessee allows faculty members with licenses to carry weapons on campus but the law does not apply to students or to the general public. Source: National Conference of State Legislatures (2018) Importance of the Issue of Campus Carry Campus carry is an issue that is ever present in the news media, and that is debated in state legislatures and on college and university campuses across the country. As more and more governors sign bills permitting firearms on campus, more and more professors submit letters of resignation or take early retirements, and more and more students transfer campuses (Friesen, 2016). The debate illustrates a disconnect between lawmakers and university administrators, faculty, staff, students, and campus police who overwhelmingly oppose the idea of allowing firearms on campus (Friesen, 2016). Furthermore, the debate on campus carry is indicative of a larger ideological debate over the goals and character of higher education. Historically, the emphasis of higher education in the United States has been on critical thinking, a fostering of the marketplace of ideas, and the encouragement of questioning, scrutinizing, and challenging orthodoxy. These are the hallmarks of a society that not only pursues liberty and justice, but that allows individuals to reach their full potential (Friesen, 2016). The 3 campus carry movement is indictive of the alternative ideological view of higher education; one that emphasizes obedience to, and maintenance of, traditional authority (Friesen, 2016). This ideology views critical thinking as a threat to authority and prefers that disagreements be resolved by force – by brawn, not by brains. In this view, liberty and security is not preserved by critique and analysis, but by encouraging everyone to carry a gun (Friesen, 2016). Adding firearms into the equation further expands the ideological divide over how the goals and character of higher education affect students and professors in the classroom. If guns are to be allowed on campus, many students and faculty report that they will be more cautious and less forthright, that they will censor their opinions, and that they will feel less safe (Friesen, 2016). Those on the other side of the debate argue that allowing guns on campus will make them feel more secure. Both proponents and opponents of campus carry laws argue that this a matter of life or death. Proponents define the life or death struggle over campus carry as a matter of security, or protection against threats, whereas opponents define it as a matter of safety, or being safe from threats (Morgan, 2018). In reality, college years are a challenging time in students’ lives. According to research, the median age for initial presentation of mental disorders is 22 with an interquartile range of 19–25 (Kessler, Amminger, Aguilar-Gaxiola, Alonso, Lee, & Ustun, 2007). This is the age at which many people are enrolled as students at colleges and universities – and away from the family and friends who know them best. Many students, even those who will not suffer from mental disorders, will experience serious stress and emotional issues as they transition into adulthood. Many professors report 4 verbal abuse and physical threats from students without guns present. If guns are introduced on campus, professors worry about students’ reactions when they miss assignments, or are caught plagiarizing or cheating, or when they fail an exam. Among the majority of those who work, study, and live on college and university campuses, it is clear that firearms are an unwelcome addition on campus (Cavanaugh et al., 2012; Friesen, 2016; Patten et al., 2013; Thompson, Price, Dake, & Teeple, 2013). Purpose of This Research The news media play an important role in the discussion on campus carry. The objective of this research is to understand how the news media frame the issue of campus carry and whether they set an agenda that affects the public’s understanding of the topic. The first goal is to ascertain public attitudes and perspectives on the subject of campus carry. The second goal is to determine how the media discuss and describe campus carry. Third, comparing public opinions on campus carry to media coverage on the same topic, this research hopes to provide an understanding of whether social factors, particularly media use, influence attitudes and perspectives on campus carry. Finally, by surveying public attitudes and perspective, and by analyzing media coverage on the topic, this research seeks to determine whether campus carry is understood as being a free speech or a gun rights issue as defined by the First or Second Amendments to the Constitution of the United States. This research begins with a review of the pertinent literature in the field of communication, as well as a review of the history and evolution of the constitutional principles at the center of this debate – the First Amendment and the Second Amendment to the United States Constitution. What follows is a discussion of the contentious 5 intersection of guns and expression. The research is then put into context, followed by a presentation of the research questions and methods, as well as the results of both a campus survey and a content analysis of news media on the subject of campus carry. This research culminates with a discussion of how the survey and content analysis results contribute to previous literature on framing, agenda-setting, and campus carry. 6 II. COMMUNICATION LITERATURE REVIEW This section presents a review of the pertinent literature in the field of communication as it relates to this study. Specifically, it discusses and compares the theories of agenda- setting, second-level agenda setting, and framing. In addition to discussing the theoretical boundaries that distinguish these paradigms, this section examines the cognitive processes involved in these areas of communication. Agenda-Setting Theories: First-Level and Second-Level In their seminal article, McCombs and Shaw (1972) suggested that the media set the public agenda. They state: In choosing and displaying news, editors, newsroom staff, and broadcasters play an important part in shaping political reality. Readers learn not only about a given issue, but also how much importance to attach to that issue from the amount of information in a news story and its position. (McCombs & Shaw, 1972, p. 176) In other words, in the messages that they transmit to audiences, the media may not exactly tell the public what to think, but they may very well tell them what to think about. The theory of framing is related to the theory of agenda-setting and was originally put forth by Goffman (1974). Framing expands and builds upon agenda-setting research by narrowing its focus on the nature of the issue being examined rather than on the particular topic. Framing theory is premised on the idea that when the media focus their attention on certain events, to the exclusion of others, they provide those events with meaning. In essence, framing theory suggests that the way in which a narrative is presented to an audience (i.e., “the frame”) influences how that audience processes the 7 information and how it interprets the meaning of the message being conveyed (Goffman, 1974). Because the way in which news information is framed has the potential to shape and influence public perceptions, attitudes, and concerns, it may very well be that the media not only tell audiences what issues to think about (McCombs & Shaw, 1972), but also how to think about those issues (Goffman, 1974). There exists a longstanding debate about whether framing theory should be considered a subset of agenda-setting known as “second-level agenda-setting.” Overtime, as agenda-setting theory evolved and grew, the focus shifted away from analyzing what the media tells its audience to think about and, instead, narrowed toward a focus on how the media tells its audience to think about an issue. In other words, subsequent research focused on the attributes that describe and depict the agenda issues/objects and the cognitive components associated with them. These cognitive components can include descriptions of characteristic as well as the tones (e.g., positive, neutral, or negative) of these characteristics. According to McCombs, Llamas, Lopez-Escobar, & Rey (1997), issue salience corresponds with the first level of agenda setting (i.e., telling the audience what to think about), whereas attribute salience corresponds with second-level agenda setting or framing (i.e., telling the audience how to think about said issue). It is the second level of agenda-setting that considers how the agenda behind the media’s presentation of attributes affects public opinion (McCombs & Evatt, 1995). It also seems that the audience’s psychological distance from an object or issue plays an appreciable role in the media’s ability to shape the audience’s opinion on that object or issue. Research on second-level agenda setting, has shown that on some issues, when using attribute agendas in the news with which the audience had low psychological 8 distance, the media was able to advance compelling and persuasive arguments for the salience of those issues in the public agenda (Ghanem, 1997). In other words, second- level agenda-setting distinguishes itself from traditional agenda-setting because of its focus on attribute salience, rather than issue salience, as well as its consideration of the public’s attribute agenda as an important variable of analysis. While McCombs, Shaw, and Weaver (2014) have generally argued that framing theory is a subset of agenda-setting theory, and one that functions as a second-level media effect, disagreement has been voice by Schefuele (2000). In arguing the opposite view, Scheufele (2000) asserts that there are distinct theoretical boundaries that divide framing and agenda-setting, that they operate by way of different cognitive processes (accessibility as opposed to attribution), and that they correspond to distinct audience perceptions and interpretations (i.e., perceptions regarding the importance of the issue rather than the interpretation of what the news issue means). However, Weaver (2007) maintains that second-level agenda setting and framing both focus on how, rather than which, issues or objects are represented in the media; they are both concerned with the most significant and prominent aspects of the descriptions and themes of those issues or objects; and they are both focused and concerned about the ways of thinking about issues or objects, rather than which issues or objects are being thought of. Yet, Scheufele and Tewksbury (2007) argue that framing is more nuanced and developed than second-level agenda setting. In their view, while both framing and second-level agenda setting are concerned with attribute salience, framing includes a wider array of cognitive processes. As stated by Scheufele and Tewksbury (2007) “framing differs significantly from these accessibility-based models [i.e., agenda setting]. 9 It is based on the assumption that how an issue is characterized in news reports can have an influence on how it is understood by audiences” (p. 11). In other words, it is the difference between whether audiences think about an issue at all based upon how prominently it was reported in the media, as opposed to what they think about an issue because of how the media presented the topic. Despite the debate, there exists consensus that both agenda-setting and framing theory study how the media draw their audience’s attention to specific topics and, thereby, set an agenda. It is generally agreed that framing theory employs a deeper examination of agenda setting by the news media and finds that the way in which information is communicated creates a compelling frame for the issue or object presented. At its core, framing involves both selection and salience by the media (Entman, 1993). As stated by Entman (1993), “to frame is to select some aspect of a perceived reality and make them more salient in a communicating text, in such a way as to promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation for the item described” (p. 52). Stated differently, news frames highlight certain aspects of communicated information on a topic and, in so doing, elevate them in importance – or salience – making them “more noticeable, meaningful, and memorable to audiences” (Entman, 1993, p. 53). Framing Theories Typically, media frames define problems, diagnose causes, make moral judgments, suggest remedies for said problems, and make predictions of likely effects (Entman, 1993; Gamson, 1992). Framing research suggests that news frames work to indicate how audiences interpret issues and events and, therefore, frames can exert a 10 significant influence on audience beliefs, perceptions, attitudes, and behaviors (Entman, 1993; Tewksbury & Scheufele, 2009). While Ghanem (1997) argued that the media is able to advance compelling and persuasive arguments for the salience of issues in the public agenda when an audience’s psychological distance from the subject is low, others argue that news frames seem to exert the strongest influence on issues where the audience is not already well informed or cognitively engaged (Iyengar, 1991; Kahneman & Tversky, 1984; Zaller, 1992). Simply put, framing is the way in which sources of communication define and construct communicated information (Goffman, 1974). According to Goffman (1974), framing is an instinctive and inescapable component of human communication; people apply their own frames to the way in which they communicate and receive communications, and people interpret and comprehend what is happening in the world around them through these frameworks. Frame analysis contends that human beings classify their experiences, and develop beliefs, perceptions, attitudes, and behaviors according to previously established guiding frames of reference (Goffman, 1974). Of particular interest to this study is how people form their primary frameworks – the lenses and filters through which they interpret the world. According to Goffman (1974), there are two distinctions that must be recognized and acknowledged within primary frameworks: one is natural and the other is social. Both distinctions play a vital role in how individuals interpret information and apply their own experiences, to understand the wider social context of an issue. The principal difference between the two is a matter of functionality. Natural frameworks identify incidents as uncontrollable occurrences that do not attribute causation to any social forces. They are understood as 11 “undirected, unoriented, unanimated, unguided” events (Goffman, 1974, p. 22). Conversely, social frameworks view incidents as socially engineered occurrences “that incorporate the will, aim, and controlling effort” of social players, such as human beings (Goffman, 1974, p. 22). Social frameworks build upon natural frameworks and, together, they create the frames that influence and determine how information is interpreted, processed, understood, and communicated. Goffman’s (1974) guiding assumption was that human beings use such frameworks regularly, and that they do so both consciously and subconsciously. To better understand how the media frame the debate on campus carry, it is important to understand what are, arguably, the conflicting rights at the core of this issue. Understanding the conflicting rights at the core of the issue is important because laws permitting the carrying of guns on college and university campuses present existing or potential conflicts between the rights guaranteed in the First and the Second Amendments. Campus carry laws raise two important yet competing constitutional claims – the First Amendment right to free speech and academic freedom versus the Second Amendment right to bear arms for the purpose of self-defense (Lewis, 2017). These rights are considered to be incompatible in higher education, which is viewed as a special and protected environment (Lewis, 2017). It is argued that honoring Second Amendment rights and allowing “firearms in classrooms and academic buildings will have a chilling effect on academic speech in many different ways, including students’ in- class expression, professor and student engagement, professor and professor interaction, and administrator and professor relations” (Lewis, 2017, p. 2111). Consequently, not honoring campus carry laws may violate the Second Amendment right to self-defense, 12 while honoring them may violate First Amendment protections of free speech and academic freedom. The First and Second Amendments to the Constitution of the United States have both been the subject of varied interpretation and debate throughout the nation’s history, ones that persist into present day. Therefore, Chapter Three of this dissertation presents a relevant history of the First Amendment, including an overview of how the legal theories and standards addressing the freedom speech and expression, specifically seditious libel, academic freedom, and symbolic speech, have evolved over time. Chapter Four presents a relevant history of the Second Amendment, including an overview of how the legal theories and standards addressing the right to bear arms in the United States have shifted since the 1970s. Chapter Five expands upon those discussions by addressing how the campus-carry debate has forced the contentious intersection of guns and expression. 13 III. REVIEW OF LAW: THE HISTORY AND EVOLUTION OF THE FIRST AMENDMENT OF THE UNITED STATES CONSTITUTION First Amendment This chapter presents an overview of the history and evolution of the First Amendment and discusses the constitutional principles that it protects. It begins with a discussion of the early history of seditious libel and the ratification of the First Amendment, followed by a discussion of seditious libel during the early days of the republic and through the twentieth century. It is followed by a discussion of the protection of both academic freedom and symbolic speech under the First Amendment. Introduction In A Worthy Tradition: Freedom of Speech in America, Harry Kalven (1988) wrote, “American law in its pursuit of subversives has by no means always been able to avoid shifting its concern from the content of the speech to the character of the speaker” (p. 4). Kalven’s (1988) remark illustrates the government’s challenge in creating and implementing legal rules that are grounded in an objective identification of specific harms during times of great political controversy and division. Historically, separating the content of speech from the character of the speaker has been no simple task. Over time, the U.S. Supreme Court expanded the scope of political speech in the U.S. to protect both democratic dissent and disagreement (Kalven, 1988). Generally, it is impermissible for government actors to regulate speech on the basis of its content or the viewpoint of the speaker (Kalven, 1988). This distinction is relevant to the present study because campus carry laws have the potential to violate both the right to academic freedom and the right to symbolic speech. Many classroom discussions at the college and 14 university level involve controversial political topics. Students, faculty, and staff who have been asked to censor or limit their discussions on such topics as a result of their university’s compliance with campus carry laws, argue that their academic freedom has been restricted (DeBrabander, 2016; Lewis, 2017). Conversely, some who support campus carry, and who wish to openly carry firearms on campus, argue that restrictions on their ability to do so have the potential to violate their right to symbolic speech (Kopel, 2014; Klukowski, 2014; Vasek, 2014). The challenges inherent in separating the content of such speech from the character of the speaker trace their roots to pre- Revolutionary America and the law of seditious libel (Amar, 2006; Eldridge, 2005; Levy 1960, 1985; Nelson, 1959; Solomon, 2016; Stone, 2004, 2007; Chafee, 1920). Early History of Seditious Libel and the Ratification of the First Amendment Seditious libel came ashore in America as part of the English common law adopted by the colonies. Parliament made seditious libel a crime in 1275 (Levy, 1985; Nelson, 1959). At the time, it was a crime to disseminate any slanderous or false news that might cause a discord to grow between the King and his people. Seditious libel was a crime because it might have caused a future disturbance, not because it actually did. Yet, the idea of free speech begins to unfold in, and traces its legal roots to, the English Bill of Rights (1689), which provided freedom of speech and debate in Parliament without lawmakers having to answer for it in any court (Levy, 1985; Nelson, 1959). In regards to the printed word, in the 1600s, the Crown established an elaborate licensing system as a way of censoring written work deemed offensive to government, or religion, and thereby preventing publication of dissent altogether (Levy, 1985; Nelson, 1959). Sir William Blackstone (1765, p. 66), in Commentaries on the Laws of England, 15 stated that “Freedom of the press consists in laying no previous restraints upon publications, and not in freedom from censure for criminal matter when published.” In so doing, he endorsed the idea that freedom of the press meant nothing more than the right of printers to publish without prior censorship (i.e., prior restraint). While at the time of the American Revolution only one state constitution explicitly protected the freedom of speech, the founding generation used both print and speech to protest against the British. Despite undefined legal protections, and restrictive limits on both speech and press, the founding generation engaged in healthy and robust debate and protest. They engaged in every possible means of dissent available: in newspapers, pamphlets, songs, sermons, speeches, poems, plays, letters, petitions, and liberty trees (Levy, 1960, Solomon, 2016). They debated and protested as if seditious libel did not exist. They cherished the freedom of expression as a basic requirement of a self-governing society, and they saw it as their duty to assess the performance of those chosen to represent them. Their efforts worked to expand the public sphere of political speech (Habermas, 1962). They rebelled against the tyranny of laws punishing political dissent by rising up in protest and rejecting the repressive laws designed to silence them. Their actions helped to create the American concept of the freedom of expression. The broad freedom of expression and vigorous political culture enjoyed by contemporary Americans was born in the 1600s. It began with the tax protest of Puritan preacher John Wise in 1687, it continued with the imprisonment of printer and journalist John Peter Zenger for seditious libel, and the dissent against British standing armies and the imposition of British taxes such as the Stamp Act in 1765 (Levy, 1960, Solomon, 16 2016). It reached its culmination during ratification of the Constitution in the late 1780s, particularly the ratification of the First Amendment in 1791(Levy, 1960, Solomon, 2016). In the early days of the colonies, under the law of seditious libel, it was a crime to criticize the government, its policies or officials, and thereby reduce the public esteem in which they were held (Amar, 2006; Blackstone, 1987; Eldridge, 2005; Levy 1960, 1985; Nelson, 1959; Solomon, 2016; Stone, 2004, 2007; Chafee, 1920). While the law of the colonies during the founding period made it a criminal offense for citizens to publish information critical of the government, very few grand juries (juries of one’s peers) were willing to indict. Politicians desperate to see their critics prosecuted asked state general assemblies to intervene, and even they largely refused (Eldridge, 2005; Solomon, 2016). In 1768, when the Royal Governor of Massachusetts, Francis Bernard, asked his state’s general assembly to prosecute Boston Gazette publishers, Benjamin Edes and John Gill, for their criticisms against taxation without representation in Parliament, the general assembly not only refused but, quoting one of Cato’s Letters1, stated that the freedom of speech and “the liberty of the press are great bulwarks of the liberty of the people” (Eldridge, 2005; Nelson, 1959; Solomon, 2016; Chafee, 1920). It followed that, as guardians of the people, it was the general assembly’s incumbent duty to defend and maintain those rights. In fact, the First Amendment guarantees of freedom of expression came into being amid raucous speech throughout the country; the ratification process involved with the Constitution and the Bill of Rights is still the most extensive political debate in the 1 Cato’s Letters were a series of essays, written anonymously by two English pamphleteers, using the pseudonym Cato. The central argument of the letters was that criticism of government was not possible without the freedom of the press. The letters are considered to be some of the most politically influential essays of the 18th century. 17 nation’s history. The First Amendment, as penned by James Madison, reads: “Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press; or the right of the people peaceably to assemble, and to petition the Government for a redress of grievances” (U.S. Const. am. 1, 1791). Interestingly, the First Amendment makes no mention of seditious libel. Seditious Libel During the Early Days of the Republic During the Revolution, Founding Fathers John Adams and Thomas Jefferson, were arduous critics of seditious libel laws and any restraints on speech or of the press; however, they adjusted course once elected president (Adams, 2011; Solomon, 2016). During Adams’ administration, in an effort to strengthen national security in an undeclared, quasi-war with France, the Federalist Congress passed the Alien and Sedition Acts of 1798 (merely seven years after the ratification of the First Amendment). The Alien and Sedition Acts allowed the administration to imprison and deport non-citizens whom they found dangerous and hostile, and to prosecute and jail many of the president’s critics and political opponents. Americans repudiated the practice and went so far as to reelect the jailed political opponents while they were still in their prison cells (Solomon, 2016). The Acts expired on Adams’ last day in office. When Jefferson succeeded Adams, he pardoned those who were convicted under the Acts and Congress repaid their fines. Jefferson condemned Adams’ actions but, once he became president, he was quickly tempted to punish his own critics (Solomon, 2016). While Madison believed that all seditious libel actions were destructive to a self- governing society, Jefferson understood the First Amendment to mean that Congress (i.e., 18 the federal government) could make no laws abridging the freedoms of speech and the press, but that the individual states were free to do so (Hunt, 1900). Jefferson was also adamant that he did not want to protect the publication of false information (Hunt, 1900). Prosecutions brought after the Alien and Sedition Acts were done so under state law. If a state used its libel law to punish dissent, the defendant could contest the law’s validity only under the state constitution, and not under the First Amendment. That did not change until 1925 when the Supreme Court of the United States held that the freedom of speech and the freedom of the press were protected from abridgment by the states through the Due Process Clause of the Fourteenth Amendment, which had been ratified, after the Civil War, in 1868 (Gitlow v. NY, 1925). Sedition in the Twentieth Century While the efforts of the founding generation were instrumental in expanding the notions of liberty and freedom across the nation, seditious libel was not easily eliminated from the American legal system. The United States Supreme Court had little opportunity to expound on the meaning of the freedom of speech in the nineteenth century (Rabban, 1997). However, in the twentieth century, the Court did address seditious behavior on several occasions, including in the cases of Patterson v. Colorado (1907), Schenck v. U.S. (1919), Abrams v. U.S. (1919), Whitney v. California (1927), Dennis v. U.S. (1951), New York Times Co. v. Sullivan (1964), and Brandenburg v. Ohio (1969). In 1907, in the case of Patterson v. Colorado, the Court considered the legality of contempt charges that had been brought against Patterson, a newspaper publisher and a former U.S. Senator, for publishing material that accused Colorado judges of being biased and for upending election results. Patterson argued that truth should be a defense. 19 In using the “bad tendency” test, the Court upheld the charges. The bad tendency test, which was rooted in English common law, allowed for speakers to be punished if their speech had a propensity to harm the welfare of the public. Justice Oliver Wendell Holmes declared that the First Amendment was to be interpreted as meaning nothing more than the limited, old-English view articulated by Blackstone (Patterson v. Colorado, 1907). Holmes wrote that the objective of the First Amendment was to guard against prior restraints, not to prevent punishment for publications that may harm the public welfare (Patterson v. Colorado, 1907). During WWI, Congress enacted espionage and sedition laws, which allowed for the prosecution of those who opposed the war, as well as other assorted radicals. Before 1919, the bad tendency test, as formulated in Patterson v. Colorado (1907), was used to decide whether speech could be criminalized. The bad tendency test was seemingly overturned in Schenck v. U.S. (1919), when the stricter principle of “clear and present danger” was introduced. Defendants Charles Schenck and Elizabeth Baer were charged with the crime of violating the Espionage Act of 1917 because they obstructed the recruitment of men who were eligible to enlist in World War I by distributing flyers that urged resistance (Schenck, 1919). A unanimous Court, in an opinion by Justice Holmes, concluded that the defendants could be punished because they intended to commit a crime and their leafleting posed a clear and present danger of succeeding with that crime. In Schenck (1919), Justice Holmes famously likened Schenck’s actions to falsely shouting “fire” in a theater, which would also be unprotected under the First Amendment. Numerous people were convicted under the standard set by Holmes’ clear and present 20 danger test for merely protesting government policies, but who were accused of provoking illegal action (Schenck, 1919). Schenck’s (1919) clear and present danger test was eventually replaced by the stricter “imminent lawless action” test as articulated in Brandenburg v. Ohio (1969). In the meantime, the Court went back to using the bad tendency test when deciding First Amendment cases. At the start of the Court’s next term, only eight months after the Schenck (1919) decision, the Court again used the bad tendency test in Abrams v. U.S. (1919). In Abrams, the Court upheld the conviction of Russian immigrants who published and distributed leaflets calling for a general strike in ammunition plants in protest of the U.S. sending troops to Russia, and for otherwise advocating revolutionary, anarchist, and socialist/communist views. Again, the question before the Court was whether the application of the Espionage Act in this way infringed upon the rights protected by the free speech clause of the First Amendment. The majority of the Court ruled that it did not and emphasized that protections on speech are lower during wartime (Stone, 2004, 2007). This time, however, Justice Holmes, joined by Justice Louis Brandies, dissented arguing that the First Amendment protects the right to disagree with the government’s objectives. Holmes reverted back to the clear and present danger test that he articulated in Schenck (1919) and said that he failed to see a clear and present danger of immediate evil or that the defendants intended to create such a danger. In the absence of such danger, he now believed that speech should be protected (Abrams, Holmes dissenting, 1919). In his dissent in Abrams (1919), Holmes moved toward a more expansive concept of freedom of speech, closer to the one articulated by James Madison more than a century 21 earlier. Holmes asserted that Congress cannot prohibit all efforts to influence the people. He further argued that dissent should not be suppressed because the ultimate good is found in a marketplace of ideas where the inherent competition of the market assures that the best ideas win. According to Holmes, even ideas that shake the foundation of the country should be protected unless they come extremely close to inciting immediate illegal action (Abrams, 1919). The case of Whitney v. California (1927) marked a decision by the Court where mere political speech was deemed unprotected. In this case, Charlotte Whitney was convicted for simply being associated with the Communist Party, in violation of the California Criminal Syndicalism Act of 1919. In an unanimous decision, the Court held that although she did not commit a crime, per se, Whitney’s relationship with the Communist party represented a bad tendency and was thus unprotected (Whitney, 1927). In Dennis v. U.S. (1951), the Court upheld the conviction of eleven Communist Party leaders for petitioning for socialist reforms in violation of the Smith Act, which criminalized teaching and advocating the overthrow of the United States government. The defendants argued that the Act violated their First Amendment rights. In upholding the convictions, the Court held that there is a distinction between merely teaching communist philosophies and actively advocating those ideas (Dennis, 1951). In the Dennis case, such advocacy had the potential to threaten the government, and the probability of that threat being successful was not a necessary element to justify restricting the freedom of speech (Dennis, 1951). In 1964, in the case of New York Times Co. v. Sullivan – 173 years after the ratification of the First Amendment – the Court finally explicitly rejected seditious libel 22 in America. Sullivan, a police commissioner in Montgomery, Alabama, sued the New York Times for publishing a political advertisement that included false statements of fact about the actions of the police during a civil rights protest. Justice William J. Brennan, for a unanimous court, held that the First Amendment provides strong safeguards for the press when public officials sue for libel – regardless of whether the allegedly defamatory statements are false. Justice Brennan clarified that the First Amendment does protect false defamatory statements so long as they were made innocently (New York Times Co., 1964). However, the First Amendment does not protect intentional lies or falsehoods made with reckless disregard of whether the statements were true or not. The Court never had the opportunity to consider the constitutionality of the Alien and Sedition Acts before they expired with the Adams administration in 1801, but Brennan stated that its attack on liberty, “has carried the day in the court of history” (New York Times Co., 1964). According to Justice Brennan, the principal meaning of the First Amendment is “a profound national commitment to uninhibited, robust, and wide-open debate on public issues” (New York Times Co., 1964). Justice Brennan further stated that there is no requirement that such political debates bear any resemblance to polite language (New York Times Co., 1964). Justice Brennan’s words regarding impolite language were put to the test a few years later in the case of Brandenburg v. Ohio (1969). In the Court’s decision in Brandenburg v. Ohio (1969), Dennis (1951) was distinguished, Whitney (1927) was overruled, the bad tendency test was finally overturned, and the clear and present danger test was replaced by the “imminent lawless action” test. The imminent lawless action test is the standard currently employed to define the limits of the freedom of speech. 23 Brandenburg was a leader in the Klu Klux Klan and his speech at a Klan rally garnered him a conviction under an Ohio criminal syndicalism law. In an unanimous decision, the Court held that the Ohio criminal syndicalism law violated Brandenburg’s constitutional right to free speech. Under the new imminent lawless action test, the Court declared that speech is unprotected by the First Amendment if the speaker’s intent is to incite a violation of the law that is both imminent and likely (Brandenburg, 1969). In Brandenburg’s case, the Ohio law prohibited the teaching and advocacy of doctrines while disregarding whether that teaching and advocacy would, in actuality, incite imminent lawless action. Unlike the Dennis (1951) case, the court held that failing to distinguish between advocacy and incitement resulted in an overly broad Ohio law that was in violation of the First Amendment (Brandenburg, 1969). Over the course of the twentieth century, U.S. Supreme Court rulings regarding sedition and seditious libel evolved to coincide with the earlier ideas of English pamphleteer and historian, John Milton (1644), who emphasized the value of free speech and expression as an integral tool for discovering truth. To Milton, free speech and expression included the revolutionary idea that the publication of criticism and dissent should not be censored or punished but, rather, should be tolerated and encouraged (Chemerinsky & Gillman, 2017). According to Milton (1644), censoring or punishing dissent would have the effect of discouraging learning and undermining the ability to understand truth. This is a particularly relevant threat to academic freedom on college and university campuses. 24 Academic Freedom Concerns regarding treasonous, seditious, and subversive behavior also carried over into restrictions on the employment of teachers based upon their political views. During the McCarthy Era, which began in the 1940s and lasted through the 1950s, some employers required that employees, in particular public employees such as teachers, sign oaths and affirmations declaring that they were not past or present members of subversive groups. Others were asked to comply with subpoenas to investigate their supposed subversive behavior. Those who refused to comply were fired or were found in contempt of court. Many appealed these decisions and the cases that made their way to the U.S. Supreme Court established the concept of constitutionally protected academic freedom. These cases include Adler et al. v. Board of Education of the City of New York (1952), Wieman v. Updegraff (1952), Sweezy v. New Hampshire (1957), and Keyishian v. Board of Regents of the University of New York (1967). In 1947, when President Harry Truman issued an executive order directing for loyalty investigations into federal employees, several state legislatures followed suit and passed similar laws regarding state employees. One such law, the Feinberg Law, was a New York state statute which was intended to prevent members of subversive groups, in particular members of the Communist Party, from gaining or keeping employment in public schools. In the case of Adler et al. v. Board of Education of the City of New York (1952), Adler, a high school mathematics teacher, on behalf of the New York State Teacher’s Union, contested the law and the issue eventually reached the U.S. Supreme Court. The Court rejected the idea that the Feinberg Law violated the defendants’ freedom of speech and assembly (Adler et al., 1952). The majority of the Court ruled that 25 while individuals have “the right under our law to assemble, speak, think, and believe as they will”, the Court’s previous decision in United Public Workers v. Mitchell (1947) established that they did not have the “right to work for the state in the school system under their own terms” (Adler, 1952). The Court further held that schools were sensitive places and that the state had a right to exercise its police powers to protect them (Adler, 1952). While the majority of the Court held to uphold the validity of the Feinberg Law, Justice Hugo L. Black’s and Justice William O. Douglas’ dissents marked the first mention of academic freedom in a U.S. Supreme Court case. In his dissent, Justice Black stated that the state law was nothing more than the government trying “to mold people into a common intellectual pattern” (Adler, Black dissenting, 1952). In his view, “government should leave the mind and spirit of man absolutely free” (Adler, Black dissenting, 1952). In Justice Douglas’ dissent, joined by Justice Black, he categorically denied that the state has the power “to place its employees in the category of second-class citizens and to deny them freedom of thought and expression” (Adler, Black and Douglas dissenting, 1952). Furthermore, Douglas stated that, “the very threat of such procedure is certain to raise havoc with academic freedom…there can be no academic freedom in such environments” (Adler, Black and Douglas dissenting, 1952). Douglas’ and Black’s logic would carry the day and much of the majority of the Court’s reasoning in the Adler case would be abandoned fifteen years later in the case of Keyishian v. Board of Regents of the University of New York (1967). Before the U.S. Supreme Court considered the Keyishian case, it heard two other cases on the subject of academic freedom. The first, Wieman v. Updegraff (1952), 26 decided in the same year as the Adler case, involved a state-imposed loyalty oath that required professors in the state of Oklahoma to swear that they had never been members of subversive or communist groups. When professors at one state college refused to sign the oath, a taxpayer sued to estop the college from paying professors’ salaries. In a unanimous ruling, the Court held that Oklahoma’s loyalty-oath law violated the professors’ due process rights under the Fourteenth Amendment (Wieman, 1952). In their concurring opinion, Justices William O. Douglas and Felix Frankfurter stressed the importance of safeguarding academic freedom and the profession of teaching under the constitutional protections of the First Amendment (Wieman, Douglas and Felix dissenting, 1952). A few years later, the U.S. Supreme Court, in the case of Sweezy v. New Hampshire (1957), formally established constitutional protections for academic freedom. In the early 1950s, the New Hampshire legislature passed the Subversive Activities Act of 1951, a law that authorized the New Hampshire Attorney General to investigate subversion. Failure to cooperate with the Attorney General meant hefty fines and imprisonment for up to twenty years. Sweezy was a Marxist economist, as well a contributor to, and editor of, socialist magazines. After he conducted a guest lecture on Marxism at the University of New Hampshire, the Attorney General sought to discover Sweezy’s lecture notes and demanded that he answer questions regarding his beliefs on socialism and the advocacy of Marxism. Citing his First Amendment right to the freedom of expression, Sweezy refused to comply, and was found in contempt of court. The U.S. Supreme Court reversed Sweezy’s conviction in what would become the landmark case for academic freedom. Given the totality of the circumstances, the Court 27 found that the Attorney General’s mandate was overly broad, the government’s interest was unclear, and the investigation lacked direction (Sweezy, 1957). In his concurrence, Justice Frankfurter, formerly a professor at Harvard Law School, joined by Justice John Marshall Harlan, focused on the need to protect the principles of academic freedom from government interference (Sweezy, Frankfurter and Harlan dissenting, 1957). Specifically, they stated that, “a free society [depends] on free universities” and if universities are not protected from government intervention and interference, the consequences will be disastrous to the sacrality of the “intellectual life” of the university (Sweezy, 1957). In one of the last “Communist cases” of the 1950s and 1960s, the U.S. Supreme Court solidified the protection of academic freedom, under the First Amendment, in the case of Keyishian v. Board of Regents of the University of New York (1967). In this case, a New York state law prohibited state employees from being members of groups and organizations that were considered treasonous or seditious, or otherwise advocated for the overthrow of the United States government. The State University of New York took things one step further and required that all of its employees sign an oath that they did not belong to the Communist Party. Those who refused to sign the oath, including Keyishian, an English professor, had their employment with the university terminated. The U.S. Supreme Court held that the New York state law was too vague and overbroad, and that it infringed on the First Amendment rights of public servants, specifically teachers, to be honest and forthright to further an educational system that was as free and open as possible (Keyishian, 1967). The Court reiterated that government regulation of First Amendment rights can only be done with “narrow specificity” and that allowing such vague and overbroad laws to stand not only violated teachers’ First 28 Amendment rights, but was also detrimental to academic freedom, the life of the university, and education in general (Keyishian, 1967). Symbolic Speech In the 20th century, the United States Supreme Court extended the concept of freedom of speech far enough that it moved under a larger umbrella of protections that are often referred to as the freedom of expression. First Amendment protections were extended to include speech that was not spoken word in the strictest sense, including self- expressive creative work (Volokh, 2009). Three centuries earlier, while the founding generation defined the freedom of speech in large part through print, they also engaged in a great deal of public oration and symbolic speech. In fact, no single event in United States history gave more meaning to the freedom of expression than the ratification of the Constitution (Levy, 1960, Solomon, 2016). As legal scholar Akhil Reed Amar has noted, it was significant to the meaning of freedom of expression that “the very act of constitutional ordainment occurred in and through a regime of boisterous, virtually uncensored speech” (2012, p. 66). Even those who were opposed to ratification expressed their opposition by publicly burning copies of the Constitution (Levy, 1960, Solomon, 2016). They did so with the confidence that such symbolic speech was well within their rights of expression. Symbolic speech was an essential part of the founding generation’s dissent from Britain. Flags, liberty poles, liberty trees, and effigies all served as symbols that were full of meaning and possessed a power that written and spoken word did not. While dissenters published argumentative letters and critical newspaper articles, symbolic forms of political expression were more accessible to more people. Symbolic speech had the 29 potential to communicate complex arguments in simple and unambiguous ways. Symbolic speech had the potential to garner immediate attention, and to invite immediate responses and participation. This made symbolic speech not only popular, but also democratic. Perhaps most important to a self-governing democratic society, symbolic speech instigated more speech. In the years following the American Revolution, citizens, as well as judges, understood that the freedom of expression included symbols in addition to printed and spoken word (Volokh, 2009). Legal cases regarding First Amendment protection of symbolic speech would eventually reach the U.S. Supreme Court. These cases included, Stromberg v. California (1931), Tinker v. Des Moines Independent Community School District (1969), and Texas v. Johnson (1989). The issue of First Amendment protection of symbolic speech first came before the U.S. Supreme Court in 1931 in the case of Stromberg v. California. At the center of this cases was a 1919 California statute that prohibited the public display of red flags, which were considered a symbol of the Communist Party and a call for dissent. Stromberg, a summer camp teacher, as well as other employees at the camp were arrested for displaying a red flag. In Stromberg, the Court found the State of California’s ban on red flags to be unconstitutional and in violation of both the First and Fourteenth Amendments (Stromberg, 1931). Stromberg (1931) is considered a landmark First Amendment case because it is viewed as one of the first decisions in which the Court used the Fourteenth Amendment to incorporate a First Amendment protection (i.e., symbolic speech or expressive conduct) from infringement by a state. 30 In another landmark case, Tinker v. Des Moines Independent Community School District (1969), the U.S. Supreme Court defined the First Amendment rights of students in U.S. public schools by addressing symbolic speech. In Tinker, students were suspended for wearing black armbands in protest of the Vietnam War. The speech regulation at issue in Tinker was “based upon an urgent wish to avoid the controversy which might result from the expression, even by the silent symbol of armbands, of opposition to this Nation’s part in the conflagration in Vietnam” (Tinker, 1969). The Court stated, “It can hardly be argued that either students or teachers shed their constitutional rights to freedom of speech or expression at the schoolhouse gate” (Tinker, 1969). Ruling in favor of the students, the Court stated that the First Amendment applies to public schools and that administrators must demonstrate constitutionally valid reasons for regulating speech in classrooms (Tinker, 1969). In doing so, the Court articulated a substantial disruption test, that has become known as the Tinker Test, and which is still used by courts today to determine whether a school’s disciplinary actions violate students’ First Amendment rights. In the two-prong substantial disruption test articulated in Tinker, the Court held that for school officials to justify censoring speech, they (1) “must be able to show that [their] action was caused by something more than a mere desire to avoid the discomfort and unpleasantness that always accompany an unpopular viewpoint,” and (2) that the speech or expression of the student(s) would “materially and substantially interfere with the requirements of appropriate discipline in the operation of the school” (1969). In applying this test, the Court held that the students, by wearing armbands, did not cause a 31 disruption and that their activity represented constitutionally protected symbolic speech (Tinker, 1969). In 1989, in the case of Texas v. Johnson, the U.S. Supreme Court reiterated that the First Amendment specifically disallows the abridgment of speech and that its protection extends to symbolic speech. In this case, the defendant was charged under a Texas state law, that prohibited vandalizing or desecrating venerated objects, after he set fire to an American flag during a protest. The Court considered the question of whether the First Amendment protected non-speech acts, since Johnson was convicted of flag desecration rather than verbal communication, and, if so, whether Johnson’s burning of the flag constituted expressive conduct protected by the First Amendment (Johnson, 1989). In citing Stromberg (1931) and Tinker (1969), the Court held that, under the circumstances, Johnson’s flag burning “constituted expressive conduct, permitting him to invoke the First Amendment” and that such symbolic speech was protected under the First Amendment of the U.S. Constitution (Johnson, 1989). In Texas v. Johnson (1989) the Court rejected “the view that an apparently limitless variety of conduct can be labeled ‘speech’ whenever the person engaging in the conduct intends thereby to express an idea.” However, the Court did recognize that certain conduct can be “sufficiently imbued with elements of communication to fall within the scope of the First and Fourteenth Amendments” (Johnson, 1989). When determining whether specific conduct possesses enough elements of communication to trigger the First Amendment, the Court suggested asking if “an intent to convey a particularized message was present, and whether the likelihood was great that the message would be understood by those who viewed it” (Johnson, 1989). To clarify, the 32 standard is not just a matter of the subjective intent of the “speaker”, but also of the objective understanding of the audience/“receiver” of that message. When analyzing the audience/“receiver”, and whether the “speaker’s” message was unambiguous and apparent, a reasonable person standard is to be applied (Johnson, 1989). In Texas v. Johnson (1989), the Court determined that although, generally, the government can more easily restrict expressive conduct than written or spoken word, the government cannot prohibit conduct simply because it contains expressive elements (Johnson, 1989). Today, proponents of campus carry laws suggest that denying them the right to carry firearms on college and university campuses violates their First Amendment right to express themselves, including by way of symbolic speech. Opponents of campus carry laws also cite to the First Amendment and argue that allowing guns on campus has a silencing effect that serves as a prior restraint and limits their academic freedom. Others contend that prohibiting guns on campus is a violation of their Second Amendment right to keep and bear arms. As such, the next chapter discusses the history and evolution of the Second Amendment to the U.S. Constitution. 33 IV. REVIEW OF LAW: THE HISTORY AND EVOLUTION OF THE SECOND AMENDMENT OF THE UNITED STATES CONSTITUTION Second Amendment Whereas the U.S. Supreme Court did not explicitly reject seditious libel, or formally protect symbolic speech or academic freedom, under the First Amendment until the mid-twentieth century, it was not until the twenty-first century that the Court recognized an individual right to gun ownership under the Second Amendment. While the gradual change in the interpretation of the First Amendment was due to evolving standards, the abrupt change in the interpretation of the Second Amendment was due to a calculated redefinition of the law, which was enshrined in the 2008 U.S. Supreme Court case of District of Columbia v. Heller. What follows is a discussion of the redefinition of the Second Amendment. The discussion includes an examination of the legislative intent behind the amendment’s ratification, the plain wording of the text, the precedent that courts have set in interpreting the amendment’s meaning, and the path that lead to the U.S. Supreme Court’s current interpretation of the law. Introduction The issue of whether the Second Amendment (U.S. Const. am. 2, 1791) protected an individual right to own a gun came before the U.S. Supreme Court four times prior to the Court’s District of Columbia v. Heller decision in 2008. Unlike the Court’s vacillating on First Amendment protections, on each of those four occasions, the Court ruled that the Second Amendment did not recognize a right to individual gun ownership (U.S. v. Cruikshank, 1876; Presser v. Illinois, 1886; Miller v. Texas, 1894; U.S. v. Miller, 1939). However, by the time that the Court heard the Heller case in 2008, a strong 34 national momentum had grown for an alternative interpretation of the law. In Heller (2008), in a majority opinion written by Justice Antonin Scalia, the Court decided that the Constitution bestows a right to own a gun for self-defense in the home. In so finding, the U.S. Supreme Court overturned two centuries of precedent. The Second Amendment was adopted as part of the Bill of Rights in 1791, but only as recently as 2008 – 217 years after it was adopted – did the Court find that there exists an individual right to gun ownership (D.C. v. Heller, 2008). The Second Amendment to the U.S. Constitution is one sentence long; it states: “A well regulated militia, being necessary to the security of a free state, the right of the people to keep and bear arms, shall not be infringed.” (U.S. Const. am. 2, 1791). For the majority of the nation’s history, the Second Amendment received little notice and was largely insignificant (Waldman, 2014; Yassky, 2000). Today, it has become synonymous with American gun culture and with opposition to gun control. The amendment’s muddled wording and its odd grammar and syntax stand out from the rest of the Constitution. Two centuries later, judges, lawyers, and even the general public continue to debate its commas and clauses. During the first 200 years of the country’s existence, judges consistently concluded that the Second Amendment authorized states to form militias – what today is known as the National Guards (Efficiency in Military Act, 1903). Repeatedly, the Court and the individual states unanimously ruled that the “right to bear arms” did not extend to civilians, but only to those while serving in a well-regulated militia. It was not until 2008 that Justice Scalia, who was endorsed and well-rated by the National Rifle Association (NRA), stated that the Second Amendment protects a civilian’s right to keep a handgun 35 in their home for the purpose of self-defense (Heller, 2008; Murphy, 2014). To most of those who study the Second Amendment, this was a highly distorted interpretation of the law (Coyle, 2013; Epstein, 2008; Giffords Law Center, 2013; Konig, 2009; Kozuskanish, 2009; Posner, 2008; Siegel, 2008; Stevens, 2014; Sunstein, 2008; Waldman, 2014; Wilkinson, 2009). No such right exists; but special interest groups, such as the NRA, have been tremendously effective in persuading people to believe that it does (Finn, 2013). The consequences have been profound. To understand the history and evolution of the Second Amendment, it is important to examine the legislative intent behind its ratification, the plain wording of the text, and the precedent that courts have set in interpreting its meaning. Legislative Intent and the Ratification of the Second Amendment James Madison is credited with being the author of most of the Constitution, particularly the Bill of Rights. At the time of the Constitution’s ratification, specific amendments were proposed by only six of the thirteen states (Bogus, 2002; Breyer, 2005; Hunt, 1900; Waldman, 2014). In the end, the Constitution was ratified without the Bill of Rights and the debate concerning amendments swiftly subsided. However, had it not been for the peculiarities of one congressional contest in the first election, involving Patrick Henry’s attempt to undermine James Madison’s run for Senate in Virginia, and the debates that followed, a bill of rights might easily have fallen off of the national agenda (Hunt, 1900; Waldman, 2014). It is difficult to ascertain an individual’s motivation without understanding his or her circumstances and values. This may be as, or more, important than knowing his or her professed reasons for taking certain action. Madison did not explicitly state his 36 motivation in writing the Second Amendment. However, when it came time to ratify the Constitution, there was growing fear and trepidation that a national standing army would pose a significant threat to the security of the separate and individual states (Hunt, 1900; Spitzer, 2000; Waldman; 2014). Additionally, southern states were concerned that the constitutional powers bestowed upon the newly-formed government could be used to deprive the states of an armed militia, which they had come to rely upon to keep slaves under control. According to several legal scholars, a significant part of Madison’s motivation in writing the Second Amendment was to assure the south, generally, and his constituents in Virginia, specifically, that the slave system would not be subverted in this way (Bogus, 1998, 2002; Hadden, 2001; Waldman, 2014; Weir, 1969). In the debates between the Federalists and the Anti-Federalists during the Constitutional Convention, the issues of slavery and slave revolt were never far below the surface (Bogus, 1998, 2002; Hadden, 2001; Hunt, 1900; Schwartz, 1971; Waldman, 2014; Weir, 1969). Patrick Henry, the former governor of Virginia who is famously remembered for saying “Give me liberty, or give me death!”, saw the Constitution as tyrannical and worried that it gave Congress too much power. In his dramatic and gifted oratorical fashion, he warned that, “they will take your [slaves] from you!” – although he used far less politically-correct language (Schwartz, 1971). Likewise, another delegate of the Constitutional Convention, George Mason, worried that Congress could call up the state militias for federal purposes thereby disabling his state, Virginia. In refusing to sign the Constitution, Mason argued that such actions by the federal government should be banned without the consent of the state legislatures because they would leave southern states defenseless against slave insurrections (Schwartz, 1971). Mason worried that 37 Congress would make state militias incompetent by disarming them and by neglecting to provide for them, and that states would not be able to do so on their own because the exclusive right would belong to Congress (Schwartz, 1971). During the Constitutional Convention, Henry and Mason expressly accused Madison of putting forth a Constitution that would empower the federal government to subvert the slave system (Bogus, 1998, 2002; Hadden, 2001; Hunt, 1900; Schwartz, 1971; Waldman, 2014; Weir, 1969). In response, Madison proffered that the new Constitution would assure that the militias would be properly armed and that both Congress and the states would have the power to arm them. (Bogus, 1998, 2002; Hadden, 2001; Schwartz, 1971; Waldman, 2014; Weir, 1969). Therefore, the following year when Madison joined the First Congress, he had strong motivation to resolve the issue. The problem was two-fold: he needed to appease concerns that a national standing army would threaten the security of the separate and individual states, and he needed to assure constituents in southern states that the federal government would not undermine the slave system by depriving the states of armed militias (Waldman, 2014). From Madison’s two- fold problem, the Bill of Rights were born. The Bill of Rights, the first ten amendments to the Constitution of the United States, imposed limits on the power of the new federal government. In drafting the Second Amendment, Madison did make explicitly clear that it was modeled on the English Bill of Rights (Hunt, 1900; Schwartz, 1971). In the English Bill of Rights, keeping and bearing arms were clearly understood to have a military meaning and meant the defense of the public, not the defense of self (Blackstone, 1987; Schwoerer, 2000). 38 In late eighteenth century America, serving in the militia, and therefore owning a rifle while in the militia, were compulsory and universal for white men aged 16-60 (Hunt, 1900; Spitzer, 2000; Waldman, 2014). Additionally, the first draft of what would become the Second Amendment had a conscientious objector clause for religious sects suggesting military purpose (Hunt, 1900; Spitzer, 2000; Waldman, 2014). Quakers (pacifists who could shoot for food or to kill vermin, but not hunt for sport or wage war), who dominated the state of Pennsylvania, had pushed for the exemption (Hunt, 1900; Waldman, 2014). This clause required that conscientious objectors hire someone to take their place in their state’s militia, again suggesting military purpose (Hunt, 1900; Waldman, 2014). The second draft made clear that the purpose of a militia was to prevent and preclude the establishment of a standing army (Amar; 2006; Hunt, 1900; Spitzer, 2000; Stevens, 2014; Waldman, 2014). In the legislative record of the time, there is no mention, whatsoever, of a right to bear arms for self-defense, or for sport or hunting, or for any purpose other than service in the militia. Plain Wording of the Text According to Michael Waldman (2014), president of the Brennan Center for Justice at New York University School of Law, the eloquent men who wrote “We the People” and the First Amendment did the country no favors when they drafted the Second Amendment. One reason that the Second Amendment was largely ignored for so long is that it is difficult to decipher. While generation after generation has tried to make sense of the stuttering text, there are a few known facts to keep in mind. It is known that the debate among the Framers revolved around the survival and role of the state militias (Amar; 2006; Hunt, 1900; Spitzer, 2000; Stevens, 2014; Waldman, 2014). It is also 39 known that original drafts made clear that militia service was universal and that militiamen were to keep and bear their own firearms (Hunt, 1900; Spitzer, 2000; Waldman, 2014). While the meaning of the preambulatory clause, “A well-regulated militia being necessary”, is known, what was the significance of placing it at the beginning of the amendment? Was it a mere throat clearing before the declaration of the right (Waldman, 2014)? Did it limit the right? Or did it explain the meaning of the right? It must be noted that in the eighteenth century, “preambles” or “explanatory clauses” at the beginning of legal declarations had force: they were to be read as limiting what came after (Linguistic Research Center, 2012; Jones v. Walker, 1791; Waldman, 2014). Linguists point to the Framers’ familiarity with Latin and call the first clause an “ablative absolute” that provides the conditions under which the rest of the sentence is to be considered valid (Linguistic Research Center, 2012; D.C. v. Heller, Stevens dissenting, 2008). At a time of intense public argument about constitutional principles, the preamble aimed to communicate with ordinary Americans in their town meetings, coffee houses, and taverns. The fact is, the Second Amendment is the only one of the ten in the Bill of Rights that has an explanatory clause of any kind (U.S. Const. am. 1-10, 1791). The First Amendment does not say, “Robust debate being necessary to sound public policy, Congress shall make no law…” (Waldman, 2014, p. 61). Therefore, the preamble in the Second Amendment must be taken seriously. The Second Amendment is militia specific. One’s right to keep and bear arms is protected when, and only when, serving in the militia; remember, there was no standing military at the time of this nation’s founding. Again, the Second Amendment is the only place in the Bill of Rights, or even the greater Constitution, that has a qualifying clause – 40 “A well-regulated militia being necessary, …” (U.S. Const. am. 2, 1791). Once a standing army had been established, well-regulated militias were no longer necessary. Since the qualifying clause of the amendment is moot, the operative clause, “…the right of the people to keep and bear arms, shall not be infringed”, should be considered null and void (U.S. Const. am. 2, 1791). Second Amendment Precedent and Interpretation For more than 200 years after the ratification of the Second Amendment, federal judges unanimously and consistently ruled that the right protected by the Second Amendment was limited in two important ways. First and foremost, the right applied only to those when keeping and bearing arms for militia purposes (Stevens, 2014; Wilkinson, 2009). Second, despite the accepted understanding that the amendment could limit the power of the federal government to regulate the use or ownership of firearms, it did not impose any such limits on the power of state or local governments (Stevens, 2014; Wilkinson, 2009). In the 1800s, as the country grew and changed, it became more individualistic and less compelled by a duty to serve – especially in a militia. In the Jacksonian Era, gun violence rose sharply especially in the south and in the west (O’Meara, 1890). States began passing the first gun-control laws – focused specifically on concealable weapons (knives and pistols), not muskets (O’Meara, 1890). This was the first time in the country’s history that some Americans argued that the “right to bear arms” protected an individual’s right to own a gun. Those arguments were largely unpersuasive and, in the early 1800s, state courts generally held that “the right to keep and bear arms” referred exclusively to militias, and not to an individual right (Cornell, 2006). 41 In 1820, a Kentucky state court overturned a law that barred the carrying of concealed weapons (Bliss v. Commonwealth of Kentucky, 1820). In that case, the defendant had hidden a knife in a cane. However, it must be noted that Kentucky stood apart from other state court rulings on the subject. Contemporaneously, the supreme court in Tennessee upheld a state law that prohibited the carrying of concealed weapons (Aymette v. State, 1840), and an Arkansas state court held that the Second Amendment, and similar provisions in that state’s constitution, only protected militias (State v. Buzzard, 1842). Two models emerged, but the “Arkansas Doctrine”, limiting the constitutional right to the militia, and only the militia, became the standard interpretation (Cornell, 2006). The Civil War dramatically increased gun ownership in the country, while the militia all but vanished (O’Meara, 1890). Contemporaneously, courts ruled that the Bill of Rights only limited actions taken by the federal government and the U.S. Supreme Court reaffirmed that the Second Amendment did not create or bestow an individual right to gun ownership (United States v. Cruikshank, 1876). Ten years later, in Presser v. Illinois (1886), the U.S. Supreme Court held that the state of Illinois could decide who was in their militia and that a state could not ban guns to the degree that it interfered with the federal government’s military needs. In Miller v. Texas (1894), a criminal defendant argued that the state’s law prohibiting the carrying of weapons violated the Second Amendment. Once again, the U.S. Supreme Court disagreed. In 1903, Congress passed the Efficiency in Militia Act, which created the National Guard and allocated the federal funds necessary to train part-time soldiers. Despite these actions by Congress, the feared national standing army did not truly materialize until 42 World War I. Over time, gun ownership and gun rights evolved as the country spread west; guns were commonplace in the west, but the crowded cities of the east were a far less hospitable environment for an armed population (Waldman, 2014). With the rise of urban sprawl, farmhands and immigrants moved to the cities where the weapons of choice were knives and clubs (Asbury, 1928). In the 1800s, municipal police departments were established to impose order (Uchida, 2010). It was a tumultuous time, characterized by class and labor conflicts, as well as political violence. Presidential assassinations were the epitome of political violence; Lincoln was shot and killed in 1865, Garfield in 1881, McKinley in 1901, and his successor, Theodore Roosevelt, was shot (but survived) in 1912. By 1900, there were as many people crammed into New York City as there had been in all of America when the Constitution was ratified (Uchida, 2010). In 1911, firearm deaths in that city rose by 50 percent, prompting the medical examiner to push for stricter gun laws in the state (Czitrom, 1991; Jacobs; 2002; Kenneth & Anderson, 1975; Welch, 2008). New York’s new gun control law required a license, approved by the local sheriff, to own a firearm and made it a felony to concealed carry the weapon outside of the home (Jacobs; 2002; Kenneth & Anderson, 1975). Other states followed with similar revolver acts (Jacobs; 2002; Kenneth & Anderson, 1975). During prohibition, new weapons that were designed for use during World War I, as well as cars and paved highways, allowed gangsters to wreak havoc cross state lines. During the Great Depression, armed bank robbery and bootlegging were highly profitable ventures. It was not until the constitutional revolution of the New Deal that the national government finally took shape and assumed an active role in criminal justice policies 43 (Ackerman, 1991). In an effort to combat the growing problems associated with gangsters and guns, the National Firearms Act of 1934 – the first federal gun legislation – easily swept through, and was passed by, Congress. It levied heavy taxes on the types of guns used by gangsters; it required that sawed-off shotguns and machine guns be registered, and it prohibited them from being transported across state lines (National Firearms Act, 1934). The National Rifle Association (NRA) – then a sportsman’s club – backed the plan (Winkler, 2011). Congress’s federal crackdown on guns was immensely popular across the country. The year the law passed, Hollywood embraced a voluntary censorship code – everyone was on board: no more enormously popular gangster movies. Instead, Hollywood turned to making movies about government agents (G-men) pursuing gangsters (Surette, 2014). In 1939, in US v. Miller, the United States Supreme Court upheld the 1934 National Firearms Act – by far its most direct examination of the Second Amendment in its first two centuries. In Miller (1939), the U.S. Supreme Court unanimously held that Congress possessed the authority to prohibit and criminalize the possession of a sawed-off shotgun since that type of weapon had no legitimate or reasonable connection to the efficiency, the competence, or the maintenance of a “well- regulated Militia.” In the 1960s and the 1970s, urban violence began to climb as an influx in migration north happened to coincide with the significant decline in the availability in industrial jobs in that part of the country (Frum, 2000). The focus on guns reached its pinnacle with the back-to-back assassinations of Martin Luther King, Jr. and Robert F. Kennedy in the spring of 1968. The publicization of these political acts of violence lead 44 to the Gun Control Act of 1968, a law that provided for the first federal licensing system of gun dealers, that banned the import of military-style weapons, and which prohibited those deemed dangerous from owning guns. The NRA stayed silent (Winkler, 2011). In 1969, Richard Nixon, the “law and order” president, appointed conservative Chief Justice Warren Burger to the U.S. Supreme Court. After his tenure on the Court came to an end, Burger (1991) famously stated, “the Second Amendment has been the subject of one of the greatest pieces of fraud – I repeat fraud – on the American public by special interest groups that I have ever seen in my lifetime.” A fraud that was largely perpetrated by the NRA. In 1981, the assassination attempt against president Ronald Reagan, in which his press secretary, James Brady, was shot and permanently disabled, lead to the passage of the Brady Handgun Violence Prevention Act, colloquially known as the Brady Bill, in 1994. The Brady Bill imposed the requirements of a background check, as well as a five- day waiting period, before purchasing a firearm (Brady Handgun Violence Prevention Act, 1994). That same year, President Bill Clinton also won passage of what became known as the “Assault Weapon Ban” as part of a larger crime bill (Public Safety and Recreational Firearms Use Protection Act, 1994). The “Assault Weapon Ban” expired ten years later, in 2004, during the George W. Bush administration. In 2005, Congress passed, and President Bush signed into law, the Protection of Lawful Commerce in Arms Act (2005), a law providing broad immunity from lawsuits for gun manufacturers and sellers; they responded by shifting their resources to support the NRA (Waldman, 2014). 45 The Path to Heller Prior to 2008, the Supreme Court of the United States considered the issue of whether the Second Amendment to the United States Constitution protected an individual right to own a firearm four times and never once found a right to individual gun ownership (U.S. v. Cruikshank, 1876; Presser v. Illinois, 1886; Miller v. Texas, 1894; U.S. v. Miller, 1939). According to Waldman (2014, p. 94), “there was no more settled view in constitutional law than that the Second Amendment did not protect an individual right to own a gun.” Lower courts agreed – if they even considered cases on the subject. Alternative views were seen as frivolous and fringe positions. But, by the time that the Court heard District of Columbia v. Heller in 2008, a strong momentum had grown for the alternative view. A major factor that figured into the redefinition of the Second Amendment was a massive accumulation of scholarship that argued against the traditional interpretation of the amendment. These articles insisted that the historians, the legislators, and the courts of the past 200 years had gotten it wrong. They relied on originalism, a theory of law and constitutional interpretation that states that the only way to interpret the Constitution is to ask what the Founding generation would have thought its terms meant. The validity of originalism depends upon its approach to history and its use of historical evidence; it should be as accurate and as unbiased a review of the material available as is possible (Rakove, 2002). One political scientist examined a century’s worth of law review articles on the Second Amendment, from the time the reviews first began to be indexed in 1888 to 1999 46 (Spitzer, 2000). He found that up until 1960, every single law review article2 written on the subject of the Second Amendment determined that the amendment did not guarantee an individual right to gun ownership (Spitzer, 2000). The first author of a law review article to suggest otherwise was a student (Hays, 1960). Exponentially, other articles began to echo the alternative, revisionist view. One law professor found that 60% of these revisionist authors were directly employed by, or represented, the NRA (Bogus, 2002). Since the late 1970s, the NRA has sponsored legal seminars and has paid millions to these authors; it has offered tens of thousands of dollars in contest prizes for “best essays”, paid a lawyer $15,000 to write harsh book reviews, and even provided for a million dollar endowment for a Patrick Henry professorship in constitutional law and the Second Amendment at George Mason University (Bogus, 2002; Finn, 2013; Waldman, 2014). The writing in these articles and books is very circular in nature; the authors use the same materials and quotes, quotes are taken out of context or misquoted all together, and the authors cite to one another (Bogus, 2002; Waldman, 2014). Soon, even prominent liberal and progressive constitutional law professors and scholars, such as Akhill Reed Amar (2006) of Yale and Columbia Law Schools, began to endorse these views. However, it is important to understand that this revisionist wave came from lawyers and law professors, and not from historians. Historians, who spent their careers studying the meaning and context of the Founding Fathers’ words and actions, held to their 2 The unusual nature of law review articles should be addressed. Unlike other scholarly works, they are not peer-reviewed by subject-matter experts. They are chosen for publication, and edited, by students. Yet, courts cite them frequently and, when it comes to the interpretation of gun rights, these revisionist articles have played a crucial role. 47 longstanding view that the Framers’ concerns about the militia are what gave birth to the Second Amendment (Waldman, 2014). The 1960s also marked the rise of the judicial right. Since that time, conservative jurists, relying on originalism, have dominated federal courts. As a result, originalism has come to control the jurisprudence on firearms more than on any other major topic. All the while, and until the mid-1970s, the Republican platform supported gun control (Waldman, 2014). In 1975, as the NRA began to focus its attention on the Second Amendment, then-governor Ronald Reagan gave their rhetoric a significant charge when he stated that the Second Amendment is clear that there is no room for gun control advocacy (Winkler, 2011). When Reagan ran for president in 1980, the Republican platform proclaimed that they believed in the inherent right of every citizen to keep and bear arms and they denounced the federal registration of firearms under the National Firearms Act (Winkler, 2011). That same year, for the first time ever, the NRA bestowed candidate Reagan with an official presidential endorsement (Winkler, 2011). By 1994, when President Bill Clinton signed the Brady Bill and the “Assault Weapon Ban” into law, the NRA had shifted its focus from hunting and sport shooting, to challenging the legitimacy of the government (Waldman, 2014). The NRA’s power to elect presidents, and the judges whom they appoint, began to shift the mechanisms of government. The individual rights argument was not only advancing among scholars, lawyers, and judges, it was also winning in the forum of public opinion. Libertarian lawyers targeted the nation’s strictest gun law, one that had been passed by the local government in the District of Columbia in 1976. That law prohibited individuals from keeping loaded handguns in their homes without a trigger 48 lock. District of Columbia resident, Dick Heller, was a security guard who sought the right to be able to carry his work-issued revolver back to the high-crime neighborhood where he resided. His case reached the U.S. Supreme Court and a litany of briefs were filed, including by linguists who wrote to explain the preamble in the Second Amendment (Waldman, 2014). The arguments before the Court highlighted the high degree to which originalism had triumphed. Few points were made about current gun laws, or the toll of gun violence in modern society, or legislative history, or even precedent – all the things that prior courts had relied upon to make major decisions. Even the queries from the Justices focused heavily on colonial America and British history of the 1600s (Heller, 2008). In the Court’s 5-4 decision in Heller (2008), Scalia divided the Second Amendment into two segments, drawing a line between its prefatory clause (i.e., “A well regulated militia, being necessary to the security of a free state,…”) and its operative clause (i.e., “…the right of the people to keep and bear arms, shall not be infringed.”). Then, in his argument, he proceeded to completely skip over the prefatory clause, just as they do in their mural in the lobby of the NRA headquarters in Fairfax, Virginia (Waldman, 2014). In his opinion, Scalia stated that “bear” means to carry and that “state” means government generally, rather than what it means in every other part of the Constitution (e.g., referring to individual states such as Virginia, etc.) (Heller, 2008). Scalia asserted that, for colonists, the right to have guns was fundamental; it was a natural right (Heller, 2008). According to Scalia, the Second Amendment did not create a new right; rather, it acknowledged an existing one (Heller, 2008). 49 Scalia’s opinion did grapple with the fact that the Second Amendment was drafted largely as a response and reaction to the fear voiced by many Founders that state militias would be disarmed by the central government. However, the opinion did not address the connection between that fear and the decision to respond with an amendment. Instead, the opinion thumbs through the state constitutions of the late 1700s, only one of which explicitly protected arms for self-defense at the time that the Second Amendment was written and ratified (Heller, 2008). In his opinion, Scalia ignored the prefatory clause of the Second Amendment entirely and applied a literal interpretation of the operative clause. The result was a U.S. Supreme Court ruling to overturn two-hundred years of precedent on the Second Amendment without acknowledging that it had done so or distinguishing itself from prior cases. By engaging in a highly-selective reading of historical texts, Scalia seemed to be arguing out of both sides of his mouth; at one point acknowledging that “like most rights, the right secured by the Second Amendment is not unlimited”… and then concluding that the Second Amendment “elevates above all other interests the right of law-abiding, reasonable citizens to use arms in defense of hearth and home” (Heller, 2008). In so doing, Scalia, steered away from the originalism that he had long championed. For 22 years prior to the decision in Heller (2008), Scalia had the opportunity to consider federal gun laws in several cases. The Court’s previous reluctance to find an individual right to handgun ownership was not an accident, nor an oversight, nor the result of laziness. It reflected a judicial consensus. What changed in those 22 years was the make-up of the Court and the power and influence of the NRA. 50 Justice John Paul Steven’s dissent in the Heller (2008) decision powerfully laid out the historical record. There, the writings of Madison were considered, as well as the debates over the Second Amendment, and what it meant to the men who ratified it. Stevens quoted jurist Joseph Story, from an 1833 treatise, as focusing the attention of the Second Amendment exclusively on the militia (Heller, Stevens dissenting, 2008). Stevens argued that the militia continue to be the only protected party. According to Justice Stevens, as a matter of states’ rights, and to protect their sovereignty, the “Second Amendment was adopted to protect the right of the people of each of the several states to maintain a well-regulated militia” (Heller, Stevens dissenting, 2008). Justice Stevens reiterated, “that there is no indication that the Framers of the Second Amendment intended to enshrine the common-law right of self-defense in the Constitution”; it is simply not what the amendment was about (Heller, Stevens dissenting, 2008). He further stated that, “even if the arguments on both sides were equally balanced, respect for the well-settled views of all our predecessors on this Court, and for the rule of law itself, would prevent most jurists from endorsing such a dramatic upheaval in the law” (D.C. v. Heller, Stevens dissenting, 2008). In his impassioned dissent, Stevens made it clear that he was incensed that the majority of the Court could so flippantly disregard 200 years of precedent. In the views of the dissenting Justices in the Heller (2008) decision, the Second Amendment is about states’ rights, not individual rights (Heller, Breyer and Stevens dissenting, 2008). In their summation, the Second Amendment is a gun control amendment; after extensively studying the legislative intent, the text, as well as the precedent and the legal history of the Second Amendment, they fail to see it any other 51 way (Heller, Breyer and Stevens dissenting, 2008). In later books written on the subject, Justice Stephen Breyer and Justice Stevens voiced that the Court’s decision in D.C. v. Heller (2008) should be overturned (Breyer, 2010; Stevens, 2014). They advocated that state and federal legislators, not judges, should make decisions about what types of firearms should be made available to private citizens and under what circumstances (Breyer, 2010; Stevens, 2014). In addition, according to Justice Stevens, serious thought should be given to amending the Second Amendment to make it unambiguous and to assure that it conforms with the original intent of its drafters (Stevens, 2014). Legal analyst Jeffrey Toobin (2012) put it well when he stated that Scalia translated an eighteenth-century right to military weapons into a twenty-first century right to handguns. That supposed right to handguns finds itself at the center of the debate on campus carry. What follows is a discussion of how the campus carry debate has forced the contentious intersection of guns and expression. 52 V. THE CONTENTIOUS INTERSECTION OF GUNS AND EXPRESSION This chapter addresses how the campus-carry debate has forced the contentious intersection of guns and expression. It begins with a discussion of the undeniable power and influence of the National Rifle Association. It examines the importance of distinguishing whether carrying guns on college and university campuses falls within the scope of First Amendment protected expression. It concludes with a statement on the incompatibility of campus carry with the legal principles at the core of both the First Amendment and the Second Amendment. The Undeniable Power and Influence of the NRA A multi-decade, concerted effort by the National Rifle Association (NRA) pumped a tremendous amount of money and resources into this country’s legal and political systems to change the longstanding view that the Second Amendment protects one’s right to bear arms only while serving in a well-regulated militia. In the 1970s, the NRA began to hold seminars encouraging legal scholars to adopt the alternative view that the Second Amendment protects an individual right to own a gun. By sponsoring essay contests and awarding tens of thousands of dollars, and by endowing a million dollar Second Amendment chair at George Mason University, NRA-manipulated legal thought translated into a mountain of law review articles that urged courts to take a highly libertarian view of the Second Amendment (Finn, 2013; Waldman, 2014). Beginning with Ronald Regan in 1980, the NRA began endorsing conservative presidential candidates, who would go on to appoint conservative judges, to further change the legal interpretation on the subject (Winkler, 2011). In 2008, the conditions were ripe for Supreme Court Justice Antonin Scalia to reverse the long settled view in 53 constitutional law that the Second Amendment did not protect an individual right to own a gun (Heller, 2008). This new individual rights perspective has not only prevailed among legal scholars, lawyers, and judges, it has also come to dominate public opinion. Today’s NRA often urges the public to challenge the very legitimacy of government (Luo, 2017; Waldman, 2014). In stating that, “it’s not about guns; it’s about freedom”, the NRA, and its members, argue that it is necessary to protect the right to individual gun ownership so that, if needed, the government can be overthrown (Kopel, 2014; Luo, 2017; NRA, 2018). The NRA, many of its sympathetic legal scholars, as well as many gun owners argue that openly carrying a gun in public is a deterrent to would-be shooters, a form of defensive gun use whereby gun owners can protect themselves should violence erupt, and a form of symbolic political speech (Kopel, 2014; Klukowski, 2014; Vasek, 2014). Although openly carrying firearms on campus may be viewed as expressive or symbolic speech, legal precedent has established that laws regulating speech, which are neutral toward the content of the speech, do not violate First Amendment protections (United States v. O’Brien, 1968). In fact, if laws such as those that establish college and university campuses as gun free zones are justified by a significant government interest, and are narrowly tailored and unrelated to the suppression of speech, they are likely to be upheld (United States v. O’Brien, 1968). Therefore, arguments in favor of recognizing campus carry as a form of protected symbolic speech fail a standard reading of the precedent established in this area of constitutional law (United States v. O’Brien, 1968). 54 Distinguishing Whether Campus Carry Falls Within the Scope of First Amendment Protected Expression It is important to distinguish whether carrying guns on college and university campuses falls under the larger umbrella of protections that are referred to as the freedom of expression because, historically, it has been more difficult to limit the First Amendment than it has the Second Amendment (Horwitz, 2016). While gun-rights activists argue that one of the reasons that they carry openly displayed weapons is for the purpose of symbolic speech (e.g., to educate the public on gun rights), others argue that it is nothing more than a form of intimidation; the message being that the gun carrier is prepared to kill (Horwitz, 2016). The U.S. Supreme Court has often protected symbolic speech under the First Amendment, including displaying a red flag (Stromberg v. California, 1931), the wearing of armbands (Tinker v. Des Moines Independent Community School District, 1969), and political protests that have involved flag burning (Texas v. Johnson, 1989) (Volokh, 2009, 2016). Nevertheless, the argument that exercising the right to carry a gun is a form of protected communication does not pass legal muster. Post-Heller (2008), courts have rejected the idea that firearms have communicative connotations (Burgess v. Wallingford, 2013; Nordyke v. King, 2012), as well as the assertion that carrying a firearm is a protected form of speech (Enos v. Holder, 2012; Wortman v. U.S., 2015). The U.S. Supreme Court has held that the First Amendment does not protect speech that a reasonable audience would consider intimidating (Brandenburg, 1969). According to the Court, protecting such “speech” contradicts the First Amendment’s protections of the “freedom of speech” and the “right of the people to peacefully assemble” (U.S. Const. am. 1, 1791). If guns on campus are seen as a form of 55 intimidation, if the expression takes on an insurrectionist tone – one that argues against the legitimacy of government, or if the message is viewed as the gun carrier’s willingness to kill, a strong argument could be made that the audience understands that the speaker intends to incite a breach of the peace or a violation of the law that are both imminent and likely. In this case, the expressive speech would not be protected by the First Amendment. Furthermore, the actual language of the Second Amendment, and the U.S. Supreme Court decisions interpreting that language, unequivocally state that gun rights are limited in range and scope. Even in his decision in District of Columbia v. Heller (2008), Justice Antonin Scalia stated that the rights secured by the Second Amendment are not unlimited. Specifically, he stated that no part of the Heller opinion should question or cast doubt on the longstanding laws prohibiting the carrying of firearms in sensitive places such as schools (Heller, 2008). The U.S. Supreme Court has long viewed schools as sensitive places which should be protected (Adler, 1952). Likewise, it has viewed restrictions on teachers’ rights to expression as detrimental not only to academic freedom, but also to the life of the university and education in general (Keyishian, 1967). In the case of Sweezy v. New Hampshire (1957), the U.S. Supreme Court recognized that academic freedom, and its connection to free speech at public universities and colleges, is a prime exercise of the constitutional principles embodied in the First Amendment. Ten years later, in Keyishian v. Board of Regents of the University of New York (1967), the Court stated that there is a deep commitment in this county to safeguard academic freedom, making academic freedom a special concern of the First Amendment. 56 According to some First Amendment scholars, “given the nature of academic inquiry, only an open, robust and critical environment for speech will support the quest for truth” (Hall, 2002, para. 1). Others have argued that by including academic freedom as a right protected by the First Amendment, the Court has elevated academic freedom to the level of a constitutional freedom (Wolcott, 2017). Some go as far as to say that any limits on academic freedom are a violation of First Amendment rights (Kitrosser, 2017). In the end, despite a shared legislative history of both being written by James Madison in 1791, and having interpretations that have evolved over time, the Second Amendment right to bear arms, and the First Amendment right to free speech, are largely incompatible. This is especially true of the campus environment which cherishes open, robust, and critical debate, as well as academic freedom. Therefore, it holds that allowing for the possession of firearms on college and university campuses, known as campus carry, is incompatible with the legal principles at the core of both the First Amendment and the Second Amendment. 57 VI. CONTEXT OF RESEARCH This chapter discusses the context of the present research. It presents an overview of the history and social science research surrounding the campus carry debate. It offers an explanation of the prior research studies on campus carry as well as where the present study fills a gap in the literature. It concludes with a discussion of why the states of Oregon and Washington are populations of interest in the present study. History and Research on the Campus Carry Debate With the recent spike in shootings on college campuses across the U.S., the NRA and gun-rights activists have pushed states to eliminate gun-free zones and to allow people to conceal and/or open-carry guns on campus. Those in favor of campus-carry laws argue that the Second Amendment grants a broad entitlement on gun owners (Fennell, 2009; Kopel, 2014; Klukowski, 2014; Moody, 2014; Vasek, 2014). Proponents of campus-carry laws and policies argue that allowing people to possess firearms on public university and college campuses will deter would-be shooters and enable students, faculty, and staff to better defend themselves and others if there is a shooting on campus (Fennell, 2009; Kopel, 2014; Klukowski, 2014; Moody, 2014; Vasek, 2014). Opponents of campus-carry laws argue that permitting guns on college and university campuses will create unsafe environments (Cavanaugh, et al., 2012; DeBrabander, 2016; Lewis, 2017; Patten, et al., 2012; Smith, 2012; Wolcott, 2017). They worry that people will draw their weapons in response to merely perceived or minor threats, or when angry, or that guns might accidentally discharge, causing injury (DeBrabander, 2016). Others argue that guns in the classroom have the potential to undermine, and pose a threat to, free speech because guns may have a silencing effect 58 and discourage speech, discussion, and debate in classrooms (Cavanaugh, et al., 2012; DeBrabander, 2016; Lewis, 2017; Patten, et al., 2012; Smith, 2012; Wolcott, 2017). Therefore, they argue, it follows that campus-carry laws and policies have the potential to hinder academic freedom and to encourage de facto censorship (Barnes, 2017; Horwitz, 2016; Kitrosser, 2017; Lewis, 2017; Smith, 2012; Wolcott, 2017). Prior Research Studies on Campus Carry Prior studies have examined student and/or faculty opinions and attitudes toward private citizens carrying concealed weapons on college and university campuses (Cavanaugh et al., 2012; Patten et al., 2013). In one study, the vast majority of students at two universities expressed their extreme unease with the idea of allowing individuals to carry firearms on campus (Cavanaugh et al., 2012). In a study of more than 2,100 students, staff, and faculty across two campuses, 70% of respondents were opposed to the possibility of legally carrying concealed firearms on campus (Patten et al., 2013). Generally, the thought of allowing guns on campuses decreases the sense of safety for the majority of both students and faculty (Patten et al., 2013). Another study of nearly 800 faculty members showed that the vast majority (98%) already felt safe on their campuses and they did not support (94%) individuals carrying concealed firearms on their campuses (Thompson, Price, Dake, & Teeple, 2013). Interestingly, students were slightly more likely to approve of concealed handguns off campus than on campus, suggesting that they consider the campus environment to be unique with respect to the carrying of concealed handguns (Cavanaugh et al., 2012). 59 Gap in the Literature on Campus Carry While a considerable amount of information exists on media framing of gun violence in general, there is virtually no information on how the media frame the specific debate on campus carry or how it sets an agenda on the issue. Likewise, while prior studies have surveyed student and faculty opinions and attitudes toward carrying concealed weapons on campus, none have examined whether there is a correlation between media use and audience perceptions, on campus, regarding the issue. By addressing these gaps in the literature, it is the goal of this research to better understand how the media frame the campus carry debate, and whether they set an agenda that influences the salience of attitudes on campus toward the issue. Populations of Interest in this Study As stated earlier, research has shown that on some issues, when using attribute agendas in the news with which an audience had low psychological distance, the media was able to advance compelling and persuasive arguments for the salience of those issues in the public agenda (Ghanem, 1997). Therefore, of particular interest to the current study are the University of Oregon (UO) and Central Washington University (CWU). Oregon is a mandatory campus carry state, meaning that guns are allowed on public college and university campuses. The University of Oregon attempted to ban the carrying of guns on its campus but, in 2011, the Oregon Court of Appeals overturned the UO’s ban, stating that only the state legislature could regulate the sale, use, and possession of firearms (Cramer, 2014). The UO has reimposed the ban on students and employees through contract, however, these contractual prohibitions are not applicable to those visiting campus. According to Cramer (2014), “if an insane person walks onto a University of 60 Oregon campus with a rifle, he will not be technically outside the law (until he makes threats or opens fire)” (p. 412). However, students, faculty, or staff who defend themselves or others with a firearm are subject to expulsion or termination (Cramer, 2014). Washington is an institutional campus carry state, in which the possession of concealed firearms is prohibited at public colleges and universities, “unless prior written approval has been obtained from the university chief of police, or any other person designated by the president of the university” (WAC, 478-124-020). Washington state law does, however, allow state universities to provide for on-campus firearms storage facilities, which allows individuals to travel to and from the facilities with firearms in their possession (WAC, 106-124-700). There has been a concerted effort by some on Central Washington University’s main campus in Ellensburg, Washington, to allow for the concealed carry of firearms at all times. However, at the time that this research was conducted, the possession of concealed firearms, other than en route to and from on- campus firearms storage facilities, is prohibited at CWU. Republican lawmakers have also introduced bills in the Washington State Legislature that would allow qualified adults to carry firearms in schools. What follows are the research questions and the research methods in this study, which surveyed students, faculty, and staff from both the University of Oregon and Central Washington University. Respondents were asked questions designed to illicit demographic information, as well as perceptions and attitudes on a variety of subjects, including campus carry. This study also executed a content analysis of newspaper articles 61 on the subject of campus carry. The goal of this study was to evaluate how individual framing of campus carry compared to media framing of campus carry. 62 VII. RESEARCH QUESTIONS AND RESEARCH METHODS Introduction This study used two methods (a cross-sectional survey, plus a content-analysis of information extracted from the news media) to collect empirical data in order to evaluate respondents’ attitudes toward, and how the media frame, the debate on campus carry. Based on a comparison of the quantitative data, the goal of this research was to ascertain to what extent do the perceptions and attitudes of the survey respondents align with the framing of campus carry in the news media. In an effort to minimize social desirability bias, survey respondents were not asked outright if they support or oppose campus carry. Instead, support or opposition was inferred from survey questions that did ask them to predict their attitudes in three given situations. Based on commonly discussed concerns in the literature on campus carry, specifically, respondents were asked whether they would (1) feel more or less secure on campus, whether they would (2) change their behavior in the classroom, and whether they would (3) refrain from engaging in controversial classroom discussions if individuals were legally allowed to carry firearms on campus. These three measures thus served as the three dependent variables in this study. Research Questions and Hypothesis The following research questions and hypothesis guided this study: RQ1: Based on the analysis of the survey data, to what extent do the demographics and perceptions of the respondents predict their attitudes toward campus carry? H1: One or more demographics and/or perceptions of the respondents are significant predictors of their attitudes toward (1) Feeling Secure on Campus; (2) 63 Changing Behavior in Classroom; and (3) Refraining from Engaging in Controversial Classroom Discussions. In particular, those who support the First Amendment right to the freedom of speech are less likely to support campus carry, whereas those who support the Second Amendment right to bear arms are more likely to support campus carry. RQ2: How does a descriptive content analysis of information extracted from newspaper articles describe the following themes: (1) Feeling Secure on Campus; (2) Changing Behavior in Classroom; (3) Refraining from Engaging in Controversial Classroom Discussions (4) the First Amendment; (5) the Second Amendment; (6) Terminology of Campus Carry; (7) Discussion of Campus Carry, and (8) Debate on Campus Carry. RQ3: Based on a comparison of the survey and content analysis data, to what extent do the perceptions and attitudes of the survey respondents align with the framing of campus carry in the media? No hypotheses were tested in association with RQ2 and RQ3 because these questions were not addressed by examining the statistical relationship between dependent and independent variables. The descriptive content analysis involved an exploration of the frequencies of themes, but it did not involve hypothesis testing using inferential statistics. Exploratory studies do not require hypotheses. It is not possible to test hypotheses deductively using the results of a descriptive analysis (William & Trochim, 2020). Survey Sampled were students, faculty, and staff at two Northwest United States universities, Central Washington University (CWU) and the University of Oregon (UO). 64 Students are defined as those enrolled in either a Law and Justice (LAJ) course or a Communication (Comm) course at CWU, or those enrolled in either a School of Journalism and Communication (SOJC) or Legal Studies course at UO. Faculty are defined as deans, professors, adjuncts, or lecturers in these programs. Staff are defined as those who work in another roll in these departments, such as secretaries, assistants, counselors, advisors, and information technology associates. While these populations served as a purposive sample, mixed method sampling, of approaching some respondents during class time to take the survey, while emailing others, was used in this study. Students (n = 435, response rate of 71%) in UO’s School of Journalism and Communication (SOJC) were approached in person and asked to take the survey during class time. Students (n = 46, response rate of 18%) and faculty/staff (n = 20, response rate of 83%) in CWU’s Law and Justice Department (LAJ) as well as the students (n = 24, response rate of 35%) and faculty/staff (n = 12, response rate of 71%) in CWU’s Department of Communication (Comm) were emailed and asked to take a 10-12 minute survey titled, “Crime, Security, and Freedom.” Faculty/staff (n = 28, response rate of 25%), in UO’s School of Journalism and Communication (SOJC), as well as students (n = 44, response rate of 10%) and faculty/staff (n = 12, response rate of 15% respectively) in UO’s School of Law’s Legal Studies (LS) undergraduate program were emailed and asked to take the same survey. Those who were emailed the survey were emailed three times: the initial request, followed up by two reminder emails. The samples were collected via an online Qualtrics survey. In all, a total of 593 initial survey responses, out of a potential 1610, were collected. After accounting for missing values (i.e., absent survey response data, recorded 65 as blank cells in the SPSS data editor) and outliers (i.e., cases with unusually large or small values) 517 survey responses were used in the actual statistical analysis. This is an overall response rate of 32%. These particular samples were chosen for three reasons. First, the samples were chosen because of the political and legal prominence of the campus carry debate on both campuses. Second, the School of Journalism and Communication, and the Legal Studies program at the UO, and the departments of Law and Justice and Communication at CWU were chosen for this study because they teach, study, and conduct research on crime (in particular gun violence), the media, and constitutional law – including both the First and Second Amendments. The third reason was for convenience; the principle researcher is a professor in the CWU Law and Justice department and a student in the UO School of Journalism and Communication program. As such, the researcher had access to all of these target groups. Via survey, the following information was collected: • Age • Perceptions regarding crime and • Gender victimization rates • Marital Status • Concerns about campus violence • Whether the respondent has • Confidence in police preventing children campus crime • Ethnicity • Firearm ownership • National background • Comfort with guns • Official residence • Valid concealed handgun license • Community where currently (if answered in the affirmative, reside asked follow-up questions on • Political ideology whether they would carry if • Amount, and type, of media legally allowed) consumed • Law Enforcement Experience • Religious background • Military Experience • Level of education 66 Content Analysis of Survey Responses Nearly 90% of the initial survey respondents (n = 532) reported receiving the majority of their information on current events and crime/victimization rates from the media/internet. Therefore, a content analysis was conducted of respondents’ replies to a request to list all of the television networks, newspapers, magazines, radio stations, video games, and internet sites that they regularly engage with. Based on these responses, a list was compiled of the most commonly mentioned newspaper, television broadcast, and radio broadcast media outlets. At first, the list was limited to newspaper, television broadcast, and radio broadcast media outlets because they were among the most commonly cited sources of news and information. While many respondents cited social- networking websites, such as Facebook and Twitter, as a main source of news and information, those sites were not included in the larger agenda-setting and framing content-analysis of this study because these platforms do not generate media, they merely produce a forum where media can be shared. For the purpose of this immediate study, only the newspaper sources were coded. Since print, broadcast, and radio sources all heavily rely on the associated press (AP) for their information, coding only newspapers serves as a global measure of media output on the issue of campus carry. According to Lacy, Watson, Riffe, & Lovejoy (2015), to minimize individual subjectivity, researchers should draw from the literature and previous studies to gather search terms that offer more than face validity. Keyword searches should also have content validity; they do so by measuring different aspects and components of the same concept. Search terms should be precise in identifying and selecting relevant content, and they should be the primary focus of the found articles. Based on the literature and 67 previous studies, the search terms in this research included “guns on campus”, “campus carry”, and “concealed carry on campus”. Limiting the focus solely to newspaper articles, a cursory search of each newspaper’s online archive was conducted using the terms “guns on campus”, “campus carry”, and “concealed carry on campus” to determine which of the most commonly named newspaper sources from the campus survey discussed the issue of campus carry in the three year period between May 23, 2014 and May 23, 2017. Using May 23, 2014 as a starting point allowed for the coding of articles during the time period when the campus- carry debate reignited and received considerable media attention after the Isla Vista shooting, which occurred on the University of California—Santa Barbara campus on May 23, 2014. May 23, 2017 was used as the end point because the last of the survey responses were collected at that time. Using these terms and parameters, if the most commonly named newspapers failed to address the issue, or if they mentioned it in only one or two stories, they were excluded from the list. The final list included six newspaper sources (The Guardian, The New York Times, The Oregonian, The Seattle Times, The Washington Post, and USA Today). It should be noted that while the two newspapers local to each surveyed community, The Register Guard in Eugene, Oregon and The Daily Record, in Ellensburg, Washington were frequently mentioned in survey responses, these sources did not discuss the topics during the selected three-year period, indicating that the debate on campus carry had largely been put to rest on these two campuses. Therefore, these two local newspapers were excluded from the coding. Of the six newspapers that were included in the coding, 68 three are national papers, two are regional papers that correspond to the regions where the universities are located, and one is an international paper. Justification for Using Multivariate Statistical Methods to Test H1 The Statistical Package for the Social Sciences (SPSS v. 24.0) software was used to address the research questions and to test the associated hypotheses. This section presents two reasons why multivariate methods were used to analyze the data collected in this study. First, when a large amount of data has been collected, it is a challenge to interpret that data to obtain a realistic picture of the entire situation, and to make appropriate conclusions. Univariate and bivariate tests do not always portray a realistic picture of the entire phenomenon being studied, and may lead to misleading conclusions, because they only operate on a small amount of data at one time (Hair et al., 2010). For example, a univariate test which only looks at the effect of one independent variable on one dependent variable, ignores the confounding effects of other variables in the same data set. Confounding variables are third variables that correlate (positively or negatively) with both the dependent variable and the independent variable so that the results do not reflect a realistic relationship between two variables under study (Andrada, 2007). A bivariate test only looks at the association between two variables, but does not take into account the confounding variables that act as mediators or moderators of that association (Waliczek, 1996). Because the survey data in this research study includes a large number of demographics, as well as perceptual and media consumption variables, multivariate analysis controlled for the effects that these variables may have on the relationship between the primary independent and dependent variables (Hair et al. 2010). 69 Second, when multiple inferential statistical tests are conducted on one set of data at the conventional α = .05 level of statistical significance, there is an elevated probability of making Type I errors. A Type I error is defined as falsely assuming that the p-value indicates statistical significance, when in fact, the results may not be statistically significant due to random chance (Duffy, 2010). The probability of a Type I error when conducting one test = .05, implying a 5% chance that the statistical inference is incorrect. The probability of a Type I error when more than one test is interpreted in one study is 1- (1-.05) k where k = the number of tests (Hair et al., 2010). Because the p-values obtained from at least 30 separate inferential tests needed to be interpreted to test the stated hypotheses of this study using univariate or bivariate tests, the probability of making a Type I error = 0.785 (i.e., over three quarters of the statistical inferences may be incorrect). Accordingly, multivariate statistical tests were considered more appropriate than univariate and bivariate tests for the purpose of analyzing the large amount of data collected in this study. Multivariate statistical analysis refers to measuring the relationships between multiple variables (at least three) at one time. Multivariate techniques attempt to model reality when the effects of multiple factors on one or more dependent variables need to be evaluated. For example, feeling secure on campus may be related to numerous factors, including the perceptions of the respondents toward different social issues, the time spent with media, as well as the socio-demographic characteristics of the respondents (e.g., age, gender, educational level, and political party). Multivariate statistical analysis provides more power to identify relationships between variables, and protects against the elevation of Type I errors, when, as in this study, the researcher 70 hypothesizes that multiple independent variables are predictors of one or more dependent variables (Hair et al., 2010). Analysis of Survey Data - Testing H1 The multivariate method of data analysis applied in this study to test H1 was multiple linear regression. H1 was tested by constructing three multiple linear regression models, defined by the generalized equation: Y = b 0 + b1X1 + b 2X2 + ... + bkXk Where: Y = the dependent variable (i.e., the outcome that the researcher wants to predict); b0 = a constant or baseline value; and b1, b2...bk = the unstandardized partial regression coefficients for k independent (predictor) variables (X1, X2...Xk). An unstandardized partial regression coefficient is not the same as a bivariate correlation coefficient. Each value of b represented the amount by which the score for the dependent variable changed if the score for the independent variable changed by one unit, assuming that the effects of all the other independent variables in the same model were partialled out or held statistically constant (Rawlings et al., 2013). Table 2 lists the three dependent variables, measured in the survey using 7-point Likert scales. The lower the score (closer to 1) the more secure the respondents felt that they would feel on campus, the more the respondents agreed that they would change their behavior in the classroom, and the more they agreed that they would refrain from engaging in controversial classroom discussions if their university allowed individuals to legally carry firearms on campus. The higher the score (closer to 7) the less secure the respondents felt that they would feel on campus, the less the respondents agreed that they would change their behavior in the classroom, and the less they agreed that they would 71 refrain from engaging in controversial classroom discussions if their university allowed individuals to legally carry firearms on campus. Table 2 Dependent Variables in Multiple Linear Regression Dependent variable Survey Item: Please rate your response to the following statement: If my university allowed individuals to legally carry firearms on campus, I would: 1. Feeling Secure on 1 = Much more secure on campus Campus 2 = More secure on campus 3 = Somewhat more secure on campus 4 = Neither more secure nor less secure on campus 5 = Somewhat less secure on campus 6 = Less secure on campus 7 = Much less secure on campus 2. Changing Behavior in 1 = Strongly Agree Classroom 2 = Agree 3 = Somewhat Agree 4 = Neither Agree nor Disagree 5 = Somewhat Disagree 6 = Disagree 7 = Strongly Disagree 3. Refraining from 1 = Strongly Agree Engaging in Controversial 2 = Agree Classroom Discussions 3 = Somewhat Agree 4 = Neither Agree nor Disagree 5 = Somewhat Disagree 6 = Disagree 7 = Strongly Disagree One limitation of this study is that negative or reversed Likert scales were used throughout most of the original survey, where lower scores indicated more security and more agreement and higher levels meant less security and less agreement. This required that non-intuitive, double-negative reasoning be applied to evaluate the meaning of the coefficients. The scores for the Likert scales were not later reversed because, if they were reversed, the results of the descriptive and inferential statistical analysis would be 72 different from those obtained if the scores were retained in their original response format. For example, the factor structure of a questionnaire and the descriptive statistics for the variables are compromised when Likert scales are reversed (Suarez-Alvarez et. al., 2018). Reversing Likert scales is referred to as a “questionable practice” and therefore it was not practiced in this study (Suarez-Alvarez et al., 2018, p. 149). Table 3 lists the 11 predictor variables used in this study. Six of the predictor variables were ranked using ordinal level (5- to 11-point) scales. Also included were the socio-demographic characteristics of the respondents, assumed to be confounding variables, measured at the categorical level (age, gender, university affiliation, political party, and education) using dummy binary variables, where 1 = member of the specified category and 0 = not a member of the specified category (Rawlings et al., 2013). For over 50 years, there has been a debate in the literature regarding whether or not Likert scales are ordinal level variables (appropriate for use in non-parametric statistical analysis) or if they are interval variables (appropriate for use in parametric statistical analysis) (Agresti, 2010). Carifo and Perla (2008) have concluded that, “those who hold the ordinalist view of Likert scales rarely mention the abundant empirical findings about Likert scales…a variety of studies have shown that the Likert scale format produces empirical interval level data at the scale level” (p. 1151). Therefore, the researcher was justified to analyze the dependent variables measured with Likert scales in this study using parametric statistics. 73 Table 3 Predictor Variables in Multiple Linear Regression Predictor Survey Item Score X1 Constitutional Q49: The freedom of 1 = Strongly Agree Freedom: speech is an inalienable 2 = Agree Right of right guaranteed by the 3 = Somewhat Agree Freedom of Constitution that should 4 = Neither Agree nor Disagree Speech/First not be subject to 5 = Somewhat Disagree Amendment censorship laws or 6 = Disagree policies. 7 = Strongly Disagree X2 Constitutional Q48: To own and bear 1 = Strongly Agree Freedom: arms is an inalienable 2 = Agree Right to Bear right guaranteed by the 3 = Somewhat Agree Arms/Second Constitution that should 4 = Neither Agree nor Disagree Amendment not be subject to gun 5 = Somewhat Disagree control laws or policies. 6 = Disagree 7 = Strongly Disagree X3 Time Spent Q65: In a typical day, 1 = Less than 1 hour with Media how much time do you 2 = 1-2 hours spend engaged with 3 = 2-3 hours media and social media 4 = 3-4 hours (Newspapers, 5 = 4-5 hours Magazines, Television, 6 = 5-6 hours Internet, Radio, Video 7 = 6-7 hours Games, etc.) 8 = 7-8 hours 9 = 8-9 hours 10 = 9-10 hours 11 = More than 10 hours X4 Crime Q35: How concerned 1 = Very Concerned Concerns are you about the 2 = Concerned current crime and 3 = Somewhat Concerned victimization rate? 4 = Neither Concerned nor Unconcerned 5 = Somewhat Unconcerned 6 = Unconcerned 7 = Very Unconcerned X5 Law Q38: Do you have 1 = Definitely yes Enforcement confidence in law 2 = Probably yes enforcement preventing 3 = Might or might not and combating crime? 4 = Probably not 5 = Definitely not 74 Table 3 continued Predictor Survey Item Score X6 Comfort Level Q69: Please rate your 1 = Very Comfortable around comfort level around 2 = Comfortable Firearms guns/firearms: 3 = Somewhat Comfortable 4 = Neither Comfortable nor Comfortable 5 = Somewhat Uncomfortable 6 = Uncomfortable 7 = Very Uncomfortable X7 Gender Q4: What is your 0 = Female; 1 = Male gender? X8 Affiliation with Q18: Which of the 0 = Student University following best describes 1 = Faculty/Staff your affiliation with the University? Faculty X9 Age Q7: What is your age? 1 = 18 to 25 years 2 = 26 to 35 years 3 = 36 to 45 years 4 = 46 to 54 years 5 = ≥ 55 years X10 Political Party Q14: Which of the 1 = Republican; 0 = Other following best describes your political affiliation? X11 Education Q13: What is your 1 = Associate Degree; 0 = Other educational background? Statistical Significance and Correcting for Type I Errors The conventional method of interpreting the result of a single inferential statistical analysis (e.g., a multiple regression model) is to declare that the result is statistically significant if p < .05, implying that there is a 5% probability that the result was due to a Type 1 error. In the testing of a statistical hypothesis, a Type 1 error is the rejection of a true null hypothesis, or a “false positive” conclusion. Yet, according to the guidelines for 75 the interpretation of p-values issued by the American Statistical Association (ASA), p < .05 does not provide objective evidence to test a hypothesis (Wasserstein & Lazar, 2016). The guidelines issued by the ASA explicitly state that “p-values do not measure the probability that the studied hypothesis is true, or the probability that the data were produced by random chance alone” (Wasserstein & Lazar, 2016, p. 131). It is the ASA’s position that p-values are neither a statement about the truth of a null hypothesis, nor about the probability that the observed data was produced by random chance. The guidelines also state that “scientific conclusions and business or policy decisions should not be based only on whether a p-value passes a specific threshold” (Wasserstein & Lazar, 2016, p. 131). In other words, using bright-line rules (such as p < .05) to justify results of data analysis can lead to inaccurate inferences and, subsequently, to poor decision making. The guidelines further note that “a p-value, or statistical significance, does not measure the size of an effect or the importance of a result”, meaning that smaller p-values do not necessarily imply large or important effects, and vice versa (Wasserstein & Lazar, 2016, p. 132). Overall, the guidelines assert that “by itself, a p-value does not provide a good measure of evidence regarding a model or hypothesis”, implying that a p- value, without context, provides limited information (Wasserstein & Lazar, 2016, p. 132). It is for these reasons that the ASA dictates that data analysis should not rest solely on the calculation of a conventional p-value when other approaches are not only feasible, but may be more appropriate. The practical implication of the ASA guidelines is that all researchers in the twenty-first century must “move to a world beyond p < .05” (Wasserman, Schrim, & Lazar (2019). One reason why the traditional p < .05 criterion was not applied in the 76 current study was that three multiple regression models were constructed, using one set of survey data. Therefore, the probability of making a Type 1 error was 1 - (1 – 0.05) 3 = .143. (i.e., a 14.3% chance rather than a 5% chance). As a result, the Bonferroni correction (a post hoc test) was applied in this study to avoid Type I errors when testing H1. The Bonferroni correction is commonly regarded by social science statisticians as the simplest, and most conservative, approach for controlling the Type I error rate, and is widely applied by many researchers (Abdi, 2007; Kaur & Sttolfzfus, 2017). Therefore, in this study, the p-value to indicate statistical significance was reduced from the conventional α = .05 level to α = .05/k where k = the number of inferential tests performed on one set of data. Therefore, the decision to test H1 was to assume that a β coefficient for a predictor variable was significantly different from zero (and it was therefore a significant predictor of the dependent variable) if p < .016 rather than p < .05. As a result, each p-value was interpreted by applying a strictly dichotomous decision- making framework. Each p-value could be either significant or not significant (Hoekstra, Finch, Kiers, & Johnson, 2006; Hurlbert & Lombardi, 2009; McShane & Gal, 2017). There could be no “almost significant findings” and there could be no “borderline significance” or “marginal significance”. Sample Size The sample size is one of the most important issues in regression analysis. If the sample size is too small then there is insufficient statistical power, and Type II errors may occur. In the testing of a statistical hypothesis, a Type II error is the non-rejection of a false null hypothesis, or a “false negative” conclusion. This implies that some, if not all, 77 of the predictors will be falsely declared to be not statistically significant even though, in fact, they are statistically significant in the population from which the sample was drawn. When the sample size drawn from the same population is increased, then those predictors that were falsely declared to be not significant in the small sample will become significant (Kaur & Stoltfuz, 2017). To avoid Type II errors in multiple regression analysis, the “rule of thumb” is that there must be at least 10 to 20 cases per predictor variable (Rawlings et. al., 2013). In this study there were 11 predictor variables (see Table 3) therefore, the sample size should be at least 10 x 11 =110 to 20 x 11 = 220. The actual sample size in this survey (excluding missing values) was N = 517, which was large enough to provide ample statistical power to avoid Type II errors. Statistical power is defined as the probability that an inferential statistical test will detect an effect when there is an effect to be detected (Kaur & Stoltfuz, 2017). If statistical power is high, then there is a reduced probability of making a Type II error, or concluding there is no effect when, in fact, there is an effect (Kaur & Stoltfuz, 2017). Power analysis was conducted using G*Power v.3.1 software to determine if the sample size was adequate to achieve sufficient power to avoid Type II errors (Fauld, Erdfelder, Buchner, & Lang, 2009). The input parameters were a small effect size (R2 = 0.15); the statistical significance level, applying the Bonferroni correction (α = .016); the sample size (N = 517) and the number of independent variables (k = 11). The achieved power (0.99) was very high, therefore there was very little or no chance of making Type II errors. 78 Effect Size Many published results of scientific research are reputed to be false due to incorrect interpretation of p-values (Nuzzo, 2014; Halsey, Curran-Everett, Vowler, & Drummond, 2015). Some scholars go as far as to assert that p-values provide “fickle”, “unreliable”, and “irreproducible” evidence to test hypotheses (Nuzzo, 2014; Halsey, Curran-Everett, Vowler, & Drummond, 2015). Consequently, p-values were not the only statistics used to draw scientific conclusions or to make decisions in this study. Additional statistics, which were not related to the p-values, were also computed in order to interpret the results. One such statistic was effect size. Unlike p-values, effect sizes indicate the extent to which the results can be applied in practice, in order to make decisions and develop policies that may result in social change. (Vacha-Haase, 2001; Hill & Thompson, 2004; Konstantopoulos & Hedges, 2008; McMillan & Foley, 2011). An effect size may reflect an important finding even when a p-value is not statistically significant (Sullivan & Feinn, 2012). The advantages of effect sizes are that, unlike p- values, they do not reflect the influence of random chance, and they are not a function of the sample size. Effect sizes reflect practical significance because, unlike p-values, they estimate the strength of the relationships between the dependent and independent variables (Kirk, 1996). The effect size for each predictor variable in a multiple linear regression model was indicated by the relative magnitude of the partial regression coefficient. The overall effect size of each multiple linear regression model was indicated by the coefficient of determination (unadjusted R2), which measured the proportion of the variance in the dependent variable explained by all of the included predictor variables (k = 11 in this 79 study). However, the unadjusted R2 value was biased because for every additional predictor/independent variable added to a multiple regression model, the R2 value increases, due to random chance. Consequently, a model with many predictors appears to have a better effect size than a model with a few predictors. If, as in the current study, there are many predictors, the results are compromised by the random noise in the data. This condition is known as overfitting the model and produces a misleadingly high R2 value (Rawlings et al., 2013). For this reason, the effect size used in this study was the adjusted R2 value. This is a modified version of R2 that has been adjusted to take into account the number of predictors in the model. The adjusted R2 is always lower than the unadjusted R2, and was essential to estimate the effect size of the models in this study because they included a large number of predictor variables. The interpretation of the effect size was as follows: Adjusted R2 = .05 was assumed to be the “minimum” effect size to indicate practical significance; Adjusted R2 = .25 was assumed to reflect a “moderate” effect, and Adjusted R2 = .64 was assumed to reflect “strong” effect (Ferguson, 2009). Content Analysis of Newspaper Articles Content analysis is a method of studying documents to examine patterns of communication. The advantage of using content analysis to examine a social phenomenon is its non-invasive nature. Unlike the analysis of survey data, content analysis does not rely on asking individuals to respond to questions (Neuendorf, 2016). Content analysis was applied in this study to determine the presence or absence of latent themes extracted from 143 media sources, specifically newspaper articles. The 143 coded articles are a result of separate searches for the terms “campus carry”, “guns on campus”, 80 and “concealed weapons on campus” in The Guardian, The New York Times, The Oregonian, The Seattle Times, USA Today, and The Washington Post newspapers. The articles were searched, collected, and coded through Nexis Uni (formerly LexisNexis Academic). All of the articles in the six newspapers that discussed “campus carry”, “guns on campus”, and “concealed weapons on campus” in the proscribed three-year date range were coded, and the results serve as a proxy representation of the larger population of news articles on the subject. According to Entman (1993), “at its core, framing involves selection and salience.” As such, the coding asked questions that could answer what the media highlighted or made prominent in these newspaper articles on campus carry. As described earlier, typically, media frames do five things; they (1) define problems, (2) diagnose causes, (3) make moral judgments, (4) suggest remedies for said problems, and (5) make predictions of likely effects (Entman, 1993; Gamson, 1992). Therefore, every newspaper article (n = 143) in The Guardian, The New York Times, The Oregonian, The Seattle Times, The Washington Post, and USA Today, in the proscribed three-year date range, was coded to answer mostly nominal questions in these five categories. The five media framing categories, in the context of campus carry, were defined by the researcher, and coded with the help of an additional coder, as follows. To determine how the media “define the problem” of campus carry, coders analyzed whether the newspaper article discussed the First Amendment and in what terms (e.g., the freedom of speech; academic freedom; critical/open discourse; the marketplace of ideas). Likewise, they analyzed whether the article discussed the Second Amendment and in what terms (e.g., broadly as the right to bear arms; specifically as the right to an armed militia; as 81 self-defense; as symbolic political speech or expression). They also coded for terminology; did the article use the term “guns on campus”, “campus carry”, or “concealed weapons on campus”? To ascertain a “causal interpretation or to diagnose the cause of the problem,” coders analyzed whether the article mentioned the number of school shootings or other attacks on campus; the number of victims; the number of states that allow, prohibit, or are considering campus carry; or the names, or number, of colleges/universities that allow or prohibit campus carry. Next, to determine whether a “moral evaluation or judgement” was made, coders analyzed whether an article took a side in the debate on campus carry. Here, coders analyzed if the valence of an article was anti-campus carry, neutral, or pro- campus carry. To determine the overall valence of each article, every paragraph in each article was coded as either anti-campus carry, neutral, or pro-campus carry. Based on the prevailing tally count of paragraphs for each article, the article as a whole was given an overall designation of either anti-campus carry, neutral, or pro-campus carry. Anti-campus carry features in an article included source imbalances that disfavored the idea of allowing guns on campus, as well as the discussion of potential harm such as accidental shootings, censorship of self or others, supposed violations of First Amendment protections, students dropping out of school or switching campuses, faculty and/or staff who quit or changed jobs, reports of feeling less secure or changing one’s behavior on campus, and protests against campus carry. Pro-campus carry features in an article included source imbalances that favored the idea of allowing guns on campus, as well as the discussion of protests in favor of campus carry, feeling more safe and secure on campus, the ability to defend oneself and others, and the acknowledgement and 82 support of Second Amendment rights. Neutral features in articles on campus carry included reporting on the names and numbers of states, colleges, and universities that allow, prohibit, or are considering campus carry. Neutral features in articles on campus carry also included facts and figures on victims of gun violence and prior shootings and gun violence both on and off campus. In terms of “treatment recommendations or suggested remedies”, coders analyzed whether an article mentioned students dropping out or switching schools; faculty or administration who quit or changed jobs; a discussion of people who now carry (either legally or illegally) or would carry if their university allowed them to do so; or protests on campus. Lastly, to determine “responses and predictions of likely effects”, coders analyzed if an article mentioned whether, if individuals were legally allowed to carry firearms on campus, do, or would, they feel more or less secure on campus; do, or would, they change their behavior in the classroom; and/or do, or would, they self-censor or refrain from engaging in controversial discussions in the classroom. Space was provided at the end of the coding sheet for coders’ comments where they could note if anything was unusual or especially difficult to code in an article, or where they could collect meaningful quotes from an article. The presence of a specific theme in an article was coded by 1 = Yes. The absence of a specific theme was coded by 0 = No. Two types of themes can be identified from media sources by content analysis: either semantic or latent. (Neuendorf, 2016), Semantic themes were not identified, because they did not look beyond the narrow surface meaning of the text. In this study, eight latent themes, which captured the underlying concepts, ideas, patterns, assumptions, and implications of the text were identified, as follows: (1) 83 Feeling Secure on Campus; (2) Changing Behavior in Classroom; (3) Refraining from Engaging in Controversial Classroom Discussions (4) Discussion of First Amendment; (5) Discussion of Second Amendment; (6) Terminology of Campus Carry; (7) Discussion of Campus Carry, and (8) Debate on Campus Carry. Inter-Coder Reliability To begin coding, reliability checks should take place to test the reliability of the coding protocol (Wimmer & Dominick, 2014). Reliability is tested via intra-coder and inter-coder reliability tests. Intra-coder reliability means testing an individual coder’s consistency and accuracy over time (Wimmer & Dominick, 2014). This should be done when the coding process will extend over a long period of time. In this research study, the coding was completed during the span of one week, therefore, intra-coder reliability testing was not necessary. Inter-coder reliability means testing consistency and accuracy across multiple coders (Wimmer & Dominick, 2014). There should be more than one coder in a content analysis study, and it is required that inter-coder reliability tests take place and be reported. In this study, two coders were used, therefore inter-coder reliability testing was conducted and those results are reported in Table 4. According to Cohen’s seminal 1960 paper, the kappa coefficient (the portion of agreement corrected for chance between two coders who are assigning value to a variable) is a measure of reliability (Cohen, 1960). When using categorial items in one’s research, Cohen’s kappa coefficient is a statistical measure of inter-rater agreement. It is concerned with how a test’s reliability constrains its validity (Cohen, 1960). Scott’s pi (1955), a statistical measurement of inter-coder reliability for nominal data is similar to Cohen’s kappa because it also improves upon simple observed agreement by correcting 84 for chance. The calculations for expected agreement in Scott’s pi (1955) and Cohen’s kappa (1960) are slightly different in terms of how probability of random agreement is calculated. Cohen’s kappa (1960) is considered to be more informative than Scott’s pi (1955) because Scott’s pi assumes that raters have the same distribution of responses. While Cohen’s kappa (1960) measures agreement between only two raters, Scott’s pi is extended to more than two raters in the form of Fleiss’ kappa (1971), a multi-rater version of Scott’s pi statistic. Krippendorf (2004, 2018) argues that alpha is superior to kappa because it treats coders as independent, and that it is superior to pi because it adjusts for small sample sizes and can be used with multiple coders and all levels of data (nominal, ordinal, interval, and ratio). Inter-Coder Agreement Scores The researcher is aware of the limitations associated with content analysis, in particular the fact that the subjective judgement of the researchers’ classification of the information extracted from the media may not be reliable. For this reason, the researcher asked another coder experienced in content analysis to join in the coding of the newspaper articles in this study. After extensive coding practice with newspaper articles on campus carry, which were dated outside the three-year period analyzed in this study, and discussion of what variables did and did not work well, a coding instrument was finalized to capture 39 nominal variables (Appendix B). The researcher and the coder then used the finalized coding instrument to independently test code a random sample of 20 articles (a 13.9% subsample of the articles used in the content analysis) from the study. The inter-coder agreement analysis was conducted using Krippendorf’s alpha as the test statistic. Using Krippendorf’s alpha, 85 the values of alpha could potentially range from 0 to 1, where 0 = perfect disagreement and 1 = perfect agreement. Krippendorff (2004) suggested that alpha ≥ 0.80 indicates good agreement between coders, whilst conclusions based on the content analysis are still acceptable if alpha ≥ 0.68. Table 4 presents the results of an inter-coder agreement analysis on the 39 nominal variables, where Krippendorf’s alpha ranged from 0.70 to 1.00 reflecting acceptable inter-coder reliability. For the purpose of the inter-coder agreement analysis, the variables were not categorized into themes. Methodological Triangulation The final phase of the data analysis was methodological triangulation, referring to a comparison of the results obtained using more than one method in the study of the same phenomenon. By combining the results obtained from multiple observers, theories, methods, and/or empirical materials, the researcher may overcome the weakness or intrinsic biases and the errors that come from single-method, single-observer, and single- theory studies (Denzin, 2006). If the results of the survey data and the content analysis tend to converge, then the conclusions may be considered more credible (Rothbauer, 2008). 86 Table 4 Inter-Coder Agreement Variable alpha Newspaper Source 0.99 Type of Newspaper Article 0.89 Does the article mention or discuss the First Amendment? 0.90 -First Amendment: as Freedom of Speech? 0.80 -First Amendment: as Academic Freedom? 0.80 -First Amendment: as Critical/Open Discourse? 0.79 -First Amendment: as Marketplace of ideas? 0.77 -First Amendment: as Other? 0.87 Does the article mention or discuss the Second Amendment? 1.00 -Second Amendment: Broadly as the Right to Bear Arms? 1.00 -Second Amendment: Specifically, as the Right to an Armed Militia? 0.80 -Second Amendment: Self-Defense? 0.88 -Second Amendment: Symbolic Political Speech or Expression? 0.79 -Second Amendment: Other? 0.78 Terminology: how does the article refer to the campus carry issue? -Terminology: “Guns on Campus”? 1.00 -Terminology: “Campus Carry”? 1.00 -Terminology: “Concealed Weapons on Campus”? 0.80 -Terminology: Other? 0.78 Discussion on campus carry: does the article mention any of the following? -The number of school shootings or other attacks on campuses? 0.91 -The number of victims of gun violence either on or off campuses? 0.87 -The names, or number, of states that ALLOW campus carry? 1.00 -The names, or number, of states that PROHIBIT campus carry? 1.00 -The names, or number, of states that are CONSIDERING campus carry? 0.87 -The names, or number, of colleges/universities that ALLOW campus 0.91 carry? -The names, or number, of colleges/universities that PROHIBIT campus 0.80 carry? -The names, or number, of colleges/universities that are CONSIDERING 0.80 campus carry? Debate on campus carry: does the article mention any of the following? -Overall Valence: Anti-campus carry? Neutral? Pro-campus carry? 0.91 -Students dropping out or switching schools? 0.91 -Faculty or administration who quit or changed jobs? 1.00 -People who now carry either legally or illegally? 0.77 -People who would carry if their university allowed them to do so? 0.88 -Protests on campus against campus carry? 0.90 -Protests on campus for campus carry? 0 .70 87 Table 4 continued Variable alpha Does the article mention whether, if individuals are or were legally allowed to carry firearms on campus, do or would others: -Feel MORE secure on campus? 0.86 -Feel LESS secure on campus? 0.90 -Change their behavior in the classroom? 0.87 -NOT change their behavior in the classroom? 0.87 -Refrain from engaging in controversial classroom discussions? 0.87 -NOT refrain from engaging in controversial classroom discussions? 0 . 8 7 88 VIII. RESULTS OF SURVEY Screening and Cleaning of Survey Data Missing values (i.e., absent survey response data, recorded as blank cells in the SPSS data editor) and outliers (i.e., cases with unusually large or small values) had to be identified before analyzing the data because they could bias the statistical inferences (Tabachnik & Fidell, 2013). Table 5 presents the results of screening for missing values among N = 593 initial respondents to the survey. The number of cases with missing values was n = 53, representing 53/593 = 8.9% of the initial number of respondents. There were too many missing values to replace them with serial means or other statistical imputations (Tabachnik & Fidell, 2013). Table 5 Screening for Missing Values Survey Response Data Number of Cases with Missing Values Feeling Secure on Campus, Changing Behavior in Classroom, and 31 Refraining from Engaging in Controversial Classroom Discussions. Inalienable Right of Freedom of Speech and Right to Bear Arms 3 Time Spent on Media 1 Crime Concerns 1 Law Enforcement 1 Comfort Level with Firearms 2 Age 7 Gender 4 Political Affiliation 2 University Affiliation 1 Total 53 Multivariate outliers, which are unusual or extreme scores in at least two variables, were identified because they compromise statistical inferences of multiple regression analysis. Outliers occur for four main reasons: incorrect data entry; incorrect 89 responding (e.g. social desirability or extreme response bias); the data are not representative of the sample; and the data are outside the expected limits of a normal distribution. Outliers may be responsible for misleading statistical inferences, because they distort the computations of the sums of squares in parametric statistics (Osborne & Overbay, 2004). A total of 23 outliers were identified by p < .001 for Mahalanobis D2 (a statistic used to test for outliers in the “Save” option of the linear regression procedure in SPSS). These 23 cases were excluded because the respondents consistently provided scores at the lower ends of the 7-point Likert scales (mainly 1 or 2) or at the higher ends of the scales (mainly 6 or 7) whereas the other respondents used the full widths of the scales (from 1 to 7). Description of the Survey Respondents Table 6 summarizes the characteristics of the 517 respondents who provided a complete set of survey responses, excluding the missing values and outliers. Most of the respondents (n = 307, 59.4%) were female. Respondents ranged in age from 18 to 54 years old, but the vast majority (n = 435, 59.4%) were 18-25 years old, and their affiliation to the University was student (n = 466, 90.1%). The respondents reported a variety of political affiliations, of which the most frequent were Democrat (n = 263, 50.9%) or Republican (n = 84, 16.2%). The educational backgrounds of the respondents ranged from High School/GED to completion of a Doctoral degree, but the most frequent level was “Some College” (n = 215, 41.6%). The ethnic composition of the respondents was dominated by White/Caucasian (n = 335, 64.8%) with the remainder consisting of minority ethnic groups. 90 Table 6 Characteristics of Survey Respondents (N = 517) Characteristic Category n % Gender Female 307 59.4 Male 210 40.6 Age (Years) 18-25 435 84.1 26-35 31 6.0 36-45 22 4.3 46-54 14 2.7 55 or older 15 2.9 Affiliation with Student 466 90.1 University Faculty/Staff 51 9.9 Political Affiliation Democrat 263 50.9 Republican 84 16.2 Independent 79 15.1 Libertarian 20 3.9 Non-Political 63 12.2 Other 9 1.7 Education High School/GED 31 6.0 Some College 215 41.6 Associate Degree 64 12.4 Bachelor’s Degree 15 2.9 Master’s Degree 15 2.9 Doctoral Degree 29 5.6 Other 148 28.6 Ethnicity White/Caucasian 335 64.8 Mixed 50 9.7 Hispanic 46 8.9 Asian 36 7 African American 28 5.4 Pacific Islander 13 2.5 American Indian 6 1.2 Other (not specified) 3 0.6 Descriptive Statistics for Dependent Variables Table 7 summarizes the descriptive statistics (M = Mean; SD = Standard Deviation, Minimum, and Maximum) for the scores awarded for the three dependent 91 variables, measured in the survey using 7-point Likert scales. The respondents used the full ranges of each scale from 1 to 7. The highest mean score (M = 5.41, SD = 1.70) reflected that, on average, the respondents were somewhat more likely to disagree (n = 376, 72.7%) with Feeling Secure on Campus if individuals were legally allowed to carry firearms on campus. This meant that respondents would likely feel less secure if guns were permitted on campus. In contrast, the lower scores (M = 3.13, SD = 1.73 and M = 3.36, SD = 198, respectively) indicated that, on average, the respondents were somewhat more likely to agree (n = 308, 59.6%) with Changing their Behavior in the Classroom, and somewhat agreed (n = 302, 58.4%) with Refraining from Engaging in Controversial Classroom Discussions if individuals were legally allowed to carry firearms on campus. This meant that respondents would likely change how they currently conduct themselves, and restrict their classroom conversations, if guns were permitted on campus. Table 7 Descriptive Statistics for Dependent Variables in Survey (N = 517) Variable M SD Minimum Maximum Feeling Secure on Campus 5.41 1.70 1 7 Changing Behavior in the Classroom 3.13 1.73 1 7 Refrain from Engaging in 3.36 1.98 1 7 Controversial Classroom Discussions Descriptive Statistics for Predictor Variables Table 8 summarizes the descriptive statistics for six of the predictor variables, measured using 5- to 11-point scales. The respondents used the full ranges of each scale. The mean score for Freedom of Speech (M = 2.17, SD = 1.18) indicated that, on average, the respondents were more likely to agree (n = 455, 88.1%) with the constitutional freedom of speech endorsed by the First Amendment (as they were asked to agree or 92 disagree with the statement: “The freedom of speech is an inalienable right guaranteed by the Constitution that should not be subject to censorship laws or policies.”). The mean score for Right to Bear Arms (M = 4.05, SD = 1.88) indicated that, on average, the respondents were more likely to neither agree or disagree (n =54, 10.4%) with the constitutional freedom to bear arms, endorsed by the Second Amendment (as they were asked to agree or disagree with the statement: “To own and bear arms is an inalienable right guaranteed by the Constitution that should not be subject to gun control laws or policies.”). Table 8 Descriptive Statistics for Predictor Variables in Survey (N = 517) Variable M SD Minimum Maximum Freedom of Speech 2.17 1.18 1 7 Right to Bear Arms 4.05 1.88 1 7 Time Spent with Media 4.23 1.88 1 11 Crime Concerns 2.80 1.17 1 7 Law Enforcement 2.56 1.07 1 5 Comfort Level with Firearms 4.06 2.01 1 7 The mean score for Time Spent with Media (M = 4.23, SD = 1.88) indicated that, on average, the respondents (n = 136, 26.3%) spent 3 to 4 hours per day engaged with media and social media (Newspapers, Magazines, Television, Internet, Radio, Video Games, etc.). The mean score for Crime Concerns (M = 2.80, SD = 1.17) indicated that, on average, the respondents (n = 198, 38.3%) were likely to be somewhat concerned about the current crime and victimization rate. The mean score for Law Enforcement (M = 2.56, SD = 1.07) indicated that, on average, the respondents (n = 144, 27.9%) probably had confidence in law enforcement preventing and combating crime. The mean score for Comfort Level with Firearms (M = 4.06, SD = 2.01) indicated that, on average, the 93 respondents (n = 79, 15.3%) were likely neither comfortable nor uncomfortable around guns and firearms. Testing of Assumptions of Multiple Linear Regression Osborne & Waters (2002, p.1) defined the “assumptions of multiple regression that researchers should always test”. They stated that, ideally, the residual errors (i.e., the differences between the observed and predicted values should be normally distributed. However, the results of multiple regression are robust to violations of the normality assumption and, therefore, residual normality was not tested. Multiple linear regression analysis assumes that dependent variables have internal consistency reliability, an attribute that is commonly estimated using the Cronbach’s alpha coefficient (Tavakol & Dennikol, 2011). However, the estimation of Cronbach’s alpha requires the variable to be measured with multiple questionnaire items (usually at least three). In this study, the three dependent variables were each measured using single questionnaire items, and therefore their internal consistency reliability could not be estimated. If the relationship between the predictor and the dependent variables are not linear, the results of linear regression analysis may be biased. In this study, the method used to test for linearity was the visual examination of residual plots (i.e. scatterplots of the standardized residuals v. the standardized predicted values). Non-linearity was not indicated because the scatterplots did not display a curved pattern on either side of their mean (zero) value. Linear regression assumes homoscedasticity, meaning that the variance of the residual errors is equal across all levels of the independent variables. In contrast, 94 heteroscedasticity means that the variance of the residual errors is unequal across all levels of the independent variables. Tabachnick and Fidell (2013) suggest that a small level of heteroscedasticity does not have much effect on the statistical inferences, however, a high level of heteroscedasticity leads to a serious distortion of the results. A high level of heteroscedasticity was not indicated because the residual plots did not display a fan-shaped or bow-tie shaped pattern, suggesting that the variance in the dependent variable increased or decreased systematically with respect to the independent variables. Multiple regression analysis assumes that the predictor variables are not multicollinear (i.e., not strongly correlated with or dependent on each other). Multicollinearity causes inflation of the standard errors of the regression coefficients. Symptoms of multicollinearity include inaccurate regression coefficients, erroneous test statistics, and distorted p-values. SPSS computed the variance inflation factor (VIF) statistics to test for multicollinearity. Multicollinearity was assumed not to compromise the results of the multiple linear regression analysis because the VIF statistics were consistently < 10 (Yoo et al., 2014). Prediction of Feeling Secure on Campus using Multiple Linear Regression Table 9 presents the Multiple Linear Regression model to test the first part of H1: One or more demographics and/or perceptions of the respondents are significant predictors of their attitudes toward: (1) Feeling Secure on Campus. The statistics (copied directly from the SPSS output) are the unstandardized regression coefficients (b) measured in the same units as those used in the survey; the standard error of b, the standardized β coefficient (measured by transformation of the variables into Z scores), 95 the t-test statistic (b/SE), and the p-value to determine if β is significantly different from zero at p < .016 (applying the Bonferroni correction). Table 9 Prediction of Feeling Secure on Campus using Survey Data (N = 517) Predictor b SE β t p (Constant) 3.744 0.342 10.94 <.001* Inalienable Right of Freedom of Speech -0.094 0.05 -0.065 -1.89 .059 Right to Bear Arms 0.283 0.038 0.312 7.36 <.001* Time Spent on Media 0.058 0.03 0.064 1.92 .055 Crime Concerns -0.1 0.05 -0.069 -2.00 .046 Law Enforcement 0.074 0.055 0.047 1.35 .178 Comfort Level with Firearms 0.237 0.035 0.280 6.79 <.001* Gender -0.099 0.123 -0.029 -0.81 .419 Affiliation with University 0.485 0.45 0.085 1.08 .282 Age -0.145 0.148 -0.077 -0.98 .326 Republican -0.801 0.166 -0.174 -4.84 <.001* Associate Degree -0.598 0.174 -0.116 -3.43 <.001* Adjusted R2 45.1% Note: * Significant predictor (p < .016 using Bonferroni correction) Adjusted R2 was used rather than unadjusted R2 in order to take into account the elevation in R2 caused by having a large number of predictor variables in the model (Rawlings, et al., 2013). Adjusted R2 = .451, indicated that 45.1 % of the variance in the dependent variable was explained, reflecting a moderate to strong effect size. Support for The Right to Bear Arms, Comfort Level with Firearms, Republican political affiliation, and an educational qualification of having attained an Associate Degree were statistically significant predictors of Feeling Secure on Campus if individuals were legally allowed to carry firearms on campus. When a respondent’s score for Right to Bear Arms increased by one unit (e.g., from 1 = Strongly Agree to 2 = Agree, implying a lower level of agreement with the Second Amendment) then the score for Feeling Secure on Campus increased by b = .283 96 units (implying that the respondent reported that they would likely feel less secure). In other words, the less the respondent agreed with the Second Amendment right to bear arms, the less secure they would feel if guns were allowed on campus. When the respondent’s score for Comfort Level with Firearms increased by one unit (e.g., from 1 = Very Comfortable to 2 = Comfortable implying a lower level of comfort with guns and firearms) then the score for Feeling Secure on Campus increased by b = .237 units (implying that the respondent reported that they would likely feel less secure). In other words, the less comfortable the respondent was with guns, the less secure they would feel if guns were allowed on campus. If the political affiliation of the respondent was Republican (coded by 1) and not another party (coded by 0) then the score for Feeling Secure on Campus decreased by b = - .801 units (implying that the respondent reported that they would feel more secure). In other words, those who identified as Republicans were more likely to feel comfortable if guns were allowed on campus than those who identified with other political parties. If the respondent had an Associate Degree (coded by 1) and not another educational qualification (coded by 0) then the score for Feeling Secure on Campus decreased by b = - .598 units (implying that the respondent reported that they would feel more secure). In other words, those who had attained an Associate degree were more likely to feel secure on campus if guns were allowed than those had attained, or were working towards, other levels of education. The negative signs of the regression coefficient were a consequence of the negative direction of the Likert scales, going from high to low (i.e., 1 = the highest level of security and 7 = the lowest level of security). If the Likert scales had been measured in a positive direction, going from low to high (i.e. 1 = the lowest level of 97 security and 7 = the highest level of security), then the regression coefficient would have a positive sign. Support for the Second Amendment was found to be a statistically significant predictor of Feeling Secure on Campus if individuals were to be allowed to carry firearms on campus. Specifically, the survey results indicate that the less one supports the Second Amendment, the less likely they are to feel secure if guns are permitted on campus. Support for the First Amendment was not a predictor of feeling secure on campus if individuals were to be allowed to carry firearms on campus. Prediction of Changing Behavior in Classroom Table 10 presents the Multiple Linear Regression model to test the second part of H1: one or more demographics and/or perceptions of the respondents are significant predictors of their attitudes toward: (2) Changing Behavior in Classroom. Adjusted R2 was used rather than unadjusted R2 in order to take into account the elevation in R2 caused by having a large number of predictor variables in the model (Rawlings, et al., 2013). Adjusted R2 = .176 indicated that 17.6 % of the variance in the dependent variable was explained by all of the 11 independent variables, reflecting a less than moderate effect size (Ferguson, 2009). The effect size was not large, consistent with the effect size in most other studies that attempt to explain the perceptions of people toward social issues. The results of surveys in social science tend to have small effect sizes because human perceptions are very difficult to predict, and there is usually an inherently large amount of variance in a dependent variable reflecting human perceptions that cannot be explained using a limited number of predictor variables (Frost, 2019). 98 Table 10 Prediction of Changing Behavior in Classroom using Survey Data (N = 517) Predictor b SE β t p (Constant) 4.514 0.427 10.57 <.001* Inalienable Right of Freedom of Speech 0.122 0.062 0.083 1.97 .050 Right to Bear Arms -0.132 0.048 -0.143 -2.76 .006* Time Spent with Media -0.014 0.038 -0.015 -0.37 .715 Crime Concerns 0.084 0.062 0.056 1.34 .180 Law Enforcement -0.049 0.069 -0.030 -0.71 .479 Comfort Level with Firearms -0.203 0.044 -0.235 -4.66 <.001* Gender 0.055 0.153 0.016 0.36 .719 Affiliation with University 0.048 0.561 0.008 0.09 .932 Age -0.323 0.185 -0.169 -1.75 .081 Republican 0.242 0.207 0.052 1.17 .242 Associate Degree 0.238 0.217 0.045 1.10 .274 Adjusted R2 17.6% Note: * Significant predictor (p < .016 using Bonferroni correction) Support for The Right to Bear Arms and Comfort Level with Firearms were again found to be statistically significant predictors (p < .016) when applying the Bonferroni correction. When a respondent’s score for Right to Bear Arms increased by one unit (e.g., from 1 = Strongly Agree to 2 = Agree, implying a lower level of agreement with the Second Amendment) then the score for Changing Behavior in Classroom decreased by b = -.131 (implying that the respondents would probably be more likely to change their behavior). In other words, the less the respondent agreed with the Second Amendment right to bear arms, the more likely they would be to change their behavior in the classroom if guns were allowed on campus. The negative signs of the regression coefficient were a consequence of the negative direction of the Likert scales, going from high to low (i.e., 1 = the highest level of security and 7 = the lowest level of security). If the Likert scales had been measured in a positive direction, going from low to high (i.e. 1 99 = the lowest level of security and 7 = the highest level of security), then the regression coefficient would have a positive sign. When the respondent’s score for Comfort Level with Firearms increased by one unit (e.g., from 1 = Very Comfortable to 2 = Comfortable, implying a lower level of comfort with guns and firearms) then the score for Changing Behavior in Classroom decreased by b = -.203 (implying that the respondents would probably be more likely to change their behavior). In other words, the less comfortable the respondent was with guns, the more likely they would be to change their behavior in the classroom if guns were allowed on campus. The negative signs of the regression coefficient were a consequence of the negative direction of the Likert scales, going from high to low (i.e., 1 = the highest level of security and 7 = the lowest level of security). If the Likert scales had been measured in a positive direction, going from low to high (i.e. 1 = the lowest level of security and 7 = the highest level of security), then the regression coefficient would have a positive sign. Support for the Second Amendment was found to be a statistically significant predictor of Changing Behavior in Classroom if individuals were to be allowed to carry firearms one campus. Specifically, the survey results indicate that the less one supports the Second Amendment, the more likely they are to change their behavior in the classroom if guns are permitted on campus. Support for the First Amendment was not a predictor of changing behavior in the classroom if individuals were to be allowed to carry firearms on campus. 100 Prediction of Refraining from Engaging in Controversial Classroom Discussions Table 11 presents the Multiple Linear Regression model to test the third part of H1: One or more demographics and/or perceptions of the respondents are significant predictors of their attitudes toward: (3) Refraining from Engaging in Controversial Classroom Discussions. Adjusted R2 was used rather than unadjusted R2 in order to take into account the elevation in R2 caused by having a large number of predictor variables in the model (Rawlings, et al., 2013). Adjusted R2 = .217, indicated that 21.7 % of the variance in the dependent variable was explained, reflecting a less than moderate effect size. Table 11 Prediction of Refraining from Engaging in Controversial Classroom Discussions using Survey Data (N = 517) Predictor b SE β t p (Constant) 4.97 0.476 10.444 <.001% Inalienable Right of Freedom of Speech 0.051 0.069 0.03 0.739 0.460 Right to Bear Arms -0.139 0.053 -0.13 -2.595 0.010* Time Spent on Media -0.022 0.042 -0.02 -0.521 0.602 Crime Concerns 0.095 0.07 0.06 1.367 0.172 Law Enforcement -0.055 0.077 -0.03 -0.724 0.470 Comfort Level with Firearms -0.266 0.049 -0.27 -5.479 <.001* Gender 0.362 0.17 0.09 2.127 0.034 Affiliation with University -0.361 0.625 -0.05 -0.578 0.564 Age -0.226 0.206 -0.10 -1.099 0.272 Republican 0.357 0.23 0.07 1.550 0.122 Associate Degree 0.206 0.242 0.03 0.851 0.395 Adjusted R2 21.7% Note: * Significant predictor (p < .016 using Bonferroni correction) The Right to Bear Arms and Comfort Level with Firearms were again found to be statistically significant predictors (p < .016) applying the Bonferroni correction. When a respondent’s score for Right to Bear Arms increased by one unit (e.g., from 1 = Strongly 101 Agree to 2 = Agree, implying a lower level of agreement with the Second Amendment) then the score for Refraining from Engaging in Controversial Classroom Discussions decreased by b = -.139 (implying that the respondent would be more likely agree to refrain from engaging in classroom discussions). In other words, the less that the respondents agreed with the Second Amendment right to bear arms, the more likely they were to agree that allowing guns on campus would restrict their conversations. The negative signs of the regression coefficient were a consequence of the negative direction of the Likert scales, going from high to low (i.e., 1 = the highest level of security and 7 = the lowest level of security). If the Likert scales had been measured in a positive direction, going from low to high (i.e. 1 = the lowest level of security and 7 = the highest level of security), then the regression coefficient would have a positive sign. When the respondent’s score for Comfort Level with Firearms increased by one unit (e.g., from 1 = Very Comfortable to 2 = Comfortable, implying a lower level of comfort with guns and firearms) then the score for Refraining from Engaging in Controversial Classroom Discussions decreased by b = -.266 (implying that the respondent would be more likely agree to refrain from engaging in classroom discussions). In other words, the less comfortable the respondent was with guns, the more likely they were to agree that allowing guns on campus would restrict their conversations. The negative signs of the regression coefficient were a consequence of the negative direction of the Likert scales, going from high to low (i.e., 1 = the highest level of security and 7 = the lowest level of security). If the Likert scales had been measured in a positive direction, going from low to high (i.e. 1 = the lowest level of security and 7 = the highest level of security), then the regression coefficient would have a positive sign. 102 Support for the Second Amendment was found to be a statistically significant predictor of Refraining from Engaging in Controversial Classroom Discussions if individuals were to be allowed to carry firearms one campus. Specifically, the survey results indicate that the less one supports the Second Amendment, the more likely they are to restrict their conversations in the classroom if guns are permitted on campus. Support for the First Amendment was not a predictor of refraining from engaging in classroom discussions if individuals were to be allowed to carry firearms on campus. 103 IX. RESULTS OF CONTENT ANALYSIS Sources of Data for Content Analysis The content analysis involved collecting data from the 143 independent sources (i.e., newspaper articles) listed in Table 12. The 143 articles were published in six different newspapers. The most frequent sources of newspaper articles were The Washington Post (n = 36, 25.2%), as well as The Seattle Times and USA Today (equally n = 30, 21.6%). The remaining articles were published in The Guardian, The New York Times, and The Oregonian. In the end, the number of articles coded by each rater was split almost evenly; the researcher coded 50.35% (n = 72) of the articles, and the additional rater coded the other 49.65% (n = 71). Table 12 Frequencies of Newspaper Sources (N = 143) Newspaper n % The Washington Post 36 25.2 The Seattle Times 30 21.0 USA Today 30 21.0 The Guardian 22 15.4 The New York Times 16 11.2 The Oregonian 9 6.3 Table 13 lists the frequencies of the different types of newspaper articles included in the content analysis. The vast majority of the articles (n = 124, 86.4%) were news stories. The remaining articles included other (e.g. opinion), editorials, and letters to the editor. Interestingly, some of the professors, authors, and attorneys previously cited in this dissertation wrote articles on the subject of campus carry in The Washington Post. In particular, proponents, such as Eugene Volokh (2015; 2015; 2016) and David Kopel (2015), and opponents, such as Firmin DeBrabander (2015), wrote editorial, opinion, and 104 news story pieces on the subject of campus carry that were published during the period of review in this content analysis study. Table 13 Frequencies of Types of Newspaper Article (N = 143) Type of Newspaper Article n % News Story 124 86.7 Other (e.g., Opinion) 10 7.0 Editorial 7 4.9 Letter to Editor 2 1.4 Table 14 summarizes the frequency distributions of the articles that discussed the three themes or frames concerning whether, if individuals are, or were, legally allowed to carry firearms on campus, others would feel more or less secure on campus, change their behavior in the classroom, or refrain from engaging in controversial classroom discussions. Each article was coded as 1 = Yes or 0 = No. Nearly half of the newspaper articles (n = 68, 47.6%) discussed Feeling More (or less) Secure on Campus, whereas a little over one-third of the newspaper articles discussed Refraining from Engaging in Controversial Classroom Discussions (n = 51, 35.7%) and Changing Behavior in the Classroom (n = 50, 35.0%), respectively. Table 14 Frequencies of Newspaper Articles Regarding Feeling Secure, Changing Behavior, and Classroom Discussions (N = 143) Themes/Frames Frequency of Articles No Yes n % n % Feeling More Secure on Campus 75 52.4 68 47.6 Changing Behavior in Classroom 93 65.0 50 35.0 Refraining Controversial Classroom Discussions 92 64.3 51 35.7 During the time period under review, the Texas State legislature passed a law permitting firearms to be carried on public college and university campuses. It was 105 reported that faculty at a university-sanctioned presentation at the University of Houston were cautioned not to provoke students after the law was implemented. Specifically, it was recommended that faculty “(1) Be careful discussing sensitive topics, (2) Drop certain topics from your curriculum, (3) Don’t ‘go there’ if you sense anger, and (4) Limit student access off hours” (Samuels, 2016). Some University of Texas professors went as far as to file a lawsuit against the university claiming that allowing guns into classrooms has the potential to be dangerous, especially when discussions delve into emotionally and politically charged topics such as gay rights and abortion (Vertuno, 2016; Rosenwald, 2016). These professors argued that “compelling professors at a public university to allow, without any limitation or restriction, students to carry concealed guns in their classrooms chills their First Amendment rights to academic freedom” (AP, 2016). In the meantime, professors have told their students that certain topics of discussion are off limits and that most communication should take place via email. Beyond the physical injuries that could result from intentional or accidental discharges of a firearm, these professors saw guns on campus as posing another risk – a chilling effect on classroom discussions and debates. Table 15 summarizes the frequency distributions of the newspaper articles reporting three different positions on campus carry. The most frequent position (n = 74, 51.7%) of the articles was anti-campus carry. Relatively few articles (n = 14, 9.8%) endorsed the position of Pro-campus carry. The remainder were neutral (n = 55, 38.5%) on the topic of campus carry. One USA Today article captured the pro- and anti-campus carry sentiment aptly, when it stated that “supporters of the Texas law say that guns in the hands of students and professors make campuses safer from serial killers and lower the 106 risk of campus rape. Critics argue it creates a tense setting in what should be a peaceful learning environment” (Jervis, 2015). Overall, most articles reflected the idea that campus carry is a “loaded debate and a bad idea” (Jervis, 2015). Table 15 Frequencies of Newspaper Items Reporting Different Positions on Campus Carry (N = 143) Positions on campus carry Anti-campus carry Neutral Pro campus carry n % n % n % 74 51.7 55 38.5 14 9.8 Table 16 presents the number, and percentage, of articles that discussed the First Amendment (n = 58, 40.6%), as well as the number, and percentage, of articles that discussed the Second Amendment (n = 89, 62.2%). Table 17 presents a summary of the frequencies of the items classified within the themes/frames. Specifically, if an article mentioned or discussed either the First Amendment or the Second Amendment, what terminology did it use to do so? Table 17 also presents information on how the articles framed the Terminology of Campus Carry, the Discussion of Campus Carry, and the Debate on Campus Carry. Table 16 Frequencies of Newspaper Articles Discussing the First Amendment and the Second Amendment (N = 143) Themes/Frames Frequency of Articles No Yes n % n % First Amendment 85 59.4 58 40.6 Second Amendment 54 37.8 89 62.2 Note: The percentages of items within each theme do not add up to 100% because each article included more than one of the specified items 107 Table 17 Frequencies of Themes/Frames Identified by Content Analysis Theme/Fra Items Yes me n % First 1. Freedom of speech 51 35.7 Amendment 2. Academic freedom 43 30.1 3. Critical/open discourse 48 33.6 4. Marketplace of ideas 33 23.1 5. Other 13 9.1 Second 1.The right to bear arms 80 55.9 Amendment 2.The right to an armed militia 10 7.0 3. Self defense 46 32.2 4. Symbolic political speech 4 2.8 5. Other 16 11.2 Terminolog 1. Guns on campus 100 60.9 y of 2. Campus carry 107 74.8 Campus 3. Concealed weapons on campus 112 78.3 Carry 4. Other Discussion 1. School shootings or other attacks 79 55.2 of Campus 2. Victims of gun violence 80 55.9 Carry 3. States that allow campus carry 115 80.4 4. States that prohibit campus carry 37 25.9 5. States considering campus carry 54 37.8 6. Colleges/universities that allow campus carry 106 74.1 7. Colleges/universities that prohibit campus carry 35 24.5 8. Colleges/universities considering campus carry 48 33.6 Debate on 1. Students dropping out or switching schools 19 13.3 Campus 2. Faculty or administration quitting or changing jobs 40 28.0 Carry 3. People who carry (either legally or illegally) 31 21.7 4. People who would carry if allowed to do so 47 32.9 5. Protests against campus carry 66 46.2 6. Protests for campus carry 33 23.1 Note: The percentages of items within each theme do not add up to 100% because each article included more than one of the specified items. Fewer than half (n = 58, 40.6%) of the newspaper articles discussed the First Amendment. When they did, articles most frequently referred to the First Amendment as the “freedom of speech” (n = 51, 35.7%) followed by “critical/open discourse” (n = 48, 33.6%), and “academic freedom” (n = 43, 30.1%). The First Amendment was least likely to be discussed as “symbolic political speech” (n = 4, 2.8%). The articles much more 108 frequently discussed the Second Amendment (n = 89, 62.2%) and did so using the terms “the right to bear arms” (n = 80, 55.9%) and “self-defense” (n = 46, 32.2%) most often. The terminology used most often to discuss campus carry was “concealed weapons on campus” (n = 112, 78.3%) and “campus carry” (n = 107, 74.8%). When covering the topic of campus carry, newspaper articles most frequently included discussions about states that already allow campus carry (n = 115, 80.4%); the colleges/universities that allow campus carry (n = 106, 74.1%); victims of gun violence (n = 80, 55.9%); and school shootings (n = 79, 55.2%). When articles framed campus carry as a debate, they most frequently discussed protests against campus carry (n = 66, 46.2%) and people who would carry firearms if they were legally allowed to do so (n = 47, 32.9%). Whether the topic was communicated as “guns on campus”, “campus carry”, or “concealed weapons on campus” the newspaper articles analyzed in this study discussed the First Amendment and the Second Amendment in the broadest spectrum of terms. Some discussed prior shootings on campuses such as Virginia Tech. Of significant interest in these articles were states, such as Texas, which have passed laws allowing guns on campus. The legislature in Texas attracted a great deal of criticism and was viewed as insensitive because its new campus carry law went into effect on the same day as the fiftieth anniversary of the first U.S. school shooting, which took place on the University of Texas—Austin campus in 1966. This shooting, colloquially known as the “U.T. Tower Shooting”, is also credited as being one of the first mass shootings in the history of the United States. In the newspaper articles, faculty overwhelmingly rejected the idea of campus carry and they discussed taking new positions at other universities or retiring early. 109 Students discussed transferring to other schools in different states or studying abroad. Dramatic debates on both sides of the issue were also discussed. Those who supported campus carry were largely students. One student, at Virginia Tech, went on a hunger strike to bring attention to what he saw as his Second Amendment right to self-defense. Methodological Triangulation The main findings of the statistical analysis of the survey data and the content analysis data were compared. The results of the survey indicated that, on average, the respondents were somewhat more likely to disagree with Feeling Secure on Campus, they were somewhat more likely to Changing their Behavior in the Classroom, and they somewhat agreed with Refraining from Engaging in Controversial Classroom Discussions if individuals were legally allowed to carry firearms on campus. The results also showed that the two most important issues of concern to university students and staff/faculty, which significantly predicted their attitudes toward campus carry, were support for the right to bear arms endorsed by the Second Amendment, and the respondent’s level of comfort with guns and firearms. Affiliation to the Republican party and having attained an Associate degree were also found to be significant predictors of support for Feeling Secure on Campus if individuals were legally allowed to carry firearms on campus. Interestingly, support for the freedom of speech, as protected by the First Amendment, did not predict attitudes toward campus carry. In the content analysis, the most frequent position taken on campus carry, evidenced by over half of the newspaper articles, was an anti-campus carry stance (n = 74, 51.7%). Using sources that skewed toward an anti-campus carry stance seems to have been a normal journalistic practice when discussing the topic of campus carry, and not 110 based solely on commentary. One of the most frequent items in the content analysis reflecting themes and frames in the news media concerning campus carry was the discussion of the Second Amendment, specifically the right to bear arms and self- defense. As was true of the survey results, the First Amendment was considered less frequently in the media coverage of campus carry. Since both the survey respondents and the newspaper articles did not favor the idea of allowing guns on campus, and since they both seemed to indicate that campus carry was more significantly a Second Amendment issue than a First Amendment issue, there was, some alignment between the survey data and the content analysis data. 111 X. DISCUSSION AND CONCLUSION Discussion Media framing and second-level agenda setting ask audiences to think about an issue, and to do so in a certain way. Whether the media has been successful depends on the public’s attribute agenda. The present research study examined the framing and agenda- setting function of the media, which evaluates the significant similarity between featured or highlighted issue attributes in the news media and the agenda of attributes identified among the news media’s audience on the same topic. An opinion survey on the issue of campus carry, and the constitutional principles surrounding the debate, as well as a content analysis of regional, national and international newspapers that survey respondents stated that they access most frequently, indicated that, by covering certain issue aspects of the debate on campus carry as opposed to others, the news media may increase the salience of those aspects among its audience. Although it could not be definitively concluded that there was cause and effect evidence that the news media was directly influencing attitudes in this study, in the debate on campus carry, the news media, by emphasizing certain attributes of the issue, may have the potential to tell the public not only what to think about an issue, but also how to think about it. The results of this research do not preclude the idea that how an issue is characterized in the news media may have the potential to influence how that issue is understood, and evaluated, by audiences. Indeed, the way in which the news on campus carry is presented creates a frame for that information – one that involves both selection and salience. This research study supports other framing research, which indicates that news frames work to suggest how audiences should interpret issues and, therefore, frames may have the 112 potential to exert a influence on audience perceptions and attitudes (Entman, 1993; Tewksbury & Scheufele, 2009). The mixed-methods study in this research focused on attitudes towards campus carry, referring to the possession of firearms on college or university campuses in the United States. A cross-sectional survey of university students, faculty, and staff was administered to collect empirical data to determine the extent to which the demographics and perceptions of the respondents predicted their attitudes toward three dependent variables related to campus carry: (1) Feeling Secure on Campus; (2) Changing Behavior in Classroom; and (3) Refraining from Engaging in Controversial Classroom Discussions. The sample size was N = 517 respondents, of which the majority (n = 466, 90.1%) were students. The highest mean scores (based on a 7-point Likert scale) indicated that, on average, the respondents somewhat disagreed with Feeling Secure on Campus if individuals were legally allowed to carry firearms on campus. In contrast, lower mean scores indicated than, on average, the respondents somewhat agreed with Changing their Behavior in the Classroom as well as Refraining from Engaging in Controversial Discussions if individuals were legally allowed to carry firearms on campus. Multiple linear regression analysis was conducted to predict the three dependent variables using eleven predictor variables: support for the First Amendment and the freedom of speech; support for Second Amendment and the right to bear arms; the amount of time spent with media; concerns about crime; confidence in law enforcement being able to combat crime; comfort level with firearms; gender; affiliation with one’s university (school, position, and major); age; political party; and education. While 113 support of the right to bear arms, endorsed by the Second Amendment, was a statistically significant predictor of the dependent variables used to ascertain attitudes toward campus carry, the results indicated that support for the freedom of speech, endorsed by the First Amendment, was not a statistically significant predictor. This is due to the fact that, among survey respondents, support for the First Amendment was so overwhelming (n = 455, 88.1%) that the variable served as a constant. In fact, the First Amendment received so much support from survey respondents that it could not be a predictive variable yet, predictively, it was very important. Since higher significance does not automatically imply stronger predictivity, variables with strong predictivity sometimes fail to be significant (Lo, Chernoff, Zheng, & Lo, 2015). This may be the case with attitudes toward the First Amendment, which proved to be an overriding concern in this study on campus carry. In other words, while respondents were more likely to neither agree or disagree (n =54, 10.4%) with the survey statement “to own and bear arms is an inalienable right guaranteed by the Constitution that should not be subject to gun control laws or policies,” they were much more likely to agree (n = 455, 88.1%) that, “the freedom of speech is an inalienable right guaranteed by the Constitution that should not be subject to censorship laws or policies.” It seems that the idea, and general acceptance, of First Amendment rights is deeply entrenched in American society and this study clearly reinforces the strength of that belief. Therefore, while the survey indicated that those who support the Second Amendment were more likely to feel secure and comfortable with the idea of campus carry, it could not be said that those who support the First Amendment were less 114 likely to feel secure and comfortable with the idea of campus carry because nearly all of the respondents (over 88%) reported that they support the First Amendment. Law Professor, George P. Magarian (2012), illustrates this concept well when he says, “even to the extent one considers bearing arms desirable, speech remains more important for individuals and society: How many avid hunters would give up communication before they gave up hunting?” (p. 53). Magarian’s statement highlights the notion that one of the main goals of the U.S. Constitution was to produce and fortify a democratic society. In a democratic society, the ability to speak freely is viewed as more of a necessity than the ability to bear arms (Magarian, 2012). The results of this research support this notion and indicate that it is a sentiment that is pervasive in American society. The strongest predictors of the dependent variables, with moderate effect sizes, were support for the right to bear arms endorsed by the Second Amendment, the respondents’ level of comfort with firearms, affiliation with the Republican party, and an education level of having attained an Associate Degree. Unlike what was true of the First Amendment, support for the Second Amendment had extremely significant effects on respondents’ attitudes toward campus carry, indicating that, perhaps the mainstream news media does not have a strong effect on the attitudes and perceptions of supporters of the Second Amendment. The survey results indicated that the less one supports the right to bear arms under the Second Amendment, the less secure they may feel on campus, and the more likely they may be to change their behavior in the classroom or to refrain from engaging in classroom discussions, if their university were to allow individuals to legally carry firearms on campus. In other words, the less one supports the right to bear arms under the Second Amendment, the less likely they are to support the idea of campus carry. 115 Based on survey responses to a question asking respondents to list the newspapers that they regularly engage with, a descriptive content analysis of text extracted from 143 newspaper articles was also conducted. The following eight latent themes and frames were identified: (1) Feeling Secure on Campus; (2) Changing Behavior in Classroom; (3) Refraining from Engaging in Controversial Classroom Discussions (4) Discussion of First Amendment; (5) Discussion of Second Amendment; (6) Terminology used to describe Campus Carry; (7) Discussion of Campus Carry, and (8) Debate on Campus Carry. The results of the content analysis, in terms of frequency of appearance in coded articles, revealed that the news media frame articles on the subject of campus carry in distinct ways. The media define the problem of campus carry more as a Second Amendment issue than as a First Amendment issue. When framed as a Second Amendment issue, the media most frequently discuss campus carry by mentioning the right to bear arms and self-defense. When framed as a First Amendment issue, the media most frequently discuss campus carry by mentioning freedom of speech, critical/open discourse, and academic freedom. When making a causal interpretation or diagnosis of the problem of campus carry, the coded newspaper articles most frequently referenced school shootings, the victims of gun violence, and the states and colleges/universities that already allow campus carry. Based on the prevailing valence of the articles on the subject, the news media in this study made a moral evaluation or judgement by most frequently taking a position against campus carry. In fact, one staff writer for USA Today went as far as to call campus carry 116 a “loaded debate and a bad idea” clearly indicating his disagreement with the notion of allowing firearms on college and university campuses (Jervis, 2015). In terms of recommendations or suggested remedies, the news media in this study most frequently discussed protests against campus carry, people who would carry firearms on campus if they were allowed to do so, and faculty or administration who would quit or change jobs if their colleges/universities allowed individuals to legally carry firearms on campus. Lastly, when discussing responses and predictions, the news media in this study made clear that most people would feel less secure, and likely change their behavior in the classroom and refrain from engaging in classroom discussions, if their colleges/universities allowed individuals to legally carry firearms on campus. Overall, the alignment between the survey data and the content analysis data indicates a twofold conclusion. First, allowing individuals to legally carry firearms on college and university campuses has the potential to make the majority of campus communities feel less safe, be less forthright and more cautious, and more likely to censor their opinions and limit their conversations. Second, the most credible conclusion may be that campus carry is both a First Amendment issue, associated with the freedom of speech, and a Second Amendment issue, associated with the right to bear arms. Limitations There are several potential limitations to this study, beginning with the survey data. First, the convenience sample of respondents who agreed to participate in the survey contained a disproportionally high frequency of students, mostly from one major/program. On face value, this had the potential to dilute distinctions between students, faculty, and staff, and between the two universities in the sample, making it 117 difficult to show whether these subgroups are similar or distinctive in their views on campus carry. However, preliminary analysis indicates that there may be significant differences between some of the underlying populations in the survey across all three dependent variables, as well as among the predictor variables. This may give nuance to the results in this study and provide guidance for further research. Second, some of the questionnaire responses collected in the survey could be potential sources of bias. Socially desirable responding, defined as the tendency of respondents to exaggerate their good, moral, or respectable behavior and to project favorable images of themselves is a common source of bias in self-report questionnaires (Krumpal, 2013). Social desirability bias may be caused by some respondents reporting that they agree with social norms, including political issues; however, in reality, they do not necessarily agree with these standards (Persson & Solevid, 2014). Another source of bias in questionnaires is extreme responding, indicated by respondents who do not take the time and effort to respond across the full width of a Likert scale but, rather, consistently provide answers at one or another extreme end of the scale (Meisenberg & Williams, 2008). Some respondents in the survey may consistently agree with all or most of the questions measured using 7-point Likert scales (e.g., using scores of 1 or 2) whilst others may consistently disagree with all or most of the questions (e.g., using scores of 6 or 7). Third, as mentioned earlier, another limitation in this study were the negative or reversed Likert scales that were used throughout most of the original survey, where lower scores indicated more security and more agreement and higher levels indicated less security and less agreement. This required that non-intuitive, double-negative reasoning 118 be applied to evaluate the meaning of the coefficients. Again, the scores for the Likert scales were not later reversed because, if they had been reversed, the results of the descriptive and inferential statistical analysis would be different from those obtained if the scores were retained in their original response format (Suarez-Alvarez et. al., 2018). Since reversing Likert scales is referred to as a “questionable practice”, it was not performed in this study (Suarez-Alvarez et al., 2018, p. 149). Likewise, as previously discussed, the researcher was also aware of the limitations of the content analysis, namely the unavoidable subjective judgement of all researchers to interpret the data reliably. For this reason, the researcher asked another rater experienced in content analysis to independently code a percentage of the articles. An inter-rater agreement analysis was conducted, using Krippendorf’s alpha as the test statistic, and good agreement between raters was indicated across all 39 variables measured, as presented in Table 4. It is a limitation that, in comparing the survey results to the results of the content analysis, it is not possible to be definitively predictive that that which was presented in the newspaper articles regarding campus carry predicted population attitudes on the subject. The positions of the newspaper articles (for or against campus carry), as well as the debates and discussions in the articles concerning campus carry, covered many more topics and issues than were included in the survey. In this respect, the questionnaire was somewhat deficient as an instrument to measure attitudes towards campus carry. In future framing and agenda-setting studies, it may be wise to first code the media articles on a topic of interest, and then to use that information as a guide in constructing the survey instrument. It makes sense to first gain an understanding of what it being presented in the 119 news media, and how, before surveying populations to accurately determine whether their attitudes and perceptions mirror those presentations. Likewise, most of the variables measured using Likert scales in the survey were different from the latent themes extracted from the media by the content analysis, both in terms of measurement level, and subject matter. While the Likert scales in the survey were ordinal, most of the content analysis was based on nominal scales. Therefore, it was not possible to conduct an extensive or rigorous comparative analysis to explore the extent to which all of the perceptions of the survey respondents reflected all of the views expressed in the media. Finally, this research studied only one medium, newspapers. Therefore, the results of this study cannot be extrapolated to all media. However, it should be noted that while newspaper readership has fallen significantly in recent years, many, especially older adults, still engage with physical newspapers. Although they do not, generally, generate their own news information, the widespread popularity of social media provides a platform for sharing digital newspaper articles, which appeals to younger audiences. While there are some instances of news being generated on Twitter, traditional news organizations continue to be the dominant source of news and to play a significant role in shaping the news agenda on social media platforms. Conclusion Despite these limitations, this study helps to fill the research gaps regarding how individuals and the media frame the issue of campus carry. Specifically it answers the question of whether campus carry is a First Amendment issue or a Second Amendment issue. The answer is that this is both. That being the case, while proponents of campus 120 carry assert that exercising the right to carry a gun is a form of communication, since the U.S. Supreme Court decision in District of Columbia v. Heller (2008), lower courts have rejected the idea that firearms have communicative connotations (Burgess v. Wallingford, 2013; Nordyke v. King, 2012). They have also rejected the notion that carrying a firearm is a protected form of speech (Enos v. Holder, 2012; Wortman v. U.S., 2015). These arguments fail the standard interpretation of the precedent set in this area of First Amendment law (U.S. v. O’Brien, 1968). Furthermore, the actual language of the Second Amendment, and the U.S. Supreme Court decisions interpreting that language, make it clear that gun rights are limited in both range and scope. Justice Antonin Scalia, in his opinion in Heller (2008), stated that the rights secured by the Second Amendment are not unlimited. Specifically, he noted that nothing in the Heller opinion should question the longstanding laws forbidding the carrying of firearms in sensitive places such as schools (Heller, 2008). In the end, the Second Amendment right to bear arms, and the First Amendment right to free speech, are largely incompatible – especially in the campus environment which cherishes open, robust, and critical debate, as well as academic freedom. The U. S. Supreme Court cases on the subject of academic freedom (Sweezy v. New Hampshire, 1957; Keyishian v. Board of Regents, State Univ. of N.Y.,1967) and symbolic speech in schools (Tinker v. Des Moines Independent Community School District, 1969), make clear that state actors, whether they be state legislatures or school officials, may not abridge teachers’ or students’ constitutionally protected First Amendment rights to free speech or expression. When a professor tells students that certain topics are off limits, or when an institution such as the University of Houston tells 121 faculty to drop certain subjects from their curriculum, that has the potential to be a violation of a constitutionally protected negative liberty. Negative liberties being defined as the freedom from external restraints on one’s actions. However, if a student, a non- state actor, interferes with someone else’s ability to express themselves, then that is likely not a violation of a constitutionally protected liberty. Rather, it is what is known as a positive liberty, or an internal constraint placed upon oneself. Balancing the values involved in the campus carry debate is no simple task. As the survey results in this study indicate, the freedom of expression is one of the most deeply rooted liberties in American society. Yet, personal safety, autonomy, and security are also fundamental values that are central to the American way of life. Professor Nicholas Miller addresses the difficulty involved in the balancing of such rights when he states: To say a decision that involves such competing values is easy means ignoring years of history and experience during which our community has learned to appreciate all these important values. To say it is easy is to remember the individual rights that we personally value most; and to misremember the community of rights of which they are a part — a community of give-and-take which makes our society a civil and fair place to live for people of all faiths, beliefs, and values (2018). Whenever such rights clash, a fair and just resolution must involve a careful and nuanced balancing of the values and interests involved. In the end, it may very well be that allowing for the possession of firearms on college and university campuses is 122 incompatible with the legal principles at the core of both the First Amendment and the Second Amendment. The issue of whether guns on campus are a Second Amendment right, or whether they are a violation of the First Amendment, may one day reach the U.S. Supreme Court. How the Court is likely to decide remains to be seen. The fact that the founding generation engaged in widespread symbolic speech does not necessarily mean that they would have approved of carrying firearms on college and university campuses, centuries later, as a legitimate form of protest, or even as a means of educating the public. Scalia once famously declared that the Constitution is not a living document; “it is dead, dead, dead!” (Scalia & Garner, 2012). With newly appointed Justices on the Court, only time will tell if Scalia’s originalist interpretation of the dead Constitution will live on. What is known is that, while the debate about the competing constitutional rights at the core of the campus carry debate may never be resolved, Founding Fathers and former presidents James Madison and Thomas Jefferson both served on the Board of Visitors for the then-newly created University of Virginia. In the minutes from its October 4, 1824 meeting, which set forth the operation of the university, the Board specifically dictated and approved that no students should use liquor or keep weapons of any kind on school grounds (Richards, 2016). This discussion and vote by Thomas Jefferson, the main author of the Declaration of Independence, and James Madison, the main author of the United States Constitution – in particular the First Amendment and the Second Amendment – indicate not only great historical precedent for college and university campuses to be treated as protected sanctuaries of learning, but also as gun free zones. 123 Future Research Frame analysis argues that individuals classify their experiences according to their previously established guiding frames of reference (Goffman, 1974). If follows that news frames seem to exert the strongest influence on issues where the audience is not already well informed or cognitively engaged (Iyengar, 1991; Kahneman & Tversky, 1984; Zaller, 1992). It also seems that the audience’s psychological distance from an object or issue plays an appreciable role in the media’s ability to shape the audience’s opinion on that object or issue (Ghanem, 1997). Therefore, it follows that if individuals already harbor strong opinions on a subject, those opinions are not likely to change based on media framing. As such, future research ideas include examining the effect of media framing and agenda-setting regarding the topic of campus carry on those who are not already familiar or well-versed on the debate, or those that are not directly affected by the issue. This will likely include surveying the general public as opposed to campus communities, as well as those who do not have strong, preconceived notions and opinions on either the First Amendment or the Second Amendment. It would be interesting to compare those results to the findings of the present research study. Likewise, given some of the limitations encountered with the survey and coding instruments in the present study, in future framing and agenda-setting studies, it would be wise to first code the media articles on a topic of interest, and then to use that information as a guide in constructing the survey instrument. Lastly, since the survey respondents in this study reinforced the idea that there is not much variation in support of the First Amendment, research should be conducted to 124 test the limits of that support. For example, should controversial speech receive the same First Amendment protections as other forms of protected speech? Do respondents consider First Amendment rights to be unrestrictable? If not, where should the line be drawn? These could serve as longitudinal studies that track the limits of support for the First Amendment over time. 125 XI. APPENDICES Appendix A: Survey Instrument Start of Block: Informed Consent Q1 Please read the following information about this study and click the "I accept" button at the bottom of your screen if you are interested in participating. The following survey is part of a research study of perceptions of crime, freedom, and security. As a member of the University community, you have been selected to participate in this study. This web-based survey will take approximately 10-12 minutes to complete. By choosing to participate you will help expand the knowledge about perceptions of crime and communication. Your decision to participate is strictly voluntary and involves only the risk of mild discomfort sometimes associated with thinking about crime and current events. You are free to answer all, some, or none of the questions on the survey. You may withdraw from participating at any time and to do so you simply close your Internet browser. Declining to participate will involve no penalty to you. You must be 18 years or older to participate in this survey. If you submit a survey, your responses are recorded without any personal identifiers, so your responses are completely anonymous. The survey is being sent to approximately 600 individuals. Data will be stored on a secure server and can only be accessed by the research team involved in the study. No guarantees can be made regarding the interception of data sent via the Internet by any third parties. Reasonable and appropriate safeguards have been used in the creation of the web-based survey to maximize the confidentiality and security of your responses; however, when using information technology, it is never possible to guarantee complete privacy. If you have questions about your rights as a research subject, you may contact: Research Compliance Services at the University of Oregon at 541-346-2510 or ResearchComplaince@uoregon.edu -or- Human Subjects Review Council, Central Washington University, at 509-963-3115 or HSRC@cwu.edu You may ask questions about the research by contacting the principal investigator, Krystal E. Noga-Styron, at keensdata@gmail.com Please click "I accept" if you wish to participate. 126 o I accept (1) o I decline (2) End of Block: Informed Consent Start of Block: Age Q6 Are you 18 years old or older? o Yes (1) o No (2) Skip To: End of Survey If Are you 18 years old or older? = No Q7 What is your age? o 18-25 (1) o 26-35 (2) o 36-45 (3) o 46-54 (4) o 55 or older (5) End of Block: Age Start of Block: Affiliation with the University 127 Q18 Which of the following best describes your affiliation with the University? ▢ Student (1) ▢ Faculty (2) ▢ Administration (3) ▢ Staff (4) Display This Question: If Which of the following best describes your affiliation with the University? = Student Q18 If you are a student, which course are you currently enrolled in? o A Journalism and Communication course at the University of Oregon (1) o A Law and Justice course at Central Washington University (2) o A Law course at the University of Oregon (3) o A Communication or Journalism course at Central Washington University (4) o Other (6) Display This Question: If Which of the following best describes your affiliation with the University? = Faculty Q66 With which university are you employed? o Central Washington University (1) o University of Oregon (2) 128 Display This Question: If Which of the following best describes your affiliation with the University? = Administration Q67 With which university are you employed? o Central Washington University (1) o University of Oregon (2) Display This Question: If Which of the following best describes your affiliation with the University? = Staff Q68 With which university are you employed? o Central Washington University (1) o University of Oregon (2) End of Block: Affiliation with the University Start of Block: Education/Politics 129 Q13 What is your educational background? o I did not graduate from High School or obtain a GED (1) o High School or GED (2) o Some College (3) o Associates Degree completed (4) o Bachelor of Arts or Bachelor of Sciences completed (5) o Some graduate work (6) o Masters degree completed (7) o Some Doctoral work (8) o Doctoral degree completed (9) Q14 Which of the following best describes your political affiliation? o Democrat (1) o Republican (2) o Independent (3) o Libertarian (4) o Not Political (5) o Other (6) ________________________________________________ Q70 Whom did you vote for in the last presidential election? ________________________________________________________________ 130 End of Block: Education/Politics Start of Block: Media Exposure Q34 In the space provided, please list all of the Television Networks, Newspapers, Magazines, Radio Stations, Video Games, and Internet sites that you regularly engage with. ________________________________________________________________ Q65 In a typical day, how much time do you spend engaged with media and social media (Newspapers, Magazines, Television, Internet, Radio, Video Games, etc.) o Less than 1 hour (1) o 1-2 hours (2) o 2-3 hours (3) o 3-4 hours (4) o 4-5 hours (5) o 5-6 hours (6) o 6-7 hours (7) o 7-8 hours (8) o 8-9 hours (9) o 9-10 hours (10) o More than 10 hours (11) End of Block: Media Exposure Start of Block: Crime Concerns 131 Q35 How concerned are you about the current crime and victimization rate? o Very concerned (1) o Concerned (2) o Somewhat Concerned (3) o Neither concerned nor unconcerned (4) o Somewhat Unconcerned (5) o Unconcerned (6) o Very Unconcerned (7) Q37 Do you feel that the current NATIONAL crime and victimization rate is up or down as compared to recent years? o Up (1) o Somewhat up (2) o Neither up nor down (3) o Somewhat down (4) o Down (5) 132 Q39 Do you feel that the current LOCAL crime and victimization rate is up or down as compared to recent years? o Up (1) o Somewhat up (2) o Neither up nor down (3) o Somewhat down (4) o Down (5) Q40 Do you feel that the current CAMPUS crime and victimization rate is up or down as compared to recent years? o Up (1) o Somewhat up (2) o Neither up nor down (3) o Somewhat down (4) o Down (5) Q46 Would you be surprised to learn that crime and victimization rates have generally DECREASED in recent years? o Definitely yes (1) o Probably yes (2) o Might or might not (3) o Probably not (4) o Definitely not (5) 133 Q47 Would you be surprised to learn that crime and victimization rates have generally INCREASED in recent years? o Definitely yes (1) o Probably yes (2) o Might or might not (3) o Probably not (4) o Definitely not (5) End of Block: Crime Concerns Start of Block: Information regarding crime Q41 Where do you obtain your information regarding the current crime and victimization rate? Please select all that apply. ▢ Media and/or Internet (1) ▢ Police Blotter/Police Log (2) ▢ Friends, Family, and/or Associates (3) ▢ Community and/or Religious Institutions (4) ▢ Personal Experience (5) ▢ Other (6) ________________________________________________ Display This Question: If Where do you obtain your information regarding the current crime and victimization rate? Please s... = Media and/or Internet 134 Q42 If you obtain your information regarding the current crime and victimization rate from the Media, what type of sources are you using? ▢ Network Television (1) ▢ Cable Television (2) ▢ Local Newspapers (3) ▢ National Newspapers (4) ▢ Weekly Magazines (5) ▢ Monthly Magazines (6) ▢ FM Radio (7) ▢ AM Radio (8) ▢ Public Radio (9) ▢ The Internet (10) ▢ Other (11) ________________________________________________ Display This Question: If Where do you obtain your information regarding the current crime and victimization rate? Please s... = Community and/or Religious Institutions Q43 If you obtain your information regarding the current crime and victimization rate from Community and/or Religious Institutions, in the space provided please list the Community and/or Religious Institutions to which you belong. ________________________________________________________________ 135 End of Block: Information regarding crime Start of Block: Law Enforcement Q37 Do you have experience in the military or law enforcement? ▢ I have military experience. (1) ▢ I have law enforcement experience (2) ▢ I do not have military or law enforcement experience (3) Q38 Do you have confidence in law enforcement preventing and combating crime? o Definitely yes (1) o Probably yes (2) o Might or might not (3) o Probably not (4) o Definitely not (5) 136 Q60 Do you believe that the rate of officer-involved shootings is generally up or down as compared to recent years? o Up Dramatically (1) o Up (2) o Somewhat Up (3) o Neither Up nor Down (4) o Somewhat Down (5) o Down (6) o Down Dramatically (7) Q63 Please rate the veracity (truth) of the following statement: The police disproportionately use deadly force on unarmed black men? o Strongly agree (1) o Agree (2) o Somewhat agree (3) o Neither agree nor disagree (4) o Somewhat disagree (5) o Disagree (6) o Strongly disagree (7) 137 Q62 Do you believe that the rate of law enforcement officers being assaulted or killed is generally up or down as compared to recent years? o Up Dramatically (1) o Up (2) o Somewhat Up (3) o Neither Up nor Down (4) o Somewhat Down (5) o Down (6) o Down Dramatically (7) End of Block: Law Enforcement Start of Block: Weapon or Security Device 138 Q44 Do you own any of the following weapons or security devices? Please select all that apply. ▢ Gun/Firearm (1) ▢ Knife (2) ▢ Taser or Stun-Gun (3) ▢ Mace or Pepper Spray (4) ▢ Baseball Bat or similar item (5) ▢ Home Alarm System (6) ▢ Car Alarm System (7) ▢ Other (8) ________________________________________________ Display This Question: If Do you own any of the following weapons or security devices? Please select all that apply. = Gun/Firearm Q49 Do you have a valid concealed handgun permit/license? o Yes (1) o No (2) Display This Question: If Do you have a valid concealed handgun permit/license? = Yes Q50 How often do you carry your gun/firearm and in which environments? ________________________________________________________________ 139 Display This Question: If Do you have a valid concealed handgun permit/license? = Yes Q51 Would you carry a gun/firearm on campus if you had the legal right to do so? o Yes (1) o Maybe (2) ________________________________________________ o No (4) Q69 Please rate your comfort level around guns/firearms: o I am very comfortable around guns/firearm (1) o I am comfortable around guns/firearms (2) o I am somewhat comfortable around guns/firearms (3) o I am neither comfortable nor uncomfortable around guns/firearms (4) o I am somewhat uncomfortable around guns/firearms (5) o I am uncomfortable around guns/firearms (6) o I am very uncomfortable around guns/firearms (7) End of Block: Weapon or Security Device Start of Block: Constitutional Freedoms 140 Q49 Please rate your response to the following statement: The freedom of speech is an inalienable right guaranteed by the Constitution that should not be subject to censorship laws or policies. o Strongly agree (1) o Agree (2) o Somewhat agree (3) o Neither agree nor disagree (4) o Somewhat disagree (5) o Disagree (6) o Strongly disagree (7) Q48 Please rate your response to the following statement: To own and bear arms is an inalienable right guaranteed by the Constitution that should not be subject to gun control laws or policies. o Strongly agree (1) o Agree (2) o Somewhat agree (3) o Neither agree nor disagree (4) o Somewhat disagree (5) o Disagree (6) o Strongly disagree (7) End of Block: Constitutional Freedoms Start of Block: Campus Carry 141 Q51 Please rate your response to the following statement: If my university allowed individuals to legally carry firearms on campus, I would feel: o Much more secure on campus (1) o More secure on campus (2) o Somewhat more secure on campus (3) o Neither more secure nor less secure on campus (4) o Somewhat less secure on campus (5) o Less secure on campus (6) o Much less secure on campus (7) Q52 Please rate your response to the following statement: If my university allowed individuals to legally carry firearms on campus, it would change my behavior in the classroom. o Strongly agree (1) o Agree (2) o Somewhat agree (3) o Neither agree nor disagree (4) o Somewhat disagree (5) o Disagree (6) o Strongly disagree (7) Display This Question: If Please rate your response to the following statement: If my university allowed individuals to leg... = Strongly agree 142 Q53 If your university allowed individuals to legally carry firearms on campus, how would it change your behavior in the classroom? ________________________________________________________________ Display This Question: If Please rate your response to the following statement: If my university allowed individuals to leg... = Agree Q54 If your university allowed individuals to legally carry firearms on campus, how would it change your behavior in the classroom? ________________________________________________________________ Display This Question: If Please rate your response to the following statement: If my university allowed individuals to leg... = Somewhat agree Q55 If your university allowed individuals to legally carry firearms on campus, how would it change your behavior in the classroom? ________________________________________________________________ 143 Q50 Please rate your response to the following statement: If my university allowed individuals to legally carry firearms on campus, I would refrain from engaging in controversial classroom discussions. o Strongly agree (1) o Agree (2) o Somewhat agree (3) o Neither agree nor disagree (4) o Somewhat disagree (5) o Disagree (6) o Strongly disagree (7) End of Block: Campus Carry Start of Block: Gender/Marital Status Q4 What is your gender o Female (1) o Male (2) o Other (3) ________________________________________________ 144 Q9 What is your marital status? ▢ Single (1) ▢ Married (2) ▢ Domestic Partnership (3) ▢ Cohabiting (4) ▢ Separated (5) ▢ Divorced (6) ▢ Widowed (7) ▢ Other (8) Q58 Do you have children? o Yes (1) o No (2) End of Block: Gender/Marital Status Start of Block: Race/Ethnicity 145 Q10 What is your race or ethnicity? Please select all that apply. ▢ American Indian or Alaska Native (1) ▢ Asian or East Indian (2) ▢ Black or African American (3) ▢ Hispanic (6) ▢ Native Hawaiian or Other Pacific Islander (4) ▢ White or Caucasian (5) End of Block: Race/Ethnicity Start of Block: Residence Q11 Which US state or territory is your official residence? ▼ Alabama (1) ... I reside outside of the United States (52) Q55 In which US state or territory do you currently reside? ▼ Alabama (1) ... I reside outside of the United States (52) 146 Q12 Which of the following best describes the community where you currently live? ▢ Rural (1) ▢ Suburban (2) ▢ Urban (3) End of Block: Residence Start of Block: Thank you Q57 Thank you for taking our survey. If you have any questions or concerns about this study please contact Krystal Noga- Styron at keensdata@gmail.com For questions about your rights as a research subject, you may contact: Research Compliance Services, University of Oregon, at 541-346-2510 or ResearchCompliance@uoregon.edu -or- Human Subjects Review Council, Central Washington University, at 509-963-3115 or HSRC@cwu.edu If you would like a copy or summary of the results please e-mail Krystal Noga-Styron at keensdata@gmail.com and results will be sent to you by December 31, 2017. TO PROTECT YOUR PRIVACY, PLEASE CLOSE THE BROWSER. End of Block: Thank you 147 Appendix B: Coding Instrument Start of Block: Coders and Articles Coded Q1 Reviewer Initials o KNS (1) o MJ (2) Q2 Newspaper Source o The Guardian (1) o The New York Times (2) o The Oregonian (3) o The Seattle Times (4) o USA Today (5) o The Washington Post (6) Q3 Article Headline ________________________________________________________________ Q4 Date of Article (entered as 2-digit month, 2-digit day, and 4-digit year, e.g, 01/01/2020) ________________________________________________________________ 148 Q18 Type of Newspaper Article: o News Story (1) o Editorial (2) o Letter to Editor (3) o Other (4) ________________________________________________ End of Block: Coders and Articles Coded Start of Block: Defining the Problem Q5 Does the article mention or discuss the First Amendment? o Yes (1) o No (2) Display This Question: If Does the article mention or discuss the First Amendment? = Yes Q6 How does the article mention or discuss the First Amendment? Yes (1) No (2) Freedom of Speech (1) o o Academic Freedom (2) o o Critical/Open Discourse (3) o o Marketplace of ideas (4) o o Other (5) o o 149 Q7 Does the article mention or discuss the Second Amendment? o Yes (1) o No (2) Display This Question: If Does the article mention or discuss the Second Amendment? = Yes Q8 How does the article mention or discuss the Second Amendment? Yes (1) No (2) Broadly as the Right to Bear Arms (1) o o Specifically as the Right to an Armed Militia (2) o o Self-Defense (3) o o Symbolic Political Speech or Expression (4) o o Other (5) o o 150 Q9 Terminology: how does the article refer to the campus carry issue? Yes (1) No (2) Guns on Campus (1) o o Campus Carry (2) o o Concealed Weapons on Campus (3) o o Other (4) o o End of Block: Defining the Problem Start of Block: Causal Interpretation or Diagnosis of the Cause of the Problem 151 Q10 Does the article mention or discuss any of the following regarding the discussion on campus carry? Yes (1) No (2) The number of school shootings or other attacks on campuses (1) o o The number of victims of gun violence either on or off campuses (2) o o The names, or number, of states that ALLOW campus carry (3) o o The names, or number, of states that PROHIBIT campus carry (4) o o The names, or number, of states that are CONSIDERING campus carry (5) o o The names, or number, of colleges/universities that ALLOW campus carry (6) o o The names, or number, of colleges/universities that PROHIBIT campus carry (7) o o The names, or number, of colleges/universities that are CONSIDERING campus carry (8) o o End of Block: Causal Interpretation or Diagnosis of the Cause of the Problem Start of Block: Moral Evaluation or Judgement 152 Q11 What side, if any, does the article take in the debate on campus carry? Quotes that evidence anti- or pro-campus carry positions may be collected here in the space provided. o Anti-campus carry (1) ________________________________________________ o Neutral (2) ________________________________________________ o Pro-campus carry (3) ________________________________________________ End of Block: Moral Evaluation or Judgement Start of Block: Treatment Recommendations or Suggested Remedies Q12 Does the article mention or discuss any of the following regarding the debate on campus carry? Yes (1) No (2) Students dropping out or switching schools (1) o o Faculty or administration who quit or changed jobs (2) o o People who now carry (either legally or illegally) (3) o o People who would carry if their university allowed them to do so (5) o o Protests on campus against campus carry (4) o o Protests on campus for campus carry (6) o o End of Block: Treatment Recommendations or Suggested Remedies Start of Block: Responses and Predictions of Likely Effects 153 Q15 Does the article mention whether, if individuals are (or were) legally allowed to carry firearms on campus, do (or would) others: Yes (1) No (2) Feel MORE secure on campus? (1) o o Feel LESS secure on campus? (2) o o Change their behavior in the classroom. (3) o o NOT change their behavior in the classroom. (4) o o Refrain from engaging in controversial classroom discussions. (5) o o NOT refrain from engaging in controversial classroom discussions. (6) o o End of Block: Responses and Predictions of Likely Effects Start of Block: Coders Comments Q16 Was there anything unusual or difficult to code on this sheet? ________________________________________________________________ Q17 If there are any additional quotes or statements which should be recorded, do so here: ________________________________________________________________ End of Block: Coders Comments 154 REFERENCES CITED Abdi, H. (2007. Bonferroni and Sidak corrections for multiple comparisons. In N.J. Salkind (ed.) Encyclopedia of Measurement and Statistics. CA: Thousand Oaks. Abrams v. U. S., 250 U.S. 616 (1919). Ackerman, B. (1991). We the People, Volume I: Foundations. Cambridge: Bellknap Press of Harvard University. Adams, J. (2011). A Dissertation on the Canon and Feudal Law. In C.F. Adams (Ed.) The Works of John Adams, Second President of the Unites States. Cambridge: Cambridge University Press. Adler et al. v. Board of Education of the City of New York, 342 U.S. 483 (1952). Agresti, A. (2010). Analysis of ordinal categorical data. New York: Wiley. Amar, A. R. (2006). America’s Constitution. New York: Random House. Amar, A. R. (2012). America's Unwritten Constitution: The Precedents and Principles We Live By. New York: Basic Books. Andrada, C. (2007). Confounding. Indian Journal of Psychiatry, 49, 129-131. Asbury, H. (1928). Gangs of New York: an Informal History of the Underworld. New York: Alfred A. Knopf. Associated Press. (2016, August 1). Texas professors sue to repeal campus gun law. The Oregonian. Retrieved from: https://www.oregonlive.com/today/2016/08/texas_ professors_sue_to_repeal .html Aymette v. State, 21 Tenn. 154 (1840). Barnes, A. T. (2017). Guns and academic freedom. Gonzaga Law Review, 53, 47-90. Blackstone, W. (1987). Commentaries on the Laws of England. In P. B. Kurland & R. Lerner, The Founders’ Constitution. Indianapolis: Liberty Fund. Bliss v. Commonwealth of Kentucky, 12 Ky. 90 (1822). Bluman v. Federal Election Commission, 800 F. Supp. 2d. 281 (2011). Bogus, C. T. (1998). The hidden history of the Second Amendment. U.C. Davis Law Review, 31, 309-410. 155 Bogus, C. T., (2002). The history and politics of Second Amendment scholarship: a primer. In C.T. Bogus, ed., The Second Amendment in Law and History: Historians and Constitutional Scholars on the Right to Bear Arms. New York: New Press. Bowker, D.W., & Randerson, P.F. (2007). Practical data analysis workbook (3rd ed.). Harlow, UK: Pearson Education. Brady Handgun Violence Prevention Act, 107 Stat. 1536 (1994). Brandenburg v. Ohio, 395 U.S. 444 (1969). Breyer, S. (2005). Active liberty: interpreting our democratic constitution. New York: Random House. Breyer, S. (2010). Making our democracy work. New York: Alfred A. Knopf. Buckley v. Valeo, 424 U.S. 1 (1976). Burger, W. (1991). Warren Burger "2nd Amendment Fraud" - 1991 PBS News Hour. Retrieved: https://www.youtube.com/watch?v=Eya_k4P-iEo&feature=youtu.be Burgess v. Town of Wallingford, 13-2369-cv (2d Cir. 2014). Carifio, J. & Perla, R. (2007). Ten common misunderstandings, misconceptions, persistent myths and urban legends about Likert scales and Likert response formats and their antidotes. Journal of Social Science, 3. 106-116. Carifio, J. & Perla, R. (2008). Resolving the 50-year debate around using and misusing Likert scales. Medical Education, 42, 1150-1152. Cavanaugh, M. R., Bouffard, J. A., Wells, W., & Nobles, M. R. (2012). Student attitudes toward concealed handguns on campus at 2 universities. American Journal of Public Health, 102(12), 2245-2247. Center for Homeland Defense and Security. (2019). K-12 School Shooting Database. Retrieved from: https://www.chds.us/ssdb/ Chafee, Z. (1920). Freedom of Speech. New York: Harcourt, Brace, and Howe. Chemerinsky, E. & Gillman, H. (2017). Free Speech on Campus. New Haven: Yale University Press. Citizens United v. F.E.C., 558 U.S. 310 (2009). 156 Cohen, J. (1960). A coefficient of agreement for nominal scales. Educational and Psychological Measurement, 10, 37-46. Cornell, S. (2008). A Well Regulated Militia: the Founding Fathers and the Origins of Gun Control in America. London: Oxford University Press. Coyle, M. (2013). The Roberts Court: the Struggle for the Constitution. New York: Simon and Schuster. Cramer, C. E. (2014). Guns on campus: a history. Academic Questions, 27(4), 411-425. Czitrom, D. (1991) Underworld and underdogs: Big Tim Sullivan and metropolitan politics in New York, 1889-1913. The Journal of American History, 78(2), 536- 558. Debrabander, F. (2016, March 4). How guns censor college classrooms. The Atlantic. Retrieved from http://www.theatlantic.com/education/archive/2016/03/the-steep- cost-of-allowing-guns-in-the-college-classroom/472296/ Dennis v. United States, 341 U.S. 494 (1951). Denzin, N. (2006). Sociological Methods: A Sourcebook. (5th ed.) Piscataway, NJ: Aldine Transaction. District of Columbia v. Heller, 554 U.S. 570 (2008). Duffy, S. (2010). Random numbers demonstrate the frequency of Type I errors: Three spreadsheets for class instruction. Journal of Statistics Education, 18, 1-18. Efficiency in Militia Act, 32 Stat. 775 (1903). Eldridge, L. (1995). Before Zenger: truth and seditious speech in colonial America, 1607- 1700. American Journal of Legal History, vol. 39, no. 3, 337-358. Eldridge, L. (1995). A Distant Heritage: the Growth of Free Speech in Early America. New York: New York University Press. English Bill of Rights (1689). 1 William & Mary Sess. 2 c 2. Enos v. Holder, 855 F. Supp. 2d 1088 (E.D. Cal. 2012). Entman, R. M. (1993). Framing: toward clarification of a fractured paradigm. Journal of Communication, 43(4), 51-58. Epstein, R. (2008). A structural interpretation of the Second Amendment. Why Heller is (probably) wrong on Originalist Grounds. Syracuse Law Review, 59, 171-184. 157 Esponiti-Vinzi, E, Trinchera, L., & Amato, S. (2010). Handbook of partial least squares. Springer Handbooks of Computational Statistics. Berlin Heidelberg: Springer- Verlag. Everytown For Gun Safety. (2019). Gunfire on School Grounds in the United States. Retrieved from https://everytownresearch.org/gunfire-in-school/#8283 Faul, F., Erdfelder, E., Buchner, A., & Lang, A.-G. (2009). Statistical power analyses using G*Power 3.1: Tests for correlation and regression analyses. Behavior Research Methods, 41, 1149-1160. Fennell, R. (2009). Concealed carry weapon permits: a Second Amendment right or a recipe for disaster on our nation’s campuses? Journal of American College Health, 58(2), 99-100. Ferguson, C.F. (2009). An effect size primer: a guide for clinicians and researchers. Professional Psychology: Research and Practice, 40, 532-538. Finn, P. (2013, March 13). NRA money helped reshape gun law. The Washington Post. Fleiss, J. L. (1971). Measuring nominal scale agreement among many raters. Psychological Bulletin, 76(5), 378-382. Friesen, S. J. (2016). Why the guns-on-campus debate matters for American higher education Retrieved from: http://theconversation.com/why-the-guns-on- campus-debate-matters-for-american-higher-education-63495 Frum, D. (2000). How we got here: the 70’s the decade that brought you modern life—for better or worse. New York: Basic Books. Gamson, W., A. (1992). Talking Politics. New York: Cambridge University Press. Ghanem, S. I. (1997). Filling in the tapestry: The second level of agenda setting . In M.E. McCombs, ed. , D. Shaw,, ed. & D. Weaver (Eds.), Communication and democracy (pp. 309–330). Mahwah: Lawrence Erlbaum. Giffords Law Center to Prevent Gun Violence. (20 May, 2013). Post-Heller Litigation Summary. Retrieved: https://lawcenter.giffords.org/gun-laws/the-second- amendment/post-heller-litigation-summary/ Gitlow v. New York, 268 U.S. 652 (1925). Goffman, E. (1974). Frame analysis – an essay on the organization of experience. New York: Harper & Row. Gun Control Act of 1968, 82 Stat. 1213-2 (1968). 158 Gwet, K. (2002). Kappa statistic is not satisfactory for assessing the extent of agreement between raters. Statistical Methods for Inter-Rater Reliability Assessment Series, 1, 1-6. Habermas, J. (1989). Structural Transformation of the Public Sphere; an Inquiry into a Category of Bourgeois Society. (T. Burger & F. Lawrence, Trans.). Boston: MIT Press.) Original work published 1962). Hadden, S. E. (2001). Slave Patrols: Law and Violence in Virginia and the Carolinas. Cambridge: Harvard University Press. Haenlein, M., & Kaplan, A. (2004). A beginner’s guide to partial least squares analysis. Understanding Statistics, 3, 283-297. Hair, J., Anderson, R. E., Babin, B. J., Tatman, R. L., & Black, W. C. (2010). Multivariate data analysis. Upper Saddle River, NJ: Prentice Hall. Hall, K. L. (2002, September 13). Free speech on public college campuses overview. First Amendment Center. Retrieved from http://www.firstamendmentcenter.org/ free-speech-on-public-college-campuses Halsey, L.G., Curran-Everett, D., Vowler, S.L., & Drummond, G.B. (2015). The fickle p value generates irreproducible results. Nature Methods, 12, 179-185. Hays, S. (1960). The right to bear arms, a study in judicial misinterpretation. William and Mary Law Review, 2, 381-406. Hill, C.R., & Thompson, B. (2004). Computing and interpreting effect sizes. In: J.C. Smart (Ed.) Higher education handbook of theory and research. Volume XIX. pp. 175-196. Boston, MA: Klewer. Hoekstra, R., Finch, S., Kiers, H. A. L., & Johnson, A. (2006). Probability as certainty: Dichotomous thinking and the misuse of p-values. Psychonomic Bulletin & Review, 13, 1033–1037. Horwitz, D. (2016). Open carry: open-conversation or open threat? First Amendment Law Review, 15, 96-120. Hunt, G. (1900). The Writing of James Madison,. New York: G. P. Putnam’s Sons. Hurlbert, S.H., & Lombardi, C.M. (2009). Final collapse of the Neyman-Pearson decision theoretic framework and rise of the neoFisherian. Annales Zoologici Fennici, 46, 311- 349. Iyengar, S. (1991). Is anyone responsible? Chicago: University of Chicago Press. 159 Jacobs, J. B. (2002). Can Gun Control Work? London: Oxford University Press. Jamieson, S. (2004). Likert scales: how to (ab)use them. Medical Education, 38, 1117- 1118. Jervis, R. (2015, October 18). Voices: Sex toys, firearms and other campus hazards. USA Today. Retrieved from https://www.usatoday.com/story/news/nation/2015/10/18/ voices-sex-toys-firearms-campus-carry/73977530/ Jones v. Walker, 13 F. Cas. 1059 (1791). Kahneman, D., & Tversky, A. (1984). Choice, values, and frames. American Psychologist, 39, 341-350. Kalven, H. (1988). A Worthy Tradition: Freedom of Speech in America. New York: Harper and Row. Kaur, P., & Stoltfuz, J. (2017). Type I, II, and III statistical errors: A brief overview. Biostatistics, 3(2), 268-270. Kenneth, L., & Anderson, J. L. (1957). The Gun in America: the Origins of a National Dilemma. Westport: Greenwood. Kessler, R.C., Amminger, G. P., Aguilar-Gaxiola, S., Alonso, J., Lee, S. & Ustun, T., B. (2007). Age of onset of mental disorders: A review of recent literature. Current Opinion in Psychiatry, 20(4), 359-364. Keyishian v. Board of Regents, State Univ. of N.Y., 385 U.S. 589 (1967). Kirk, R.E. (1996). Practical significance: a concept whose time has come. Educational and Psychological Measurement, 56, 746-759. Kitrosser, H. (2017) Free Speech, Higher Education, and the PC Narrative. Minnesota Law Review, 101, 1987-2064. Klukowski, K. A. (2014). Making Second Amendment law with First Amendment rules: the five-tier free speech framework and public forum doctrine in Second Amendment jurisprudence. Nebraska Law Review, 93(2), 429-494. Konig, T. (2009). Why the Second Amendment has a preamble: original meaning and the political culture of written constitutions in Revolutionary America. U.C.L.A. Law Review, 56, 1295-1342. Konstantopoulos, S., & Hedges, L. V. (2008). How large an effect can we expect from school reforms? Teachers College Record, 110(8),1613–1640. 160 Kopel, D. B. (2014). The First Amendment guide to the Second Amendment. Tennessee Law Review, 8, 417-478. Kozuskanich, N. (2009). Originalism in a digital age: an inquiry into the right to bear arms. Journal of the Early Republic, 29(4), 585-606. Krippendorff, K. (2004). Content analysis: An introduction to its methodology (2nd ed.). Thousand Oaks, CA: Sage. Krippendorff, K. (2018). Content Analysis: An Introduction to Its Methodology. Thousand Oaks: Sage. Krumpal, I. (2013). Determinants of social desirability bias in sensitive surveys: a literature review. Quality & Quantity. 47 (4), 2025–2047. Lacy, S., Watson, B. R., Riffe, D., & Lovejoy, J. (2015). Issues and best practices in content analysis. Journalism & Mass Communication Quarterly, 92(4), 791-811. Levy, L. W. (1960). Legacy of Suppression: Freedom of Speech and Press in Early American History. Cambridge: Harvard University Press. Levy, L. W. (1985). Emergence of a Free Press. Oxford: Oxford University Press. Lewis, S. (2017). Crossfire on compulsory campus carry laws: when the First and Second Amendments collide. Iowa Law Review, 102, 2109-2144. Lindeløv, J.K. (2019). Why SPSS is dying. Retrieved from: https://lindeloev.net/spss-is- dying/ Linguistics Research Center, University of Texas at Austin, (2012). The Second Amendment: our Latinate constitution. Retrieved: https://sites.utexas.edu/lrc /2012/12/26/the-second-amendment-our-latinate-constitution/ Lo, A., Chernoff, H., Zheng, T. & Lo, S. (2015). Why significant variables aren’t automatically good predictors. Proceedings of the National Academy of Sciences of the United States of America. doi: 10.1073/pnas.1518285112 Luo, M. (2017, August, 11). How the NRA manipulates gun owners and the media. The New Yorker. Retrieved from: https://www.newyorker.com/news/news-desk/how- the-nra-manipulates-gun-owners-and-the-media Magarian, G. P. (2012). Speaking Truth to Firepower: How the First Amendment Destabilizes the Second, Texas Law Review, 91, 49-57 (2012). 161 McCombs, M. E., & Evatt, D.(1995). Los temas y los aspectos: explorando una nueva dimension de la agenda setting [Objects and attributes: exploring a new dimension of agenda setting]. Comunicacion y Sociedad, 8(1), 7-32. McCombs, M. E., Llamas, J. P., Lopez-Escobar, E., & Rey, F. (1997). Candidate's images in Spanish elections: Second-level agenda-setting effects. Journalism & Mass Communication Quarterly, 74(4), 703–717. McCombs, M., & Shaw, D. (1972). The agenda-setting function of the mass media. Public Opinion Quarterly, 36(2), 176-187. McCombs, M. E., Shaw, D. L., & Weaver, D. H. (2014). New directions in agenda- setting theory and research. Mass Communication & Society. 17, 781–802. McMillan J.H., & Foley, J. (2011). Reporting and discussing effect size: Still the road less travelled? Practical Assessment, Research & Evaluation, 16, 1-11. McShane, B.B., & Gal, D. (2017). Statistical significance and the dichotomization of evidence. Journal of the American Statistical Association, 12 (159), 885-908. Meisenberg, G., & Williams, A. (2008). Are acquiescent and extreme response styles related to low intelligence and education? Personality and Individual Differences, 44 (7), 1539–1550. Merriam, S. (2014). Qualitative research: A guide to design and implementation. San Francisco, CA: Jossey-Bass. Miller, N. (2018, February 4). Modern American rights claims: An avoidable clash of absolutes. The Hill. Retrieved from https://thehill.com/opinion/judiciary/372199- modern-american-rights-claims-an-avoidable-clash-of-absolutes Miller v. Texas, 153 U.S. 535 (1894). Milton, J. (1644). Areopagitica [Pamphlet]. London: British Library Collection. Missouri v. Holland, 252 U.S. 416 (1920). Moody, J. (2014). Safety first: how arming Mississippi teachers would protect students. Mississippi Law Journal, 83, 1319-1341. Morgan, J. (2018). Difference Between.net. Retrieved from: http://www.differencebetween.net/language/words-language/difference-between- safety-and-security/ Murphy, B. A. (2014). Scalia: a Court of One. New York: Simon and Schuster. 162 National Conference of State Legislatures. (2018). Guns on Campus: Overview. Retrieved from: https://www.ncsl.org/research/education/guns-on-campus- overview.aspx National Firearm Act, 48 Stat. 1236 (1943). Nelson, H. L. (1959). Seditious libel in Colonial America. American Journal of Legal History, 3(2), 160-172. Neuendorf, K.A. (2016). The Content Analysis Guidebook. Thousand Oaks, CA: Sage. New York Times v. Sullivan, 376 U.S. 254 (1964). Nordyke v. King, 681 F.3d 1041 (2012). Nuzzo, R. (2014). Scientific method: statistical errors. P values, the 'gold standard' of statistical validity, are not as reliable as many scientists assume. Nature, 506, 150- 152. O’Meara, J. (1890). Concealed weapons and crimes. Overland Monthly, 2, 11-15. Osborne, J W. & Overbay, A. (2004). The power of outliers (and why researchers should always check for them). Practical Assessment, Research & Evaluation, 9(6). Retrieved from http://PAREonline.net/getvn.asp?v=9&n=6 Osborne, J.W, & Waters, E. (2002). Four assumptions of multiple regression that researchers should always test. Practical Assessment, Research & Evaluation, 8, 1-7. Oswald, M.E., & Stefan, G. (2004). Confirmation bias. In R.F., Pohl. (Ed.). Cognitive illusions: a handbook on fallacies and biases in thinking, judgement and memory (pp. 79-96). Hove, UK: Psychology Press. Patten, R., Thomas, M. O., & Wada, J.C. (2012). Packing heat; attitudes regarding concealed weapons on college campuses. American Journal of Criminal Justice, 38, 551-569. Patterson v. Colorado, 205 U.S. 454 (1907). Persson, M. & Solevid, M. (2014). Measuring political participation - testing social desirability bias in a web-survey experiment. International Journal of Public Opinion Research, 27, 98-102. 163 Pohl, J. (2016, December 5). Campus attack reignites concealed carry debate. Coloradoan.Retrieved from http://www.coloradoan.com/story/news/2016/12/05/ campus-attack-reignites-concealed-carry-debate/93498200/ Posner, R. (27, Aug., 2008). In defense of looseness. The New Republic. Retrieved: https://newrepublic.com/article/62124/defense-looseness Presser v. Illinois, 116 U.S. 252 (1886). Protection of Lawful Commerce in Arms Act, 119 Stat. 2095 (2005). Public Safety and Recreational Firearms Use Protection Act, part of the Violent Crime Control and Law Enforcement Act, 108 Stat. 1796 (1994). Rabban, D. M. (1997). Free Speech in its Forgotten Years. London: Cambridge University Press. Rakove, J. N., (2002). The Second Amendment: the highest state of originalism. In C.T. Bogus, ed., The Second Amendment in Law and History: Historians and Constitutional Scholars on the Right to Bear Arms. New York: New Press. Rawlings, J.O., Pantula, S.G., & Dickey, D.A. (2013). Applied regression analysis: a research tool. New York, NY: Springer. Richards, D. (2016, May 3). Georgia governor vetoes campus-carry bill. USA Today. Retrieved from: https://www.usatoday.com/story/news/nation- now/2016/05/03/georgia-campus-concealed-guns-bill/83889954/ Rosenwald, M. S. (2016, July 30). Guns go to college: Everything you need to know about campus carry. The Washington Post. Retrieved from: https://www. washingtonpost .com /news /local/wp /2016/07/30/everything-you-need-to-know- about-campus-carry/ Rothbauer, P. (2008). Triangulation. In: L. Given, (Ed.). The SAGE encyclopedia of qualitative research methods, 2, 892-994. Thousand Oaks, CA: Sage. Samuels, A. (2016, July 7). 3 Univ. of Texas professors sue to block new campus carry law. USA Today. Retrieved from: https://www.usatoday.com/story/college/2016 /07/07/3-univ-of-texas-professors-sue-to-block-new-campus-carry-law/37419429/ Scalia, A. & Garner, B. A. (2012). Reading Law: the Interpretation of Legal Texts. Minneapolis: West Publishing. Schenck v. United States, 249 U.S. 47 (1919). 164 Scheufele, D. (2000). Agenda-setting, priming, and framing revisited: another look at cognitive effects of political communication. Mass Communication & Society, 3(2), 297–316. Scheufele, D. A., & Tewksbury, D. (2007). Framing, agenda-setting, and priming: the evolution of three media effects models. Journal of Communication, 57(1), 9-20. Schwartz, B (1971). The Bill of Rights: a Documentary History. Broomall: Chelsea House Publishers. Schwoerer, L.G. (2000). To Hold and Bear Arms: The English Perspective. Chicago- Kent Law Review, 76(1), 27-60. Scott, W. (1955). Reliability of content analysis: The case of nominal scale coding. Public Opinion Quarterly,17, 321-325. Siegel, R. B. (2008). Dead or Alive: Originalism as Popular Constitutionalism in Heller. Yale Faculty Scholarship Series, Paper 1133. Retrieved: http://digitalcommons. law.yale.edu/fss_papers/1133 Smith, T.N. (2012). To conceal and carry or not to conceal carry on higher education campuses, that is the question. Journal of Academic Ethics, 10, 237-242. Solomon, S. D. (2016). Revolutionary Dissent. New York: St. Martin’s Press. Spitzer, R. J. (2000). Lost and found: researching the Second Amendment. Chicago-Kent Law Review, 76, 349-401. State v. Buzzard, 4 Ark. 18 (1842). Stevens, J. P. (2014). Six Amendments: how and why we should change the Constitution. Boston: Little, Brown Company. Stone, G. R. (2004). Perilous Times: Free Speech in Wartime. New York: W. W. Norton & Company. Stone, G. R. (2007). War and Liberty: An American Dilemma 1790 to the Present. New York: W. W. Norton & Company. Stromberg v. California, 283 U.S. 359 (1931). Suarez-Alvarez, J., Pedrosa, I., Lozano, L.M., Garcia-Cueto, E., Cuesta, M., & Muniz, J. (2018). Using reversed items in Likert scales: A questionable practice. Psicothema, 30, 149-158. 165 Sullivan, G.M., & Feinn, R. (2012). Using effect size—or Why the p value is not enough. Journal of Graduate Medical Education, 4(3), 279-282. Sunstein, C. (2008). Second Amendment Minimalism: Heller as Griswold, Harvard Law Review, 122, 246-274. Surette, R. (2014). Media, Crime, and Criminal Justice. Boston: Cengage. Sweezy v. New Hampshire, 354 U.S. 234 (1957). Tabachnick, B.G., & Fidell, L.S. (2013). Using multivariate statistics (6th ed.). Boston, MA: Allyn & Bacon. Tavakol, M., & Dennick, R. (2011). Making sense of Cronbach's alpha. International Journal of Medical Education, 2, 53-55. Tewksbury, D., & Scheufele, D. A. (2009). News framing theory and research. In J. Bryant & M. B. Oliver (Eds.), Media effects: advances in theory and research (3rd ed., pp. 17-33). New York: Routledge. Texas v. Johnson, 491 U.S. 397 (1989). Tinker v. Des Moines School Dist., 393 U.S. 503 (1969). Thompson, A., Price, J. H., Dake, J., & Teeple, K. (2013). Faculty perceptions and practices regarding carrying concealed handguns on university campuses. Journal of Community Health, 38, 366-373. Toobin, J. (2012). The oath: the Obama White House and the Supreme Court. New York: Random House. Trochim, M.K. (2010). Research methods knowledge base. Retrieved from: https://socialresearchmethods.net/kb/ Uchida, C. G. (2010). The development of American police: a historical overview. In R. G. Dunham & J. P. Alpert, eds., Critical Issues in Policing: Contemporary Readings. Long Grove: Waveland Press. United Public Workers v. Mitchell, 330 U.S. 75 (1947). U.S. Constitution, Amendment 1, 1791. U.S. Constitution, Amendment 2, 1791. U.S. v. Cruikshank, 92 U.S. 542 (1876). 166 U. S. v. Miller, 307 U.S. 174 (1939). U.S. v. O’Brien, 391 US 367 (1968). U.S. News and World Report. (2019). The Latest: House Votes to Allow Concealed Carry On Campuses. Retrieved from: https://www.usnews.com/news/best- states/west-virginia/articles/2019-02-27/the-latest-house-votes-to-allow- concealed-carry-on-campuses Vacha-Haase, T. (2001). Statistical significance should not be considered as one of life's guarantees. Effect sizes are needed. Educational and Psychological Measurement, 61, 219-224. Vasek, B. (2015). Rethinking the Nevada Campus Protection Act: future challenges and reaching a legislative compromise. Nevada Law Journal, 15, 389-430. Vertuno, J. (2016, August 2). Texas professors sue over guns on campus before class begins. The Seattle Times. Retrieved from: https://www.seattletimes.com/nation- world/texas-professors-sue-over-guns-on-campus-before-class-begins/ Volokh, E. (1998). The commonplace Second Amendment. New York University Law Review, 72(3). Waldman, M. (2014). The Second Amendment: a Biography. New York: Simon & Schuster. Waliczek, T. M. (1996). A primer on partial correlation coefficients. New Orleans: Southwest Educational Research Association. Washington Administrative Code, 478-124-020, Conduct on campus code—Prohibited conduct, 2013. Washington Administrative Code, 106-124-700, Firearms, explosives, dangerous chemicals—Restrictions, 1994. Wasserstein, R.L., & Lazar, N.A. (2016). The ASA's Statement on p-values: Context, process, and purpose. The American Statistician, 70, 129-133. Wasserstein, R., Schirm, A.L., & Lazar, N.A. (2019). Moving to a world beyond “p < 0.05”. The American Statistician, 73, 1-9. Weaver, D. H. (2007). Thoughts on agenda setting, framing, and priming. Journal of Communication, 57(1), 142-147. 167 Weir. R. M. (1969). The harmony we were famous for: an interpretation of pre- revolutionary South Carolina politics. The William and Mary Quarterly, 26(4), 473-501. Welch, R. F. (2008). King of the Bowery: Big Tim Sullivan, Tammany Hall, and New York City from the Gilded Age to the Progressive Era. Madison: Farleigh Dickinson University Press. Wieman v. Updegraff, 344 U.S. 183 (1952). Wilkinson, J. H., III. (2009). Of guns, abortions, and the unraveling rule of law. Virginia Law Review, 95(2). Wimmer, R. D., & Dominick, J. R. (2014). Mass Media Research: an Introduction. 10th Ed. Boston: Wadsworth. Winkler, A. (2011). Gunfight; the Battle Over the Right to Bear Arms in America. New York: W.W. Norton & Company. Wolcott, C. M. (2017). The chilling effect of campus carry: how the Kansas campus carry statute impermissibly infringes upon the academic freedom of individual professors and faculty members. Kansas Law Review, 65, 875-911. Wortman v. U.S., Case No. 5:14-cv-04567-PSG (N.D. Cal. May. 13, 2015). Yassky, D. (1999). U.S. v. Emerson, Amicus Curiae Brief, U.S. Court of Appeals, Fifth Circuit. Retrieved: http://www.potowmack.org/yass.html Yassky, D. (2000). The Second Amendment: structure, history, and constitutional change. Michigan Law Review, 99, 598-668. Yoo, W., Mayberry, R., Sejong, B., Singh, K., He, Q., & Lillard, J.W. (2014). A study of effects of multicollinearity in the multivariable analysis. International Journal of Applied Science and Technology, 4(5), 9-19. Young, J. (2007). Statistical errors in medical research - a chronic disease. Swiss Medical Weekly, 137, 41-43. Zaller, J. R. (1992). The nature and origins of mass opinions. New York: Cambridge University Press. 168