USF Libraries
USF Digital Collections

An analysis of print media reporting of established religions and new religious movements

MISSING IMAGE

Material Information

Title:
An analysis of print media reporting of established religions and new religious movements
Physical Description:
Book
Language:
English
Creator:
Griffin, Lonnie F
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
sociology
media logic
violence
descriptor
theme
angle
education
Dissertations, Academic -- Sociology -- Masters -- USF   ( lcsh )
Genre:
government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: This thesis reports findings from a content analysis of 720 newspaper articles and 3,052 newspaper article cases focused on the issue of print media bias. Sunday editions of three major newspapers were drawn from the six-year period 1998-2003 for analysis. Prior research has uncovered print media bias in reporting of religious groups, and this thesis examines the substance of those claims pertaining to both established religions and new religious movements. Research findings show that established religions and their members are typically described in favorable or neutral terms, while new religious movements and their members are consistently described with pejorative terms. However, specific established religion members received the overwhelming majority of negative religion member descriptors. Articles focusing on established religion members were found to contain the bulk of visual aides accompanying the articles. Newspaper articles discussed incidents of violence by and/or against specific religious groups of both types of religion with a high frequency. Also, newspaper article themes and angles were found to be important for conveying the content of the articles. Additionally, an appendix is included that analyzes the treatment of religion, established religions and their members, and new religious movements and their members in sociology textbooks.
Thesis:
Thesis (M.A.)--University of South Florida, 2004.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Lonnie F. Griffin III.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 119 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001469374
oclc - 55644049
notis - AJR1128
usfldc doi - E14-SFE0000276
usfldc handle - e14.276
System ID:
SFS0024971:00001


This item is only available as the following downloads:


Full Text

PAGE 1

An Analysis of Print Media Reporting of Established Religions and New Religious Movements by Lonnie F. Griffin, III A thesis submitted in partial fulfillment of the requirements for the degree of Master of Arts Department of Sociology College of Arts and Sciences University of South Florida Major Professor: James Cavendish, Ph.D. Danny Jorgensen, Ph.D. Christy Ponticelli, Ph.D. Date of Approval: March 30, 2004 Keywords: sociology, media logic, violence, descriptor, theme, angle, education Copyright 2004, Lonnie F. Griffin, III

PAGE 2

Dedication It is with great joy that I dedicate this th esis to my parents, my wife Rose, and the memory of George B. Ward, Sr.

PAGE 3

Acknowledgments I am indebted to a number of people who have contributed to the successful completion of this thesis. Certainly foremost among them is my committee chair, Dr. James Cavendish, who has given inordinate amounts of his time and energy to aid and assist me during the course of my graduate studies. I have learned a great deal from Dr. Cavendish, and I appreciate the respect, concern, and mentoring he has shown me. I am also grateful to Dr. Danny Jorgensen and Dr. Christy Ponticelli who graciously agreed to be members of my thesis committee. Their comments and questions certainly helped to focus the direction of my research and the final thesis. Although a number of faculty members have encouraged me in conducting this research, I am particularly indebted to Dr. Maralee Mayberry for providing me with the opportunity to pursue research that culminated in Appendix One. I am also indebted to Dr. Donileen Loseke who graciously provided her unpublished research and guided the development of my initial thesis proposal. For generating my interest in the sociology of religion, I am forever indebted to Rev. Neil Gray.

PAGE 4

Table of Contents List of Tables ii List of Figures iv Abstract v Chapter One: Introduction 1 Statement of Purpose 2 Literature Review 2 Prior Research 3 Justification for Research 5 Theoretical Issues 7 Hypothesis and Prior Findings 11 Methodological Issues 13 Established Religions and New Religious Movements 23 Chapter Two: Research Findings: Articles 24 Research Findings: Article Cases 37 Results and Discussion 48 Policy Implications and General Conclusions 57 References 61 Appendices 64 Appendix A: A Textbook Analysis of Religion 65 Appendix B: Characteristics of Prestige Newspaper Sample 101 Appendix C: Coding Protocol for Print Media Reporting of Religion 108 i

PAGE 5

List of Tables Table 1: Religions of the World by Percentage of Population, 1996 23 Table 2: Cross Tabulation Between Prestige Newspapers and Primary Article 24 Table 3: Cross Tabulation Between Newspaper Article Prominence and Primary 25 Table 4: Cross Tabulation of Origin of Article and Primary Article Subject 26 Table 5: Means of Article Word Counts by Primary Subject of Articles 28 Table 6: Lambda Results of Primary Subject of Article and Article Word Count 28 Table 7: Article Primary Subject and Number of Accompanying Visual Aides 29 Table 8: Cross Tabulation Between Theme and Primary Article Subject 30 Table 9: Cross Tabulation of Primary Article Angles and Primary Article Subjects 32 Table 10: Cross Tabulation Between Violence by Religious Groups and Primary 34 Table 11: Cross Tabulation Between Violence Against Groups and Members 36 Table 12: Gamma Results of Violence by Groups and Violence Against Groups 36 Table 13: Cross Tabulation Between Violence by Religion and Religion Group 40 Table 14: Cross Tabulation Between Violence Against Religion Groups 41 Table 15: Gamma Results of Violence by Groups and Violence Against Groups 42 Table 16: Frequency of Religion Group Descriptors 43 Table 17: Frequency of Religion Member Descriptors (1) 43 Table 18: Frequency of Religion Member Descriptors (2) 43 Table 19: Religion Group Descriptors and Frequency Between Established 46 Table 20: Religion Member Descriptors and Frequency Between Religion Types (1) 47 Table 21: Religion Member Descriptors and Frequency Between Religion Types (2) 48 ii

PAGE 6

Table 22: Frequency of Established and New Religion Groups and Members 77 Table 23: Mean of Ratios of Religious Groups and Members for Textbook Sample 78 Table 24: Frequency of Actual and Potential Violence By and Against Groups 80 Table 25: Prestige Newspaper Frequency and Percentage of Sample Articles 101 Table 26: Cross Tabulation Article Prominence and Prestige Newspaper 102 Table 27: Cross Tabulation Article Origin and Prestige Newspaper 104 Table 28: Word Count of Newspaper Article Information 105 Table 29: Newspaper Article Information About Visual Aides 106 iii

PAGE 7

List of Figures Figure 1: Chapter Features for Chapters on Religion and Other Social Institutions 76 Figure 2: Textbook Frequency of Actual and Potential Violence by Religion 82 Figure 3: Textbook Frequency of Actual and Potential Violence Against Religion 85 Figure 4: Mean Frequencies of Angle Categories in Textbook Sample 87 Figure 5: Mean Frequencies of Textbook Angles for Established and New Religions 89 Figure 6: Mean Frequencies of Textbook Angles for Established and New Religions 90 Figure 7: Mean Frequencies of Group Descriptors for Textbook Sample 92 Figure 8: Mean Frequencies of Group Descriptors According to Religion 93 Figure 9: Mean Frequencies of Member Descriptors for Textbook Sample 96 Figure 10: Mean Frequencies of Member Descriptors According to Religion 97 iv

PAGE 8

An Analysis of Print Media Reporting of Established Religions and New Religious Movements Lonnie F. Griffin, III ABSTRACT This thesis reports findings from a content analysis of 720 newspaper articles and 3,052 newspaper article cases focused on the issue of print media bias. Sunday editions of three major newspapers were drawn from the six-year period1998-2003 for analysis. Prior research has uncovered print media bias in reporting of religious groups, and this thesis examines the substance of those claims pertaining to both established religions and new religious movements. Research findings show that established religions and their members are typically described in favorable or neutral terms, while new religious movements and their members are consistently described with pejorative terms. However, specific established religion members received the overwhelming majority of negative religion member descriptors. Articles focusing on established religion members were found to contain the bulk of visual aides accompanying the articles. Newspaper articles discussed incidents of violence by and/or against specific religious groups of both types of religion with a high frequency. Also, newspaper article themes and angles were found to be important for conveying the content of the articles. Additionally, an appendix is included that analyzes the treatment of religion, established religions and their members, and new religious movements and their members in sociology textbooks. v

PAGE 9

Chapter One Introduction The print media and religious organizations have a symbiotic relationship. The print media is an integral part of the American mass communication system (Wright 1997). As part of the mass communication system, the print media requires news items worthy of publication. News items and stories are required that will capture the attention of the reading public, yet inform and express the public consciousness of the reading public if possible (Altheide 1991 and 1995; Hoover 1998; ODonnell 1993). Reporting of religion is, and has proven to be, a topic of interest to the reading public (Haynes 2000). However, with disturbing frequency, media coverage of new or nontraditional religions has been challenged by scholars as inflammatory, distorted, and infused with disparaging stereotypes (Wright, p. 101). Religious organizations have a spiritual and moral message they are concerned with sharing with other people, and especially for established religions, the print media has been a vehicle for conveying their messages and humanitarian deeds to the general public (Haynes 2000; Hoover 1998). Historically, established religions and their members could expect deferential print media coverage. However, As dominant religious institutions have declined in legitimacy so has the tendency to treat them (and indeed, the whole of the religion story) with deference (Hoover, p. 28). For new religious movements, the print media is often viewed as either a tool for religious legitimation or as an enemy to both feared and confronted (Richardson and van Driel 1997; Wright 1997). 1

PAGE 10

Because the print media, which is business oriented and often corporate sponsored, follows the trends in the marketplace, increasing news coverage of religion has focused on new religious movements and deviance in religion, and reporting of these movements tends toward the sensational, emotional, and judgmental (Hill, Hickman, and McLendon 2001; Richardson and van Driel 1988 and 1997; Wright 1997). Additionally, Recognition that religion is now more newsworthy than at any time in post-war American history has increased both the salience and amount of news coverage devoted to it in major news organizations (Hill, Hickman, and McLendon, p. 24). It is this context of increased attention and reporting of all religious groups that concerns my thesis. Statement of Purpose The purpose of my thesis is to examine and analyze how three prestigious newspapers, The New York Times, The Washington Post, and The Los Angeles Times, portray and describe two types of religious institutions, established religions and new religious movements, in their newspaper articles. I also want to examine and analyze how these newspaper articles employ media logic in presenting and transmitting information concerning established religions and new religious movements. Literature Review My literature review will center on establishing and signifying the relevance of my thesis. Specifically, journal articles, papers, and books are used as references and guides to provide not only a theoretical and methodological foundation, but also an operationalization of variables and concepts, and a justification for my thesis. 2

PAGE 11

Prior Research Concerning print media reporting of religion, Hoover states, By the latter half of the 19 th century, the world of religion had gradually become used to the idea that it might be scrutinized by the secular press (p. 20). Over time, the coverage of religion in the print media evolved from relatively quiet and noncontroversial coverage, presenting religion without subjecting it to intense scrutiny, to the emergence of religious controversies arising in the 1970s and onward, focusing on religious beliefs, practices, and groups critically (Hoover 1998). Coinciding with the start of this critical phase of print media reporting of religion, scholars of religion continued to study and research the published newspaper articles dealing with religion and journalists responsible for covering the religion beat (Hoover 1998; ODonnell 1993; Wright 1997). At the same time that the print media has become more critical of religion, social scientists have taken notice and initiated studies of the treatment of religion in the mass media (Haynes 2000; Hoover 1998). This newfound interest in the reporting of religion and the journalists themselves also came to be evaluated in terms related to new religious movements (Hill, Hickman, and McLendon 2001; Hubbard 1998; Richardson and van Driel 1988 and 1997). Much research focuses on print media journalists coverage of and attitudes toward established religions and new religious movements (Hill, Hickman, and McLendon 2001; Hoover 1998; ODonnell 1993; Richardson and van Driel 1988 and 1997; Wright 1997). In terms of coverage, these researchers have found that established religions were given primacy by journalists during the establishment era of the middle 20 th century, and even in the present, religion is considered to be parochial, which means 3

PAGE 12

that the religion beat pertains to and covers the religious interests of the people locally (Hoover, p. 58). Because established religions dominate society in regards to number of members, resources, and prestige in comparison to new religious movements, this parochial notion of religious coverage tends to report established religions and members over more marginalized new religious movements and members. A problem does arise concerning print media coverage of religion, for Stuart Wright states, Religious illiteracy among reporters is an acute problem recognized within the professional ranks of journalism (p. 101). Wright is referring to illiteracy as it concerns having a good understanding of what established religions entail, which leads him to speculate about how great the problem must be in relation to journalists understanding of new religious movements (Wright, p. 102). Researchers have also examined print media journalists attitudes and biases in reporting of new religious movements. Anti-new religious movement sentiment has been documented in a number of studies (Hill, Hickman, and McLendon 2001; Hubbard 1998; Richardson and van Driel 1988 and 1997; Wright 1997), and Basically, these studies confirm the general observations about the way media treats deviant groups: they are generally discredited and delegitimized (Richardson and van Driel, p. 119). Some journalists undoubtedly find nontraditional religious practices to be strange or culturally offensive, which can lead to negative reporting, and the print media is solely concerned with the portrayal of new religious movements only when conflicts are at issue (Wright, p. 103). Hill, Hickman, and McLendon have studied and documented print media bias against new religious movements using content analysis of newspaper and wire service articles, and they report that more mainstream religious groups are typically described in 4

PAGE 13

neutral or favorable terms, while new religious movements are consistently described in pejorative language (p. 24). Justification for Research My thesis was constructed as a project to replicate similar research and contribute new research findings in the area of print media bias as it concerns religious groups and members. From prior research findings, it is apparent that the notion of a value-free print media is in question if not outright challenged (Hill, Hickman, and McLendon 2001; Hoover 1998; Richardson and van Driel 1988 and 1997; Wright 1997). To determine if these prior research findings are still valid, in light of recent events such as clergy sexual abuse scandals, religious groups millennial and post-millennial activities, and the September 11, 2001 terrorist attacks, my thesis is centered on gathering data drawn directly from print media sources over the last six years and analyzing it in relation to established and new religions and their members. The extant literature provides several justifications for the importance of researching the coverage and extent of print media bias in relation to religious groups and members. Firstly, there are civic reasons tied directly to the First Amendment. As echoed in ODonnells article, Haynes states, The First Amendment guarantees freedom of the press and religious liberty, but it doesnt require that the press treat religion fully or even fairly (p. 84). If religious groups and members are given inadequate or poor media coverage, their mutual understanding and respect is not fostered, and members of religious groups are disadvantaged in the body politic and denied a voice (Haynes, p. 85). Secondly, print media bias may influence or reinforce peoples perceptions, for there exists: 5

PAGE 14

a more widespread perception that somehow new religious movements are not really part of the religious mainstream, emphasizing as they do millennial expectations, material and physical benefits, miraculous interactions with the spirit world, social and political activism, rejection of familial ties in the forging of new community and institutional relationships, and other things deemed not properly within the domain of religious thought and practice. (Hubbard, p. 60) Additionally, David Altheide states, Leaders and policymakers are very sensitive to mass media presentations about problems and issues, for leaders and policymakers give serious attention to media legitimated issues that can influence constituents and shy away from more marginal and unpopular issues (1995, p. 137). This raises political implications for fair treatment of religious groups that have been given biased treatment in the print media. Altheide also notes, Another reason why media messages are so important for policymakers is that they often learn about issues from news sources (1995, p. 138). Thirdly, the print media is often not objective and passive in reporting social conflicts. Some researchers believe that the media serves the ideological interests of the dominant, hegemonic groups in society and promotes their approach to issues (Richardson and van Driel 1988 and 1997). Established religions are dominant religious groups in society, and by coding the print media sample for evidence of bias, it is possible to gauge the level of objectivity and identify the ideology expressed in a print media article. Lastly, in its most insidious manifestation, media bias against religious groups and members is an important issue for research due to the fact that these print media articles can be distorted, exaggerated and then repeated, fomenting collective anxiety and fear which contribute, in extreme situations, to destructive actions such as mob violence (Wright, p. 104). Also problematic is the litigious culture of America. There 6

PAGE 15

are large numbers of false claim lawsuits generated against religious groups due to increased media reporting of civil or criminal court proceedings against these religious groups (Glassner, p. 37). Although these same reasons motivate me in doing this study, I am additionally motivated by three other factors. I want to see if recent religion related events, such as the clergy sexual abuse scandals, religious groups millennial and post-millennial activities, and the September 11, 2001 terrorist attacks, have led to the emergence of new media bias against mainstream denominations. Specifically, I want to examine whether the reporting of religious inspired violence and socially constructed villains, such as Islamic terrorists and Catholic pedophile priests, has increased since the time of these events. If so, has the reporting of violence and villains increased at the same rate among news stories of established religions as among news stories of new religious movements? Theoretical Issues This thesis is located within the symbolic interactionist perspective. This perspective maintains that reality is socially constructed through a process of interaction through which symbols are interpreted and given meaning. As Altheide states: The renewed interest in culture and the symbolic systems and processes through which social order is constructed, constituted, defined, interpreted, and enacted calls for an expansive perspective incorporating the processes of communication, interpretation, and meaning. (1995, p. 100-101) The symbolic interactionist perspective informs media logic. This thesis will employ elements of David Altheides theory of media logic by applying this theoretical framework to the issue of print media bias concerning established and new religions and their members. Specifically, media logic is an 7

PAGE 16

explanation of how the general guiding assumptions and principles that govern the interaction between audience members, a medium (technology), and a subject or topic interact (1995, p. 182). Altheides theory focuses on the media themselves, as organizations that have developed their own techniques, logic, procedures, and rationale for engaging in rational-bureaucratic communication (1985, p. 12). Material is continually selected in the modern (and postmodern) age on the basis of what can be done with it, whether it has the necessary requirements for the medium, and its purpose (e.g. commercialism, entertainment) (1995, p. 182). Concerning what is newsworthy, Altheide states, Is not the particular act or deed itself so much as what can be done with it, how it corresponds to the criteria of accessibility, visual quality, drama, action, audience relevance, and ease with which it can be encapsulated and given a thematic unity (1985, p. 20). Specifically, when a media logic is employed to present and interpret institutional phenomena, the form and content of those institutions are altered (1991, p. 9-10). Using this perspective allows an examination and analysis of print media articles in which descriptors, visual aids, news angles, and news themes delineate types of religions, either established or new, and their coverage, either positive or negative. To capture the specific usage and meaning of concepts and variables contained within my thesis requires adequate definitions, or operationalization. The essential concepts of David Altheides theory of media logic that are used throughout this thesis are described below. Media Logic. In general terms, media logic consists of a form of communication, the process through which media present and transmit information. Elements of this form include the various media and the formats used by these media (Media Worlds, p. 9). 8

PAGE 17

Format. Formats are rules and procedures for defining, selecting, organizing, and presenting information and other experiences (Media Power, p. 9). Format becomes a framework or a perspective that is used to present as well as interpret phenomenon (1995, p. 182). Medium. Newspapers are a medium. Altheide states, A medium is any social or technological procedure or device that is used for the selection, transmission, and reception of information (Media Worlds, p. 10). Manner. Manner refers to the way subject matter is operationally presented. Manner is the specific instance being presented, the example, the identifiable feature of the story line, what the message is about (Media Power, p. 20). Themes. Themes are general statements or images that may be used to connect one report to another to provide unity to the topics being presented (Media Worlds, p. 141). Angles. Angles are like themes, except they are used to illustrate certain themes that may or may not be made explicit (Media Worlds, p. 142). The angle facilitates placing unique occurrences in a broader context and, in a sense, rendering meaning by association (Altheide and Johnson, p. 68). David Altheide has developed these theoretical concepts of media logic over a twenty-five year period. His four books included in my thesis explain, develop, and utilize this theory. Altheide is mostly concerned with using his theory to analyze the television medium, but his theory is hypothetically applicable to any medium and my thesis will provide a demonstration of this theorys practical application to the print media, newspaper, medium. Concerning the use of media logic in relation to religion, Altheide has conducted research on the televangelism phenomenon utilizing media logic 9

PAGE 18

(Media Worlds), and he has conducted ethnographic research on organized religion that focused on the role of counselors connected with the Billy Graham Evangelical Association (Altheide and Johnson). Also pertinent to this thesis is Altheides use of media logic to research and analyze media reporting of terrorists and terrorism (Media Power and 1995). Of interest to social movement scholars and researchers familiar with the concept of frame is Altheides format concept. David Snow and Robert Benford note that the concept of frame: refers to an interpretive schemata that simplifies and condenses the world out there by selectively punctuating and encoding objects, situations, events, experiences, and sequences of actions within ones present or past environment. (Snow and Benford, p. 137) The utility of the frame concept is evident in everyday life and academia, and many disciplines apply this concept to practical concerns: sociology, psychiatry, the humanities, and cognitive psychology (Snow and Benford, p. 136). For Altheide, format is different from the concept of frame because format refers to the way communication is organized (1985, p. 14). Altheide states that: formats are metacommunication statements, or rules for the recognition, organization, and presentation of information and experience. Their power resides in employing a logic that joins the manner of specific acts to a temporal and spatial order. (1995, p. 38) Concerning formats, social order and social control are predicated on communication forms and processes (Altheide 1995). Additionally, I will provide analysis of the coded data in relation to the concept of ideology. The concept of ideology was raised in Richardson and van Driels 1988 and 1997 articles; however, while utilized in relation to the sociology of education, Jean 10

PAGE 19

Anyons conception of ideology is an excellent one and it is easily operationalized to suit the purposes required for this thesis. Anyon states: Ideology is defined here as an explanation or interpretation of social reality which, although presented as objective, is demonstrably partial in that it expresses the social priorities of certain political, economic, or other groups. Ideologies are weapons of group interest; they justify and rationalize; they legitimate group power, activities, and needs. (Anyon, p. 363) Defined in this way, it becomes possible to analyze and explore the newspaper articles for evidence of ideologies being presented in reference to established and new religions and their members, and if ideological evidence is present, an analysis of the ideologies will be conducted focusing on their justifications and rationalizations. Hypothesis and Prior Findings Are established religious groups and new religious movements subject to different treatment in news coverage (Hill, Hickman, and McLendon, p. 26)? This research question, and its focus on inequality, provided the basis for my formation of specific hypotheses. In the formation of my hypotheses for this thesis, relevant research findings from other studies have been integrated into this research project. These hypotheses are grounded in previous scholarship that has revealed anti-new religious movement bias (Hill, Hickman, and McLendon 2001; Hoover 1998; Hubbard 1998; ODonnell 1993; Richardson and van Driel 1988 and 1997; Wright 1997), and these hypotheses seek to replicate and add new findings to the social scientific study of religion in this area. I hypothesize that print media reporting of new religious movements will be more negatively biased, critical, and prone to use negative descriptors than reporting that is concerned with established religions. This hypothesis is based upon prior research 11

PAGE 20

findings that have documented negative bias, critical treatment, and negative descriptors used in discussion of new religious movements (Hill, Hickman, and McLendon 2001; Richardson and van Driel 1988 and 1997; Wright 1997). I also hypothesize that print media articles focusing on new religious movements will: receive less prominent placement, have lower word counts, have fewer, if any, accompanying photographs, and center more around a theme of violence, either potential or actual. This hypothesis stems from the research findings of Hill, Hickman, and McLendon (2001), and this hypothesis is also based on testing elements of David Altheides theory of media logic concerning the importance of formats (1985; 1995; Altheide and Johnson 1980; Altheide and Snow 1991). I hypothesize that the print media articles will employ, with a high frequency, themes and angles other than general information for both established and new religions. This hypothesis is based on Altheides prior research using media logic where multiple themes and angles other than general information were used by news mediums to capture audience attention (1985; 1995; Altheide and Snow 1991). I hypothesize that established religions will receive more coverage, mentions of specific groups and members, than new religious movements and members in the print media articles on religion. This hypothesis is based on the findings of prior research collected for this thesis related to bias (Hill, Hickman, and McLendon 2001; Hoover 1998; Hubbard 1998; ODonnell 1993; Richardson and van Driel 1988 and 1997; Wright 1997). I hypothesize that the conflict theme, the violence angle, and actual and potential violence by/against religious groups and members will occur with a high frequency and be an important theme, angle, and issues discussed in the newspaper articles on religion. This 12

PAGE 21

hypothesis stems from the research findings of Hill, Hickman, and McLendon (2001) and David Altheide (1995). Methodological Issues For analysis of the data, I followed a classical content analysis model. Non-numerical written records of communication are reduced to quantitative data. This method is applicable for this inquiry because previous similar research has utilized this method (Hill, Hickman, and McLendon; Richardson and van Driel 1988), it is non-reactive, and it avoids any ethical considerations. The random sample to be examined for this study will include newspaper articles that explicitly meet this definition. Due to the constraints posed by a research project of this scope, purposive sampling is also used. Riffe, Lacy, and Ficos book proved to be a valuable resource for providing concise information on these types of research methods, specifically in the areas of selecting a sample, creating a coding protocol, and operationalizing variables (1998). It is necessary to operationalize the specific definitions for the items that are the focus of this research project. Prestige newspapers are self-defined as national in scope and they possess international news-gathering ability (Riffe, Lacy, and Fico, p. 86). Furthermore, prestige newspapers have circulations in the millions and are important for dictating what is news for other news organizations (Altheide and Johnson; Hill, Hickman, and McLendon 2001). For my thesis, The New York Times, The Washington Post, and The Los Angeles Times have been selected as proper representations of prestige newspapers. This selection was considered proper for replication purposes because previous studies used these same prestige newspapers (Hill, Hickman, and McLendon 2001; Hoover 1998; Richardson and van Driel 1988). 13

PAGE 22

Newspaper articles are defined as all non-advertising matter in a newspaper. Consistent with Hill, Hickman, and McLendons study (2001), this includes all staff-produced news articles found in the first and local sections, but excluding editorial pages, op ed pages, reader opinions, sports, routine business data, and society news. It may include relevant features produced by local staff reporters and syndicated and wire services articles relevant to the issue of religion being analyzed (Hill, Hickman, and McLendon 2001). The Hill, Hickman, and McLendon article was the impetus for my interests concerning this issue of print media bias. Originally, I planned to conduct a replication of their study with a purposive sample drawn from different time periods other than the six-month sample they employed; however, I soon generated ideas that would allow me to conduct research that would embark on new territory. Essentially, I designed a coding protocol that would code for data similar to the data coded for by Hill, Hickman, and McLendon (2001). By this, I mean that my coding protocol would code for religious group and member descriptors, references to potential or actual violence, average number of words per article, and total number of articles per source (Hill, Hickman, and McLendon 2001). There is a noticeable lack of sociological theory accompanying the research articles in my references; however, due to my inclusion of elements of sociological theory, I am able to expand my research project and code for more variables. My research adds to Hill, Hickman, and McLendon (2001) and other literature in this area because I ground it in sociological theory. The introduction of new variables, made possible for study due to the inclusion of elements of David Altheides theory of media logic, opened 14

PAGE 23

up several unexplored areas for analysis: prominence of the newspaper article, the inclusion or exclusion of accompanying visual aids, and the newspaper articles usage of themes and angles. These new variables provide opportunities for further analysis and serve to expand the usage of David Altheides theory of media logic to an analysis of religious groups and members in the print media. For my own data set, I explored a much larger segment of time than Hill, Hickman, and McLendon (2001) do and coded for a larger cross-section of religion related articles. I extended my data set over a six-year period starting in 1998 and ending in 2003. This was done for several reasons: to capture religion related articles leading up to the millennium just as Hill, Hickman, and McLendon (2001) did; to examine print media coverage of established and new religions and their members over an extended period of time; to trace the frequency of certain socially constructed villain descriptors and their accompanying news themes and angles; and to provide a contemporary accounting of religion reporting. For the six years of my sample, I randomly sampled three months from each year for a total of eighteen months. I achieved this by dividing each year into three four-month sections, January through April, May through August, and September through December, and I then assigned each month in the sections a value of one through four. I rolled a four-sided die to randomly generate the specific month in each section I coded. This was done to prevent clustering of the data and ensures adequate dispersion of the months sampled for each year. For the eighteen months sampled, I gathered prestige newspaper articles from seventy-two different Sundays. Within these Sunday editions, I read and coded every article dealing with religion groups and members. In the interest of 15

PAGE 24

collecting as large a sample of relevant newspaper articles as possible in a timely manner, I employed purposive sampling to sample Sunday editions of my prestige newspapers. The Sunday edition not only serves as a forum for discussing news items presented in articles earlier in the week, but it is also generally the largest newspaper printed during the week, thereby making it an ideal forum for the presentation of more, and longer, news articles. For data recording and analysis, I used SPSS, the Statistical Package for the Social Sciences. I chose SPSS due to its ease of use and its ability to store, analyze, and display large amounts of data. I used ProQuest, an on-line, full-text newspaper provider, to generate the articles for my sample. After randomly determining the month to be selected from a section in a given year, I consulted a calendar to gather the precise days of that month that each Sunday fell on. With the full dates gathered, month/day/year, I was able to search each specific Sunday in the ProQuest database. ProQuest allows its users to conduct specific word searches of multiple newspapers at the same time for a given date. By using ProQuest, I was able to read the full-text of Sunday newspaper articles and collect my sample data set of 720 relevant newspaper articles, and I was able to conduct keyword searches to search the articles for descriptors relating directly to my specific coding protocol, which produced a second data set of 3,052 specific cases. Additionally, ProQuest provides accompanying information relevant to the article: newspaper name, article title, author, location in the paper, word count, and visual aids and quantity. Concerning operationalization for established religions and new religious movements, Religion in the Modern World (1996) by Steve Bruce uses a religious 16

PAGE 25

typology where established religious groups in America are classified as denominations and new religious movements are classified as cults. For clarification, it should be noted that the term cult, which is often used disparagingly, is virtually synonymous and interchangeable with the more value-free term new religious movement (Bruce 1996). Using these typologies, established religious groups possess and/or manifest the characteristics of the denomination, and new religious movements possess and/or manifest all or many, due to their highly individualistic nature, of the characteristics of the cult. The use of religious typologies is helpful for gaining understanding of what essential characteristics constitute established religious groups and new religious movements, and the use of religious typologies provides the important religious, social, and political distinctions that are lacking without including typologies. Established religions, or denominations, are religious bodies with a professional clergy. They are relatively undemanding, large, inclusive religious bodies associated with the comfortable classes, and while people may join, many people are members because they were born into the group (Bruce, p. 75). New Religious Movements, or cults, are small, loosely knit religious groups that are highly individualistic, meager in resources, often short-lived, unable to command members obedience, and lacking a sharply defined and exclusive belief system (Bruce, p. 82). In their investigation of religious group and member descriptors, Hill, Hickman, and McLendon (2001) found 19 terms used, more than once, to describe religious groups and 24 terms used, more than once, to describe members of religious groups (p. 29). Operating under the premise that all descriptive terms in an article were important 17

PAGE 26

because they offered the intellectual context within which journalists and readers evaluated the religious group or news story that was the dominant subject of the news article, all terms used as descriptors for religious groups and members in the sample of articles are recorded, and specific attention is concentrated on the frequencies with which favorable, neutral, and unfavorable terms were used to describe religious groups and their members (Hill, Hickman, and McLendon, p. 29). Concerning specific descriptors, I coded for the same specific descriptors as used by Hill, Hickman, and McLendon (2001). For religious groups, I used cult, sect, movement, group, church, religion, and an other category, and for religious members, I used terrorist, follower, member, and an other category (Hill, Hickman, and McLendon, p. 29). However, I decided to leave out one descriptor from the group category, denomination, and one descriptor from the member category, pilgrim, due to their absence or low frequency (Hill, Hickman, and McLendon, p. 29). My coding protocol looked for specific, unique descriptors that were not coded by Hill, Hickman, and McLendon (p. 29). Among them are the group descriptor faith and the member descriptors militant, adherent, pedophile, worshiper, leader, and an official religious title. I also included descriptors that did not appear in their sample, the group descriptor congregation and the member descriptors pedophile, leader, and an official religious title. These inclusions were made to record more descriptors and limit the frequency of the other category responses. Due to the time frame in which these articles appeared, I assumed that some of these descriptors would appear. Specifically, I was interested in coding for the member descriptors terrorist and pedophile. Due to the relatively recent social construction of these two villain 18

PAGE 27

descriptors (Loseke 2003 and Unpublished; Glassner 1999), I believe that my research project is longitudinal enough to capture the emergence of these descriptors in the prestige newspaper articles. Mass media reporters, editors and producers are well aware that social problems sell in the mass media market place (Loseke 2003, p. 9), and because of this phenomenon, I believe that these two villain descriptors will frequently appear in the sample of newspaper articles. Additionally, Loseke notes that there is an absence of research surrounding the issue of how stories, newspaper articles, about these villain characters, religious group members, are produced (Loseke, Unpublished), and I believe that my research project has the theoretical and methodological capability to make an initial foray into this area. The coding protocol used for this thesis is included in Appendix Three: Coding Protocol for Print Media Reporting of Religion. Based on my operationalization of components of David Altheides theory of media logic (Media Worlds, p. 9), I developed four exclusive categories of themes and eight exclusive categories of angles to code for in the prestige newspaper articles (Media Worlds, p. 141-142; Altheide and Johnson, p. 68). The primary article theme was designated by what was the primary topic of discussion for religion, religious groups, and religious members, and the primary article angle was designated by what incidents were used to illustrate the articles discussion of religion, religious groups, and religious members. The primary article themes used for my thesis are: conveying general information/news, ideological discussion, conflict, and sex. The theme of conveying general information/news was used to capture all articles that primarily focused on informing readers and the dissemination of information (Altheide 1995 and 19

PAGE 28

Altheide and Snow; Hoover 1998; ODonnell 1993). The conflict theme emerged from David Altheides research findings using media logic (Altheide and Johnson; Altheide and Snow; 1985; 1995). The ideological discussion theme emerged from prior research findings (Richardson and van Driel 1997; Wright 1997). The sex theme emerged from an initial test protocol and prior research (Loseke 2003 and Unpublished). These four themes were also selected due to their ability to provide continuity by connecting articles containing these themes with other similar articles. The theme of ideological discussion refers to articles that employed rationalizations and justifications of the social priorities and beliefs of established and new religions and their members to legitimate their groups power, activities, and needs (Anyon, p. 363). The theme of conflict relates to articles that focus on opposition, disharmony, warfare, and/or violence involving established and new religions and their members (Altheide 1995). The theme of sex refers to articles that centered on sexuality, sexual activity, and/or sexual abuse (Loseke 2003 and Unpublished; Glassner 1999). The primary angles used for this thesis are: official pronouncements, practice of worship, information, promotion, apocalyptic and/or millennial beliefs, clergy as victimizers, violence, and an other angle category. I developed these specific, unique angles by drawing on ideas generated by prior research and by refining an initial test protocol (Anyon 1979; Altheide 1995; Altheide and Johnson; Altheide and Snow; Hill, Hickman, and McLendon 2001; Hoover 1998; Loseke 2003 and Unpublished; ODonnell 1993). Angles are important from a media logic standpoint because they provide evaluation, through the manner of presentation, to the material being presented. By providing an implicit or explicit illustration of a theme, angles aid readers of these 20

PAGE 29

articles in placing these articles into a broader context and associate meaning between similar articles employing these angles (Media Worlds, p. 142; Altheide and Johnson, p. 68). The official pronouncements angle refers to articles focusing on the officially sanctioned dogma, rules, and/or announcements related to religious groups or their members. The practice of worship angle refers to articles focusing on the rites, general beliefs, and rituals practiced by religions, religious groups, and their members. The information angle refers to articles devoted to education and presenting, discussing, and/or explaining items, concepts, and/or religious groups and their members (Altheide 1995 and Altheide and Snow; Hoover 1998; ODonnell 1993). The promotion angle refers to articles that explicitly or implicitly promote religion in general, some aspect or feature of religion, and/or a specific religious group or members of a religious group (Anyon, p. 363). The apocalyptic and/or millennial beliefs angle specifically illustrates beliefs of this sort, such as a final battle between good and evil or the return of Christ or a messianic figure at the millennium. I wanted to code for this angle because this sample spans the turn of the millennium, is a component of many new religious movements, and relates to the Hill, Hickman, and McLendon (2001) research. For article angles focusing on instances in which clergy members or religious leaders were the perpetrators of crimes against church members or other persons, the clergy as victimizers angle was used to capture articles that employed this angle. An example could be the discussion of clergy members sexually abusing their church members (Loseke 2003 and Unpublished; Glassner 1999). The violence angle refers to articles that are concentrated on violence, 21

PAGE 30

actual or potential, and an example could be an illustration of the Jewish Holocaust (Altheide 1995; Hill, Hickman, and McLendon 2001). The other coding category was utilized as a variable to capture any article angle that was not applicable to the other variables. For religious group and member descriptors, I coded each instance of a specific descriptor mentioned in relation to a specific religious group or member, established or new, occurring in the newspaper articles on religion. Bruce (1996) discusses religious typologies and terms in his book, which clarifies the status of specific, unique group descriptors and the information these descriptors convey. The group descriptors utilized for this research project can be grouped according to whether they convey legitimation: church, religion, congregation, neutrality: faith, movement, group, or illegitimacy: cult, sect (Bruce 1996; Hill, Hickman, and McLendon 2001). The meaning that these group descriptors convey is based on prior research, specifically Hill, Hickman, and McLendon (2001), which has used this grouping. For religious member descriptors, the religious group member descriptors utilized for this thesis can be grouped according to whether they convey legitimation: an official religious title, leader, worshiper, neutrality: adherent, follower, member, or illegitimacy: terrorist, militant, pedophile (Hill, Hickman, and McLendon 2001; Loseke 2003 and Unpublished; Glassner 1999). The meaning that these member descriptors convey is based on prior research, specifically Hill, Hickman, and McLendon (2001), which has used this grouping. 22

PAGE 31

Established Religions and New Religious Movements It is essential to this study that the number of members and groups of established religions and new religious movements be accounted for so claims of overrepresentation, under representation, and bias, if present in the data set, can be made. In Anthony Giddens, Mitchell Duneier, and Richard Appelbaums Introduction to Sociology, Fourth Edition (2003), these authors present U.S. Bureau of the Census data on religions of the world by percentage of the population, 1996 (p. 545). These data are presented in Table 1 below. For this study, the findings presented for North America, which is the location of our prestige newspaper sample, will be included as well as the worldwide findings, since these prestige newspapers have international reporting capability. Concerning the number of religious groups, There are more than 280 religious denominations in the United States, and Today, there are about 3,000 cults in the United States (Thio, p. 308 and 311). Table 1: Religions of the World by Percentage of Population, 1996 Religion North America Worldwide Established Religions Total 91.9 % 79.2 % Catholicism 25.5 % 16.9 % Protestantism 41.0 % 7.0 % Russian / Greek Orthodox 2.2 % 3.8 % Judaism 2.0 % 0.2 % Islam 1.9 % 19.4 % Hinduism 0.5 % 13.7 % Buddhism 0.3 % 5.6 % New Religions Total 0.3 % 1.8 % Source: U.S. Bureau of the Census, 1998a, Table 1333. 23

PAGE 32

Chapter Two Research Findings: Articles This chapter will discuss research findings focused on the prestige newspaper articles and article cases as the units of analysis. For research findings concerned with the essential characteristics of the prestige newspaper sample, Appendix B: Characteristics of Prestige Newspaper Sample contains research findings related to this sample. Table 2 Table 2: Cross Tabulation Between Prestige Newspapers and Primary Article Subjects Primary article subjec t Total Established religion New religions Established members New religion members Pape r name New York Times Count 10 11 275 10 306 % withinPrimaryarticle 47.6% 73.3% 42.2% 30.3% 42.5% Washington Post Count 8 1 179 8 196 % withinPrimaryarticle 38.1% 6.7% 27.5% 24.2% 27.2% Los Angeles Times Count 3 3 197 15 218 % within Primaryarticle 14.3% 20.0% 30.3% 45.5% 30.3% Total Count 21 15 651 33 720 % withinPrimaryarticle 100.0% 100.0% 100.0% 100.0% 100.0% % of Total 2.9% 2.1% 90.4% 4.6% 100.0% 24 depicts the results of a cross tabulation between the three prestige newspapers and the primary subject of article variable collected for the 720 articles of this sample. All three newspapers are very similar to each other in their individual reporting of established and new religions and their members as shown by the percentages, and in many instances the frequency, of the articles. Interestingly, articles primarily concerned with discussing a

PAGE 33

religion, either established or new, constitute only 5 % of the sample total, and both types of religion appear with similar low frequencies in these newspapers. A significant difference is noted in the frequencies and percentages of articles primarily concerned with discussing religious group members. 90. 4 % of the sample, or 651, articles focused on established religion members, and only 4.6 % of the sample, or 33, articles focused on new religious movement members, which results in an 85.8 % point difference. These research findings demonstrate that these prestige newspapers are very similar in their reporting of religion and predominately focus their articles on established religion members. Table 3: Cross Tabulation Between Newspaper Article Prominence and Primary Article Subject Primary article subjec t Total Established religion New religion Established members New religion members Article prominence Inside page Count 17 11 480 26 534 % withinPrimarysubject 81.0% 73.3% 73.7% 78.8% 74.2% % of Total 2.4% 1.5% 66.7% 3.6% 74.2% Section page one Count 3 3 92 1 99 % within Primarysubject 14.3% 20.0% 14.1% 3.0% 13.8% % of Total .4% .4% 12.8% .1% 13.8% Front page Count 1 1 79 6 87 % withinPrimarysubject 4.8% 6.7% 12.1% 18.2% 12.1% % of Total .1% .1% 11.0% .8% 12.1% Table 3 depicts a cross tabulation between the prominence of the newspaper article and the primary subject of the article. As the cross tabulation shows, the prestige newspapers of the sample are nearly uniform in their placement of articles discussing the 25

PAGE 34

two types of religions in their newspapers. For the articles discussing religion group members, both religion group members had similar percentages of articles located in the inside pages of a newspaper, 73.7 % for established religions and 78.8 % for new religions. For articles located on a section page one, established religions had 14 % of their total articles presented on section page ones compared to only 3 %, or one, of the new religion member articles, which produces a noteworthy 11 % point difference. For articles located on the front page of a newspaper, established religion member articles had 12 % of their total presented on the front page and new religion members had 18 % of their total presented on the front page, which produces a minor, but interesting, 6 % point difference. These research findings demonstrate that newspaper articles about both types of religion receive very similar treatment in their location in the newspapers, but there are some noteworthy variations in the location of religious group member articles. Table 4: Cross Tabulation of Origin of Article and Primary Article Subject 26 Primary subjec t Total Established religion New religion Established members New members Article origin Paper's staf f reporter Count 17 9 401 16 443 % within subject 81.0% 60.0% 61.6% 48.5% 61.5% religion reporter Count 1 12 13 % withinsubject 4.8% 1.8% 1.8% international reporter Count 1 4 116 5 126 % withinsubject 4.8% 26.7% 17.8% 15.2% 17.5% A wire service Count 2 1 101 11 115 % withinsubject 9.5% 6.7% 15.5% 33.3% 16.0% Other bureau or newspaper Count 1 21 1 23 % withinsubject 6.7% 3.2% 3.0% 3.2%

PAGE 35

Table 4 depicts a cross tabulation between the origin, or source, of the newspaper article and the primary subject of the article. While similar for many article origin variables, the cross tabulation for the two types of religions produced a 21 % point difference between them regarding the percentage generated by a newspaper staff reporter, and a 22 % point difference is present between the percentages of these types of religions and articles generated by international reporters. Empty cells are found for established religions, the other bureau or newspaper variable, and new religions, the religion reporter variable. For established and new religion members, there is a 13.1 % point difference between these subject variables and articles written by newspaper staff reporters, and there is a 17.8 % point difference for articles generated by a wire service. New religious movement members have a missing value for articles written by religion reporters. With 401 articles, newspaper staff reporters generated by far more articles about established religion members than any other origin. These research findings demonstrate that newspaper staff reporters produced the bulk of these articles and were trailed distantly by newspaper international reporters. Noteworthy differences were discovered among the types of religion and origin of articles, but the small numbers of cases and missing data diminish these findings. For established and new religion members, noteworthy differences between these group members were found for wire service and staff reporter generated articles, and new religion members had an interesting finding of 33.3 % of all articles originating with a wire service. While identifiable religion reporters contributed a very minor amount of articles, it is interesting that none of those thirteen articles had new religion groups or members as the subject. 27

PAGE 36

Table 5 depicts the means of the word counts of the newspaper articles by the primary subject of the article. Since Appendix Two discusses the article sample, the total will not be discussed here. Newspapers articles primarily reporting on established religions had a mean of 1,014, and newspaper articles primarily reporting on established religion members had a mean of 1,032, which are very close to the sample mean. New religious movements had the largest mean article word count for a primary subject of an article at 1,196, and new religious movement members had the smallest mean article word count at 834, which is a moderately large variation for both new religion variables away from the sample mean. Table 5: Means of Article Word Counts by Primary Subject of Articles Primary article subject Mean N Std. Deviation Established religion 1014.10 21 747.423 New religious movement 1196.20 15 945.151 Established religion members 1031.71 651 701.704 New religious movement members 833.88 33 558.403 Total 1025.56 720 703.050 Table 6 depicts a lambda value for the variables primary subject of newspaper article and newspaper article word count. As Appendix Two reports, the mean word count for the articles of this sample is 1,026 with a minimum of 61 and a maximum Table 6: Lambda Results of Primary Subject of Article and Article Word Count Value Asymp. Std. Error Approx. T Approx. Sig. Nominal by Nominal Lambda Symmetric .061 .010 5.626 .000 Primary articlesubject Dependent .652 .070 5.617 .000 Article lengthDependent .004 .002 1.736 .083 Goodman /Kruskal tau Primary articlesubject Dependent .804 .009 .501 Article lengthDependent .004 .000 .627 28

PAGE 37

of 5919. For this lambda finding, the relationship between the primary subject of an article and the word count, or length, of an article can be summarized by the extremely interesting value of .652. This means that for purposes of predicting the primary subject of a newspaper article, we can make about 65 % fewer errors by knowing the word count of an article than we would make without knowing this information. Table 7 depicts the primary subjects of the articles and the numbers of accompanying visual aides that are found with them. For the twenty-one articles about established religions, fifteen visual aides accompanied these articles, photographs were the dominant visual aides employed, and no charts and/or graphs were discovered. New religious movements account for fifteen articles in the sample, yield nine visual aides with photographs the dominant majority, and have no information boxes. Established Table 7: Article Primary Subject and Number of Accompanying Visual Aides Primary articlesubject Photograph InfoBox Chart/Graph Other/Map Established religion Mean 1.82 1.00 1.00 N 11 2 2 New religiousmovement Mean 1.14 1.00 1.00 N 7 1 1 Established religionmembers Mean 1.96 1.00 1.11 1.00 N 361 12 19 36 New religion members Mean 1.50 1.00 1.00 N 16 1 1 Total Mean 1.92 1.00 1.10 1.00 N 395 15 20 40 religion members dominated the sample with 651 articles focused on them, and these articles had the largest number of accompanying visual aides at 428. Established religion members were the only article subjects to have all four visual aides used in their articles. Photographs are again the dominant visual aides, but interestingly, the other/map variable 29

PAGE 38

has a sizeable second largest number. Eighteen visual aides were found in the thirty-three articles primarily concerned with new religion members, and photographs were the dominant visual aides used to accompany these articles with no charts and/or graphs used to illustrate these articles. The research findings show that photographs were predominate visual aides used to accompany these articles. Of interest, new religious movement member articles had the lowest percentage of articles accompanied by a visual aide, which was calculated at 55 %. Table 8 depicts a cross tabulation between the four newspaper article themes and the primary article subjects. The general information/news theme had the largest frequency at 324 articles. While the three other primary article subjects had large percentages of articles, new religion member articles, at 6 %, had a noteworthy 27 % Table 8: Cross Tabulation Between Theme and Primary Article Subject Subject Total Established religion New religion Established members New religion membe r Theme General Information Count 11 5 306 2 324 % withinsubject 52.4% 33.3% 47.0% 6.1% 45.0% Ideological discussion Count 1 3 29 3 36 % withinsubject 4.8% 20.0% 4.5% 9.1% 5.0% Conflict Count 9 7 260 28 304 % withinsubject 42.9% 46.7% 39.9% 84.8% 42.2% Sex Count 56 56 % withinsubject 8.6% 7.8% Total Count 21 15 651 33 720 % withinsubject 100.0% 100.0% 100.0% 100.0% 100.0% % of Total 2.9% 2.1% 90.4% 4.6% 100.0% point difference with new religion articles and a 46 % point difference with established religion articles. The ideological discussion theme occurred with the least frequency in 30

PAGE 39

the sample, and established religion members had the largest frequency of articles with this theme. Interestingly, articles with new religions or their members had the highest percentage of articles that employed this theme with 20 % of the new religion articles using this theme. The conflict theme had the second largest total with 304 articles employing this theme. For new religious movement member articles, there is an extremely noteworthy 45 % point difference between these articles and articles discussing established religion members. Interestingly, the sex theme appeared only in those articles discussing established religion members. With a moderate frequency of 56, this theme accounted for 8.6 % of all the established religion member articles. These research findings show that the general information/news theme was the most widely used theme and ideological discussion was the least. Significantly, new religious movement member articles did not employ the general information/news theme, and new religious movement and member articles have larger percentages of their articles employing the ideological discussion theme. Conflict was the second most frequently used theme accounting for a large percentage of each subjects articles, and new religious movement members had a significantly large 45 % point difference with established religion members. The sex theme only applied to established religion members. Table 9 depicts a cross tabulation of the primary article angles and the primary article subjects of the sample. Table 9 presents only those angles that were represented in more than five articles. The official pronouncements angle occurred with a low frequency of only thirty articles. Twenty-eight of the thirty articles whose angle was an official pronouncement were articles about established religion members. Only one of 31

PAGE 40

these articles was about new religion members. The practice of worship angle has similar percentages for established religion and member articles, high teens, and for new religions and members, mid-twenties, which produces an 8 % point difference in both cases. With a frequency of 225, the information angle has the largest frequency, and Table 9: Cross Tabulation of Primary Article Angles and Primary Article Subjects Subject Total Establishedreligion New religion Established members New religion members Primary angle Official pronounce Count 1 28 1 30 % withinsubject 4.8% 4.3% 3.0% 4.2% Worship Count 4 4 103 8 119 % withinsubject 19.0% 26.7% 15.8% 24.2% 16.5% Information Count 12 9 197 7 225 % withinsubject 57.1% 60.0% 30.3% 21.2% 31.3% Promotion Count 3 95 1 99 % within subject 14.3% 14.6% 3.0% 13.8% Clergy as victimizers Count 1 66 3 70 % withinsubject 6.7% 10.1% 9.1% 9.7% Violence Count 1 1 161 10 173 % withinsubject 4.8% 6.7% 24.7% 30.3% 24.0% 30 % of the established religion member articles use this angle. For information, the religion groups were similar with both having around 60 % of their articles use this angle. However, the religion member articles, at 30 % and 21 %, used this angle significantly fewer times. There is a 9 % point difference between established and new religion member articles for information, and interestingly, new religion member articles utilize the smallest percentage of this angle. 32

PAGE 41

For the primary article angle of promotion, established religion and member articles had the same percentage of 14 %, and significantly, an 11 % point difference exists between new religion members and both established religion categories. Interestingly, new religious movements did not have any articles employ this angle. Apocalyptic/Millennial beliefs had the lowest frequency of any of the angles with four articles and will not be reported. For the clergy as victimizer angle, no articles with this angle were found for established religions, and only one new religion article was found to have this angle. While both established and new religion member articles had similar percentages within the primary article subject, established religion member articles had the predominate frequency of these articles. Established religion members had 66 articles with this angle compared to three new religion member articles with this angle. Interestingly, 9 % of the established religion member article total consists of this angle. Interestingly, the violence angle was the second most used article angle in this sample with a frequency of 173 articles. Violence as an article angle was used only once for each religion group, but both established, 25 %, and new religion, 30 %, members had a high percentage of articles utilizing this angle. The other primary angle category was not used. These research findings show that information was the most used article angle, and the apocalyptic/millennial beliefs angle was used in a negligible amount of articles. The official pronouncements angle was predominated by established religion member articles, and an 8 % point difference was found between the established religions and members and the new religions and members for the practice of worship angle. The information angle percentages were similar and large for both religion groups, but there 33

PAGE 42

was significantly fewer usage of this angle for articles focusing on religion members and new religion members had the smallest percentage of these articles. Interestingly, new religions did not have any articles employ the promotion angle, and new religion member articles had an 11 % point difference with both categories of established religion articles. For the clergy as victimizer angle, established religion member articles predominate this angle with these articles representing 9 % of the established religion total. The violence angle had the second highest frequency, appeared only one time for both religion groups, and consisted of a high percentage of the articles discussing religion members of both groups. Table 10 depicts a cross tabulation of violence by religious groups and the primary article subject variable. Significantly, 360 articles, or 50 % of the sample, discussed violence, potential or actual, committed by these religion groups and/or their Table 10: Cross Tabulation Between Violence by Religious Groups and Primary Article Subject Subject Total Established religion New religion Established members New members Violence by groups Potential violence Count 3 2 66 6 77 % withinPrimarysubject 42.9% 33.3% 20.2% 30.0% 21.4% % of Total .8% .6% 18.3% 1.7% 21.4% Actual violence Count 4 4 261 14 283 % withinPrimarysubject 57.1% 66.7% 79.8% 70.0% 78.6% % of Total 1.1% 1.1% 72.5% 3.9% 78.6% Total Count 7 6 327 20 360 % withinViolence bygroups 1.9% 1.7% 90.8% 5.6% 100.0% % of Total 1.9% 1.7% 90.8% 5.6% 100.0% 34

PAGE 43

members. There were 77 articles, or 11 % of the sample, that discussed the potential for violence by religious groups, and each religion pair, groups and members, has a 10 % point difference. Interestingly, there was a 23 % point difference in the potential for violence between established religions and their members. There were 283 articles, or 39 % of the sample, that discussed actual violence committed by a religious group and/or their members, and each religion pair, groups and members, has a 10 % point difference. While all percentages within the primary subjects were very high for actual violence, established religion members, with 80 % of their articles, had the largest. These research findings are evidence that a discussion of actual and potential violence by religion groups and members are an important characteristic of these articles. Actual violence was discussed with more frequency than potential violence, and there was an interesting finding of a 23 % point difference between the potential for violence of established religions and their members Table 11 depicts a cross tabulation of violence against religious groups and members and the primary article subject. Significantly, 388 of the articles, or 54 % of the sample, discussed violence, potential or actual, committed against religious groups and/or their members. While 16 % of the articles discussed the potential for violence against religious groups and/or members, very little variation is present among the religion groups and members for the percentage within the primary article subject. Interestingly, articles that discussed actual violence against established and new religions and their members consisted of over 70 % of all the articles within the primary article subjects. Significantly, 40 % of the new religion articles discussed actual violence committed against these groups. 38 % of the established religion member articles discussed actual 35

PAGE 44

violence committed against these members, and 61 % of the new religious movement member articles discussed actual violence committed against them. These research findings are evidence that a discussion of actual and potential violence committed against religion groups and/or members is an important characteristic of these articles. Actual violence committed against religion groups and/or members is discussed in three articles for every one article discussing potential violence, and there were significantly large percentages of articles that discussed actual violence against religion groups and members. Table 11: Cross Tabulation Between Violence Against Groups and Members and Article Subject Subject Total Established religion New religion Established members New members Violence agains t Potential violence Count 2 2 102 6 112 % withinPrimarysubject 25.0% 25.0% 29.5% 23.1% 28.9% % of Total .5% .5% 26.3% 1.5% 28.9% Actual violence Count 6 6 244 20 276 % withinPrimarysubject 75.0% 75.0% 70.5% 76.9% 71.1% % of Total 1.5% 1.5% 62.9% 5.2% 71.1% Total Count 8 8 346 26 388 % withinViolenceagains t 2.1% 2.1% 89.2% 6.7% 100.0% % of Total 2.1% 2.1% 89.2% 6.7% 100.0% Table12 depicts a gamma value for the variables violence committed by religious groups and/or members and violence committed against religious groups Table 12: Gamma Results of Violence by Groups and Violence Against Groups Value Asymp. Std. Error Approx. T Approx. Sig. Ordinal by Ordinal Gamma .864 .044 6.843 .000 N of Valid Cases 276 36

PAGE 45

and/or their members. Article discussion of violence against groups is an extremely powerful predictor of article discussion of violence by groups, as these data illustrate. The gamma value of +.864 indicates that our ability to accurately predict if an article contains violence by religious groups is improved by an extremely strong 86 % if we know if the newspaper article contains news of violence against religious groups. Research Findings: Article Cases This section will discuss research findings focused on the 3,052 individual cases of religion groups and members that were coded from the prestige newspaper articles as the unit of analysis. For this data set, there were 2,900 cases, or 95 %, of established religions and/or members coded, and there were 152 cases, or 5 %, of new religious movements and/or members coded. The established religion percentage of cases is only slightly higher than the percentage of established religion members in North America, but the established religion percentage of cases is 16 % greater than the worldwide percentage of these religions populations (Giddens, Duneier, and Appelbaum, p. 545). Extremely significant is the new religious movements percentage of cases, which is seventeen times greater than their percentage of North American membership and three times greater than their worldwide population (Giddens, Duneier, and Appelbaum, p. 545). These research findings present strong evidence that the prestige newspapers of the sample have cases in their articles concerning established religions and their members in proportion to their North American population and in moderately greater proportion than their worldwide population. Extremely significant is the prestige newspapers reporting of new religious movements and/or their members. These religions and/or their members were the cases in these newspaper articles an extremely disproportionate percentage 37

PAGE 46

compared to their actual percentage of the North American and the worldwide population. There were over fifty different specific established religions and/or their members coded for in this sample. Specific established religions and/or their members that were coded more than thirty-five times in this sample are: Baptist (55), Buddhism (41), Catholicism (876), Christianity (317), Episcopalian (91), Hinduism (44), Islam (654), Judaism (313), Church of Jesus Christ, Latter-Day Saints (35), Pentecostalism (49), Protestantism (63), Russian Orthodox Church (35). Catholicism, with 876 cases, had the largest frequency of established religions and/or members for this sample, and at 29 % of the cases, Catholicism was very close to its North American population percentage, 26 %. Extremely significant, Islam, with 654 cases, is 21 % of the cases, which is close to Islams worldwide population but eleven times greater than Islams North American population. Also extremely significant, Judaism and/or Jews, which accounted for 10 % of the case sample, are five times greater represented than their North American population, and this percentage of Judaism and/or Jewish cases in these prestige newspaper articles were fifty times greater than their worldwide percentage of population. Additionally, both Hinduism and Buddhism approximated their percentages of North American percentage of population, but this means that these established religions percentages of cases are extremely low when compared to their worldwide percentages of population. Extremely noteworthy, the Russian Orthodox Church and the Greek Orthodox Church are .02 % of the article cases, and this is a significantly smaller percentage of the article cases than these established religions proportion of the North 38

PAGE 47

American and worldwide population. Protestantism taken as a percentage of the case sample by itself is .02 %, which would be a significantly low percentage. However, Protestantism consists of many denominations, and when we add the additional frequencies from the sample of established religions with over 35 cases that identify with Protestantism, the results produce a percentage of cases of 16 %, which is a greater percentage of cases than the worldwide population. While this percentage of cases is significantly lower than the North American population, this does not account for the dozens of Protestant established religions and/or members that are in the sample, but were not included due to their case frequency. There were over thirty different specific new religious movements and/or their members coded for in this sample. Specific new religious movements and/or their members that were coded more than once in this sample are: Branch Davidians (5), Doomsday Cult (3), Ethical Culture (4), Falun Gong (35), Heavens Gate (6), Holy Spirit Movement (2), I Am (8), Jesus People (2), Lords Resistance Army (2), Nation of Islam (15), NeoPaganism (2), Opus Dei (6), Peoples Temple (5), Raelians (8), Silver Shirt Legion (3), Ten Commandments (10), Unification Church (4), Voodoo (3), World Church of the Creator (8), World Message Last Warning (2). Falun Gong, with 35 cases or 23 % of the new religious movement and/or member case sample, had the largest frequency and percentage of new religious movements and/or member cases for this sample. Interestingly, several new religious movements that were defunct during the time period of this sample were coded, like Heavens Gate, and some of these new religious movements have been defunct for several decades. Examples of these long defunct new religious movements are the Jesus People and Peoples Temple of the 1960s and 1970s, 39

PAGE 48

the I Am movement of the 1940s, and the Silver Shirt Legion of the 1930s. Continued newspaper reporting of these new religious movements, which represent a good cross-section of notorious groups, serves to keep a highly negative image of new religious movements in the news and before the public. Table 13 depicts a cross tabulation between the potential for violence and actual violence with established and new religions and/or their members. Interestingly, the total of 763 cases of violence by religious groups and/or their members constitutes 25 % of the sample total of 3,052 cases. Also interesting is the significantly greater frequency and percentage of cases of actual violence by religion groups and members than potential violence. The cases of potential for violence by religious groups and/or their members and cases of actual violence by religion groups and/or their members are similar Table 13: Cross Tabulation Between Violence by Religion and Religion Group RELGROUP Total established religion or member new religion or member VIOLENBY Potential violence Count 184 14 198 % within RELGROUP 26.0% 25.0% 26.0% % of Total 24.1% 1.8% 26.0% Actual violence Count 523 42 565 % within RELGROUP 74.0% 75.0% 74.0% % of Total 68.5% 5.5% 74.0% Total Count 707 56 763 % within RELGROUP 100.0% 100.0% 100.0% % of Total 92.7% 7.3% 100.0% percentages within the religion groups. These research findings suggest that prestige newspaper articles present about equal percentages of individual cases of potential for violence and actual violence by both religion groups and members; however, a 40

PAGE 49

noteworthy finding is that one-out-of-every four cases discussed actual violence or the potential for violence with the significant emphasis placed on actual violence. Table 14 depicts a cross tabulation between potential and actual violence against religion groups and established and new religions and/or their members. Interestingly, the total of 717 cases of violence against religious groups and/or their members constitutes 23.5 % of the sample total of 3,052 cases. Also significant is the much greater frequency Table 14: Cross Tabulation Between Violence Against Religion Groups and Members RELGROUP Total established religion o r member new religion o r member VAGAINST Potential violence Count 211 8 219 % within RELGROUP 31.9% 14.3% 30.5% % of Total 29.4% 1.1% 30.5% Actual violence Count 450 48 498 % within RELGROUP 68.1% 85.7% 69.5% % of Total 62.8% 6.7% 69.5% Total Count 661 56 717 % within VAGAINST 92.2% 7.8% 100.0% % of Total 92.2% 7.8% 100.0% and percentage of actual violence cases against established and new religions and/or their members rather than potential violence cases. There is a significant 18 % point difference between established religions and new religions and/or their members for cases of potential violence, and there is a significant 18 % point difference between new religions and established religion and/or their members for cases of actual violence. Extremely significant, new religions and/or their members have 86 % of their total cases dealing with violence depicted as actual violent acts. These research findings suggest that cases of violence against established religion groups and members are common, cases of actual 41

PAGE 50

violence are stressed over cases with the potential for violence, and significant differences are found between the two religion groups and their members. Established religions and/or their members have a much greater percentage of cases with the potential for violence than new religions, and new religions and/or their members have a much greater percentage of cases reporting actual violence, an extremely high 86 %, than established religions and/or their members. Table 15 presents a gamma value for the variables violence committed by religious groups and/or members and violence committed against religious groups and/or members. Case discussion of violence by groups is an extremely powerful predictor of article discussion of violence against groups, as these data illustrate. The gamma value of +.917 indicates that our ability to accurately predict if an article contains violence against religious groups is improved by an extremely strong 92 % if we know if the newspaper article contains violence by religious groups. This is strong evidence that violence initiated by religious groups is reported as being answered by violence against religious groups. Table 15: Gamma Results of Violence by Groups and Violence Against Groups Value Asymp. Std. Error Approx. T Approx. Sig. Ordinal by Ordinal Gamma .917 .025 9.216 .000 N of Valid Cases 366 Table 16 depicts the frequency of religion group descriptors in the case sample. The religious group descriptor cult was recorded 38 times and appears in only 1.2 % of the case sample total. Sect was recorded 55 times and appears in only 1.8 % of the case sample total. Movement was recorded 172 times and appears in a modest 5.6 % of the case sample total. Group was recorded 510 times and appears in a moderate 16.7 % of the case sample total. The religious group descriptor church appeared the most 42

PAGE 51

frequently in the case sample total with a frequency of 1,156, or 37.9 %. Interestingly, the descriptor religion appears only 290 times and appears in a modest 9.5 % of the case sample total. Faith was recorded 349 times and appears in a modest 11.4 % of the case sample total. Congregation was recorded 278 times and appears in a modest 9.1 % of the case sample total. Table 16: Frequency of Religion Group Descriptors CULT SECT MOVEMENT GROUP CHURCH RELIGION FAITH CONGR N Valid 38 55 172 510 1156 290 349 278 Missing 3014 2997 2880 2542 1896 2762 2703 2774 Tables 17 and 18 present the frequency of religion member descriptors in the case sample. The religious member descriptor terrorist was recorded 183 times and appears in a modest 6 % of the case sample total. Militant was recorded 161 times and appears in a modest 5.3 % of the case sample total. Adherent was recorded 26 times and appears in a negligible 0.9 of the case sample total. Pedophile was recorded 21 times Table 17: Frequency of Religion Member Descriptors (1) TERRORIST MILITANT ADHERENT PEDOPHILE FOLLOWE R N Valid 183 161 26 21 132 Missing 2869 2891 3026 3031 2920 and appears in a negligible 0.7 % of the case sample total. Follower was recorded 132 times and appears in a meager 4.3 % of the case sample total. The religious member descriptor member was the most recorded at 589 times and appeared in a noteworthy 19.3 % of the case sample total. Worshiper was recorded 91 times and appeared in a Table 18: Frequency of Religion Member Descriptors (2) MEMBE R WORSHIPE R LEADE R RELTITLE N Valid 589 91 480 1693 Missing 2463 2961 2572 1359 meager 3.0 % of the case sample total. Leader was the second largest religious member descriptor with a frequency of 480 and a noteworthy 15.7 % of the case sample 43

PAGE 52

population. While the actual largest descriptor in frequency, because it is a composite variable: pope, rabbi, imam, monk, priest, etc., an official religious title will not be considered the largest single descriptor. Extremely interesting is the very large amount of diversity found among the official religious titles in this sample, for there were over seventy-five different official religious titles coded. An official religious title was recorded 1,693 times and appeared in a significant 55.5 % of the case sample total. The other variables were also coded for both groups and members. These two other variables recorded all mentions of descriptors used to describe religious groups and members that were not specifically coded. Both of these other variables, groups and members, were extremely large. Descriptors were coded for the other group variable in 81 % of the 3,052 cases, or 2,479 cases also contained an other group descriptor. Descriptors were coded for the other member variable in 75 % of the 3,052 cases, or 2,281 cases also contained an other member descriptor. Due to the extremely large numbers of cases and frequencies of other group and member descriptors, only a general summary of these findings will be reported. For established religious groups, an other group variable coded a large majority of descriptors used to describe formal titles of places of worship: temple, synagogue, mosque, chapel, cathedral, and parish. The other group descriptor variable also coded a large number of the following descriptors used in conjunction with specific established religions: community, organization, school, tradition, and institution. Additionally, a large number of other group descriptors related to practices of worship and the structures of established religions were coded: service, rules, ceremony, rituals, tradition, state, mass, belief, diocese, archdiocese, and doctrine. For new religious movements, due 44

PAGE 53

to their smaller frequency of cases in the sample there were less opportunities to code for an other group variable; however, the bulk of other group descriptors used in conjunction with a specific new religious movement were: teaching, ritual, practice, belief, suicide, and compound. An other member descriptor variable also coded a multitude of different descriptors used to describe specific established religion and new religious movement members. The bulk of established religion member other descriptors were: parishioner, victim, people, student, faithful, believer, immigrant, extremist, and official. The bulk of new religious movement other member descriptors were: faithful, demonstrators, people, protesters, believers, plaintiffs, and practitioners. Table 19 presents the frequency of religion group descriptors in the case sample by religion type. For each of the coded group descriptor variables, the frequency of each descriptor in the sample was determined for both established and new religions. Because multiple frequencies of each group descriptor were recorded in numerous cases, Table 19 depicts the total frequency, or actual number/amount, of each descriptor for both established and new religions. Table 19 reports the actual total number of counts for each group variable and how it is distributed between established and new religions, which is different from the earlier tables and their focus on the number of individual cases in the sample. For the group descriptor cult, a huge disparity exists with new religious movements described as cults 59 times, or 94 % of the total frequency, and established religions were described as cults four times, or 6 % of the total frequency. For the group descriptor sect, a noticeable disparity is present with new religious movements 45

PAGE 54

described as sects 59 times, or 72 % of the total frequency, and established religions were described as sects 23 times, or 28 % of the total frequency. For the group descriptor movement, established religions were disproportionately described as movements 175 times, or 81 % of the total frequency, and new religious movements were described as Table 19: Religion Group Descriptors and Frequency Between Established and New Religions Cult Sect Movement Group Church Religion Faith Congregation Established Religion 4 23 175 690 2881 338 423 396 New Religion 59 59 40 88 35 15 12 0 Total Frequency 63 82 215 778 2916 353 435 396 movements 40 times, or 19 % of the total frequency. For the group descriptor group, established religions were described as a group 690 times, or 89 % of the total frequency, and new religious movements were described as a group 88 times, or 11 % of the total frequency. For the group descriptor church, established religions predominated this group descriptor and were described as a church 2,881 times, or 99 % of the total frequency, and new religious movements were described as a church only 35 times, or about 1 % of the total frequency. For the group descriptor religion, established religions predominate this group descriptor and were described as a religion 338 times, or 96 % of the total frequency, and new religious movements were described as a religion fifteen times, or 4 % of the total frequency. For the group descriptor faith, established religions were described as a faith 423 times, or 97 % of the total frequency, and new religious movements were described as a faith only twelve times, or 3 % of the total frequency. Lastly, the group descriptor congregation is only used to describe established religions with 396 specific mentions of established religions as a congregation, or 100 % of the total frequency. 46

PAGE 55

Tables 20 and 21 present the frequency of religion member descriptors in the case sample by religion type. Again, for each of the coded member descriptor variables, the frequency of each descriptor in the sample was determined for both established and new religion members. Because multiple frequencies of each member descriptor were recorded in numerous cases, Tables 20 and 21 depict the total frequency, or actual number/amount, of each descriptor for both established and new religion members. Tables 20 and 21 depict the actual total number of counts for each member variable and how it is distributed between established and new religions, which is different from the earlier tables and their focus on the number of individual cases in the sample. For the member descriptor terrorist, only established religion members were described with this descriptor for 100 % of the total frequency, or 313 specific mentions. For the member descriptor militant, established religion members were predominately described as militants 278 times, or 99 % of the total frequency, and new religious movement members were described as militants only two times, or less than 1 % of the Table 20: Religion Member Descriptors and Frequency Between Religion Types (1) Terrorist Militant Adherent Pedophile Follower Established Religions 313 278 18 24 101 New Religions 0 2 8 0 49 Total Frequency 313 280 26 24 150 total frequency. For the member descriptor adherent, established religion members were described as adherents eighteen times, or 69 % of the total frequency, and new religious movement members were described as adherents eight times, or 31 % of the total frequency. For the member descriptor pedophile, only established religion members were described as pedophiles for 100 % of the total frequency, or 24 specific 47

PAGE 56

mentions. For the member descriptor follower, established religion members were described as followers 101 times, or 67 % of the total frequency, and new religious movement members were described as followers 49 times, or 33 % of the total frequency. Table 21: Religion Member Descriptors and Frequency Between Religion Types (2) Member Worshiper Leader Religious Title Established Religion 709 103 563 6593 New Religion 99 0 35 54 Total Frequency 808 103 598 6647 For the member descriptor member, established religion members were described as a member 709 times, or 88 % of the total frequency, and new religious movement members were described as a member 99 times for a much smaller 12 % of the total frequency. For the member descriptor worshiper, only established religion members were described as worshipers for 100 % of the total frequency, or 103 specific mentions. For the member descriptor leader, established religion members were described as a leader 563 times, or 94 % of the total frequency, and new religious movement members were described as a leader a paltry 35 times, or 6 % of the total frequency. For the member descriptor religious title, established religion members predominated and were described with a religious title 6,593 times, or 99 % of the total frequency, and new religious movement members were described with a religious title only 54 times, or less than 1 % of the total frequency. Results and Discussion With the research findings for the articles and article cases presented, the results of these research findings can be discussed and applied to the thesis hypotheses. The first hypothesis stated that print media reporting of new religious movements will be more negatively biased, critical, and prone to use negative descriptors than reporting that is 48

PAGE 57

concerned with established religions. As my research findings demonstrate, the negative religion group descriptors of cult and sect, while consisting of only a small 3 % of the entire article cases, were disproportionately used to describe new religious movements. 94 % of the total frequency of the group descriptor cult was used to describe new religious movements, and 72 % of the total frequency of the group descriptor sect was used to describe new religious movements. Interestingly, only established religion members were described with the highly negative member descriptors of terrorist and pedophile, and the negative member descriptor of militant was used with a total frequency of 99 % to describe established religion members. 99 % of the total frequency of the negative descriptor terrorist was used in conjunction with describing Muslims, and 96 % of the total frequency of the negative descriptor militant was used in conjunction with describing Muslims. 95 % of the total frequency of the negative descriptor pedophile was used in conjunction with Catholics. Militant appeared in the first week coded for in the article case sample and appeared throughout the entire article case data set. Terrorist first appeared in the article case sample in the first week of the third month coded for in 1998 and continued throughout the entire article case data set. Not surprisingly, the member descriptors terrorist and militant increased after September 11, 2001. Pedophile first appeared in the article case sample in the third week of the third month coded for in 2000 and continued sporadically throughout the article case data set. For the religion group descriptors that convey legitimation, these descriptors were overwhelmingly used to describe established religions. The almost exclusive, or exclusive, usage of the religion group descriptors church, with 99 % of the total 49

PAGE 58

frequency used to describe established religions, religion, with 96 % of the total frequency used to describe established religions, and congregation, with 100 % of the total frequency used to describe established religions, are extremely powerful evidence that established religions are deemed to be legitimate religious organizations by the prestige newspaper sample. The absence of these legitimating descriptors in reference to new religious movements is evidence that the prestige newspapers in this sample do not view new religious movements as legitimate religious organizations. For religion group descriptors that convey neutrality, established religions again predominate among these descriptors. However, new religious movements garner a noticeable 19 % of the total frequency of the descriptor movement and 11 % of the total frequency of the descriptor group. Only 3 % of the total frequency of the group descriptor faith was used to describe new religious movements. Because the group descriptor faith is underrepresented, the evidence suggests that the most positive descriptors new religious movements can consistently receive in prestige newspaper articles are neutral group descriptors, namely movement and group. Religious member descriptors that convey legitimation were overwhelmingly devoted to describing established religion members. The religion member descriptor worshiper was used to describe established religion members in every instance. 94 % of the time that the member descriptor leader was used, it described established religion members, and 99 % of the time that a religious title was used, it referred to members of established religions. This is extremely strong evidence that the prestige newspapers do not describe new religious movement members as legitimate religious members. 50

PAGE 59

For religious member descriptors that convey neutrality, established religions predominate among these descriptors, but new religious movement members are described with a moderate frequency. New religious movement members were described as adherents for 31 % of the total frequency of this variable and as followers for 33 % of the total frequency of this variable. New religious movement members were also described as a member in 12 % of the total frequency for this variable. This evidence supports the notion that, when prestige newspapers describe new religious movement members, the most positive types of descriptors used are the neutral descriptors, and these descriptors are used with only a moderate frequency. The research findings presented give strong support for this hypothesis and replicate prior research findings (Hill, Hickman, and McLendon 2001; Richardson and van Driel 1988 and 1997; Wright 1997). Both types of legitimating descriptors, group and member, were predominately, or exclusively, used to describe established religions or their members. Neutral descriptors, both group and member, did have a meager to moderate frequency of usage with new religious movements, and neutral descriptors can be viewed as the most positive type of description a new religious movement or member can expect to receive. As expected, new religious movements were the most likely to receive negative group descriptors; however, due to the recent social construction of religious villains (Loseke 2003 and Unpublished; Glassner 1999), established religions have come to receive the highest proportion of negative religion member descriptors. Of the established religions, Muslims were the most likely to be described as terrorists and militants, and Catholics were the most likely to be described as pedophiles. 51

PAGE 60

My second hypothesis stated that print media articles focusing on new religious movements will: receive less prominent placement, have lower word counts, have fewer, if any, accompanying photographs, and center more around a theme of violence, either potential or actual. Article research findings demonstrate that articles about new religious movements and about established religions had moderately different treatment in their placement in the prestige newspapers. However, prestige newspapers placed 5 % more of the new religious movement articles on inside pages than established religion articles. New religion articles were placed less on the first page of a section with 11 % fewer articles than established religions, and possibly due to the sensationalism associated with some new religious movements, new religious movement articles were placed on the front page of a prestige newspaper 6 % more often than established religion articles. Interesting findings were uncovered concerning word counts. Prestige newspaper articles dealing with new religious movement members received the lowest word counts, and articles discussing new religious movements had the largest word counts of the sample. Additionally, a lambda score of 65 % was generated, which means that for purposes of predicting the primary subject of a newspaper article, we can make about 65 % fewer errors by knowing the word count of an article than we would make without knowing this information. Interesting findings were uncovered concerning the visual aides accompanying the prestige newspaper articles. The research findings show that photographs were the most used visual aides accompany prestige newspaper articles. Of interest, new religious movement member articles had the lowest percentage of articles accompanied by any visual aide, which was calculated at 55 %. Established religion members were the only 52

PAGE 61

article subjects to have all four visual aides used in their articles. Established religion groups had more photographs accompanying their articles than new religious groups, and established religion member articles had a significantly greater number of photographs accompanying their articles than new religious movement member articles. Concerning the violence theme, 360 articles, or 50 % of the sample, discussed violence, potential or actual, committed by these religion groups and/or their members. Actual violence was discussed with more frequency than potential violence. Research findings suggest that cases of violence against established and new religion groups and members are common, cases of actual violence are stressed over cases with the potential for violence, and significant differences are found between the two religion groups and their members. Established religions and/or their members have a much greater percentage of cases with the potential for violence than new religions, and new religions and/or their members have a much greater percentage of cases reporting actual violence against them, an extremely high 86 %, than established religions and/or their members. These research findings tend to provide support for my second hypothesis, which is based on similar research conducted by Hill, Hickman, and McLendon (2001) and concerned with testing elements of David Altheides theory of media logic (1985; 1995; Altheide and Johnson 1980; Altheide and Snow 1991). Like the Hill, Hickman, and McLendon (2001) article, this research discovered significant evidence that the theme of violence was quite prevalent in the sample of articles from these prestige newspapers. For the article cases, new religious movements and/or their members were discussed as having more actual violence conducted against them then established religions and/or their members. 53

PAGE 62

Altheides theory of media logic is concerned with the form of communication and the process of presenting and transmitting information (Media Worlds, p. 9). Because of this concern for the form and process of presenting and transmitting information, article prominence, word count, and visual aides are important for conveying legitimation on a newspaper article subject, attracting the attention of the reader to the article, and providing supporting evidence to the article. New religious movement articles did display the hypothesized article prominence with these articles predominately located in the inside pages, and there was a meager representation of articles on the first page of a newspaper section. My analysis of the article word count supports this hypothesis with the new religious movement member articles, the bulk of the total number of articles discussing new religions, having the lowest mean word count for the sample. The research findings for visual aides, especially photographs, support this hypothesis with new religious movement member articles having the lowest percentage of articles accompanied by a visual aide, which was calculated at 55 %. Established religions had more photographs accompanying their articles than new religious movements, and established religion member articles had a significantly greater number of photographs accompanying their articles than new religious movement member articles. My third hypothesis stated that the print media articles would employ, with a high frequency, themes and angles other than general information for both established and new religions. In support of this hypothesis, general information was the theme that occurred/appeared most frequently, but it was followed in frequency by the theme of conflict. The theme of conflict occurred in only twenty fewer articles than general information for a close second. For angles, the violence angle was the second most 54

PAGE 63

used article angle in this sample with a frequency of 173 articles, or 24 % of the article sample. Additionally, the worship angle had a noticeable frequency of 119 articles, or 17 % of the article sample. This hypothesis is based on Altheides prior research using media logic where multiple themes and angles other than general information were used by news mediums to capture audience attention (1995; 1985; Altheide and Snow 1991). Clearly, after the theme of general information, the dominant theme used to illustrate these prestige newspaper articles is the theme of conflict, and the violence angle was the second most used angle in these articles. This emphasis on conflict and violence supports Altheides research that notes the extensive use of this theme and angle across various mediums (1995 and Altheide and Snow 1991). My fourth hypothesis stated that established religions would receive more coverage in terms of mentions of specific groups and members than new religious movements. This hypothesis is based on the findings of prior research collected for this thesis related to bias (Hill, Hickman, and McLendon 2001; Hoover 1998; Hubbard 1998; ODonnell 1993; Richardson and van Driel 1988 and 1997; Wright 1997). Established religions and their members were the primary subjects of the newspaper articles and the article case sample. With over fifty different specific established religions mentioned in the article case sample, established religions and their members were noticeably more represented than new religious movements and their members. Interestingly, about 20 % of the established religious denominations were coded, while only about 1 % of new religious movements were coded (Thio, p. 308 and 311). Clearly, substantial evidence supports this hypothesis. 55

PAGE 64

Lastly, my fifth hypothesis stated that the conflict theme, violence angle, and instances of actual and potential violence by/against religious groups and members would occur with high frequencies and be important items discussed in the newspaper articles on religion. This hypothesis stems from the research findings of Hill, Hickman, and McLendon (2001) and David Altheide (1995). This hypothesis is supported by research findings already presented that stressed the importance and frequency of this theme, angle, and issues. Concerning ideology, it is apparent that these prestige newspapers are engaged in explaining and interpreting social reality, and while the newspaper medium is considered to be an objective format, strong evidence has been presented that demonstrates that the prestige newspapers of this sample adopt an ideological stance. The research findings and hypotheses analyzed for my thesis present extremely strong evidence that the prestige newspaper articles express the social and spiritual priorities of the dominant religious groups in North America and the world. Although established religions are subject to negative reporting, as my thesis demonstrated, the dominant religious ideologies of established religions still garner the overwhelming majority of prestige print media reporting. Legitimating religion group and member descriptors are reserved almost exclusively for established religions and their members. Additionally, prestige newspaper articles that focus on established religions and their members have a greater frequency of visual aides, more diversity of visual aides, and longer articles measured by word count. New religious movements and their members are marginalized, trivialized, and discussed primarily with negative or neutral descriptors in these prestige newspapers. This skewed treatment in the prestige print media constructs an image of these new 56

PAGE 65

religious movements as having lowered social and spiritual status. When only 1 % of new religious movements are the subjects of prestige newspaper articles, these newspapers reinforce the supposed illegitimacy of these religious movements by focusing on the most sensational groups. Furthermore, by focusing their discussion of new religious movements and/or their members on the issue of violence, with the emphasis on actual violence, these articles strip new religious movements of religious legitimation. Policy Implications and General Conclusions My thesis has demonstrated that print media bias, favoring established religions and members and marginalizing and denigrating new religious movements and members, is still a hallmark of prestige newspaper reporting on religion groups and members. I believe that I accomplished the initial goals I had for this thesis, which were to demonstrate the utility of applying media logic to analyzing prestige newspaper articles and documenting evidence of print media bias. In light of the research findings, my thesis has been able to find support for my hypotheses. Using the prestige newspaper article samples, I was able to amass evidence that lends support to all of my hypotheses and demonstrates some of the various ways print media bias is implemented. My thesis uncovered several key findings concerning established religions and their members. The sample newspapers predominately focused their articles on established religion members. Newspaper staff reporters generated more articles about established religion members than any other origin. Established religion members were the only article subjects to have all four visual aides used in their articles, and their articles contained the largest number of accompanying visual aides. The sex theme only applied to established religion members, and established religion member articles 57

PAGE 66

predominate the clergy as victimizer angle. Islam and Judaism were significantly over represented in the newspaper article sample based on their North American population. Only established religion members were described with the highly negative religion member descriptors of terrorist and pedophile. Religion group and member descriptors that convey legitimation were overwhelmingly used to describe established religions. My thesis uncovered several key findings concerning new religious movements and their members. Compared to their actual percentage of the North American and the worldwide population, new religious movements were significantly over represented as prestige newspaper article cases; however, only 1 % of new religions were discussed. New religious movement members had 33.3 % of all articles originate with a wire service. New religious movement member articles had the lowest percentage of articles accompanied by a visual aide. Falun Gong constituted 23 % of the new religious movement and/or member case sample. The negative religion group descriptors of cult and sect were disproportionately used to describe new religious movements. Many policy implications emerge from this thesis. Obviously, the need for professionally educated and non-biased religion reporters is apparent. The reporters who cover the religion beat and religion-related events should have a working understanding of the subjects of their articles. Competence in religion must become the standard for covering this beat. The policy implications for media fairness raised by Charles Haynes should be implemented with haste to improve the quality of religion reporting: expanding opportunities for continuing education in religion, hiring more religion specialists, and adding a religious-studies requirement for journalism majors (p. 87). Reporters covering 58

PAGE 67

the religion beat, and their readers, should be cognizant of the six factors identified by Stuart Wright as contributing to media bias against new religious movements: ignorance of new religions, degree of acculturation to society by new religions, available economic resources of reporters, time constraints of reporters, biased or limited sources of information for articles, and front-end/back-end disproportionality of religion reporting (p. 104). The last factor refers to the tendency of journalists to report extensively on allegations against a new religious movement at the beginning of a crisis and to report less extensively on the resolution of the crisis (Hill, Hickman, and McLendon, p. 25). Religion reporters should pursue more diversity in their subject matter, article themes, and article angles when reporting on religion groups and members, especially as it concerns new religious movements and members. Extremely important is the need for religion reporters to recognize that their articles shape and influence the knowledge people have concerning religions and their members. Great care should be taken by all religion reporters to eliminate, or clarify, all negative descriptors when referencing new and established religions and their members. Concerning ideas for further research, I believe that this thesis has highlighted the need for sociologists of religion and communication to continue to explore the content included in prestige newspaper articles, and print media in general, as it concerns religion. As with other research, this thesis served as a replication and will need replication to determine if the research findings are supported. Larger samples of prestige newspaper articles, and if possible other print media, should be coded for bias, as demonstrated in Appendix One. The number of variables relating to the group and member descriptors should be expanded to include more descriptors, which will reduce 59

PAGE 68

the rather large findings attributed to the other category variables. This thesis could lend itself easily to a longitudinal study incorporating prestige newspapers from time periods earlier than the sample studied here. It would be interesting to code prestige newspapers from different eras to determine whether the themes, angles, and descriptors in my thesis appeared in previous decades. Lastly, research should continue to expand the scope of David Altheides theory of media logic. 60

PAGE 69

References Altheide, David. An Ecology of Communication: Cultural Formats of Control. New York: Aldine De Gruyter, 1995. Altheide, David. Media Power. Beverly Hills: Sage Publications, Inc., 1985. Altheide, David and John M. Johnson. Bureaucratic Propaganda. Boston: Allyn and Bacon, Inc., 1980. Altheide, David and Robert P. Snow. Media Worlds: In The Postjournalism Era. New York: Aldine De Gruyter, 1991. Anyon, Jean. Ideology and United States History Textbooks. Harvard Educational Review, August 1979, Volume 49:3, pages 361-383. Bigelow, William. Inside the Classroom: Social Vision and Critical Pedagogy. Teachers College Record, Spring 1990, Volume 91:3, pages 437-448. Bruce, Steve. Religion in the Modern World: From Cathedrals to Cults. Oxford: Oxford University Press, 1996. Giddens, Anthony, Mitchell Duneier, and Richard Appelbaum. Introduction to Sociology, Fourth Edition. New York: W. W. Norton and Company, 2003. Giroux, Henry. Radical Pedagogy and the Politics of Student Voice. In Henry Giroux, Pedagogy and the Politics of Hope. Oxford: Westview, pages 119-145, 1997. Giroux, Henry. Theory and Resistance in Education: A Pedagogy for the Opposition. South Hadley, Mass.: Bergin and Garvey, 1983. Glassner, Barry. The Culture of Fear: Why Americans Are Afraid of the Wrong Things. New York: Basic Books, 1999. Haynes, Charles. Faith, Fairness, and the American Press. Media Studies Journal, Fall 2000, Volume 14:3, pages 82-87. Henslin, James. Essentials of Sociology: A Down to Earth Approach, Third Edition. Boston: Allyn and Bacon, 2000. Hess, Beth, Peter Stein, and Susan Farrell. The Essential Sociologist: An Introduction. Los Angeles, Roxbury Publishing Company, 2001. 61

PAGE 70

Hill, Harvey, John Hickman, and Joel McLendon. Cults and Sects and Doomsday Groups, Oh My: Media Treatment of Religion on the Eve of the Millennium. Review of Religious Research, September 2001, Volume 43:1, pages 24-38. Hoover, Stewart. Religion in the News: Faith and Journalism in American Public Discourse. Thousand Oaks: Sage Publications, Inc., 1998. Hubbard, Jamie. Embarrassing Superstition, Doctrine, and the Study of New Religious Movements. Journal of the American Academy of Religion, Spring 1998, Volume 66:1, pages 59-92. Kendall, Diana. Sociology in Our Times: The Essentials, Third Edition. Belmont, Ca.: Wadsworth, 2002. Lindsey, Linda and Stephen Beach. Essentials of Sociology. Upper Saddle River, N.J.: Prentice-Hall, 2003. Loseke, Donileen. The Construction of Evil in Social Problems Formula Stories: Pedophile Priests, Terrorists, and the Creation of Symbolic and Social Boundaries. Unpublished paper provided through courtesy of the author. Loseke, Donileen. We Hold These Truths to be Self-Evident: Problems in Pondering the Pedophile Priest Problem. Sexualities, 2003, Volume 6:1, pages 6-14. Macionis, John. Sociology, Seventh Edition. Upper Saddle River, N.J.: Prentice-Hall, 1999. ODonnell, John. Predicting Tolerance of New Religious Movements: A Multivariate Analysis. Journal for the Scientific Study of Religion, December 1993, Volume 32:4, pages 356-365. Richardson, James T. and Barend van Driel. Journalists Attitudes Toward New Religious Movements. Review of Religious Research, December 1997, Volume 39:2, pages 116-136. Richardson, James T. and Barend van Driel. Print Media Coverage of New Religious Movements: A Longitudinal Study. Journal of Communications, Summer 1988, Volume 38:3, pages 37-61. Riffe, Daniel, Stephen Lacy, and Frederick Fico. Analyzing Media Messages: Using Quantitative Content Analysis in Research. Mahwah: Lawrence Erlbaum Associates, Inc., 1998. Schaefer, Richard. Sociology, Eighth Edition. New York: McGraw-Hill, 2003. 62

PAGE 71

Snow, David and Robert Benford. Master Frames and Cycles of Protest. In Aldon Morris and Carol McClurg Mueller, Frontiers in Social Movement Theory. New Haven: Yale University Press, pages 133-155, 1992. Snyder, Benson. The Hidden Curriculum. New York: Alfred A. Knopf, 1971. Stockard, Jean. Sociology: Discovering Society. Belmont, Ca.: Wadsworth, 1997. Teitelbaum, Kenneth. Critical Lessons from Our Past: Curricula of Socialist Sunday Schools in the United States. In Michael Apple and Linda Christian-Smith, The Politics of the Textbook. Boston: Routledge, pages 135-165, 1991. Thio, Alex. Sociology: A Brief Introduction, Fifth Edition. Boston: Allyn and Bacon, 2003. Vaughan, Ed. Sociology: The Study of Society. Upper Saddle River, N.J.: Prentice-Hall, 2001. Wright, Stuart. Media Coverage of Unconventional Religion: Any Good News for Minority Faiths? Review of Religious Research, December 1997, Volume 39:2, pages 101-115. 63

PAGE 72

Appendices 64

PAGE 73

Appendix A: (Continued) Appendix A: A Textbook Analysis of Religion Introduction Religion and education are major social institutions. Both social institutions are responsible for imparting beliefs, knowledge, and values that are considered important to the social reproduction of individuals and entire cultures. Education and religion reflect society, yet both are also involved in issues of social stability and social change. Concerning the intersection of religion and education, controversy over public education and the teaching of religion continues to be problematic. From the Scopes Monkey Trial, which put the teaching of evolution in public schools on trial, to the present debate concerning the teaching of the Ten Commandments and the place of prayer in schools, religion and education are often sites of open struggle in public K-12 education. At the college and university level, the intersection of religion and education is less problematic due to the non-compulsory aspect of religious instruction in secular schools, but religion is, and historically has been, a critical component of study for many academic disciplines: anthropology, history, philosophy, psychology, religious studies, and sociology. Religious organizations have a spiritual and moral message they are concerned with sharing with other people, and especially for established religions, the educational system has been a vehicle for conveying their beliefs and dogmas to the public (Haynes 2000; Hoover 1998; Teitelbaum 1991). For new religious movements, however, the educational system is often viewed as an enemy that does not give their beliefs and dogmas fair treatment, if they are mentioned at all (Anyon 1979; ODonnell 1993; Teitelbaum 1991). The dynamic of the interaction between education and its support and 65

PAGE 74

Appendix A: (Continued) emphasis on mainstream culture, people, and organizations and a silence or meager discussion of marginalized or minority cultures, individuals, and organizations is a documented one (Anyon 1979; Bigelow 1990; Giroux 1983 and 1997; Snyder 1971; Teitelbaum 1991). This dynamic is referred to as the hidden curriculum. The hidden curriculum manifests itself in school curriculum. As Bigelow states, One function of the school curriculum is to celebrate the culture of the dominant and to ignore or scorn the culture of subordinate groups (p. 439). Textbooks used in schools have been analyzed to demonstrate that they perpetuate a hidden curriculum (Anyon 1979; Teitelbaum 1991). Recent research on the political economy of textbook publishing in the United States has further highlighted a trend toward increased homogenization and standardization of school textbook adoption (Teitelbaum, p. 135). Anyon notes that only forty publishing houses have the means to produce a textbook profitably in the United States (p. 362), and as demonstrated by a perusal of their publishers, the same publishing houses that publish public school texts also publish the majority of texts used in colleges and universities. Although this should not be problematic, Some scholars of education have argued that what is taught in schools is not objective, but reflects the interests of certain powerful social groups (Anyon, p. 362). Concerning textbooks covering social science curricula, general tendencies representative of this aspect of the hidden curriculum have been uncovered. Teitelbaum notes, First, text and trade books reinforce a sense of tradition, culture, and history that favors the material interests and ideological perspectives of white middle-and upper-class males (p. 135). This ability to control the dissemination of information and the legitimating of certain knowledge over other types are powerful means to maintain 66

PAGE 75

Appendix A: (Continued) control and power (Anyon, p. 362). The end result of the hidden curriculum entails the transmission of the dominant cultures knowledge, ideals, and beliefs to the next generation of students. Statement of Purpose The purpose of this research project is to examine and analyze how ten, Introduction to Sociology, textbooks portray and describe the topic of religion and two types of religious institutions, established religions and new religious movements, in their specific chapters concerning religion. I also want to examine and analyze how these textbooks employ, or do not employ, media logic and the hidden curriculum in presenting and transmitting information concerning established religions and new religious movements. Additionally, I want to link the coding protocol, theory, and research design of my thesis on a print media analysis of religion to this research project concerning textbooks to demonstrate the utility of my initial research design and its potential for further application. Research Questions and Theory This research project is connected to the larger thesis that is tasked with examining and analyzing print media articles about established religions and new religious movements in relation to the theoretical components of David Altheides theory of media logic. Are established religious groups and new religious movements subject to different treatment in print media coverage? This is the research question that prompted my interests in the intersections between the media, culture, and religion. As I soon discovered, print media reporting always is selective and there is a sort of media logic whereby the media decides what is worthy of reporting (Altheide 1991 and 1985). For 67

PAGE 76

Appendix A: (Continued) this appendix, I want to examine and analyze the ten textbooks and chapters on religion according to the concepts and ideas of media logic and the hidden curriculum. This research project is located within the symbolic interactionist and conflict perspectives. Essentially, the symbolic interactionist perspective maintains that reality is socially constructed through a process of symbolic interaction. This research project will employ elements of David Altheides theory of media logic and apply this theoretical framework to the topic of religion coverage in Introduction to Sociology textbooks. Utilizing this perspective allows an examination and analysis of the textbooks and religion chapters in which descriptors, images, educational angles, and educational themes delineate the treatment of religion and religious groups, either established or new, and their coverage, either positive or negative. The conflict perspective maintains that social behavior is best understood in terms of conflict or tension between competing groups. This research project will employ elements of Henry Girouxs radical pedagogy and the discourse of textual analysis (Giroux 1983 and 1997) and Jean Anyons concept of ideology (Anyon 1979). By utilizing these theoretical tools and concepts, an examination and analysis of the hidden curriculum is possible in the sample selected for this research project. Hypothesis Is a hidden curriculum evident in Introduction to Sociology textbooks and their treatment of the topic of religion? This question guides this research project. Although prior research has revealed anti-new religious movement bias in other mediums like newspapers (Hill, Hickman, and McLendon 2001; Hubbard 1998; Richardson and van Driel 1988 and 1997; Wright 1997), no study exists that looks at new religious 68

PAGE 77

Appendix A: (Continued) movements, established religions, and their presentation in Introduction to Sociology textbooks. Furthermore, no study exists that analyzes Introduction to Sociology textbooks in light of their treatment of religion, the hidden curriculum, and media logic. This absence of research is surprising given the amount of research conducted by scholars on the hidden curriculum in schools and academia. I hypothesize that the chapters on religion found in Introduction to Sociology textbooks will compare similarly, in relation to chapter length and visual aids, with the other chapters constituting the social institutions sections of these textbooks. This hypothesis is based on determining how the format of religion chapters compares, or does not compare, with other social institution chapters. Since religion is an important social institution, I have hypothesized that the chapters on religion will compare similarly to other social institution chapters. I hypothesize that the textbook chapters on religion will employ angles other than general information. This hypothesis is based on David Altheides prior research using media logic where multiple angles other than general information were used by information mediums to capture audience attention (1985; 1995; Altheide and Snow 1991). I hypothesize that established religions will receive more coverage and mentions of specific groups and members than new religious movements in the chapters on religion. This hypothesis is based on the findings of prior research collected for the main thesis related to bias (Hill, Hickman, and McLendon 2001; Hoover 1998; Hubbard 1998; ODonnell 1993; Richardson and van Driel 1988 and 1997; Wright 1997). I hypothesize that violence, both actual and potential, by and against religious groups and members will be an issue and angle discussed in the textbooks chapters on religion. This hypothesis 69

PAGE 78

Appendix A: (Continued) stems from the research findings of Hill, Hickman, and McLendon (2001) and David Altheide (1995). I also hypothesize that the textbooks discussion of new religious movements and members will be more critical and prone to use negative descriptors than discussions of established religions and members. This hypothesis is based upon prior research findings that have documented negative bias, critical treatment, and negative descriptors used in discussion of new religious movements (Hill, Hickman, and McLendon 2001; Richardson and van Driel 1988 and 1997; Wright 1997). Definitions of Terms To capture the specific usage and meaning of concepts and variables contained within this research project will require adequate definitions. Concerning definitions for established religions and new religious movements, Religion in the Modern World by Steve Bruce utilizes a religious typology where established religious groups in America are classified as denominations and new religious movements are classified as cults (1996). For clarification, it should be noted that the term cult, which is often used disparagingly, is virtually synonymous and interchangeable with the more value-free term new religious movement (Bruce 1996). Utilizing these typologies for this research project, established religious groups possess and/or manifest the characteristics of the denomination, and new religious movements possess and/or manifest all or many, due to their highly individualistic nature, of the characteristics of the cult. The use of religious typologies is helpful for gaining understanding of what essential characteristics constitute established religious groups and new religious movements, and the use of religious typologies provides the important religious, social, and political distinctions that are 70

PAGE 79

Appendix A: (Continued) lacking without its inclusion. Additionally, concise definitions of terms used throughout this research project are also included. Established Religion. Denominations are religious bodies with a professional clergy. They are relatively undemanding, large, inclusive religious bodies associated with the comfortable classes, and although people may join, many people are members because they were born into the group (Bruce, p. 75). New Religious Movement. Cults are small, loosely knit religious groups that are highly individualistic, meager in resources, often short-lived, unable to command members obedience, and lacking a sharply defined and exclusive belief system (Bruce, p. 82). Textbook. Textbooks are social products that can be examined in the context of their time, place, and function (Anyon, p. 361). A textbook refers to specific books that hold themselves out as instructional aids, texts, for a specific body of knowledge. For the purposes of this research project, the term textbook refers to Introduction to Sociology textbooks designed for instructing students of sociology at the college or university level. Hidden Curriculum. The term hidden curriculum is used to describe the implicit demands (as opposed to the explicit obligations of the visible curriculum) that are found in every learning institution and which students have to find out and respond to in order to survive within it (Snyder 1971). Ideology. Ideology is defined here as an explanation or interpretation of social reality which, although presented as objective, is demonstrably partial in that it expresses the social priorities of certain political, economic, or other groups. Ideologies are weapons of group interest; they justify and rationalize; they legitimate group power, activities, and needs (Anyon, p. 363). 71

PAGE 80

Appendix A: (Continued) Radical Pedagogy. An approach to the pedagogical experience that recognizes and questions the different layers of meaning and struggle present in educational institutions with an aim to critique and develop new possibilities for classroom instruction (Giroux 1997, p. 122). Discourse of Textual Analysis. The purpose of this approach is to provide teachers and students with the critical tools necessary to analyze those socially constructed representations and interests that organize particular readings of curriculum materials (Giroux 1997, p. 137). This approach analyzes textbooks with a view to uncovering the meanings, contradictions, and differences present in the form and content of the text (Giroux 1997, p. 137). Media Logic. In general terms, media logic consists of a form of communication, the process through which media present and transmit information. Elements of this form include the various media and the formats used by these media (Media Worlds, p. 9). Format. Formats are rules and procedures for defining, selecting, organizing, and presenting information and other experiences (Media Power, p. 9). Medium. Textbooks are a medium. Altheide states, A medium is any social or technological procedure or device that is used for the selection, transmission, and reception of information (Media Worlds, p. 10). Manner. Manner refers to the way subject matter is operationally presented. Manner is the specific instance being presented, the example, the identifiable feature of the story line, what the message is about (Media Power, p. 20). Themes. Themes are general statements or images that may be used to connect one report to another to provide unity to the topics being presented (Media Worlds, p. 141). 72

PAGE 81

Appendix A: (Continued) Angles. Angles are like themes, except they are used to illustrate certain themes that may or may not be made explicit (Media Worlds, p. 142). Methods and Data A classical content analysis research design, where non-numerical written records of communication are reduced to quantitative data, was used. This method is applicable for this inquiry because previous similar research has utilized this method (Anyon 1979 and Teitelbaum 1991), it is non-reactive, and it avoids any ethical considerations. The purposive sample to be examined for this study includes ten Introduction to Sociology textbooks that explicitly meet the definition for textbooks. The textbooks to be utilized have all appeared in the seven-year period between 1997 and 2003. This research project extracts the specific chapters on religion included in each of these textbooks for analysis. Additionally, the textbooks chapters on other social institutions are analyzed for comparison with the chapters on religion. This research project specifically scours the sample textbooks for all mentions of religion of either type pertinent to this study and utilizes quantitative data collection. Following the same approach utilized in the larger thesis, each textbook chapter on religion is coded for the frequency of religious groups or members mentioned, the frequency and type of violence by groups and members mentioned, if at all, and the frequency of each descriptive term used in relation to religious groups and members. Additionally, each textbook chapter on religion is scrutinized as to its prominence within the text, accompanying visual aids, discussion of violence, format, themes, and angles. This information provides the data for further analysis. 73

PAGE 82

Appendix A: (Continued) Descriptive terms referring to religious groups and individual people are categories for data collection. It is necessary to code for descriptors because the terms selected for this research project carry a specific connotation: legitimation, neutrality, or illegitimacy. By recording the frequency of specific descriptors used to discuss a specific religious group and/or member, it is possible to analyze how these religious groups and/or members are described, portrayed, and socially constructed in these textbooks. Concerning operationalization, the specific descriptive terms this research project looked for regarding religious groups are: cult, sect, movement, group, church, religion, faith, congregation, and an other category. The specific descriptive terms this research project looked for regarding individuals are: terrorists, militant, adherent, pedophile, follower, member, worshiper, leader, an official religious title, and an other category. These descriptors are the same descriptors used in the larger thesis coding protocol. Textbook Research Findings The research findings concerning the first hypothesis mentioned, that the Introduction to Sociology textbooks chapters on religion will compare similarly, in relation to chapter length and visual aids, with the other chapters constituting the social institutions sections of these textbooks, will now be discussed. David Altheides theory of media logic is concerned with the way a medium, in our case a textbook, transmits information to the reader (Media Worlds, p. 9). By utilizing formats, textbooks are able to convey information, and the appearance of the formats selected for disseminating information take on a specific form (Media Power, p. 9). Formats can refer to information provided in textbooks outside the actual body of the text. Format here refers 74

PAGE 83

Appendix A: (Continued) to the prominence, length, and accompanying visual aids that are included in the chapters on religion and social institutions in the textbook sample, and this emphasis on coding the form, or format, of these textbooks is central to Girouxs discourse of textual analysis (Giroux 1997, p. 137). I selected and coded for six variables related to this usage of format: chapter prominence, chapter length, photographs, information boxes, chart and/or graphs, and any other visual aids relating to the chapters on religion and social institutions found in the sample. For chapter prominence, I coded each of the textbooks chapters on religion in relation to where that chapter was located in the text: first third of text, second third of text, and last third of text. Of the ten textbooks coded, eight were found to contain their chapter on religion in the last third of the textbook, and the other two texts were found to have their chapter on religion in the second third of the text. Because religion is a component of the social institutions section of these textbooks, and the other chapters on social institution topics were found to be grouped together with the chapter on religion, it can be concluded from the data that religion, and the chapters on topics included in the social institution sections of these textbooks, was not deemed a topic worthy of prominent inclusion in these textbooks. For the analysis of chapter length and accompanying visual aids in the chapters on religion and the other social institution chapters contained in each textbook, a figure of the relevant findings was generated. Figure 1 represents the averages for the ten chapters on religion compared with the averages found for the thirty-seven chapters on other social institutions found in this sample. The findings show that the average number of pages, photographs, information boxes, charts and/or graphs, and any other visual aids, 75

PAGE 84

Appendix A: (Continued) like maps or cartoons, is very similar for chapters on religion and the other social institutions chapters, if not actually greater. The mean number of pages for the religion chapters is 31, and 28 pages for the other social institution chapters. The mean number of Figure 1: Chapter Features for Chapters on Religion and Other Social Institutions Chapter TopicReligion ChapterOther InstitutionAverage Frequency Per Chapter403020100 PagesPhotosInfo BoxesGraphsOther Aids photographs is 11 for religion chapters, and 9 photographs for the other social institution chapters. The mean for both chapter samples concerning information boxes is the same, four. For charts and/or graphs, the mean for the religion chapters is 5.8 and for the other social institutions it is 5.9. The mean number of other visual aids is the same for both samples, three. These findings lend support to my hypothesis concerning the chapters and are evidence that religion is by no means a marginalized social institution, or topic, in the textbook sample. For my second hypothesis, I hypothesized that established religions would receive more coverage, mentions of specific groups and members, than new religious movements 76

PAGE 85

Appendix A: (Continued) and members in the chapters on religion. I coded my textbook chapters on religion to specifically count each mention of a religious group and each mention of a religious group member for both established and new religions. My findings are presented in Table 22. In the textbook sample, there was a great deal of variation among the individual texts for these variables. For established religions, a mean of 149.6 mentions of established religious groups was coded for the sample, with a standard deviation of 70.7. For established religious members, a mean of 152.4 mentions of established religious Table 22: Frequency of Established and New Religion Groups and Members for Textbooks N Minimum Maximum Mean Std. Deviation ESTRELIGION 10 56 244 149.60 70.724 ESTRELMEM 10 42 245 152.40 62.415 NRM 10 2 74 26.80 22.180 NRMMEMBE R 10 3 38 18.90 11.976 members was coded for the sample, with a standard deviation of 62.4. These large mean scores for established religions and members contrasts sharply with the findings for mentions of new religious movements and members. The data recorded for new religious movements shows a mean of 26.8 mentions of new religious movements, with a standard deviation of 22.2, and a mean of 18.9 mentions of new religious movement members was coded for the sample, with a standard deviation of 12. Established religions were mentioned a minimum of 56 times in a specific textbook to a maximum of 244 times. These data are significantly different from the findings for new religious movements where the mentions of groups fluctuated from a minimum of 2 to a maximum of 74. Established religion members were mentioned a minimum of 42 times to a maximum of 245 times among the textbooks, and significant disparity is noticeable for mentions of new religious movement members with a minimum of 3 mentions and a maximum of 38. 77

PAGE 86

Appendix A: (Continued) From these findings, it is possible to note that established religions and established religion members are mentioned with significantly greater frequency than new religious movements and new religious movement members in this textbook sample. These findings support my hypothesis, and these findings support the hidden curriculum contention that textbooks are more concerned with disseminating information pertaining to dominate culture, in this case established religion. Henry Girouxs discourse of textual analysis allows for an analysis of visible privilege and the contrast of silence on a subject (1997, p. 137). From the data, the textbooks religion chapters engage in a discourse privileging established religions and members, and although new religious movements and members are mentioned in these texts, the low frequency of mentions is tantamount to silence. To further analyze the disparity between frequency of mentions of established religions and members and new religious movements and members, it is necessary to view the data from the variables dealing with the ratios of religious groups and members. Utilizing SPSS, I was able to establish the ratio of new religious movements mentioned in the chapters on religion to established religions, and I was able to establish the ratio of new religious movement members to established religion members. As Table Table 23: Mean of Ratios of Religious Groups and Members for Textbook Sample N Minimum Maximum Mean Std. Deviation RATIO GROUP 10 2.70 48.05 18.8032 14.20073 RATIO MEMBE R 10 2.99 30.00 13.1593 8.49823 23 on mean ratios depicts, the mean ratio of mentions of new religious movements to established religions is 18.8, with a standard deviation of 14.2, and the mean ratio of mentions of new religious movement members to established religion members is 13.2, with a standard deviation of 8.5. This mean data concerning the ratios of mentions of new and established religions and members is powerful evidence that new religious 78

PAGE 87

Appendix A: (Continued) movements and members are mentioned fewer times in the textbook chapters on religion than established religions and members. From the mean data, it is possible to say that this sample of textbooks mentions a new religious movement once for about every five mentions of an established religion, and new religious movement members are mentioned once for about every eight mentions of an established religion member. The unequal coverage of new religious movements and members among the ten different textbooks will now be discussed. A remarkably large maximum ratio of 48 mentions of new religious movements to established religions in The Essential Sociologist is contrasted by the minimum ratio of 2.7 in Sociology in Our Times. The ratio of new religious movement members to established religious members ranges from a maximum of 30, again in The Essential Sociologist, to a minimum of 3 in the Essentials of Sociology by Lindsay and Beach. While none of the textbooks in the sample comes close to providing new religious movements or members an equal, or even proximate, number of mentions, The Essential Sociologist provides the largest ratios of new religious movements and members to established religions and members. Unfortunately, the ratios found for the other nine textbooks supports the assertions of the hidden curriculum argument and Girouxs discourse of textual analysis contention that differences are evident in the high frequency of the texts discussion of privileged groups, established religions and members, and the low ratio and few mentions of non-privileged groups, new religious movements and members (1997, p. 137). For my third hypothesis, I hypothesized that violence, both actual and potential, by and against religious groups and members would be an issue and angle discussed in the textbooks chapters on religion. I coded each of the textbooks chapters on religion 79

PAGE 88

Appendix A: (Continued) for the frequency of specific mentions of actual violence and the potential for violence occurring by/or against specific established religions and members and new religious movements and members. Concerning the religion chapters discussion of violence and potential for violence by/or against religious groups and members, Table 24 depicts the total frequency of mentions for each variable. For mentions of actual violence by a Table 24: Frequency of Actual and Potential Violence By and Against Groups and Members N Minimum Maximum Mean Std. Deviation ACTUAL VIOLENCE BY RELIGIONGROUP/MEMBE R 10 0 29 16.40 9.107 POTENTIAL FOR VIOLENCE BYRELIGION GROUP/MEMBE R 10 6 22 13.50 4.994 ACTUAL VIOLENCE AGAINST RELIGION GROUP/MEMBER 10 1 12 6.30 3.945 POTENTIAL FOR VIOLENCE AGAINST RELIGION GROUP/MEMBER 10 1 10 5.20 2.821 religious group, Sociology in Our Times represents the minimum score of zero and Introduction to Sociology, Fourth Edition represents the maximum score of 29; however, due to the rather frequent discussion of actual violence by religious groups and members throughout the sample, a mean score of 16.4 is found for the sample, with a standard deviation of 9 mostly attributable to the one outlier discussed. As noted earlier, the mean number of pages for the religion chapters is 31, so with a mean score of 16 for frequency of mentions of actual violence by religious groups and/or members, this works out to a mention of actual violence on every other page of the sample chapters. This appears to be a rather high frequency of mentions of actual violence because although religions and members of religions have engaged in violence, religion is traditionally and popularly regarded as advocating peace (Hoover 1998). All the textbooks in the sample discussed the potential for violence by religious groups and/or members with a minimum of six and a maximum of 22 mentions, with a 80

PAGE 89

Appendix A: (Continued) smaller standard deviation of five. The mean for this variable is still rather high at 13.5 mentions per chapter. If the means for the variables actual and potential violence by religion group/member are added together they equal 29.9, and this works-out so that a mention of actual or potential violence by religious groups and/or members occurs on average on every page but one in the chapters on religion. Considering that it is assumed that the information provided in textbooks is objective (Anyon, p. 362), it is obvious that the sample discusses violence, actual or potential, by religious groups and/or members, and this discussion of violence supports my hypothesis that actual and potential violence by religious groups and members would be mentioned in these religion chapters. I also created variables that recorded the specific frequency of mentions of violence and the potential for violence by established religions and their members and new religious movements and their members. Concerning the hidden curriculum, I wanted to find out the frequency of mentions of violence, potential and actual, attributable to established religions and members and new religious movements and members. As expected, due to their more frequent mentioning in the textbook chapters on religion, Figure 2 depicts actual violence and the potential for violence as occurring with a greater frequency to established religions and their members. There are interesting findings, however, concerning this data in relation to new religious movements. Although new religious movements were mentioned infrequently in the textbooks, all the textbooks in the sample mention examples, or an instance, of the potential for violence attributable to new religious movements or their members. Additionally, seven of the ten textbooks discuss actual violence attributed to new religious movements or their members at least twice in their chapters on religion. 81

PAGE 90

Appendix A: (Continued) Figure 2: Textbook Frequency of Actual and Potential Violence by Religion Group/Member TEXTThe Essential SociolSociology: The StudySociology: DiscoveriSociology: A Brief ISociology in Our TimSociology 8th Ed. ScSociology 7th Ed. JoIntro to Soc. 4th EdEssentials of Soc. LEssentials of Soc. JMean3020100 ACTVIOERACTVIONRPOTVIOERPOTVIONR Concerning Anyons definition of ideology, the mentioning of actual, or the potential for, violence rather frequently in light of the infrequent mentioning of new religious movements or their members in this sample is representative of the negative ideology assigned to new religious movements and their members by dominant, more powerful, social groups. Textbooks are supposed to interpret social reality objectively; however, the rather frequent mentioning of violence, or the potential for violence, in relation to infrequently mentioned new religions or members rationalizes the perceived illegitimacy of these groups by the dominant, larger society. This negative presentation of new religions and their members serves to justify the legitimacy of established religions and their members in society and their prerogative to commit, or have the potential for, violence. 82

PAGE 91

Appendix A: (Continued) For violence perpetrated against religious groups and members, Table 24 is again helpful. Actual violence against religious groups and their members was discussed between a minimum of one time to a maximum of twelve times, for a mean of 6.3 and a standard deviation of 4, for the textbooks in the sample. The potential for violence against religious groups and their members was discussed between a minimum of one time to a maximum of ten times, for a mean of 5.2 and a standard deviation of 2.8, for the textbooks in the sample. Unlike the discussion of violence perpetrated by religious groups and members, violence perpetrated against religious groups and members was discussed in every textbook in the sample. An interesting discovery is noticeable in reference to the discussions of violence by and against religious groups and members, for mentions of actual violence or the potential for violence by religious groups and members are discussed at a mean frequency 2.5 times more often than textbooks discussion of violence or the potential for violence being perpetrated against religious groups or their members. This finding is interesting in that it supports the view that the textbooks of the sample are, on average, much more concerned with presenting religious groups and members as the perpetrators of violence or the potential initiators of violence. Due to the years of publications for this sample, 1997 to 2003, the focus on violence and the potential for violence for established religious groups and members could be associated with the wave of fundamentalist religious violence and potential for violence that has captured the attention of the media and world community. References to violence and the potential for violence were numerous in relation to current fundamentalist conflicts occurring around the globe involving Muslims, Hindus, Jews, and Christians fighting among themselves. However, the references to violence or the 83

PAGE 92

Appendix A: (Continued) potential for violence by new religious movements and their members generally involved violence or the potential for violence being directed at the group members themselves. The suicides of Heavens Gate members, the murder-suicides at Jonestown, and the deaths of many Branch Davidians at their Waco compound were discussed in nearly all the textbooks, and these violent depictions of new religious movements and their members serves the dominant ideology that discredits these movements and their members as illegitimate, dangerous, and kooky (Hubbard 1998; Wright 1997). With textbooks published after the September 11 th tragedy, the presentation of Islamic fundamentalists is often focused around violence or the potential for violence, but outside of Islam, violence and the potential for violence are not expanded on in detail after a brief mention in relation to established religions and their members. And, these groups and members are generally discussed in relation to some other feature, like beliefs, practices, number of members, holy literature, and founder(s), but a reference to violence or the potential for violence may entail the only discussion of a new religious movement or member in the chapter. This disparity in the discussion of violence and the potential for violence for established religions and their members on the one hand and new religions and their members on the other is characteristic of the hidden curriculum. For the specific breakdown of mentions of violence against established and new religions and their members, it will be necessary to view Figure 3 depicting the frequency of violence against these groups and members. Due to the conflicts and potential conflicts with dominant society arising from the marginalized beliefs and practices of many new religions, I expected to find ample amounts of examples discussing violence or potential violence against new religions and their members, and Figure 3 illuminates some 84

PAGE 93

Appendix A: (Continued) interesting findings. The discussion of violence against and the potential for violence against the two religious groups and their members is highly uneven, but violence directed against established religions or their members is discussed at least once in every text and the potential for violence against established religions or their members is discussed at least once in every textbook but one. Because established religions are Figure 3: Textbook Frequency of Actual and Potential Violence Against Religion Group/Member TEXTThe Essential SociolSociology: The StudySociology: DiscoveriSociology: A Brief ISociology in Our TimSociology 8th Ed. ScSociology 7th Ed. JoIntro to Soc. 4th EdEssentials of Soc. LEssentials of Soc. JMean1086420 VIOAGERVIOAGNRMPOTAGERPOTAGNRM discussed with a higher frequency in the texts and were mentioned more in reference to perpetrating violence or potential violence, it is not surprising that established groups and members would dominate the mentions of violence or potential violence against religious groups or members. What is interesting is that in three textbooks new religions or their members receive more mentions of violence, or potential violence, directed against them. This finding could be taken as representative of the confrontations and conflicts that often 85

PAGE 94

Appendix A: (Continued) embroil marginalized new religions; however, one textbook does not mention any violence of either type directed against new religions and five textbooks clearly over-represent violence, and potential violence, against established religions. These findings again demonstrate that the dominant ideology is presented in the majority of these textbooks. Dominant religious groups, established religions and their members, receive preferential treatment and capture the bulk of discussion. Concerning the chapter angles used in these chapters on religion, I coded for eight different variables: official pronouncements, practice of worship, information, promotion, apocalyptic and/or millennial beliefs, clergy as victimizers, violence, and an other angle category. Angles are important from a media logic standpoint because they provide evaluation, through the manner of presentation, to the material being presented. From a hidden curriculum standpoint, angles can be the locus of a neutral, positive, or negative portrayal of a group or person by discussing them in relation to a specific angle. In this appendix, I broke each chapter on religion down into small units of text, which was achieved by isolating the text following each individual chapter heading, subheading, or information box. I then proceeded to analyze the portions of text utilizing the same procedure and protocol utilized in the larger thesis. For clarification, it will be necessary to discuss how each of the chapter angle variables was operationalized. Text angles focusing on the dogma, rules, and/or announcements related to religious groups or their members were coded as official pronouncements. The coding category practice of worship refers to a text angle focusing on the rites, general beliefs, and rituals of religions and religious groups members. Information refers to a text angle devoted to education and presenting, 86

PAGE 95

Appendix A: (Continued) discussing, and/or explaining items, concepts, and/or religious groups and their members. The coding category promotion is assigned to a text angle that explicitly or implicitly promotes religion in general, some aspect or feature of religion, and/or a specific religious group or members of a religious group. Apocalyptic and/or millennial beliefs refers to a text angle that deals specifically with beliefs of this sort, such as a final battle between good and evil or the return of Christ at the millennium. I wanted to code for this variable because the sample spans the turn of the millennium. For text angles focusing on instances in which clergy members or religious leaders were the perpetrators of crimes against church members or other persons, I assigned the coding category clergy as victimizers. An example could be the discussion of clergy members sexually abusing their church members. The coding category violence refers to a text angle that is concentrated on violence, actual or potential, and an example could be a discussion of the Jewish Holocaust. The other coding category was utilized as a variable to capture any text angle that was not applicable to the other variables. Figure 4: Mean Frequencies of Angle Categories in Textbook Sample OTHERAVIOLENCECLERGYVAAPMLBELAPROMOTEAINFOAPRAWORANumber of Sections WMean20100 87

PAGE 96

Appendix A: (Continued) Figure 4 represents the mean frequency for the eight angle categories utilized in this appendix. I hypothesized that the textbook chapters on religion would employ angles other than general information. I also hypothesized that violence, both actual and potential, by religious groups and members and against religious groups and members, would be an angle discussed in the textbooks chapters on religion. Because the sample consisted of textbooks, I based my hypothesis on the assumption that information would be the main angle employed in the chapters on religion, and this hypothesis is supported by the data that shows information as the angle with the greatest mean, which is about twice that of the next closest mean. As hypothesized, there was another angle discussed with a moderately high frequency, practice of worship, but I expected to see more than just this one angle. As for my hypothesis concerning the violence angle, the data shows that it occurred with the third highest mean frequency in the chapters on religion; however, the mean frequency of violence as an angle was less than five times per chapter, which is hardly a high frequency. It appears that violence is a more important issue for specific discussions of religious groups and members than as a text angle. Figure 4 also shows that official pronouncements, apocalyptic/millennial beliefs, and clergy as victimizers were rarely used text angles. Additionally, I expanded my coding protocol of the religion chapter angles to include coding for the religious groups and members, either established or new religions. This is important for it distinguishes which specific angles are being used to illustrate, or frame, the texts discussion of religious groups and their members and the frequency. As Figure 5 depicts, established religions and their members were discussed with higher frequencies than new religions and their members in these four variables: official 88

PAGE 97

Appendix A: (Continued) pronouncements, practice of worship, information, and promotion. Interestingly, new religions were not the focus of a discussion concerning official pronouncements in any of the textbooks. From these data, it is easy to see that the dominant groups, established religions, are the recipients of the bulk of the discussion. Figure 5: Mean Frequencies of Textbook Angles for Established and New Religions (1) PROMONRMPROMOESTINFONRMINFOESTRPRACTNRMPRACTESROFFPRONROFFPROERMean121086420 Figure 6 depicts the angle variables for apocalyptic/millennial beliefs, clergy as victimizers, violence, and the other category, which specifically dealt with established religions. In light of certain text units not containing any mention of a specific religious group or member, a category for no identifiable religion was coded. Obviously, the mean for no identifiable religion is predominate for this grouping of variables and is the second largest mean score after information. This finding is due to the large amount of text angle units that discussed religion and aspects of religion from a general, non-specific standpoint. Apocalyptic/millennial beliefs produced the same finding of one angle each for the entire sample. For clergy as victimizers, new religions 89

PAGE 98

Appendix A: (Continued) were compared to this angle twice for every one use of this angle with an established religion. Compared to the entire sample, this finding cannot be considered that important. Figure 6: Mean Frequencies of Textbook Angles for Established and New Religions (2) NOIDRELOTHERESTVIONRMVIOLESTRCLERGYNRCLERGYERAPOCMNRMAPOCMESTMean86420 The violence angle is interesting in that Figure 6 shows that the mean of violence as an angle is about twice per textbook for established religions and once per textbook for new religions. This is the only angle variable where new religions compare closely with established religions, but the violence angle also happens to be a negative angle. It is an interesting finding that new religions only compare closely with established religions in textbook angles in the violence category, for this alignment with text units discussing violence perpetuates and implies the dominant ideology that new religions and their members are violent, or capable of violence, and this softens the depiction of established religions and their members as violent, or capable of violence. In sum, these findings tend to support the assertions of the hidden curriculum, Girouxs discourse of textual analysis, and Anyons notion of ideology in that marginal groups in society are 90

PAGE 99

Appendix A: (Continued) overlooked in school material in favor of discussion of dominant groups, and when discussed, marginal groups are referenced negatively. The primary chapter theme is fairly easy to distinguish with the aid of the data collected for the text angles. The primary chapter theme variables refer to the primary topic of discussion concerning religious groups and religious group members for the religion chapters: conveying general information, ideological discussion, conflict, or sex. From the predominance of information as the main chapter angle utilized in all ten textbooks, I believe that this finding supports an assertion that the theme of conveying general information is accurate for all the textbooks in this sample. The theme of conveying general information was used by the textbooks of this sample to connect the chapters on religion with the other chapters in these textbooks, and this theme is representative of the theme one expects from a textbook, which is a concern for the dissemination of information. The specific information and content that is transmitted by these textbooks, however, is a point of contention. For religious group descriptors, I coded each instance of a specific descriptor mentioned in relation to a specific religious group, established or new, in the textbook chapters on religion. As discussed in the methods and data section, the group descriptors utilized for this research project can be grouped according to whether they convey legitimation: church, religion, congregation, neutrality: faith, movement, group, or illegitimacy: cult, sect. I hypothesized that the textbooks discussion of new religious movements and members will be more critical and prone to use negative descriptors than discussions that are concerned with established religions and members, and by analyzing the following data it will be possible to see if this claim is supported. 91

PAGE 100

Appendix A: (Continued) Figure 7 depicts the mean frequency of the group descriptors of the textbook sample. The greatest mean frequency, over 100, is attributable to the other descriptor variable. This is not surprising due to the extremely large pool of potential descriptors available for use in describing a religious group. In light of the fact that I was looking for Figure 7: Mean Frequencies of Group Descriptors for Textbook Sample OTHERDESCONGREGAFAITHRELIGIONCHURCHGROUPMOVEMENTSECTCULTMean120100806040200 a relatively small number of variables, the variables I did code for turned up interesting findings. The legitimation variables of religion and church and the neutral variables of group and movement had the highest mean scores, which tends to support the notion that textbooks are engaged in value-free discussion of religious groups. However, the mean frequency of cult is sizeable and actually larger than the mean frequencies of the faith and congregation variables. Additionally, the mean frequency of sect is about the same as the mean frequency for faith and larger than the frequency for congregation. Obviously, these textbooks do employ negative descriptors. To further analyze these findings it will be necessary to look at these variables in relation to the 92

PAGE 101

Appendix A: (Continued) specific religious group that they were used to describe, either established or new religions. As Figure 8 depicts, a disparity is found between the application of certain descriptors to established and new religions. For the descriptors conveying illegitimacy, cult was a descriptor used for new religions at twice the mean frequency of its use to Figure 8: Mean Frequencies of Group Descriptors According to Religion Group/Member CONGRNRMCONGRESTFAITHNRMFAITHESTRELIGNRMRELIGESTCHURCHNRCHURCHERGROUPNRMGROUPESTMOVENRMMOVEESTSECTNRMSECTESTCULTNRMCULTESTMean3020100 describe established religions. Sect, however, was used overwhelmingly to describe established religions. These findings are representative of the popular notion that deviant religious groups of an established variety tend to be called sects, and the value-free term new religious movement has already been discussed as interchangeable with the negative descriptor cult (Bruce 1996). For the neutral descriptors, faith was almost exclusively reserved for describing an established religion, and established religions were described as groups about three times more than new religions. Only in the movement variable did new religions receive a slightly larger amount of these neutral descriptors. The neutral 93

PAGE 102

Appendix A: (Continued) set of descriptors is silent on discussing new religions as a faith, partial in its application of the term group to established religions, and slightly favored new religions in its application of the descriptor movement. An explanation for why the descriptor movement was used with a slightly greater mean frequency for new religions over established religions in the sample could stem from the identification of the term new religious movement with new religions in general. For descriptors that convey legitimacy, all three were used overwhelmingly more in relation to established religions than new religions. In describing established religions, the mean frequency of church was around 28, but the mean frequency for new religions being described as a church was about twice per chapter on religion. This is a ratio of fourteen to one and represents a huge disparity. For the religion variable, the mean frequency for established religions was about 26 mentions per religion chapter, but new religions had this descriptor applied to them only about five times per chapter for a ratio difference of about five to one. New religions were described only twice in the entire sample as a congregation, but this descriptor was used with enough frequency to describe established religions that it has a mean score of three. From these group descriptor findings, it is apparent that this sample of textbooks utilizes negative descriptors of new religions to a high degree, the variable cult, and neutral descriptors, like movement, are the best descriptors that new religions can expect to read in a textbook discussion of their religion. Legitimating descriptors are predominately reserved for use in describing only established religions. Concerning the hidden curriculum, the explicit usage of descriptors that convey and bestow legitimacy or illegitimacy serves to inform students that the groups having 94

PAGE 103

Appendix A: (Continued) these various descriptors applied to them are in fact illegitimate or legitimate. The dominant ideology is served when the dominant groups, in this case established religions, are described with legitimating descriptors and minority groups, new religions, are described with descriptors that convey illegitimacy. Furthermore, the limited application of, or failure to apply, legitimating descriptors to new religions strengthens the dominant ideology that these new religions are illegitimate, and by limiting the use of legitimating descriptors for new religions, the dominant ideology justifies its reservation of applying legitimating descriptors only for real, read established, religions. For religious member descriptors, I coded each instance of a specific descriptor mentioned in relation to a specific religious group member, established or new, in the textbook chapters on religion. As discussed in the methods and data section, the religious group member descriptors utilized for this research project can be grouped according to whether they convey legitimation: an official religious title, leader, worshiper, neutrality: adherent, follower, member, or illegitimacy: terrorist, militant, pedophile. As Figure 9 depicts, the variable for an other member descriptor is the largest with a mean frequency of about 75 for the chapters on religion. As with religious groups, there are a variety of different ways to describe the members of religious groups. The absence of any usage of the variable pedophile for a religious group member is apparent. This variable, pedophile, was included to provide continuity with the larger print media study, but fortunately, none of the sample textbooks utilized this highly negative descriptor. Interestingly, outside of the variable for an official religious title, the mean frequencies for the member descriptors that were coded are often less than the mean score 95

PAGE 104

Appendix A: (Continued) totals for the religious group descriptors. This finding suggests that the sample textbooks emphasized religious groups over religious members. From the mean frequency for an official religious title, about 40, it can be concluded that these textbooks focused a Figure 9: Mean Frequencies of Member Descriptors for Textbook Sample OTHERMDOFFRELTILEADERWORSHIPEMEMBERFOLLOWERPEDOPHILADHERENTMILITANTTERRORISMean806040200 good deal of their discussions of religious group members around recognized officials affiliated with a religious group. Official religious title refers to descriptors like: pope, priest, pastor, reverend, minister, monk, guru, and nun. Mean frequencies for terrorist and militant are low, and this is due to only two and four textbooks respectively utilizing these negative descriptors. Adherent and worshiper were hardly used as descriptors, and leader was only utilized in seven textbooks. The neutral descriptors of follower and member were used with the highest mean frequency for this sample. For further clarification, it will be necessary to analyze the breakdown of each of these variables between the two different religious group members, established and new. 96

PAGE 105

Appendix A: (Continued) As Figure 10 depicts, the highly negative religious member descriptors of terrorist and militant are solely used to describe established religion members, predominately Muslims. The other negative descriptor of pedophile was not used to Figure 10: Mean Frequencies of Member Descriptors According to Religion Group/Member TITLENRMTITLEESTLEADNRMLEADESTWORSHNRMWORSHESTMEMBERNRMEMBERERFOLLOWNRFOLLOWESADHERNRMADHERESTMILITNRMMILITESTTERRORNRTERRORESMean403020100 describe members of religious groups; so established religion members carry the sole distinction of being labeled with negative, illegitimating descriptors. These are interesting findings, but the limited frequency and appearance in the sample of these descriptors limit their utility for broad analysis. For descriptors that convey legitimation, all three descriptors favor usage in describing established religion members. The descriptor worshiper is used at a three to one ratio in favor of established religion members, and leader is used to describe established religion members at a ratio of about two to one. The sample applied an official religious title about seven times more often to an established religion member than a member of a new religion. 97

PAGE 106

Appendix A: (Continued) By utilizing Henry Girouxs discourse of textual analysis, it is readily apparent that disparity exists concerning the usage of descriptors that convey legitimation in the textbook sample. The high prevalence of legitimating descriptors being used in relation to established religion members is indicative of ideology at work that reserves the legitimating descriptors for dominant religious group members. For the neutral descriptors, all three also favor usage with established religion members. Adherent was used to describe new religion members only twice in the entire sample. Concerning the member descriptor, the mean score of this descriptor is at about a two to one ratio in favor of new religions; however, upon analyzing the data, there is one outlier that skews this data and distorts a situation that should be closer to approximation. For the descriptor follower, this variable was used in favor of established religions at a ratio of about three to one. New religions again are provided some descriptive room in the neutral category; however, established religions dominate the neutral and legitimating descriptor categories. In sum, I believe that I found some evidence to support my hypothesis concerning the descriptors, but it is also apparent from the data that, although ideology and differential treatment are present in the textbook sample, established religions and members were described with negative, illegitimate descriptors. Conclusion Although data collection for this project is completed, the analysis of the data and the interpretation of the findings is at an elementary, preliminary stage. Further analysis and interpretation of these research findings will be carried out as the actual thesis itself evolves and provides this researcher greater clarity in interpretation of data. I do believe that I accomplished the initial goals I had for this research project, which were to 98

PAGE 107

Appendix A: (Continued) demonstrate the utility of applying media logic to analyzing textbooks and the applicability of my thesis coding protocol to other forms of content. This research project has attempted to provide an initial foray into the intersection of religion and education. It is hoped that this research project has shed some light on the treatment and description of the topic of religion, established religions, and new religious movements in Introductory to Sociology textbooks. In light of the research findings, this project has been able to make tentative claims based on the hypotheses. From this sample and coding protocol, I was able to amass evidence that could lend support to all of my hypotheses except for the final one concerning the descriptors. I was able to use David Altheides media logic as a theoretical base and link with the larger thesis project, and I was able to incorporate and utilize elements and methodological perspectives from Henry Giroux and Jean Anyon, which are informed by the radical pedagogical tradition and an understanding of the hidden curriculum. Concerning ideas for further research, I believe that this research project has highlighted the need for sociologists of religion and education to explore the content included in textbooks relating to the topic of religion. As with other research, this project will need replication to determine if the tentative findings are supported. Additionally, research should be conducted to expand the scope of this rather limited study. Larger samples of textbooks should be coded, and if possible other textbooks, like Social Problems texts, should be scrutinized. The number of variables relating to the group and member descriptors should be expanded to include more descriptors, which will reduce the rather large findings attributed to the other category variable. This research project could lend itself easily to a longitudinal study incorporating Introductory to Sociology 99

PAGE 108

Appendix A: (Continued) textbooks from time periods earlier than the sample studied here. It would be interesting to code textbooks from different eras to determine how the variables coded for in this study appeared, or do not appear, for previous decades. This coding protocol could also be used to analyze the treatment of religion and religious groups in other disciplines textbooks. Obviously, a hidden curriculum is present in the data uncovered in this sample, so after further replication to validate the findings, researchers should be prepared to pursue investigating the underlying causes for the differences and disparity highlighted in this study. Lastly, teachers and instructors tasked with implementing these curriculum materials must be educated in, and employ, the radical pedagogical techniques necessary for combating the hidden curriculum. 100

PAGE 109

Appendix B: (Continued) Appendix B: Characteristics of Prestige Newspaper Sample This appendix will discuss research findings focused on the prestige newspaper article sample as the unit of analysis. The eighteen-months of the sample generated seventy-two different Sundays, and it is from these Sunday prestige newspapers that articles were coded. From these seventy-two Sundays, I collected 720 different articles from the prestige newspapers of the sample. Based on the sample sizes, this yields an average of ten newspaper articles per Sunday and 3.3 articles per prestige newspaper each Sunday. However, while all prestige newspapers produced at least one article for each Sunday of the sample, the sample fluctuated from a low of four to a high of eighteen articles found per Sunday. Table 25 displays the frequencies and percentages of the Table 25: Prestige Newspaper Frequency and Percentage of Sample Articles Frequency Percent Valid Percent Cumulative Percent Valid New York Times 306 42.5 42.5 42.5 Washington Post 196 27.2 27.2 69.7 Los Angeles Times 218 30.3 30.3 100.0 Total 720 100.0 100.0 sample articles found for the three prestige newspapers. The New York Times accounted for the largest number of articles, 306 or 42.5 % of the sample, and had the highest average of articles per Sunday newspaper at 4.25. The Los Angeles Times, with 218 articles or 30.3 % of the sample, accounted for the second largest number of articles and had an average of 3.03 articles per Sunday newspaper. The Washington Post follows closely with 196 articles, or 27.2 % of the sample, and had an average of 2.72 articles per Sunday newspaper. Clearly, The New York Times predominates this data set. For the prominence of the newspaper articles, or where they were located in the newspaper, 74.2 %, or 534, of the newspaper articles were found on an inside page of the 101

PAGE 110

Appendix B: (Continued) newspaper. Fourteen percent, or 99, of the newspaper articles were located on page one of a newspaper section, and 12.1 %, or 87, of the newspaper articles appeared on the front page of the newspaper. Table 26 displays the frequency and percentages of these Table 26: Cross Tabulation Article Prominence and Prestige Newspaper Newspape r name Total New Yor k Times Washington Post Los Angeles Times Article prominence Inside page Count 250 143 141 534 % withinNewspape r name 81.7% 73.0% 64.7% 74.2% % of Total 34.7% 19.9% 19.6% 74.2% Section page one Count 31 29 39 99 % withinNewspape r name 10.1% 14.8% 17.9% 13.8% % of Total 4.3% 4.0% 5.4% 13.8% Front page Count 25 24 38 87 % withinNewspape r name 8.2% 12.2% 17.4% 12.1% % of Total 3.5% 3.3% 5.3% 12.1% Total Count 306 196 218 720 % withinNewspape r name 100.0% 100.0% 100.0% 100.0% % of Total 42.5% 27.2% 30.3% 100.0% findings. By conducting a cross tabulation of the article prominence with the sample newspapers, the location and distribution of articles sampled from each prestige newspaper can be determined. As depicted in Table 26, The New York Times has the greatest percentage of inside-page articles appearing in the prestige newspaper sample at 34.7 % of the inside-page total, so The New York Times has about a 15 % point difference between it and the other prestige newspapers. This difference is due to The New York Times having over 100 more articles than either of the other newspapers for this variable. 102

PAGE 111

Appendix B: (Continued) The Los Angeles Times, at 17.9 %, has the largest percentage of newspaper section page-one articles. With 17.4 % of its articles appearing on the front-page, The Los Angeles Times has the largest percentage of articles presented on the front page of the newspaper, and there is a 9.2 % point difference between The Los Angeles Times and The New York Times, which has the smallest percentage of front-page articles. This cross tabulation is evidence that having a greater quantity of religion-related articles for a prestige newspaper does not mean a greater quantity of articles to be found with prominent placement. For the origin, or who was responsible for writing the article, Table 27 displays the findings and percentages for the newspaper sample. The New York Times, at 67.3 %, had the largest number of articles written by newspaper staff reporters, and because it has over 100 more articles than the Washington Post for this variable, The New York Times has a 14 % point difference over this paper for the percentage of total. Only a minimal number of articles written by prestige newspaper religion reporters were coded, so pertinent observations cannot be generated for this variable. A probable explanation to account for this paucity of articles written by professional religion reporters is that these religion reporters can be identified as staff reporters, and many prestige newspaper religion reporters have their own columns and/or bylines, which precluded their use in my sample due to my coding protocol restrictions. Concerning articles written by newspaper international reporters, The Los Angeles Times has a negligible amount of these articles, and the other two prestige newspapers do not demonstrate much difference with about 25 % of their articles originating this way. For articles generated by a wire service, like Associated Press or Reuters, The New York 103

PAGE 112

Appendix B: (Continued) Times and Washington Post have similar frequencies and percentage totals. The Los Angeles Times has the highest percentage of wire service articles at 29.8 %, and this is a Table 27: Cross Tabulation Article Origin and Prestige Newspaper Newspape r name Total New Yor k Times Washington Post Los Angeles Times Article origin Newspaper's staf f reporter Count 206 105 132 443 % within Newspape r name 67.3% 53.6% 60.6% 61.5% % of Total 28.6% 14.6% 18.3% 61.5% Newspaper's religion reporter Count 1 4 8 13 % within Newspape r name .3% 2.0% 3.7% 1.8% % of Total .1% .6% 1.1% 1.8% Newspaper's international reporter Count 71 53 2 126 % within Newspape r name 23.2% 27.0% .9% 17.5% % of Total 9.9% 7.4% .3% 17.5% A wire service Count 23 27 65 115 % within Newspape r name 7.5% 13.8% 29.8% 16.0% % of Total 3.2% 3.8% 9.0% 16.0% Other bureau o r newspaper Count 5 7 11 23 % within Newspape r name 1.6% 3.6% 5.0% 3.2% % of Total .7% 1.0% 1.5% 3.2% Total Count 306 196 218 720 % within Newspape r name 100.0% 100.0% 100.0% 100.0% % of Total 42.5% 27.2% 30.3% 100.0% 104 22.3 % point difference over The New York Times, which has the lowest percentage of wire service articles. Only a minimal number of articles that originated with other news

PAGE 113

Appendix B: (Continued) bureaus and newspapers were coded, so pertinent observations cannot be generated for this variable. A probable explanation to account for the paucity of articles generated by other news bureaus and newspapers is that this sample consists of prestige newspapers and these newspapers usually dictate what is news for other news organizations (Altheide and Johnson; Hill, Hickman, and McLendon 2001). The article length variable coded for the precise number of words present in each prestige newspaper article. As Table 28 displays, the minimum amount of words in a newspaper article was 61 and the maximum amount of words for an article in this data set was 5919. The 720 articles included in this data set have a mean of 1025.56 words per article with a standard deviation of 703.05. The mean word counts for the individual prestige newspaper of the sample closely approximate about one-half of a newspaper page of text per article. Table 28: Word Count of Newspaper Article Information N Minimum Maximum Mean Std. Deviation Article length 720 61 5919 1025.56 703.050 Valid N (listwise) 720 Table 29 reports findings for the four different types of visual aids accompanying the prestige newspaper articles that were coded and counted: photographs, information boxes, charts and/or graphs, and an other category, which primarily recorded maps. Photographs were the dominant visual aides used to accompany newspaper articles. Although 325 articles did not include any photographs, there were 760 photographs found accompanying 395 articles, which produces a mean of 1.92 photographs per article. Interestingly, 82 articles were coded that had between three and the variable maximum of twenty-three photographs per article, which represents 11% of the sample. 105

PAGE 114

Appendix B: (Continued) Appearing with only fifteen different articles, information boxes were the least recorded visual aides found in this sample. With only twenty-two found, charts and/or graphs did not appear very frequently in this sample, but multiple occurrences were recorded in two newspaper articles containing two charts and/or graphs. Lastly, forty different articles contained visual aides that could be classified as an other category/map. Table 29: Newspaper Article Information About Visual Aides Photograph Info Boxes Chart/Graph Other/Maps N Valid 395 15 20 40 Missing 325 705 700 680 Mean 1.92 1.00 1.10 1.00 Median 1.00 1.00 1.00 1.00 Mode 1 1 1 1 Std. Deviation 1.829 .000 .308 .000 Minimum 1 1 1 1 Maximum 23 1 2 1 Sum 760 15 22 40 Performing a cross tabulation of the visual aid variables and the prestige newspapers uncovered interesting findings. For the photograph variable, The Washington Post only had 35 photographs, or 8.9 % of the total number of photographs. With 209, or 52.9 %, of the photographs appearing in its newspaper, The New York Times had the most photographs presented for a prestige newspaper. The Los Angeles Times had 151 of the total number of photographs accompany its newspaper articles, or 38.2 % of the total number of photographs. While there were only a negligible number of information boxes, nine of the fifteen, or 60%, were found in The Los Angeles Times. For the charts and/or graphs visual aides, The New York Times accounted for fifteen of the total of twenty-two charts and/or graphs, or 70 % of the total of charts and/or graphs. Concerning the other category/map visual aide, The New York Times and The Los Angeles Times were both similar with each newspaper having about 45 % of the percentage total for this variable. 106

PAGE 115

Appendix B: (Continued) However, The Washington Post newspaper articles only had 12.5 % of the percentage total of other category/map visual aides accompanying its articles. 107

PAGE 116

Appendix C: (Continued) Appendix C: Coding Protocol for Print Media Reporting of Religion Introduction This print media protocol is aimed at assessing the reporting of established religions and new religious movements by prestige newspapers. It examines newspaper articles and descriptive terms used in these articles to determine how the print media presents biased reporting of established religions and new religious movements and their members. The following definitions are important in selecting and analyzing the content under study. Established Religion Established religions are respectable religious bodies with a professional clergy. They are relatively undemanding, large, inclusive religious bodies associated with the comfortable classes, and, while people may join, many people are members because they were born into the group. New Religious Movement New religious movements are small, loosely knit religious groups that are highly individualistic, meager in resources, often short-lived, unable to command members obedience, and lacking a sharply defined and exclusive belief system. Prestige Newspapers Prestige newspapers are self defined as national in scope and they possess international news gathering ability. Furthermore, prestige newspapers have circulations in the millions and are important for dictating what is news for other news organizations. Newspaper Article A newspaper article focuses on religions, either established or new, and/or members of religions, either established or new. Newspaper articles are defined as all non-advertising matter in a newspaper. This includes all staff-produced news articles found in the first and local sections, but excluding editorial pages, op ed pages, reader opinions, sports, routine business data, and society news. It may include relevant features produced by local staff reporters and syndicated and wire services articles relevant to the issue of religion being analyzed. Procedure The following steps should be taken in the content analysis coding described below (V stands for variable): (A) All relevant religion articles are read to identify established religions and new religious movements and their members; (B) All descriptive terms used in the article concerning established religions and new religious movements and their 108

PAGE 117

Appendix C: (Continued) members, leaders and laity, are recorded by the coder; (C) Each article is then analyzed for specific characteristics described below. V1. Newspaper Name Code the newspapers with the associated numbers: New York Times = 1 Washington Post = 2 Los Angeles Times = 3 V2. Article Date Includes: Year, month, and day V3. Article Prominence Code the article locations with the associated numbers: Front page = 2 Section page one = 1 Inside page = 0 V4. Article origin: designated by article byline 1 = Newspapers staff reporter 4 = A wire service 2 = Newspapers religion reporter 5 = Other bureau or newspaper 3 = Newspapers international reporter V5. Article length: designated by word count Identify any visual aids accompanying article and record frequency V6. photograph V7. information box V8. chart/graph V9. other visual aid V10. Primary article subject: designated by what is most discussed in the article, frequency 1 = an established religion 3 = established religion members 2 = a new religious movement 4 = new religious movement members V11. Article discussion of violence: designated by the article discussing violent activity or potential for violent activity attributable to religious groups or members 109

PAGE 118

Appendix C: (Continued) Actual violence = 2 Potential for violence = 1 Violence not mentioned = 0 V12. Article discussion of violence: designated by the article discussing violent activity or potential for violent activity attributable against religious groups or members Actual violence = 2 Potential for violence = 1 Violence not mentioned = 0 V13. Primary article theme: designated by what is the primary topic of discussion in relation to religion, religious groups, and religious members 1 = Conveying general information/news 3 = Conflict 2 = Ideological discussion 4 = Sex V14. Primary article angle: designated by what is the primary angle used to illustrate the article 1 = Official pronouncements 5 = Apocalyptic/Millennial beliefs 2 = Practice of worship 6 = Clergy as victimizers 3 = Information 7 = Violence 4 = Promotion 8 = Other Religious group descriptors: designated by one mention of the following in relation to discussing a specific established or new religion in article and frequency V15. cult V16. sect V17. movement V18. group V19. church V20. religion V21. faith V22. congregation V23. any other descriptor Religious member descriptors: designated by one mention of the following in relation to discussing members of established or new religions in article and frequency V24. terrorist 110

PAGE 119

Appendix C: (Continued) 111 V25. militant V26. adherent V27. pedophile V28. follower V29. member V30. worshiper V31. leader V32. official religious title V33. any other descriptor V34. Newspaper article title: title of newspaper article


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001469374
003 fts
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 040524s2004 flua sbm s000|0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0000276
035
(OCoLC)55644049
9
AJR1128
b SE
SFE0000276
040
FHM
c FHM
090
HM51
1 100
Griffin, Lonnie F.
3 245
An analysis of print media reporting of established religions and new religious movements
h [electronic resource] /
by Lonnie F. Griffin III.
260
[Tampa, Fla.] :
University of South Florida,
2004.
502
Thesis (M.A.)--University of South Florida, 2004.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 119 pages.
520
ABSTRACT: This thesis reports findings from a content analysis of 720 newspaper articles and 3,052 newspaper article cases focused on the issue of print media bias. Sunday editions of three major newspapers were drawn from the six-year period 1998-2003 for analysis. Prior research has uncovered print media bias in reporting of religious groups, and this thesis examines the substance of those claims pertaining to both established religions and new religious movements. Research findings show that established religions and their members are typically described in favorable or neutral terms, while new religious movements and their members are consistently described with pejorative terms. However, specific established religion members received the overwhelming majority of negative religion member descriptors. Articles focusing on established religion members were found to contain the bulk of visual aides accompanying the articles. Newspaper articles discussed incidents of violence by and/or against specific religious groups of both types of religion with a high frequency. Also, newspaper article themes and angles were found to be important for conveying the content of the articles. Additionally, an appendix is included that analyzes the treatment of religion, established religions and their members, and new religious movements and their members in sociology textbooks.
590
Adviser: Cavendish, James
653
sociology.
media logic.
violence.
descriptor.
theme.
angle.
education.
0 690
Dissertations, Academic
z USF
x Sociology
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.276