Chapter Template

Home , The Houma Courier, Trenton (pilot boat)

2013

The Dissertation Committee for Marcus James Funk Certifies that this is the approved version of the following dissertation:

Analyzing Content Deviance in American Community Journalism Websites and Social Media

Committee:

George Sylvie, Supervisor

Maxwell McCombs

Wenhong Chen

Dominic Lasorsa

Gene Burd

Wanda Cash Analyzing Content Deviance in American Community Journalism Websites and Social Media

Marcus James Funk, M.A., B.A.

Dissertation

Presented to the Faculty of the Graduate School of

The University of Texas at Austin

in Partial Fulfillment

of the Requirements

for the Degree of

Doctor of Philosophy The University of Texas at Austin December 2013 Dedication

To my parents, Jim and Darlene, and my brother Quentin.

To my friends Reggie, Brian, Grady, Mehr, Dwindle, Kathleen, Richie, and Melissa.

To my committee and colleagues at the University of Texas at Austin, Dr. George Sylvie, Dr.

Renita Coleman, Wanda Cash, Dr. Maxwell McCombs, Dr. Wenhong Chen, Dr. Dominic

Lasorsa, Dr. Gene Burd, Dr. Kelly Kaufhold, Bill Minutaglio, and Sam Gwynne.

To my professors at Trinity University, Dr. Tucker Gibson, Dr. David Crockett, Dr. Aaron

Delwiche, Dr. Jennifer Henderson, Dr. Peter O‟Brien, and Dr. David Lesch.

Also, special thanks to Katharine Martin, David Jackson, Bruce Simmons, Dr. Thomas Young,

Gifford Nielsen, the staff of Jenn‟s Copies, and the Rotary Club of Terrell, Texas.

Thank you all, from the bottom of my heart.

Analyzing Content Deviance in American Community Journalism Websites and Social Media

Marcus James Funk, Ph.D.

The University of Texas at Austin, 2013

Supervisor: George Sylvie

Abstract: This dissertation explores deviance, operationalized through news factors, among American community weekly, community daily, large daily, and national daily newspaper websites and social media posts. Computerized quantitative analysis indicates that circulation size makes little to no significant difference concerning the publication of deviant news factors; smaller circulation sizes are significantly related to the publication of news concerning local communities, but not egalitarian news factors generally. Qualitative, structured interviews of community newspaper editors and publishers illustrate a different agenda – a clear focus for news on “regular people and routine events,” arguably egalitarianism, over news on “unusual people or extraordinary events,” arguably deviance. This indicates a need for further evaluation and development of computerized content analysis, gatekeeping theory, and the community newspaper industry. Results also suggest a need to reconsider and re-evaluate normative deviance as a concept and point to two potential theoretical developments: considering a Deviant- Egalitarian Spectrum and drastically broadening the current fringe focus of deviance research.

Table of Contents

List of Tables ...... xvi

List of Figures ...... xvii

List of Illustrations ...... xviii

CHAPTER 1: INTRODUCTION 1

1.1: Visualizing the Dissertation ...... 3

1.2: Core Concepts ...... 5

1.3 Importance & Significance of the Study ...... 6

1.4: Hypotheses ...... 8

1.5: Definition of Terms ...... 10

1.5.a: (Normative) Deviant News Factors: ...... 12

1.5.b: Social Significance Factors: ...... 12

1.5.c: Egalitarian News Factors: ...... 13

1.5.d: Media Definitions ...... 13

1.6: Assumptions...... 16

1.7: Delimitations and Limitations of Study ...... 19

1.8 Overview of Complete Dissertation ...... 20

CHAPTER 2: LITERATURE REVIEW 22

2.0: Introduction ...... 22

2.1: Cohesive Argument ...... 24

2.2: Gatekeeping Theory ...... 25

2.2.a: Gatekeeping & The Hierarchy of Influences ...... 25

2.2.b: Theoretical Limitations and Overlap with Other Models ...... 27

2.3: Deviance ...... 30

2.3.a: Conceptualizing Deviance ...... 30

2.3.b: Operationalizing Deviance ...... 32

2.3.c: Contextualizing Deviance ...... 33

2.4: News Factors & News Values ...... 35

2.4.a: News Factors ...... 35

2.4.b: News Values ...... 38

2.5: Community in the Digital Age ...... 39

2.5.a: Imagined Community ...... 40

2.5.b: Community Journalism & Imagined Community ...... 41

2.5.c: Community Journalism in Practice ...... 43

vii

2.6: New Journalisms & Social Media ...... 46

2.6.a: Community Newspapers Online: A Lack of Scholarship ...... 47

2.6.b: Social Media Writ Large ...... 48

2.7: Summary and Literature / Hypotheses Link ...... 50

CHAPTER 3: METHODOLOGY 52

3.0: Introduction ...... 52

3.1: Hypotheses ...... 53

3.2: Methodology Structure ...... 54

3.3: Computerized Content Analysis ...... 55

3.3.a: Quantitative Research Paradigm ...... 55

3.3.b: Quantitative Dataset & Sampling...... 58

3.3.c: Quantitative Data Collection ...... 61

3.3.d: Quantitative Data Processing ...... 64

3.3.e: Abandoning Magnitude ...... 65

3.3.f: Constructing and Importing Word Dictionaries ...... 67

Deviant News Factors: ...... 67

Social Significance Factors: ...... 67

Egalitarian News Factors: ...... 67

3.3.g: Quantitative Data Units ...... 68

3.3.h: DICTION 6.0 Analysis ...... 70

viii

3.3.i: Generalizable Validity ...... 722

3.4: Qualitative Structured Interviews ...... 74

3.4.a: Qualitative Research Paradigm ...... 74

3.4.b: Qualitative Dataset & Sampling...... 76

3.4.c: Qualitative Data Collection ...... 77

3.4.d: Structured Interview Amendments ...... 81

3.4.e: Qualitative Data Processing ...... 81

3.5: Time Line ...... 82

CHAPTER 4: RESULTS 83

4.0: Introduction ...... 83

4.1: Repetition of Hypotheses ...... 84

4.2: Quantitative Results ...... 85

4.2.a: Deviance and Egalitarianism on Traditional News Websites ...... 85

4.2.b: Deviance and Egalitarianism on Social Media ...... 89

4.2.c: Evaluating Quantitative Hypotheses ...... 95

4.3: Qualitative Results ...... 99

4.3.a: Newsroom Size and Characteristics ...... 99

4.3.b: Perceived Journalist / Audience Interaction ...... 1022

4.3.c: Deviant & Egalitarian News...... 104

4.3.d: Assessing The Qualitative Hypothesis ...... 108

4.4: Conclusion of Results ...... 110

CHAPTER 5: GENERAL & THEORETICAL DISCUSSION 111

5.0 Introduction ...... 111

5.1: Quantitative Results Review ...... 112

5.2: Qualitative Results Review ...... 113

5.3: Commonalities Between Quantitative and Qualitative Results ...... 114

5.4: Differences Between Quantitative and Qualitative Results ...... 117

5.4.a: Validation of Media Sociology ...... 118

5.4.b: Hyper-Sensitivity and Non-Contextualized Nature of Computerized Content Analysis Software ...... 119

5.4.c: Gap Between Online and Offline Community Journalism ...... 121

5.5: Reconsidering Normative Deviance ...... 124

5.5.a: Theoretical Discussion & Development of Normative Deviance: A Two Option Approach ...... 124

5.5.b: The Spectrum Option: Developing an Inverse Relationship Between Deviance & Egalitarianism...... 127

5.5.c: The Expansion Option: Broadening Operationalizations of Normative Deviance ...... 129

5.6: Reconsideration of Journalist-Audience Interaction and Gatekeeping Theory ...... 136

5.7: Implications for Journalists and Publishers ...... 138

CHAPTER 6: CONCLUSIONS & OPPORTUNITIES FOR FUTURE RESEARCH 141

6.0: Introduction ...... 141

6.1: Evaluating the Importance & Significance of the Study ...... 147

6.2: Summary of Major Contributions ...... 151

6.3: Opportunities for Future Research ...... 152

6.3.a: Media Sociology & Cognitive Dissonance ...... 152

6.3.b: Tilt, Sensitivity, and Proper Noun Usage in Computerized Content Analysis ...... 154

6.3.c: Online and Offline Community Journalism ...... 155

6.3.d: Inquiring About Deviance: Spectrum and Broadening ...... 157

6.4: Final Thoughts ...... 158

APPENDICES 159

Appendix 1: Map of Regional Categorizations ...... 159

Appendix 2: List of Selected Newspapers by Circulation Category ...... 160

Pacific: ...... 160

Weekly: ...... 160

Community Daily: ...... 160

Large Daily: ...... 160

Rocky Mountain: ...... 161 xi

Weekly: ...... 161

Community Daily: ...... 161

Large Daily: ...... 161

Southwest: ...... 162

Weekly: ...... 162

Community Daily: ...... 162

Large Daily: ...... 162

Great Plains: ...... 163

Weekly: ...... 163

Community Daily: ...... 163

Large Daily: ...... 163

Great Lakes: ...... 164

Small Weekly ...... 164

Community Daily ...... 164

Large Daily ...... 164

Southeast: ...... 165

Small Weekly: ...... 165

Community Daily: ...... 165

Large Daily: ...... 165

Mid-Atlantic ...... 165

Small Weekly: ...... 165

xii

Community Daily: ...... 166

Large Daily: ...... 166

Northern Region ...... 166

Small Weekly: ...... 166

Community daily: ...... 167

Large Daily: ...... 167

National Newspapers: ...... 167

Appendix 3: Newspaper Groups ...... 168

Group A: ...... 168

Group B: ...... 169

Group C: ...... 170

Group D: ...... 172

Group E: ...... 173

Appendix 4: Constructed Week Schedules ...... 175

Group A: ...... 175

Group B: ...... 175

Group C: ...... 176

Group D: ...... 176

Group E: ...... 176

xiii

Appendix 5: Constructed Week Calendar Graphic ...... 178

Appendix 6: Social Media Data Collection Schedule ...... 179

Group A: ...... 179

Group B: ...... 179

Group C: ...... 179

Group D: ...... 179

Group E: ...... 179

Appendix 7: Social Media Data Collection Graphic...... 180

Appendix 8: Structured Interview Script ...... 181

Appendix 9: Full Word Dictionaries ...... 183

1: Prominence...... 183

2: Conflict ...... 184

3: Oddity ...... 186

4: Impact ...... 186

5: Timeliness ...... 187

5.1: Recentness ...... 187

5.2: Dates ...... 187

6: Proximity ...... 188

6.1: General Proximity ...... 188

6.2: Specific Proximity ...... 188

6.2.a: Proximity A.CW ...... 188 xiv

6.2.b: Proximity A.CD ...... 189

6.2.c: Proximity A.LD ...... 189

6.2.d: Proximity A.ND ...... 190

6.2.e: Proximity B.CW ...... 191

6.2.f: Proximity B.CD ...... 191

6.2.g: Proximity B.LD ...... 192

6.2.h: Proximity B.ND ...... 193

6.2.i: Proximity C.CW ...... 193

6.2.j: Proximity C.CD ...... 193

6.2.k: Proximity C.LD ...... 194

6.2.l: Proximity C.ND ...... 195

6.2.m: Proximity D.CW ...... 195

6.2.n: Proximity D.CD ...... 195

6.2.o: Proximity D.LD ...... 196

6.2.p: Proximity D.ND ...... 197

6.2.q: Proximity E.CW ...... 197

6.2.r: Proximity E.CD ...... 197

6.2.s: Proximity E.LD ...... 198

6.2.t: Proximity E.ND ...... 198

Appendix 10: IRB Exemption Notice ...... 199

REFERENCES 201 xv

List of Tables

F1: Word Frequency Means of News Factors on Newspaper Websites by

Circulation Category ...... 87

F2: Pearson‟s Correlations of Means Comparing Circulation Category and News Factors on

Newspaper Websites ...... 88

F3: Word Frequency Means of News Factors on Facebook Pages by

Circulation Category ...... 91

F4: Pearson‟s Correlations Between Circulation Category and News Factors on Newspapers‟

Facebook Pages ...... 92

F5: Word Frequency Means of News Factors on Twitter Pages by

Circulation Category ...... 93

F6: Pearson‟s Correlations Between Circulation Category and News Factors on Newspapers‟ Twitter

Pages ...... 94

F7: Deviance Typology of Prominence and Conflict ...... 132

xvi

List of Figures

I1: Conceptual Visualization ...... 3

I2: Methodological Visualization ...... 4

xvii

List of Illustrations

I1: The conceptual visualization of the dissertation project. The study tests how media with high and

low levels of audience interaction utilize the same news factors similarly or

differently; as such, eventual analysis will compare x1 and x2, as well as y1 and y2.

...... 3

I2: The methodological visualization of the dissertation project. The study tests how media with high

and low levels of audience interaction utilize the same news factors similarly or

differently. Eventual analysis will study similarities and differences within a values

and b values; secondary analysis will contrast a(1-4) and b(1-4)...... 4

F1: ANOVA analysis of word frequency means of deviant, social significance, and egalitarian news

factors on newspaper websites by circulation category. Numbers in the

“Community Weekly,” “Community Daily,” “Large Daily” and “National Daily”

rows reflect means for all word frequency analyses per factor per circulation

category; numbers for the “ANOVA” rows reflect the ANOVA analysis. For all

columns, df = 3, N = 140, and ** indicates significance at the 0.01 level.

Timeliness and proximity were formed as an averaged composite for recentness

and dates, and general proximity and specific proximity,

respectively...... 87

F2: Pearson‟s correlation analysis of word frequency means of deviant, social significance, and

egalitarian news factors on newspaper websites by circulation category. For all

instances, N = 140 and ** indicates correlation is significant at the 0.01 level (2-

tailed)...... 88

xviii

F3: ANOVA analysis of word frequency means of deviant, social significance, and egalitarian news

factors on newspaper Facebook pages by circulation category. Numbers in the

“Community Weekly,” “Community Daily,” “Large Daily” and “National Daily”

rows reflect means for analyses per factor per circulation category; numbers in the

“ANOVA” rows reflect the ANOVA analysis. For all columns, df = 3, N = 20, *

indicates significance at the 0.05 level, and ** indicates significant at the 0.01

level. Timeliness and proximity were formed as an averaged composite for

recentness and dates, and general proximity and specific proximity, respectively.

...... 91

F4: Pearson‟s correlations analysis of word frequency means of deviant, social significance, and

egalitarian news factors on Facebook pages by circulation category. For all

instances, N = 20, * indicates correlation is significant at the 0.05 level (2-tailed),

and ** indicates correlation is significant at the 0.01 level

(2-tailed)...... 92

F5: ANOVA analysis of word frequency means of deviant, social significance, and egalitarian news

factors on newspaper Twitter feeds by circulation category. Numbers in the

“Community Weekly,” “Community Daily,” “Large Daily” and “National Daily”

rows reflect means for analyses per factor per circulation category; numbers in the

“ANOVA” rows reflect the ANOVA analysis. For all columns, df = 3, N = 20, and

* indicates significance at the 0.05 level. Timeliness and proximity were formed as

an averaged composite of scores for recentness and dates, and general proximity

and specific proximity, respectively...... 93

xix

F6: Pearson‟s correlations analysis of word frequency means of deviant, social significance, and

egalitarian news factors on newspaper Twitter pages by circulation category. For

all instances, N = 20, * indicates correlation is significant at the 0.05 level (2-

tailed), and ** indicates correlation is significant at the 0.01 level

(2-tailed)...... 94

F7: Visualization of deviance using a two-by-two typology. The X axis refers to conflict in events

and the Y axis refers to the level of prominence. News within quadrant 1, of

prominent people and conflicting events, is the most deviant; news falling in

quadrants 2 and 3 are also deviant based upon either conflict or celebrity. Only

quadrant 4, regular people and non-conflicting events, has low deviance and is

rarely considered news. …………………………………...………….132

A1: Map of Regional Categorizations. The dataset was randomly collected from eight geographic

American regions, as illustrated here. Regional categories were based upon

common cultural, political and economic features, as well as geographic

contiguousness. Within each region, five small weekly newspapers, five

community daily newspapers, five large daily newspapers and (when applicable)

one national daily newspaper were randomly selected...... 159

A5: Calendar graphic expressing the constructed week schedule for quantitative data collection. Data

was collected over a ten week period between January and March, 2013. The five

groups of 25 newspapers each are expressed here through color coordination. Red

is Group A, orange is Group B, green is Group C, blue is Group D, and purple is

Group E. Dates were selected at random using multi-sided dice; however, those

random selections were structured to ensure that dates for each group were at least

approximately a week apart to reduce the likelihood of redundant

data...... 178

A7: Color coordinated graphic representation of social media data collection...... 180

xxi

CHAPTER 1: INTRODUCTION

What is “news?” This is the central question of journalism and communication studies. One common answer in popular culture and the mass communication academy is an image of journalists as “watchdogs” (Bhattacharyya, 2011; Donohue, Tichenor, & Olien, 1995; Hanitzsch, 2005; Pinto, 2009; Protess, Leff, Brooks, & Gordon, 1985; Stone & Banning, 1997). The term suggests a dog with a press pass, rummaging through government documents and crime reports to fetch hard-hitting news and tear- jerking feature stories.

An increasing volume of scholarship, however, likens the modern journalist less to a vigilant civil servant and more to a worried caveman keeping a nervous eye out for predators and threats. On a psychological level, safety is the caveman‟s paramount concern; when a predator, or any potential threat, appears, the caveman and his clan fall into full retreat. In contemporary America, the media may serve a similar function. Rather than working to improve cultural dialogue or facilitate democracy, journalists instead serve basic survival instincts and act as an alarm for a population or a society.

Studies of deviance in news, gatekeeping theory and news factors argue that journalists are principally concerned with identifying threats and potential predators. Those threats, literal or philosophical, potentially endanger the physical safety of media audiences or the ideological hegemony of the status quo (Gans, 1979; Miliband, 1969; Paletz & Entman, 1981; Shoemaker, 1984; Shoemaker & Danielian, 1991; Shoemaker & Reese, 1996; Shoemaker & Vos, 2009; White, 1950). Journalists and media audiences derive that need for vigilance from a basic sociological desire for safety and security (Shoemaker, 1996). Put another way, by a self-help book on anxiety:

“The shape on the horizon was either a bear or a blueberry bush, and the only way to find out was to go and see for yourself. If you go off toward the vague shape often enough, eventually it turns out to be a bear, and that day you‟re the bear‟s lunch. … We‟re the children of the children of the children (and so forth) of the ones that played it safe and went back to the cave.” (Wilson & Dufrene, 2010, p30)

Media deviance studies argue that humankind still is perched at the edge of the cave, squinting off into the distance, and worried if the distant blurry shape posed an opportunity to eat or be eaten. The metaphorical caveman has evolved into a journalist, however, with the mass media tools to communicate expediently with his tribe and save the general population the risk and trouble of bear watching.

Current scholarship on deviance and gatekeeping theory lacks in two critical regards, however. Firstly, most analyses of deviance focus on distinct fringe groups or behavior. While news articles on violent crime (Boyle & Armstrong, 2009), violent protests (Boyle & Armstrong, 2009) and the Ku Klux Klan (Shoemaker, 1984) clearly qualify as “deviant,” they do little to provide a comprehensive definition of a difficult sociological term; furthermore, although news factor research has been identified as a fruitful avenue for operationalizing deviance, deviance has infrequently been explored (Shoemaker & Danielian, 1991; Shoemaker & Vos, 2009). A thorough integration of news factor research (Badii & Ward, 1980; Eilders, 2006; Galtung & Ruge, 1965; Joye, 2010; Reinemann & Schulz, 2006) and scholarship on deviance presents clear opportunity for both fields.

Secondly, gatekeeping theory assumes that media producers‟ focus on deviance develops in a vacuum. Shoemaker and Vos argue that journalists have only “modest exposure” to readers or viewers, which generates “an abstract, second-hand sense of what the audience wants from news media” (Shoemaker & Vos, 2009, p52-53); this assumption is based upon noteworthy past studies (Gans, 1979; Tuchman, 1978; White, 1950).

The assumption‟s implicit logic is that all journalists have no practical interaction with audiences, and that an open dialogue would decrease the editorial focus on deviance. This study asks: Do media with high and low degrees of audience interaction publish different amount of deviant news? And, what are editors‟ and publishers‟ perspectives on audience interaction and deviant news content?

This dissertation utilizes news factors to measure deviance, and considers circulation size and social media as barometers for audience interaction. The following sections explain the rationale for these selections.

1.1: Visualizing the Dissertation

This dissertation studies the potential editorial focus on deviance among news media with high and low degrees of audience interaction. Deviant news content will be measured through Bridges and Bridges‟ (J. Bridges, 1989; J. Bridges & L. Bridges, 1997) seven news factors; high- and low-interactivity media will be operationalized partly by circulation size and partly by comparing traditional websites and social media. This study also utilizes a quantitative, computerized content analysis of online news content and qualitative, structured interviews of community newspaper editors and publishers.

Illustration 1 conceptually visualizes these relationships. Quantitative data will ultimately compare news factors expressed at x1 and x2, or news posted on traditional websites of small and large newspapers; analysis also will compare y1 and y2, or social media posts made by small and large newspapers. Traditional websites‟ news factors also will be compared to those on social media; although this comparison speaks less fundamentally to this dissertation, findings may potentially be fruitful. (See I1.)

I1: Conceptual Visualization

I1: The conceptual visualization of the dissertation project. The study tests how media with high and low levels of audience interaction utilize the same news factors similarly or differently; as such, eventual analysis will compare x1 and x2, as well as y1 and y2.

The dissertation can be further visualized in I2, which also offers an overview of the methodology. This study analyzes news content from the traditional websites and social media pages of four kinds of American newspapers (community weekly newspapers, which are small hyper-local publications with less than 50,000 regular circulation; community daily newspapers, or daily newspapers 3

with less than 50,000 regular circulation; large daily newspapers, or what the average reader might consider metropolitan newspapers, which have greater than 50,000 regular circulation; and national newspapers, which have greater than 500,000 daily circulation) for Bridges and Bridges‟ (J. Bridges, 1989; J. Bridges & L. Bridges, 1997) news factors. How each circulation category expresses these news factors on each online platform can serve as a measurement of deviant and non-deviant news. As such, primary analysis will study similarities and differences within a(1-4) and b(1-4); secondary analysis also will contrast a values and b values. (See I2.)

I2: Methodological Visualization

I2: The methodological visualization of the dissertation project. The study tests how media with high and low levels of audience interaction utilize the same news factors similarly or differently. Eventual analysis will study similarities and differences within a values and b values; secondary analysis will contrast a(1-4) and b(1-4).

1.2: Core Concepts

This dissertation uses five core concepts of communication theory and practice.

1. Deviance as news, as when journalists focus on physical and ideological threats to the status quo out of a sociological need for stability and security (Heckert & Heckert, 2007; Shoemaker & Danielian, 1991; Shoemaker, 1996) .

2. Gatekeeping theory and the related hierarchy of influences model, which structures editorial decisions as practices influenced by subjective and objective individual and professional influences (Shoemaker & Reese, 1996; Shoemaker & Vos, 2009; White, 1950).

3. News factor research, which distills all news into discreet elements, pieces or values (Bridges, 1989; Bridges & Bridges, 1997; Galtung & Ruge, 1965; Jong Hyuk & Yun Jung, 2009; Joye, 2010).

4. American community newspapers and their websites, which are typically exhaustively locally oriented (Elliott & Greer, 2010; Hansen, 2007; Hansen & Hansen, 2011; Lauterer, 2006; Reader, 2006).

5. Social media, in particular Facebook and Twitter, which offer direct audience engagement and clear potential for local or national political participation, and have rarely been studied in conjunction with community journalism (Ahmad, 2010; Hermida, 2010; Ifukor, 2010; Johnson & Perlmutter, 2009).

1.3 Importance & Significance of the Study

This dissertation, therefore, lies at a key academic crossroads. Journalists cover deviant events and ideas partly out of a sociological need for security, and partly as a substitute for substantive dialogue with media audiences (Shoemarker & Vos, 2009).

It happens that the news factor approach, largely pioneered in Europe (Badii & Ward, 1980; Eilders, 2006; Galtung & Ruge, 1965; Joye, 2010; Kepplinger & Ehmig, 2006), can effectively measure deviance, as well as facilitate for independent theoretical development. Bridges and Bridges (J. Bridges, 1989; J. Bridges & L. Bridges, 1997) argue that seven news factors appear in a wide range of American media and each concept fits squarely into the professional routines level of Shoemaker and Reese‟s (1996) hierarchy of influences model. So this study presents an opportunity to explore deviance, gatekeeping theory, and news factors.

Furthermore, this dissertation explores a key assumption of gatekeeping theory: That some professional routines, such as reliance upon deviance and the use of news factors, result from a disconnection of journalists from media audiences. Several studies of major and national media have indicated that audience input is considerably less influential than professional norms, standards, and routines (Cassidy, 2006; Gans, 1979; Shoemaker & Reese, 1996; Shoemaker & Vos, 2009; Tuchman, 1978). Equally common, however, are studies documenting high levels of community accessibility and local focus among American community newspaper editors and publishers (Garfrerick, 2010; Hansen, 2007; Hansen & Hansen, 2011; Lauterer, 2006; Reader, 2006), and studies acknowledging the civic and open-source accessibility of social media (Hermida, 2010; Ifukor, 2010; Larsson & Moe, 2012; Lasorsa, Lewis, & Holton, 2011).

As to how reliance on deviance, as expressed through news factors, manifests itself in media with high and low rates of audience interaction, this analysis will explore that question and provide threefold benefit for the academy and the industry.

1. Analyzing the web and social media content of newspapers with diverse circulation can potentially broaden and deepen the study of deviance. If data indicate that deviance is unrelated to circulation size or audience interaction, then gatekeeping theory is supported and strengthened; if circulation size or audience interaction influence the publication of deviance, however, results could potentially argue that journalists‟ have a highly circumstantial focus on deviance.

2. Studying news factors expressly explores their position within gatekeeping theory and the hierarchy of influences; by using news factors as a barometer of deviance, they become more incorporated into current gatekeeping literature. Furthermore, applying news factors to divergent American media validates the news factor approach generally while also expanding it into less-examined American media.

3. From a practical perspective, community newspapers are surviving and thriving in an industry beset by financial decline and hardship. Exploring the news factors utilized by community newspapers could offer important insight into the root of community newspaper‟s editorial successes when compared to larger and national newspapers; that information could potentially be very helpful to media practitioners.

1.4: Hypotheses

This dissertation studies deviance, gatekeeping theory, and news factors. It also measures interaction between journalists and audience members using two measurements. The first measurement focuses on newspaper circulation size. A litany of research indicates that community newspapers, which are typically quite small, have demonstrably more interaction with their audiences than larger newspapers. This is partly due to circulation size, since smaller newspapers have a much easier time shaping editorial content to fit the needs of a smaller audience, but it is also partly because of the mission and local-first credo of community journalism (Brockus, 2009; Hansen, 2007; Hansen & Hansen, 2011; Lauterer, 2006; Reader, 2006; Smethers, Bressers, Willard, Harvey, & Freeland, 2007). The second measurement compares newspapers‟ use of news websites and social media, which have been consistently considered more interactive and participatory than older online media (C. Greer & Ferguson, 2011; Hermida, 2010; Lasorsa, et al., 2011). (See Chapter 1:1 for a visualization.)

If highly interactive media, in terms of both circulation size and traditional/social online media, retain a focus on deviance and deviant news, then gatekeeping theory‟s assumptions about the omnipresence of deviance are affirmed. Lack of such focus, however, indicates added nuance for gatekeeping theory scholarship.

This study hypothesizes that audience interaction will influence the presence of deviance and deviant news. Given their local orientation and accessibility, it postulates that community newspapers will be more focused on egalitarian news than deviant news (H1). It also postulates that community newspapers will be less focused on conflict-based deviance, or extraordinary events, but perhaps more prone to prominence-based deviance, or news on celebrities or high-profile individuals. It seems logical that local celebrities and political elites would be more accessible in small communities than larger ones; the mayor of New York City, for example, may be less accessible to the average journalist than the mayor of suburban Elgin, Texas. Similarly, a high school quarterback may give more interviews than the Baltimore Ravens quarterback (H2). It hypothesizes that despite well documented differences between small and large circulation sizes, newspapers will express more egalitarian news factors on social media pages than on traditional websites; social media are, by their very nature, more collaborative and discussion-oriented than traditional websites (H3). It further hypothesizes that, on a qualitative level, the more community newspaper editor‟s perceived interaction with audiences, both on and offline, the greater their editorial preference for egalitarian news factors over deviant news factors. 8

The methodology behind these research questions, as well as literature solidifying the foundation for particular approaches and details, will be explored partly in Chapter 1.7, and exhaustively in Chapter 3. Definitions of key terms used in these research questions will be explored in Chapter 1:4.

H1: American community newspapers will publish online news with higher rates of egalitarian news factors than deviant news factors; larger newspapers will publish online news with greater rates of deviant news factors than egalitarian news factors.

H2: American weekly and daily community newspapers will publish online news on conflict- based deviance less frequently than larger daily newspapers, but will publish prominence-based deviant news more frequently than larger daily newspapers.

H3: Regardless of circulation size, American community and daily newspapers will publish social media content with higher rates of egalitarian news than deviant news factors.

H4: The more contact a community newspaper editor has with their community and readers, the more preference will be given to egalitarian news over deviant news.

1.5: Definition of Terms

Clear definitions of terms, and concise and discreet operationalizaions, are important for any dissertation. This study relies upon a number of well known concepts, and a few that suffer from ambiguity; both are rigorously defined in this chapter.

Gatekeeping theory has a long and esteemed scholarly history (Bleske, 1991; Bowman, 2008; Cassidy, 2006; Gans, 1979; Lewin, 1947; Shoemaker & Reese, 1996; Shoemaker & Vos, 2009; Tuchman, 1978; White, 1950). It has evolved beyond David Manning White‟s (1950) original claim that newspaper editors color their content decisions with individualized objective and subjective biases; today it incorporates Shoemaker and Reese‟s (1996) hierarchy of influences model, which argues that a spectrum of influences craft and shape media content. While individuals‟ decisions can influence news creation, more powerful effects at the professional routines level influence and ultimately standardize news content to conform with institutional preferences rather than individual audience‟s tastes or needs. That routines level particularly uses deviance in news; gatekeeping theory argues that such reliance on deviance develops largely because journalists have little contact with their audiences (Shoemaker & Vos, 2009). Without dialogue articulating what news audiences really want, gatekeepers default to news about deviance; this study tests that assumption by measuring both audience interaction and editorial focus on deviance.

Gatekeeping Theory: A theory of individual and institutional media influences that argues, in part, that media are crafted through institutional reliance on news about deviance. It also stipulates that journalists rarely are in serious contact with their audiences.

Testing these two assumptions of gatekeeping theory requires articulation and theoretical development concerning deviance and distance from the audience. Gatekeeping implies a relationship between the two concepts; each must be explored in tandem to determine if the current understanding of that relationship is accurate.

Scholars frequently divide deviance into three categories: normative deviance, labeling deviance, and conscious deviance (Shoemaker, 1984; Shoemaker & Vos, 2009). The most common, normative deviance, is of interest here; labeling and conscious deviance, while not studied in this dissertation, are also defined in the interest of inclusivity.

Normative Deviance: Behavior, ideas, groups, or events are deviant when they break social rules or norms.

Labeling Deviance: Behavior, ideas, groups, or events are deviant when an individual or group calls them deviant.

Conscious Deviance: A person or group is deviant when aware that his, her, or their behavior is in some sense wrong or disapproved.

Normative deviance is best operationalized through the concept of news factors, in particular those news factors concerning prominence, conflict, and oddity (Shoemaker, 1984; Shoemaker & Vos, 2009). Factors concerning “social significance” also have been identified as supplementary factors which encourage publication of deviant news (Shoemaker, 1984); social significance factors, in effect, magnify the importance of deviant news. However, other news factors emphasizing egalitarian concepts (such as proximity and timeliness) have not been considered, and are operationalized here as “Egalitarian Factors.”

Social Significance Factors: Social significance factors relate to news factors which pertain to details of volume or scope. Not intrinsically deviant or egalitarian, they serve as descriptors of other deviant or egalitarian factors (i.e., impact, magnitude).

Egalitarian Factors: News which emphasizes ordinary occurrence, egalitarian news factors focus on tangible details and regular interaction (i.e., timeliness, proximity).

News factor research, too, has a long history. This primarily European approach to journalism scholarship was pioneered by Johan Galtung and Mari Ruge (1965), who argued that 12 discreet “news factors” were common to news articles on international crises; the more factors a potential news item contained, the more likely the item was to become published news. The oft-cited analysis has been explored in conjunction with related, and often very similar, news factor rubrics utilizing distinct factors. As Kepplinger and Ehmig (2006) noted, news factors are primarily used to explain the character, length and placement of published articles.1

News Factors: The individual, distilled factors in a news article. Pioneered by Galtung and Ruge (1965), these factors are components that can be content-analyzed in any news article (i.e., conflict, controversy, reference to elite nations).

1 Distinctions between “news factors” and “news values,” which are related but independent concepts, will be explored in Chapter 2.

This dissertation utilizes a news factor approach designed by Bridges and Bridges (J. Bridges, 1989; J. Bridges & L. Bridges, 1997). Seven news factors will be applied to online news content in American community newspaper websites. Those factors will remain largely consistent in this dissertation with their original definitions (Bridges, 1989). Each factor corresponds with ideas - originally presented by Galtung and Ruge (1965) - that since evolved. The factors also are grouped into categories of “Deviant News Factors,” “Social Significance Factors,” and “Egalitarian News Factors” to operationalize and study deviance.

1.5.A: (NORMATIVE) DEVIANT NEWS FACTORS: Deviant news factors emphasize deviance from regular society (Shoemaker & Vos, 2009). They classify aberrations from societal norms based on celebrity, conflict, or strangeness.

Prominence: Prominence refers to elite or infamous individuals, issues or institutions mentioned in an article. Prominence can be local, as in a mayor or a sports team, or non-local, as in a president or an ambassador.

Conflict: Conflict refers to open disagreement between persons, groups, animals or issues, against one another or nature. Clear, articulate opposition is required; however, conflict can be broadly defined. Conflict includes elections, sports games, crime, and severe weather.

Oddity: Oddity refers to news coverage which recognizes a rare or very unusual event or occurrence. Odd news is news because it is odd or novel, not simply an unusual detail of regular news. News about surfing dogs (Connelly, 2012), a shark falling out of the sky (H. Mitchell, 2012), or a shrubbery named in honor of Lady Gaga, a popular singer, (Gupta, 2012) all potentially qualify.

1.5.B: SOCIAL SIGNIFICANCE FACTORS: Social significance factors cannot be considered deviant or egalitarian. They are expressions of quantity or depth that enhance, augment, magnify, or devalue deviant and / or egalitarian factors in the same story; put another way, these factors cannot directly describe either deviance or egalitarianism, but their inclusion potentially enhances accompanying coverage of deviance or egalitarianism.

Impact: Impact refers to the effect or consequence of a news story, either damaging or enhancing, massive or miniscule. It is akin to intensity. An article about a freeway closure could have impact, as could coverage of cancer treatments, legislative hearings, or congressional elections. News about negligible events, like a minor tax increase or short road detours, could also qualify.

Magnitude: Magnitude refers to quantifiable or measureable detail highlighted because of its largeness or smallness, or other connection to the news story. Articles about tax increases or decreases are obvious candidates, as are stories concerning rainfall totals, temperature swings, or college graduation rates.

1.5.C: EGALITARIAN NEWS FACTORS: Egalitarian news factors consider the tangible, ordinary, and normal; they are sometimes considered “contingent conditions” (Shoemaker & Danielian, 1991). These factors are easily relatable and highly tangible to an audience because they focus on either local or current ideas, events, individuals, and institutions.

Timeliness: Timeliness refers to the currency of a news story. News content relating to an event that occurred fewer than two days prior to the publication, or forecasting a news event fewer than two days in the future, qualifies. News on a government hearing taking place in a few hours qualifies, as does a recap of an election the previous evening. While this operationalization is potentially limiting, its narrow focus allows study of truly current and timely events in news.

Proximity: Proximity refers to the local-ness of a news item. Articles that mention a location, event, individual or institution within the immediate coverage area of a newspaper (operationalized as within 20 miles) qualify as proximate.

1.5.D: MEDIA DEFINITIONS The media under scrutiny in this dissertation also deserve definition; indeed, one of the primary terms in this study, community journalism, suffers from more than a bit of ambiguity. The industry standard defines community newspapers as publications with fewer than 50,000 regular circulation (Lauterer, 2006); however, this does not offer a delineation by frequency of publication. Furthermore, from an academic perspective, community journalism as a term is typically infrequently operationalized.

Lowrey, Brozana and Mackay found that of 108 scholarly studies of community journalism between 1995 and 2005, only 65 offered direct or implied definitions of community; only about 30 offered any definition of “community journalism” (Lowrey, Brozana, & Mackay, 2008,,, p280 - 2). The study offered an inclusive theoretical definition of community journalism based on discussions with newspaper editors and academics.

“„Community news media,‟ then, are media capable of fostering the process of community …. Community journalism would (a) reveal, or make individuals aware of, spaces, institutions, resources, events, and ideas that may be shared, and encourage such sharing and (b) facilitate the process of negotiating and making meaning about community.” (Lowrey et al., 2008, p288)

Although it says nothing about cultural or demographic influences, this theoretical definition helps when considering theoretical notions of community, particularly as it relates to Benedict Anderson‟s (2006) “imagined communities,” often tied to studies of local, niche and community journalism (Anderson, 2006; Burd, 2008; Cover, 2005; Funk, 2012; S. Lewis, 2008; Reader, 2006). On a practical level, however, this definition is less helpful when planning a methodological approach to community journalism; potentially any number of online of offline media make individuals “aware of, spaces, institutions, resources, events, and ideas that may be shared” (Lowrey et al., 2008, p288), and this definition offers no distinction between traditional for-profit journalism models and civic or non-profit journalism models. This dissertation is concerned with for-profit community newspapers, however, not structural interpretations of community writ large. As such, this dissertation offers the following definitions of community journalism, followed by definitions for larger circulation categories; 50,000 regular circulation is the traditional cutoff point between small and large publications.

Community Weekly Newspapers: Weekly, local, for-profit, American newspapers with regular print circulation of fewer than 50,000 copies.

Community Daily Newspapers: Daily, local, for-profit, American newspapers with regular daily print circulation of fewer than 50,000 copies.

Large Daily Newspapers: Daily, for-profit, American newspapers with regular daily print circulation of more than 50,000 copies but fewer than 500,000 copies.

National Newspapers: Daily, for-profit, American newspapers with a regular daily print circulation greater than 500,000 copies.

Similarly, the term “social media” has evolved to encompass a wide variety of technological platforms, including so-called “micro-blogging” (Ahmad, 2010; Hermida, 2010; Lasorsa, et al., 2011) and social networking (Johnson & Perlmutter, 2009; Papacharissi, 2009; Woolley, Limperos, & Oliver, 2010). On a practical level, the most successful such tools are Facebook and Twitter; as such, the operationalized definition of “social media” for this dissertation is as follows.

Social Media: Social media are online media designed for broad, inter-cultural dialogue, discussion and participation. In this dissertation, social media is operationalized as Facebook and Twitter posts by newspapers.

Finally, methodological definitions are in order; more thorough explication of these terms will follow in the Assumptions section, but an overview is appropriate here. This dissertation uses a computerized content analysis and structured interviews with community newspaper editors.

Computerized Content Analysis: Quantitative content analysis aided by a sophisticated word- counting program called DICTION 6.0. The software scans large collections of text to determine the frequency of particular words grouped by the user (Conway, 2006; Jarvis, 2004; Krippendorff, 2004).

Word Dictionaries: Customized word groups designed by a DICTION user. These are collections of individual words, some of them quite lengthy, which the program uses to determine word frequency. For example, a word dictionary for the news factor prominence would include words such as mayor, senator, minister, CEO, and quarterback; DICTION would use that dictionary to scan for each individual word, and then pools together results for one frequency for Prominence-type language.

Structured Interviews: Structured interviews are qualitative, face-to-face, scripted interviews utilizing a pre-made list of questions and procedures to ensure reliability between interviewers. For this dissertation, such interviews will be conducted with community newspaper editors in conjunction with Institutional Review Board approval.

1.6: Assumptions

Generally speaking, this study avoids axioms of fact or substance. It does, however, make a pair of methodological assumptions while taking steps to avoid a third.

The first methodological assumption relates to computerized content analysis. This dissertation assumes that, when properly executed, computerized content analysis can be a powerful tool for processing large amounts of communication data effectively. There is some debate on this point, however, particularly because the potential of computerized content analysis schemes often is embellished or exaggerated; although this dissertation assumes computerized content analysis is reliable and effective, a brief review of literature challenging that assumption is worthwhile.

Conway (2006) ran a concurrent computerized content analysis of a traditionally coded second- level agenda setting study; he found significant disparity between the computerized and traditionally coded results, largely because the computer software recorded vastly more data than individual coders. While a human coder can distinguish between relevant and incidental mentions of the same word and know immediately if the phrase is used to describe (in this case) a gubernatorial candidate, computers considered every textual mention equally and regardless of context (Conway, 2006).

Other scholars, too, have argued that computerized content analysis programs are essentially highly sophisticated blunt instruments (Krippendorff, 2004; Nuendorf, 2002). These programs cannot process context, subtext, implicitly, nuance, association, sarcasm, or any traditional understanding of textual meaning. Studies which rely on even relatively simple associations between ideas, such as Conway‟s (2006) second-level agenda setting study of attributes associated with gubernatorial candidates, utilize texts which are too sophisticated for computerized content analysis. Computerized content analysis studies which condense the unit of analysis to single words, however, rather than associations, provide data which are considerably more valid and reliable (Aust, 2004; Ballotti & Kaid, 2000; Cho et al., 2003; Jarvis, 2004). Once that validity has been established, however, computerized content analysis software is extremely reliable.

“Unlike humans, computers are deterministic machines. They cannot not process text reliably. Computers have no sense of what they do, who their users are, or what the character strings they are processing may mean to human readers, nor are they sensitive to the shifting cultural contexts relative to which we read and understand text. Computers do not even know the difference between randomly generated character strings and words or symbols that have meaning to 16

humans … The use of computers is most appropriate for recurrent and repetitive tasks that can be conceptualized without uncertainty. Searching, coding, sorting, listing, and counting are obvious candidates.” (Krippendorf, 2004, p259-61)

Assumption 1: Provided that the units of analysis are individual words, word dictionaries can be constructed using computerized content analysis software to validly and reliably study journalism data.

Content analysis literature is divided on “inferences,” or drawing conclusions based upon patterns in studied data. Some scholars argue that content analysis serves as a benchmark for abductive, deductive and inductive claims about content creators; Krippendorf (2006, p28) likens the content analyst to the fictional detective Sherlock Holmes, stitching together a hidden mystery out of observable pieces, and other scholars share his view (Weber, 1990). Others, including this author, consider content analysis a more conservative enterprise documenting only content which is extant or absent in particular data or text (Nuendorf, 2002; Poindexter & McCombs, 2000).

Both perspectives have flaws. Content analysts who endorse inferencing will risk false or unsubstantiated conclusions, but analysts opposed to inferencing will lack any detail on the content creation process. Furthermore, any content analysis is limited to the data and text in hand; even inferential studies can offer only inferential insights into the minds and decisions of content creators or audiences. Either way, interviewing content creators directly and avoiding the need for inferencing altogether yields more insight. As such, this dissertation adopts an agnostic view toward content analysis inferences and related assumptions, using a mixed-methods approach and pairing computerized content analysis with structured, in-person interviews.

There is a related assumption here; while quantitative inferencing can be problematic, qualitative interviews are considered intrinsically subjective (Fontana & Frey, 2005). Any interview is “inextricably and unavoidably historically, politically, and culturally bound” (Fontana & Frey, 2005, p695); many researchers embrace that subjectivity, as a matter of methodology and as an important element of a study‟s findings (Kong, Mahoney, & Plummer, 2002).

This dissertation makes no such claims, and instead seeks to mitigate potential qualitative complications in order for findings to reflect what Nuendorf (2002) called “intersubjectivity,” or consistent knowledge and facts that generally can be agreed upon (Nuendorf, 2002, p11). As to qualitative interviews, typically an illustrative way to provide “texture and context” that goes beyond survey data

(Poindexter & McCombs, 2000, p271), the most reliable and intersubjective approach is the structured interview.

Through planning, scripting, and training, a structured interview can be administered intersubjectively by a number of qualified interviewers. Provided interviewers do not deviate from the script, results can generally be considered reliable. There are a few potential issues, particularly concerning the selection of interviewers. If questions are concerned with race or ethnicity, e.g., research focusing on perceptions of news on immigration or affirmative action programs, then the race of the interviewer can potentially bias a subject‟s answers; “high status” and student interviewers also can potentially generate a “response effect,” thus influencing a subject‟s responses and resulting qualitative data (Fontana & Frey, 2005, p702). Emotional aspects of an interview also are typically overlooked by the highly structured format. However, this dissertation is unconcerned with topics of race, gender, age, socio-economics, or emotion; it is restricted purely to professional editorial decisions made by community journalists, which should be largely secure ground for intersubjective structured interviews.

“The relatively minor impact of the interviewer on response quality in structured interview settings is directly attributable to the inflexible, standardized, and premeditated nature of this type of interviewing. There is simply little room for error.” (Fontana & Frey, 2005, p703)

Therefore, this dissertation assumes that structured, qualitative interviews will reveal reliable, intersubjective data on community newspaper editorial decisions concerning the Bridges and Bridges (J. Bridges, 1989; J. Bridges & L. Bridges, 1997) seven news factors.

Assumption 2: Qualitative, structured interviews will generate reliable, intersubjective data concerning community newspaper editorial decisions.

1.7: Delimitations and Limitations of Study

This dissertation is arguably comprehensive and well-grounded. It is delimited, obviously, by its subject matter; the study focuses on American community newspapers, as well as small, large and national daily newspapers, and the website and social media presence of each. It, therefore, does not consider traditional print newspaper versions or media platforms, nor does it consider the Internet or social media writ large. This is acceptable, given the dissertation‟s primary focus on community journalism. Furthermore, it is delimited by its theoretical focus on gatekeeping theory rather than a more varied theoretical lens; this, too, is acceptable, given the clear applicability of gatekeeping and its potential scholarly value.

This dissertation is not delimited or limited by regional geography or narrow sampling windows. As will be discussed further in Chapter 3 on methodology, the dissertation uses a randomly generated, nationwide sample of community newspapers, community daily newspapers, large daily newspapers and national daily newspapers; it also utilizes a constructed-week model and other precautions to ensure a randomly generated quantitative data set.

The biggest potential limitation concerns reliance upon a single methodology, in this case content analysis; inferences based upon quantitative content analysis would require significant assumptions. By adopting an additional qualitative methodology using structured interviews, this issue is avoided; however, the dataset for those qualitative interviews provides a primary limitation.

Although the quantitative dataset is constructed from a nationwide sample, the qualitative dataset draws entirely from one state. Interviews of community newspaper editors will be conducted at the Texas Press Association Midwinter Conference and Trade Show in Houston, Texas, in January 2013. While a nationwide sample is theoretically possible, logistical and financial limitations make this difficult; furthermore, although Texas is only one state, it also has a population rivaling many small nations. It has a high number of vibrant community newspapers, and as such serves as the best single state for this study; furthermore, while it is not an ideal dataset, a number of published journalism studies have focused exclusively on Texas newspapers and Texas issues (Conway, 2006; Cope, 2011; Jensen & Uddameri, 2009; Kraeplin, 2008; S. Lewis, Kaufhold, & Lasorsa, 2009; Schweitzer & Smith, 1991; J. Stewart, 2011).

1.8 Overview of Complete Dissertation

The remainder of this dissertation is outlined here. Chapter I, the introduction, has focused on general ideas, problem statements and terminological operationalizations for this study. The principle focus is the intersection of American community journalism and daily newspapers, online and on social media platforms, with news factor and gatekeeping research.

Chapter 2 will focus on the literature review. Chapter 1.3 served in many ways as a de facto introduction to Chapter 2. It introduced key literature on gatekeeping theory, news factors, community newspapers online and on social media, and broad social media research.

Chapter 3 focuses on methodology. The mixed methods approach has been explained in part in Chapter 1, but a more thorough overview is appropriate here. Data will consist partly of a computerized content analysis and partly of in-depth interviews. The former will utilize a randomly generated, nationwide sample of community newspapers, community daily newspapers, large daily newspapers, and national daily newspapers over a constructed week period in January and February, 2013. Concerning websites, data collection will utilize four constructed weeks, one for each circulation size, spread over an eight-week period; three articles will be pulled from each website on each day of analysis. There will be 40 community newspapers, 40 community daily newspapers, 40 large daily newspapers, and five national daily newspapers; the total data set will consist of 2,625 articles. Data will be processed through the computerized content analysis program DICTION 6.0. Word dictionaries will be constructed to reflect the seven news factors designed by Bridges and Bridges (J. Bridges, 1989; J. Bridges & L. Bridges, 1997) and tested equally on data from each circulation size.

Concerning social media, 30 Facebook and Twitter posts will be downloaded from the social media pages of each of the 125 newspapers; one-time data collection will prevent redundancy in the dataset, given that many newspapers update social media news irregularly. DICTION will be used to analyze this data as well, with slightly different news factors and word dictionaries.

Because content analysis is limited, at best, to inferences about content creators, in-depth interviews expand the study dramatically. They also allow for more comprehensive consideration of gatekeeping theory. Qualitative, in-depth interviews will be conducted during the Texas Press Association Midwinter Conference and Trade Show in Houston, Texas, in January 2013; interviews will target

community newspaper editors throughout Texas and inquire about opinions and usage of news factors online and on social media.

Chapter 4 will focus on quantitative and qualitative results. Chapter 5 will focus on discussion of the quantitative and qualitative results, as well as opportunities for further research.

CHAPTER 2: LITERATURE REVIEW

2.0: Introduction

This dissertation seeks to study gatekeeping theory‟s focus on deviance and news factors. Gatekeeping theory stipulates that journalists have little practical interaction with their readers and audiences, and instead write primarily for one another (Shoemaker & Reese, 1996; Shoemaker & Vos, 2009); without substantive conversation with media audiences about news interests or preferences, journalists default to reliance on news on deviance, or aberrant events which are disconnected from ordinary life because of either conflict or celebrity (Jong Hyuk, 2008; Shoemaker & Danielian, 1991).

A critical assumption here has never been tested: How might the editorial focus on deviance, and types of deviance, fluctuate among media with varying interaction levels with readers and audiences? Gatekeeping theory assumes all media are equally disconnected from audiences, and equally reliant upon deviance as a substitute for honest interaction (Shoemaker & Reese, 1996; Shoemaker & Vos, 2009); this may well be nuanced or false.

As such, this dissertation analyzes deviant and egalitarian news across two dimensions of audience interaction. The first relates to media circulation size, which has a clear association with audience interaction: American community newspapers have a long history of connectivity with local readers, institutions, ideas, issues and events (Bishop, 2009; Brockus, 2009; Dill & Wu, 2009; Hansen & Hansen, 2011; Lauterer, 2006; Reader, 2006; Smethers, et al., 2007); community newspapers offer ideal comparison with larger metropolitan and national media which have been the traditional focus of gatekeeping theory and deviance (Lewin, 1947; Shoemaker & Vos, 2009; White, 1950).

The second dimension contrasts traditional websites with social media. Facebook and Twitter, relatively new entrants to the World Wide Web, are considered highly interactive and participatory by nature (Hermida, 2010; Lasorsa, et al., 2011). Newspapers‟ social media pages serve a similar function, in this sense, as community newspapers. They are an interactive, participatory media which offer prime comparison to traditional, non-participatory media like traditional websites.

These community newspapers and social media may – or may not – rely upon deviant news as gatekeeping theory suggests. This dissertation hypothesizes that audience interaction influences a newspapers‟ publication of deviant news to a degree, and seeks to test its hypotheses through both 22

computerized content analysis and qualitative structured interviews. Deviant and egalitarian news, as well as social significance, will be measured using news factors, a primarily European approach that gatekeeping scholarship has mentioned as a measure of deviance (Badii & Ward, 1980; Eilders, 2006; Galtung & Ruge, 1965; Harcup & O'Neill, 2001; Joye, 2010; Östgaard, 1965).

2.1: Cohesive Argument

This literature review, therefore, is designed to weave together central concepts and key studies in and around the core of this dissertation. It will focus primarily on five principal sections:

1. Gatekeeping Theory

2. Deviance

3. News Factors

4. Imagined Community & Community Newspapers in the Digital Age

5. Social Media & American Community Newspapers

These sections together illustrate the core cohesive argument of this dissertation: Gatekeeping theory stipulates, without proper testing, that journalists default to news about normative deviance because they have no substantive interaction with media audiences. This analysis measures, both quantitatively and qualitatively, journalists‟ interaction and dialogue with audiences as well as their use of deviant, social significance, and egalitarian news factors.

Such a study requires: explication of past research on deviance, work traditionally focused on fringe groups rather than robust theoretical development; news factor research, studies traditionally focused on European and international news rather than American journalism; community and community newspapers, highly locally focused publications that have never been studied with deviance in mind; and social media, also highly conversational and deliberative content that has rarely been analyzed in conjunction with community newspapers or deviance.

2.2: Gatekeeping Theory

Gatekeeping theory has a long and highly tangible history; however, it relies on key untested assumptions. The most noteworthy is an assumption that because journalists have “modest exposure to their audience” (Shoemaker & Vos, 2009, p52), a focus on normative deviance becomes an institutional default (Jong Hyuk, 2008; Pritchard & Hughes, 1997; Shoemaker, 1984; Shoemaker & Danielian, 1991). This assumption deserves proper testing and vetting, but first requires a fuller understanding of gatekeeping theory.

2.2.A: GATEKEEPING & THE HIERARCHY OF INFLUENCES Gatekeeping theory was originally pioneered by Kurt Lewin (1947), a government-contracted academic attempting to persuade housewives to purchase unconventional meat during World War I food rationing; he found that housewives were the gatekeepers, so to speak, determining what food was granted placement on the dinner table, and what was rejected. Any number of objective and subjective “forces” influenced the overall process as well. A woman‟s decision to accept or reject lengua2 for a beef stew was one “gate,” and one force might be the price of the food; another force might be if the meat were accidentally dropped in the parking lot, thus earning inadvertent exclusion from dinner (Shoemaker & Vos, 2009).

Later, David Manning White adopted the theory to communication research. His case study of an iconic “Mr. Gates,” a Cold War-era wire editor at a major newspaper, revealed that wire articles were accepted for and rejected from eventual publication based on a variety of objective and subjective forces. Some forces were arguably professional, such as the newspaper covering that particular topic too frequently, or not having space for an article or (in at least one case) accidentally misplacing the news copy; others were clearly subjective, such as considering a reporter or issue “too red3” (Lewin, 1950; Shoemaker & Vos, 2009; White, 1950). White modeled gatekeeping as a pinball machine of sorts, with individual gatekeeping decisions serving as the flippers which rejected an article or allowed it to fall into publication (Shoemaker & Vos, 2009, p16).

2 Spanish for “beef tongue” 3 Communist; White was writing at the height of the Cold War. 25

This model is clearly centered on the individual; subsequent theorizing began to question if Mr. Gates acted alone, in a sense, or if he simply reinforced the preferred standards and practices of his profession. Replications, adaptations and further theorizing of White‟s analysis (Gans, 1979; Gieber, 1956; Hirsch, 1977; Sigal, 1973) found more synchronous industry standards of professional news content that went beyond Mr. Gates‟ personal preferences.

“Hirsch said the reasons offered by the wire editor in White‟s study for rejecting stories were primarily based on professional norms – commonly held views in the news industry about whether a story is newsworthy. He concluded that these norms, which could also be determined routine forces, were a better explanation for the decisions made, rather than White‟s original conclusion that the wire editor was highly subjective.” (Cassidy, 2006, p8)

Gatekeeping ultimately evolved to incorporate Shoemaker and Reese‟s (1996) hierarchy of influences model, which argues that a spectrum of influences weighs upon media content creation. At the bottom, and more tangible, end of the hierarchy are individual and communication routine levels, while higher influences include organizational structure, extra-media, social institution, and social system influences. Of critical interest to journalism studies are the individual and communication routine levels.

A number of studies indicate that journalists behave more as professional journalists than as individual journalists, and that a socialized understanding of journalistic standards and values exists among media practitioners. For example, professional practices and an emphasis on reporting “in the public interest” remain an integral part of journalism in the Internet era (Bowman, 2008); routine influences, in particular “peers on staff” and “journalistic training,” are highly influential indicators of professional role conceptions among print and online journalists, while personal demographics are largely inconsequential (Cassidy, 2006). Network television news in the 1970s was a product formed by a vast array of different professionals which effectively homogenized the product (Epstein, 1973), female gatekeepers largely craft editorial content using traditional journalism practices irrespective of their gender (Bleske, 1991), and community newspaper editors in early primary states focus uniformly on candidate visits rather than broader elections or political policy (Funk, 2013). Such examples underscore the critical assumption studied here: as with these professional norms, an institutional focus on deviance has become entrenched in newsrooms because of a lack of open dialogue between journalists and audiences.

Such professional routines ultimately craft and shape news items more than the creativity or insight of individual journalists. Gatekeeping theory places two stipulations on this content creation

process, however. First, gatekeeping theory argues that news organizations are attracted to deviance, or aberrations from normal events and typical individuals; second, the theory claims that content is created largely independently of the audience.

“Individual journalists have modest exposure to their audience. Journalists no doubt encounter readers, listeners, or viewers on occasion, and news items increasingly come with an email address that allows the audience to send comments to reporters. However, Gans [1979] and others … show that journalists have only an abstract, second-hand sense of what the audience wants from the news media. Journalists instead rely on well established routines to produce news it is believed will appeal to the intended audience.” (Shoemaker & Vos, 2009, p.52-53)

It seems logical that without sustained direct contact with media audiences, a focus develops on deviant events, ideas and behaviors, as well as a variety of celebrity figures (Jong Hyuk, 2008; Pamela Shoemaker, 1984; Shoemaker & Reese, 1996; Shoemaker & Vos, 2009); because journalists are creatures of routine, professional routines develop around this focus on deviance, and those routines in turn trump the efforts and characteristics of individual journalists and serve to homogenize media content (Bleske, 1991; Bowman, 2008; Cassidy, 2006; Gans, 1979; Shoemaker & Vos, 2009; Tuchman, 1978). One potential counter argument exists, however: Because media audiences consist largely of normal people, so to speak, it might be doubtful that their ordinary routines would generate news. Deviance and audience content, therefore, may ultimately have a spurious relationship.

These questions deserve study. The relationship between audience interaction and deviance has not yet been tested, leaving a critical deficit concerning both gatekeeping theory and the study of deviance.

2.2.B: THEORETICAL LIMITATIONS AND OVERLAP WITH OTHER MODELS It is worth noting that, all in all, gatekeeping theory has relatively shallow theoretical depth. Arguably unlike news factor research (Bridges & Bridges, 1997; Galtung & Ruge, 1965; Harcup & O'Neill, 2001; Joye, 2010; Kepplinger & Ehmig, 2006), agenda setting (McCombs, 2004; McCombs & Shaw, 1972) and framing (Entman, 2003; Reese & Lewis, 2009; Semetko & Valkenburg, 2000), gatekeeping theory lacks substantive predictive power. Indeed, Shoemaker and Vos argue that “Regrettably, theorizing about gatekeeping has not been in large supply” (Shoemaker & Vos, 2009, p11).

The theory seems more reactive than proactive, in a sense. Even early studies of gatekeeping focused on decisions that had already been made, not decisions that could or would be made later (Gieber, 1956; Hirsch, 1977; Sigal, 1973; White, 1950). More recent scholarship documenting the homogeneity of 27

modern media is similarly limited; documenting news media‟s formulaic nature is intriguing, but noting institutional patterns does little to predict future news coverage, apart from the observation that news will be covered similarly by similar news institutions (Bleske, 1991; Bowman, 2008; Cassidy, 2006; Epstein, 1973; Funk, 2013; Shoemaker & Reese, 1996).

Gatekeeping theory‟s more predictive cousins, structural pluralism theory and the community structure approach, use similar theoretical lenses to examine media audiences. Structural pluralism theory argues that demographic characteristics in a particular media audience influence journalists‟ editorial decisions and affected news content (Donohue, Olien, & Tichenor, 1985; Olien, Donohue, & Tichenor, 1978). The model has been used to explore demographic influence over environmental news (Jeffres et al., 2011) and local public affairs blogging (Watson & Riffe, 2011); adaptations by John Pollock and an army of undergraduate researchers also have identified demographic influences over news about childhood obesity (Briones, Catona, Keefe, Zimbaldi, & Pollock, 2008), gay rights (E. Mitchell, Pollock, Schumacher, & de Zutter, 2006), and Islam (Pollock, Piccillo, Leopardi, Gratale, & Cabot, 2005).

“Based on classic sociological theories of Durkheim and others, the community structural pluralism studies of Tichenor, Donohue, and Olien at the University of Minnesota have pioneered a research paradigm and nurtured generations of graduate students and scholars in the conviction that macrosocial forces play a key role in journalists‟ coverage of critical events, especially those affecting the interests of local political and economic elites (see Tichenor et al., 1973, 1980, 1995; Donohue, Olien, & Tichenor, 1985, 1989; Jeffres et al., 2000).” (Pollock, 2007, p23)

Proper analysis of structural pluralistic and community structure influences, however, requires measurement of community interests and demographics; such analysis goes typically beyond the scope of gatekeeping theory, which rarely surveys or considers audiences. The community structure approach has been called an intellectual opposite of agenda setting; community structure looks at bottom-up influences over media content, while inter-media agenda setting considers top-down influences from larger to smaller media (McCombs & Funk, 2011). The same could be said of the distinction between structural pluralism and gatekeeping theory. The former is concerned with bottom-up influences over media producers, while the latter is more concerned with the top-down decisions and homogenizing mechanics of editors and journalists.

Here the study of deviance and news factors offers clear potential for gatekeeping theory. Studies of deviance argue that journalists are prone to react to perceived physical or ideological threats to the status quo and report them (Arpan & Tuzunkan, 2011; Boyle & Armstrong, 2009; Jong Hyuk, 2008;

Pritchard & Hughes, 1997; Shoemaker & Danielian, 1991); news factor research identifies a number of discreet elements common to virtually any news story, and argues that the more factors present in a potential news item, the more likely the item is to receive journalist‟s attention (Badii & Ward, 1980; Eilders, 2006; Galtung & Ruge, 1965; Joye, 2010; Martin, 1988; Östgaard, 1965). Both approaches offer predictive power in ways that traditional gatekeeping theory lacks. Furthermore, exploring the intersections between gatekeeping theory, deviance and news factors offers to confirm or disconfirm one of gatekeeping theory‟s critical assumptions about deviance and audience interaction.

2.3: Deviance

Exploring those intersections, however, requires a full understanding of deviance and studies of deviance in news. Normative deviance, for example, has a relatively broad theoretical definition that is inconsistent with the bulk of relevant analyses focused on fringe groups and radical ideologies. Exploring deviance in detail allows the operationalization of normative deviance to be fully articulated and potentially reconsidered, which in turn allows for a more fundamental understanding of gatekeeping theory and modern journalistic practice.

2.3.A: CONCEPTUALIZING DEVIANCE Deviance, as a term, suffers from a degree of ambiguity. In a popular context, the word is arguably associated with sex offenders, fringe groups and extreme philosophies; in sociology, there is a running debate on the definition of deviance, including fairly broad conceptualizations of “absolutist definitions” and “definitional approaches” (Little, 2007), “positive deviance” (Heckert & Heckert, 2007) and “normative” and “reactivist” perspectives of deviance (Pontell, 2007). Constructing a specific and concise definition and operationalization, too, is key. Studies of linguistics and semantics point to the complex relationships between words, labels, communication and identity; discreet definitions are effectively the building blocks of human thought, and relate directly to how humans perceive and judge the world around them. Embler (1962) wrote that a lack of reliable and relevant information is scarce, and that ultimately, individuals believe what they want to believe – beliefs formulated by powerful, evolving associations between words, ideas, and personal identity. Clearly defined terms therefore become critical to the scientific method.

“… names become tools without which civilization, to say nothing of human thought itself, would be very nearly impossible. Yet even here one can be misled into thinking he knows something when he doesn‟t. Unless one has some kind of plan or purpose for the use of his knowledge, some imagination, in short, the memorization of names can be a waste of time.” (Embler, 1962, p220)

Other studies of semantics have documented the power of words in identity formation and cultural forces. For example, Enghels and Jansegers (2013) considered the complex semantics and implications of the verb “sentir(e)” in Spanish, French, and Italian; Candel (2013) used semantics and secondary education using models to consider relationships between student experiences and theoretical abstractions. Other studies have considered lexical semantic typology (Goddard, 2012), semantic maps

(Regier, Khetarpal, and Majid, 2013) and the semantic evolution of Persian loan words (Shariq, 2013). Each case demonstrates how language can be a springboard for divergent definitions, and in turn, diverse identities and ideologies.

In contrast, the term has a relatively specific definition in communication research. Although media ostensibly are objective and neutral, evidence suggests that media often instead preserve and defend prominent ideologies and the status quo (Boyle & Armstrong, 2009; Daley & James, 1988; Gans, 1979; Shoemaker, 1984; Shoemaker & Vos, 2009; White, 1950). Shoemaker (1984) argues that such deviance is rooted in studies of “social control” (Lauderdale & Estep, 1980; Miliband, 1969; Paletz & Entman, 1981), and that media vary coverage of political groups in direct relation to these groups‟ perceived deviance. For journalists and readers, this focus on potentially threatening ideas, events, groups and agendas reflects a very basic – if not primal – urge to preserve personal safety.

“People routinely survey their environments for things that are deviant or unusual because they pose potential threats. These can be as common as a car darting in front of someone on a busy street to less frequent threats like invading armies. People may be alert to the possibility that their children may become drug users, or that they may be criticized in the workplace or for deviant political ideas which may change the nature of society. The difference between professional information gatherers such as journalists and the rest of us is that journalists‟ surveillance is institutionalized and sanctioned, whereas we generally survey the environment for our more informal and personal purposes. Journalists fulfill people‟s innate desire to detect threats in the environment, keep informed about the world, and devise methods of dealing with those threats, whether real or potential.” (Shoemaker, 1996, p32)

News coverage of this “innate desire” has evolved into content about three principle forms of deviance: normative deviance, labeling deviance, and conscious deviance (Shoemaker, 1984; Shoemaker & Vos, 2009). They are defined below:

Normative Deviance: Behavior, ideas, groups, or events are deviant when they break social rules or norms.

Labeling Deviance: Behavior, ideas, groups, or events are deviant when an individual or group calls them deviant.

Conscious Deviance: A person or group is deviant when they are aware that their behavior is in some sense wrong or disapproved.

Of these, analyses of normative deviance are the most common in communication research. Aggressive or extreme tactics among abortion protestors, for example, were the strongest predictor of news coverage in elite media between 1960 and 2006; the long-term political positions of pro-life or pro- 31

choice groups were less determinant of news coverage than short-term action, particularly actions which threatened day-to-day stability and the status quo (Boyle & Armstrong, 2009). Also, Alaskan newspapers responded to the nomination of a socialist environmental commissioner through a “highly ideological” lens that considered socialism a threat to political order, despite normative claims of objectivity and professionalism (Daley & James, 1988). Too, news coverage of an initial deviant event, such as child molestation charges among Catholic clergy members, can spark an inter-media agenda-setting “trigger” that leads to more coverage of related deviance in similar issues (Breen, 1997). Even in an experimental setting, perceived deviance of a news item played a significant role in online news selection; personal involvement with an issue, however, significantly mitigated the authority of deviance in news selection (Jong Hyuk, 2008). Additionally, the “social significance” of a news event, or its importance, impact or consequence, is positively associated with news coverage and deviance (Shoemaker & Danielian, 1991).

“As soon as [readers] are exposed to stories with different levels of deviance, deviant stories describing unexpected events draw uncontrolled attention by eliciting an orienting response. … After exposure, the audience members realize that the deviant stories should be learned for good surveillance of their environment. The motivated members concentrate and elaborate on the stories with a special attention …. As a result, audience members are likely to select deviant stories rather than less deviant stories and, if multiple deviant stories are presented, they are likely to select the most deviant story first.” (Jong Hyuk, 2008, p52-52)

Each case associated deviance with frequency and negativity of coverage. Normative deviance made an event or group more likely to see news coverage, but the more demonstrated deviance, the more negative the news coverage. Shoemaker‟s (1984; 1996) model of deviance effectively and logically adapts a simple sociological concept of surveillance into communication research.

2.3.B: OPERATIONALIZING DEVIANCE Although deviance can be conceptualized with some ease, at least as it applies to communication and journalism research, operationalizing the term is more complicated. Normative deviance provides a more specific model for deviance than broader concepts in sociological texts, but it still requires tangible application. Scholars have approached that application, and that operationalization, in a variety of ways.

Jong Hyuk (2008), for example, used a pair of coders to determine how online news articles fit into a four-pronged scale of deviance based on statistical deviance, social change deviance and normative deviance. Another study operationalized deviant news “as events that are scheme incongruent or events that are unexpected” (David, 1996); yet another considered the “volume” and “valence” of news coverage

(Breen, 1997). Survey research also operationalized deviance through three questions based on three items: similarity to most Americans, amount of advocated change, and how close a particular editor felt to the group (Shoemaker, 1984).

The most effective operationalizations use news factors to express deviance. Shoemaker and Vos (2009) argue that “news attribute lists” serve as benchmarks for the cognitive construct of “newsworthiness” (Shoemaker & Vos, 2009, p25); Shoemaker has used news factors in other considerations of deviance (Shoemaker & Danielian, 1991). Furthermore, although research focused explicitly on news factors infrequently mentions deviance, findings generally reflect news which could easily fit the normative deviance model (Galtung & Ruge, 1965; Harcup & O'Neill, 2001; Joye, 2010; J. Lewis & Cushion, 2009).

“Laws and norms define the boundaries of the civilized world. Inside the boundaries is civilization, society as it is supposed to exist. The outside is deviance, a world full or norm and rule breaking, some minor and some fully evil. Events happening outside of the boundaries are more likely to become news items.” (Shoemaker & Vos, 2009, p24)

News factors have not been used extensively for establishing deviance, however; indeed, few methodologies have been used extensively, given that the research explicitly focused on deviance is relatively limited. Shoemaker and Danielian (1991) established news factors as “newsworthiness indicators,” arguing that factors such as oddity, prominence, conflict and sensationalism qualify as indicators of normative deviance; however, there is limited implementation of this approach.

2.3.C: CONTEXTUALIZING DEVIANCE The subject matter of deviance research also deserves mention. The operationalization of normative deviance is potentially quite broad. Indeed, Shoemaker‟s (1996) conceptualization spans a gulf between routine traffic accidents and rare military invasions, and her newsworthiness indicators (Shoemaker & Danielian, 1991) encompass potentially a variety of deviant events; news based on prominent individuals may or may not produce the same kinds of deviant news as articles focused solely on conflict, for example. However, in practice, research on deviance has focused more on fringe movements than broader definitions of deviance.

Studies of protest groups are common. Shoemaker (1984) surveyed journalists about a large number of groups, from the relatively innocuous (League of Women Voters, Sierra Club) to the politically intolerant (KKK, Nazi Party). Graphic photography of protest movements has been studied 33

(Arpan & Tuzunkan, 2011), as has coverage of aggressive abortion protestors (Boyle & Armstrong, 2009). Other studies on deviant behavior, including molestation scandals in the Catholic church (Breen, 1997) and graphic crime news (Jong Hyuk, 2008) also have been analyzed.

It is reasonable to assume a tacit association has developed between the term deviance and groups such as the KKK and Nazi Party. While these organizations would be considered deviant by almost any objective definition, it is important to note that the operationalization of normative deviance is not limited to fringe groups. Particularly if deviance is established by news factors, it is reasonable to argue that deviance spans a much larger spectrum than the difference between abortion protestors and graphic homicides. That spectrum has not yet been explored, but its potential has clearly been established.

Because of these fringe associations, the researcher considered substituting a new term for deviance. “Aberration,” in particular, is a word barely considered in a communication context (Hoffer & Butterfield, 1976). However, given the scholastic history of the word “deviance,” and its clear conceptual fit within this dissertation, the word was retained; furthermore, sociological definitions of deviance are too broad and unwieldy for this research, and words such as “aberration,” “strangeness” or “unusualness” are simply too broad for practical implementation.

Therefore, while it is worth noting the limiting previous subject matters of deviance research, it is equally fruitful to identify the potential for much more inclusive applications of the term. Limiting the definition of deviance to radical and fringe groups is unfairly restrictive, however. Obviously a majority of American news is not focused on the KKK or Nazi party (Shoemaker, 1984), and even broader operationalizations concerning violent crime and demonstrations (Arpan & Tuzunkan, 2011; Breen, 1997; Jong Hyuk, 2008) limit the applicability of deviance to broader spectra of media and subject matter. Furthermore such a radically focused definition would certainly lead to flawed analysis concerning community journalism; simply put, the KKK is unlikely to have a major presence in the average American hamlet. Even if national media are drawn to coverage of small fringe groups, this radical focus is simply too far beyond the reliably hyper-local focus of community newspapers (Garfrerick, 2010; Hansen & Hansen, 2011, 2012; Lauterer, 2006). As such, this dissertation offers an opportunity to reconsider deviance writ large alongside its role within broader gatekeeping theory.

2.4: News Factors & News Values

News factors offer a strong method of operationalizing deviance (Shoemaker & Vos, 2009). News factors research argues that every news article contains discreet factors which can be distilled and quantified (Galtung & Ruge, 1965; Reinemann & Schulz, 2006); those factors have previously been considered deviant and socially significant (Shoemaker, 1984; Shoemaker & Danielian, 1991; Shoemaker, 1996), and this dissertation introduces egalitarian news factors. A full understanding of news factors as a general approach, however, is necessary before using the approach as a foundation for an analysis of deviance.

2.4.A: NEWS FACTORS The study of news factors fits squarely into this focus on news routines. Gatekeeping theory argues that the composite characteristics, or the individual pieces, of a news item ultimately are less relevant than the subjective editorial decisions surrounding their publication (Shoemaker & Vos, 2009). However, news factors research focuses on condensing news articles into common, discreet pieces; it has a history all its own that ultimately lends itself well to the study of deviance.

The news factor approach stems from a pivotal study by Johan Galtung and Mari Ruge (1965), who analyzed coverage of three overseas crises in four Norwegian newspapers and outlined 12 “news factors” common to crisis coverage: frequency, threshold, unambiguity, meaningfulness, consonance, unexpectedness, continuity, composition, reference to elite nations, reference to elite people, reference to persons, and reference to something negative. They found that the more factors a potential news item contained, the more likely that item would receive coverage (Galtung & Ruge, 1965).

“While it seems to be beyond dispute that selection has to take place in order to reduce the complexity of the world surrounding us, the criteria for this [selection] process are a subject of continuing debate. … Unlike the other approaches, news value research examines the content characteristics of the mass media. It is assumed that events have certain characteristics that make them newsworthy. These features – as perceived by journalists – are referred to as news factors.” (Eilders, 2006, p5-6)

The iconic Galtung and Ruge (1965) study has been widely replicated. Harcup and O‟Neill (2001) found the majority of those 12 factors applied to daily news coverage in United Kingdom newspapers; unambiguity was particularly common, although the framework had some difficulty describing entertainment news, references to sex or animals, or news motivated by a photograph (Harcup

& O'Neill, 2001). Joye (2010) refocused Galtung and Ruge‟s (1965) study from crisis news to disaster news in Flemish newspapers and found proximity, geographically and culturally, as the predominant motivator for coverage of European or western disasters and marginalized news coverage of Asian, African or Latin American disasters (Joye, 2010). Kepplinger and Ehmig (2006) argued that news factors could serve predictive purpose rather than simply offering post-production illustrative detail, while Lewis and Cushion (2009) found breaking news so prevalent on 24-hour British television news networks that factors of “unpredictability” often usurped “predictability;” breaking news can be banal, in a sense, as long as it was current first and foremost. Of note, too, are considerations of Islamic religious values serving as news factors in Arab newspapers (Elliott & Greer, 2010; Mowlana, 1996).

Many studies ultimately focus on or highlight one primary news factor (Joye, 2010; Kepplinger & Ehmig, 2006; J. Lewis & Cushion, 2009; Pamela Shoemaker, 2006); alternate sets of factors have evolved as well. Schulz (1976) designed a news factor rubric which emphasized national centricity, regional centricity, geographical proximity, political proximity, cultural proximity, and ethnocentricity, among others. Badii and Ward (1980) focused on significance, normality, prominence and reward; Corrigan (1990) considered significance, vitality and conflict, human interest, timeliness, prominence, consequence, and proximity; Östgaard (1965) focused on simplication, identification, and sensationalism; and, Gladney (1996) expanded the list to 18 factors including “accuracy,” “good writing” and “visual appeal” (Gladney, 1996). Finally, as previously mentioned, Bridges and Bridges (J. Bridges, 1989; J. Bridges & L. Bridges, 1997) studied proximity, prominence, timeliness, impact, magnitude, conflict and oddity.

It also is worth mentioning that news factor research is distantly related to some framing research. Semetko and Valkenburg (2000) found five prominent frames in news coverage of a European political conference: attribution of responsibility, conflict, human interest, economic consequences, and morality. Their framework has been adapted and replicated (De Vreese, Peter, & Semetko, 2001; Holt & Major, 2010; Matthes, 2009; Semetko & Valkenburg, 2000). News factor analyses and framing studies have independent histories and clear distinctions; however, there is a curious degree of overlap worth noting here.

The principles behind the news factor tradition are sound. There is clear academic consensus that news can be distilled into common factors, and that in most cases, the presence of those factors in a news item increases the likelihood of its publication. There also is consensus that news factors fit into the

professional routines level of the hierarchy of influences model (Shoemaker & Reese, 1996; Shoemaker & Vos, 2009); further theoretical development, however, does less to explain how those factors can predict of future news coverage and how editors perceive them (Shoemaker & Danielian, 1991). While studies show news articles can be distilled into discreet news factors, these primarily European studies simply mesh into the professional routines section of the hierarchy of influences.

But news factors offer additional utility. Gatekeeping theory stipulates news factors are characteristics subject to subjective routine-level editorial decisions; the news factor approach typically documents these factors, but rarely digs into deeper theoretical territory into why those news factors are so routinely chosen. Gatekeeping theory, too, acknowledges a professional media preference for deviance – stories outside the boundaries of normal, typical life. This dissertation argues that news factors can serve as a quantitative tabulation of deviant and egalitarian news factors.

The study of news factors is compelling in its own right, and has never been applied to American community newspapers. Since community newspapers elicit a great deal of audience interaction (Funk, 2013; Hansen, 2007; Hansen & Hansen, 2012; Lauterer, 2006; Reader, 2006), such community newspapers provide a key test for gatekeeping theory‟s assumption of audience distance, discussed in 1.2.a and 1.2.c. The closest community application was a study of daily newspapers in Oklahoma more than 30 years ago (Badii & Ward, 1980). And, although Bridges and Bridges (J. Bridges, 1989; J. Bridges & L. Bridges, 1997) considered newspaper circulation size when crafting their national datasets, that circulation was not considered an independent variable; too, they studied only daily publications, not weeklies.

Furthermore, the study of news factors has only infrequently analyzed any American media. While some domestic studies have used the approach (Badii & Ward, 1980; Bridges, 1989; Bridges & Bridges, 1997; Corrigan, 1990), the approach is far more common in the United Kingdom (Firmstone, 2008; Harcup & O'Neill, 2001; J. Lewis & Cushion, 2009), mainland Europe (Eilders, 2006; Joye, 2010; Kepplinger & Ehmig, 2006; Östgaard, 1965; Schulz, 1976) and elsewhere (Elliott & Greer, 2010; Jong Hyuk, 2009; Jong Hyuk & Yun Jung, 2009; Mowlana, 1996). American studies, by contrast, are more commonly focused on news values and news production more generally (Gans, 1979; Johnson & Kelly, 2003; Shoemaker, 2006; Tuchman, 1978).

2.4.B: NEWS VALUES Exploration of news factors requires delineation between “news factors” and “news values,” which are similar but distinct. In a sense, news factors studies are a sub-set of news values studies. News factors are the discreet pieces or components of a news story, while news values relate to qualitative decisions made by journalists and editors to determine what is, and is not, news. A prime scholarly example is Johnson and Kelly‟s (2003) online survey of online editors at traditional newspapers which indicates broad principles of interpretive function, disseminator function and adversary function are relatively consistent among print and online editors (Johnson & Kelly, 2003); these “functions” refer to qualitative ideas, not quantitative factors buried inside published news. Other studies include broader analyses of news selection (Gans, 1979; Shoemaker & Reese, 1996; Tuchman, 1978).

While tangentially related, these concepts have key distinctions; given their terminological similarity, it is important to distinguish between them. News factor research must include factors – the research must be interested in the pieces comprising specific news items, the discreet factors present in any particular article. Broader analysis of the news construction process concerns news values.

2.5: Community in the Digital Age

Gatekeeping theory‟s assumption that interaction between journalists and media audiences, and journalists‟ editorial use of deviance, are inversely related is rooted in an assumption that journalists have “modest exposure to their audience” (Shoemaker & Vos, 2009, p52). Nowhere is this assumption less true than in American community newspapers. A wide range of scholarship detailed in this section outlines the highly interactive, community-centric and locally focused nature of small weekly and daily newspapers. Such publications offer an ideal environment to test gatekeeping theory‟s assumption concerning audience interaction and deviant news. Understanding the background and examples of community newspapers‟ accessible and hyper-local nature is therefore highly relevant.

The term “community journalism” has a dual definition in the mass communication academy. On the one hand, it is a colloquial term that refers to small, locally-focused newspapers which are often highly traditional and located in rural or suburban areas; this conceptualization is clear, concise, and quite common (Bowd, 2011; Funk, 2013; Givens, 2012; Hansen & Hansen, 2011; Hume, 2005; Lauterer, 2006; Martin, 1988). On the other hand, this traditional definition does little to consider conceptual or theoretical definitions of community. Understanding theoretical community is critical to journalism for writers and academics. Gene Burd, writing in 1979, argued that a broader understanding of community would be needed as technology and media continued to change – long before the Internet escalated the pace of media evolution (Burd, 1979).

“Definitions of community become crucial to journalistic training, practice and performance. In fact, the separation of community from any type of journalism may be a contradiction. The assignment of the term „community‟ to mean only journalism of small communities may reflect an anti-urban, anti-big city bias among American writers and thinkers.” (Burd, 1979, p3)

This is especially true in the digital age. All communities are intangible, flexible, inherently ideological entities imagined by their members (Anderson, 2006). The degree of imagination and flexibility, however, fluctuates based upon the population and physical size of any particular community.

This fluctuation remains true in the Internet age. Smaller communities, with smaller media, tend to imagine their communities to a lesser degree than larger media and larger communities. It is important to note, however, that the process of imagined community can no longer be devoted exclusively to physical communities. Online and ideological communities of all varieties also must be considered with the same logic. 39

Indeed, the concept of community is more malleable now than it has ever been before. This complexity surrounding community needs exploration before community journalism can be operationalized.

2.5.A: IMAGINED COMMUNITY The classic definition of imagined community comes from Benedict Anderson (2006), who wrote that every community exists primarily in the hearts and minds of its members. Holding that community together and crafting that identity‟s features and characteristics are local media. Before print media, and in particular the printing press, community associations were limited to visible physical spaces and devoted almost exclusively to ideas and characteristics common to individuals in direct contact with one another; after local media took strong root in a society, however, media began establishing borders and identities around areas and concepts which were considered local and foreign. Local media began delineating which places, ideas, and individuals were related to a local community, and which were not, through slow but persistent repetition. These associations were particularly significant on a linguistic level, and served to standardize local dialects of German, English, or Spanish into broader amalgamations (Anderson, 2006).

To Anderson, the cornerstone of that imagined community was print capitalism. To local media, homogeneity and an imagined community were profitable constructions; generating audience loyalty to a place lead by proxy to loyalty to its newspaper. The simplest way to imagine that profitable community, step by step, was slowly developing associations between a place, a medium, and particular ideas. That‟s why the most essential part of a newspaper (or any medium) remains the header, or banner, atop the page which imagines the community and publication in tandem (Anderson, 2006).

“The date at the top of the newspaper, the single most important emblem on it, provides the essential connection – the steady onward clocking of homogenous, empty time. Within that time, „the world‟ ambles sturdily ahead. … which made it possible for rapidly growing numbers of people to think about themselves, and to relate themselves to others, in profoundly new ways.” (Anderson, 2006, p33-36)

From there, Anderson argued, local identity lead to regional and national identity and nationalism. Associations of millions of people, separated by great distances, were held together by artificial ideas and an imagined sense of collective belonging. It would be simply impossible, he argued, for every American to meet every other American in person; however, through media and imagined community, all Americans share a similar sense of national community. 40

The notion of community remains most concrete, however, among small communities and small media. Although it is important to acknowledge the intangible nature of community, it is equally important to recognize the relatively concrete natures of local community and community journalism.

2.5.B: COMMUNITY JOURNALISM & IMAGINED COMMUNITY Traditional, print community journalism has long held a “relentlessly local” focus (Lauterer, 2006). Small American newspapers run more locally focused content than larger publications (Bridges & Bridges, 1997; Funk, 2010); their obituaries hold special reverence for local readers (Hume, 2005); and they are sorely missed when they disappear (Smethers et al, 2007). There also is evidence that modern newspaper use promotes “social cohesion,” or a sense of collective belonging among readers (Yamamoto, 2011). However, in most cases, journalists articulate “local news” through basic geography; this works well for, say, a town of 800 residents, but becomes inflexible as populations grow.

Reader (2006) interviewed journalists at small and large newspapers and noted that all journalists ultimately substitute an imagined community for a lack of personal knowledge of their readers, but that the “process of “imagining” the community is more difficult and more prone to detachment in large, pluralistic communities, and somewhat easier and more prone to attachment in small, homogenous communities” (Reader, 2006, p855). Communities still are imagined, only to a lesser degree; there also is some ambiguity on how much, exactly, local community is imagined and enforced by local media (Reader, 2006).

For example, researchers at the universities of Alabama and Virginia Tech gathered a group of community journalists and academics to ask a simple question: What does the phrase “community journalism” mean? Not surprisingly, there was no immediate consensus. The term meant different things to different editors, in part because of differences in newspaper size and editorial mission (Lowrey, Brozana & Mackay, 2008). The group ultimately decided that community journalism is a shared process which structures facilities, institutions and spaces for collective sharing.

“Community journalism is intimate, caring and personal; it reflects the community and tells its stories; and it embraces a leadership role,” the group decided (Lowrey, Brozana & Mackay, 2008, p276). This definition finds middle ground among diverse ideologies; more importantly, however, it establishes a key middle ground between professional practice and Anderson‟s highly theoretical definition of imagined community. The Lowrey, Brozana and Mackay model acknowledges that community is partly

an artificial construction; “the community” is not fully defined, and claims of it as “intimate, caring and personal” require a complicit audience. However, the model does not assert that community is entirely constructed or artificial; adopting a leadership role is different than espousing values or safeguarding ideology. The model implies that there is a cohesive, intrinsic community which also is intimate, caring and personal, and that the community newspaper is connected to those values and that place (Lowrey, Brozana & Mackay, 2008, p276).

Put another way: The community newspaper is part community imagination, part community connection. Community journalists retain those connections through proximity and contact with local community members as part of their essential mission, thus reinforcing this middle ground between the imagination of an artificial community and a reflection of a tangible community. Journalists at larger newspapers simply cover too much ground, and too many audiences, to develop similar relationships.

“Australian country newspapers‟ contribution to interaction, communication and empowerment in their circulation areas is underpinned by their emphasis on „localness.‟ … Journalists at country newspapers are usually readily accessible to their audiences, something described by one tri- weekly journalist as building integrity into the process of communicating the news.” (Bowd, 2011, p76-85)

More illustrative examples lie among local media in semi-autonomous regions of Spain and France. Galicia, Basque and Catalonia each have local Spanish dialects and historical independence from Madrid; editors of their local newspapers saw a daily tension between preserving that local identity, and those dialects, and conforming to professional industry standards (S. Lewis, 2008). Similarly, journalists at Corsica‟s largest newspaper, Corse-Matin, are torn between remaining officially agnostic toward charged, occasionally separatist tensions about the island‟s semi-autonomous status and supporting a “„regional‟, folkloric and depoliticised” Corsican cultural identity (Richardson, Huckerby & Williams, 2008, p574).

Like any small or regional community, local identity already was more tangibly rooted in the community; Basques had a sense of being Basque, for example, that was at least partly grassroots and not fully constructed by the media. The concept of Spain or France writ large, however, was more artificial precisely because those areas are so much larger (S. Lewis, 2008; Richardson, Huckerby, & Williams, 2008). The same can be said for urban American environments. Metropolitan cities have diversified geographically, racially, and culturally, creating a “new urban geography” that needs new, more narrowly targeted media to serve community needs (Burd, 2008).

Lowrey, Brozana and Mackay also noted, however, that the study of community journalism rarely acknowledged this dynamic. They studied 108 scholarly articles on community journalism between 1995 and 2005 and found that only about 30 even defined “community journalism” at all – and another 30 suggested or directly stated that community is fundamentally tied to physical location. This, they argued, is an oversimplification. Physical space is certainly an integral part of imagined community, particularly a small community; however, it is only part of that imagination. “Negotiated shared meaning,” and each of the important functions community newspaper editors identified, also are an inseparable part of the equation.

“„Community news media,‟ then, are media capable of fostering the process of community as depicted previously. Community journalism would (a) reveal, or make individuals aware of, spaces, institutions, resources, events, and ideas that may be shared, and encourage such sharing and (b) facilitate the process of negotiating and making meaning about community.” (Lowrey, Brozana & Mackay, 2008, p288).

In a sense, scholars and community journalists must recognize that they are, in fact, imagining a social community as much as a physical one. The process remains simpler, however, for community journalists; small communities require less artificial imagination than cities or nations, as Anderson (2006) also noted.

2.5.C: COMMUNITY JOURNALISM IN PRACTICE That focus on traditional geographic community, however, has worked quite well for community journalists. In 2006, nearly 98 percent of American newspapers had regular circulation of less than 50,000; those publications reached a total of 108.9 million cumulative readers every week. Nearly half that circulation was tapped by 7,865 weekly newspapers, including a variety of community newspapers, alternative presses and ethnic- or sexuality-oriented niche publications (Lauterer, 2006, p6-7). Recent Pew data indicate that nearly 72 percent of adults enjoy and rely heavily upon local newspapers, and that surprisingly diverse groups of readers are attached to local newspapers (Miller, Purcell & Rosensteil, 2012); data also indicate that small-town residents in particular have heavy reliance on, and attachment to, traditional newspapers (Miller, Rainie, Purcell, Mitchell, & Rosenstiel, 2012). These community newspapers are scattered throughout American communities, from metropolitan neighborhoods to isolated rural outposts; most have long since cornered the news markets in their communities, primarily through a focus on local news, institutions and individuals.

“It‟s the kind of journalism practiced by newspapers where the readers can walk right into the newsroom and tell an editor what‟s on their minds. It‟s the kind of newspaper that covers the town council, prints the school lunch menus, leads the sports page with the high school football game, tells you who visited Aunt Susie last week and runs photos of proud gardeners holding oddly shaped vegetables. … Though such papers are small, their impact is huge.” (Lauterer, 2006, p3)

By focusing exhaustively on a particular community, community newspapers have generated a reliable product with a steady advertising base. They‟ve also provided intriguing locally focused counter- examples to major national media focused more and more on conflicts and campaigns.

When Hurricane Katrina ravaged New Orleans and the Louisiana coast in 2005, local and regional media were considerably more likely to quote relief workers and survivors than national media; they also provided information on how to receive disaster aid and where to find help or loved ones (Dill & Wu, 2009); similarly, when a tornado demolished West Liberty, Ky., and its newspaper, The Licking Valley Courier, survivors rallied around the publication as a symbol of the entire community (Hansen & Hansen, 2012).

“Such a crisis brings the nature of often unspoken and taken-for-granted relationships to the surface and makes them visible. … In the present case, both the newspaper itself and the community experienced a disaster and were in crisis mode and each may play a significant role in the other‟s recovery.” (Hansen & Hansen, 2012, p8)

Martin (1988) found that a local newspaper in Pennsylvania quoted more sources, and a greater variety of sources, during tense education budget negotiations than the regional metropolitan newspaper‟s coverage of the same event (Martin, 1988). Community newspaper editors are more accessible to the public (Bowd, 2011; Lauterer, 2006); they receive positive feedback for offering more local content (Hansen & Hansen, 2011); they generally shy away from Associated Press and wire news content, particularly at very small newspapers (Funk, 2010); local obituaries are historically considered particularly important to local audiences (Hume, 2005), and although one study indicates that social announcements have declined in community newspapers, it also notes an increase of published community and church-based announcements (Givens, 2012). American community journalism also has been suggested as a model for developing local journalism in China (Lauterer, 2012).

Furthermore, editors at small and community newspapers generally are more concerned with maintaining a pro-community attitude and positive local relations than adhering to broader, more rigid professional standards (Reader, 2006). This perspective can arguably lead to a “softer” focus on hard 44

news (Bishop, 2009; Harry, 2001; Jeffres & Dobos, 1983), as can community newspaper‟s reliance upon local, potentially impressionable advertisers (Donohue, et al., 1985). However, this perspective also builds a great deal of audience loyalty and a fundamental connection to the local community.

Community journalism, therefore, provides an ideal opportunity to test gatekeeping theory‟s assumption concerning audience interaction and deviance. Community newspapers are close to community readers; therefore, studying community journalists‟ publication of deviant news, as well as qualitatively analyzing their perspective on deviant and egalitarian news factors, speaks to the core of this study.

2.6: New Journalisms & Social Media

Much of this study, however, is focused on social media. It is logical to assume that a community newspaper‟s print focus would be extended to its online presence, but this notion further complicates operationalizations of imagined community. These complexities surrounding online community must be explored before analysis can begin.

If scholars of community journalism are remiss to mention imagined community, then academics analyzing new media must be particularly careful. Community journalism, by its very nature, assumes and imagines an easily identifiable community; other communities, particularly those online, require a more flexible understanding. This is true even for logical communities, such as the social media platforms of traditional print community newspapers.

Put another way: It is no longer logical to assume that “community” is limited to physical space. Online or niche communities are no less imagined, or artificial, than traditional communities.

Concerning offline media, Reader (2007) identified an imagined community among listeners of, and feedback to, National Public Radio. NPR‟s weekly “letters” segment reads audience feedback to previous stories on air. Despite clear gatekeeping forces determining which letters would be accepted or excluded, the service was designed to be inclusive and build camaraderie among NPR‟s audience. It worked. “Imagining community is as much a process of journalism as a product of it, a process that is largely defined by the sociology of the news profession” (Reader, 2007, p655). A community of audience members for a radio news program is not the same as a community of residents in a particular town, nor is it precisely what Anderson (2006) had in mind when he formulated his theory; however, the same concepts apply.

On a larger scale, newspapers within the digital world Second Life held some community- imagining ability as well (Brennen & dela Cerna, 2010). Newspapers in the popular online world were very much rooted in the world‟s identity, and based much of their coverage on massively multiplayer online issues. The Metaverse Messenger, Alphaville Herald and The Second Life Newspaper were each very cognizant of community needs and happenings within the world, and used jargon and language clearly tailored to computer-savvy readers; however, they also helped articulate a community identity within the digital realm, even if that mediated identity was based partly on pre-existing community characteristics. 46

“The creation and maintenance of community is a fundamental concern of all the journalism practiced in Second Life. All three newspapers include frequent extended discussion of legal and technical issues and community concerns as well as provide detailed information to help residents navigate through Second Life.” (Brennen & dela Cerna, 2010, p550)

The same could be said for homosexual media and communities, or any number of non- geographic or ideologically based media audiences. Gay media typically would not be considered “local,” as “gay” pertains more to lifestyle than place; in fact, that social identity is displaced over wide geographic areas. That wide geographic focus gives those media even greater authority in imagining community and belonging; in this case, Cover (2005) argues, that imagined community is specifically focused on a sense of non-geographic community, belonging and self-identification. He sees “community as an identifiable group held together by a set of symbols, rituals, institutionalized behaviours and norms, whether enacted through specific social spaces such as media audiencehood or geographically-local public space” (Cover, 2005, p 115).

This more flexible articulation of community, and imagined community, shifts the focus beyond physical space and into social and ideological realms. Here, arguably, Anderson‟s (2006) notion of imagined community holds even more authority. Without a tangible geographical identity to serve as a benchmark, media potentially have a blank canvas upon which to imagine a particular non-local community. In Lewis‟ (2008) Basque study, newspaper editors partly based their imagined communities on pre-existing regional identity; homosexual or niche publications could potentially start from scratch.

Online communities imagined by websites such as “Tu Diabetes” are completely artificial; an online health community may have no obvious geographic roots, and the website designers have a great deal of authority in determining what constitutes such a community (Arduser, 2011). A community of Wikipedia authors and editors is highly technologically deterministic; without the website, there would be no imagined community surrounding the website (Pentzold, 2011). Similarly, a YouTube community surrounding the Madeline McCann kidnapping case, a controversial and highly-covered child kidnapping investigation in Portugal, is highly imagined and disassociated from any particular place (Kennedy, 2010).

2.6.A: COMMUNITY NEWSPAPERS ONLINE: A LACK OF SCHOLARSHIP Once broad online communities are explored, community journalism online deserves particular consideration. As with any other media, community newspapers face pressures to evolve and digitize;

while there has been ample discussion of so-called “citizen journalism” and community newspapers (Brockus, 2009; Burroughs, 2006; S. Lewis, et al., 2009), literature on community newspapers‟ transition toward digital media is more limited (C. Greer & Ferguson, 2011; Lowrey, et al., 2008).

There also is evidence that the audiences for a community newspaper‟s print and online versions are different. Hansen (2007) found that one community newspaper in London, Ky., had an online audience that was younger, more affluent, more educated and more geographically displaced than its print audiences; online audiences were largely ex-residents, many of them local high school graduates who had moved to new areas but wanted to retain a connection to the London area; Sylvie and Chyi (2007) also found that local readers constituted fewer than 50 percent of most newspaper‟s online audiences, while the remainder was devoted to varying degrees of “long-distance readership.”

But these studies do not expressly analyze online content in American community newspapers; in fact, the researcher was unable to find any study which contrasted print and online content written by community newspapers. Given the turbulent changes and vast potential offered by online and new media (Hermida, 2010; Lasorsa, et al., 2011), as well as evidence that online and print audiences differ, the composition of online, community newspaper content is absolutely worth considering.

Therefore, community newspapers provide an ideal media for the study of gatekeeping theory. Their local orientation offers an intriguing test for gatekeeping‟s focus on deviance and distance from the audience. Deviant news factors may, or may not, remain present in community newspapers despite that connection to local audiences; thus, deviance and audience disconnection may be either a fundamental or circumstantial component of gatekeeping theory.

2.6.B: SOCIAL MEDIA WRIT LARGE Like community journalism, social media present great opportunity for interaction between audience members and media producers; interaction between media producer and media audiences is easily accessible, thus offering an opportunity to further test gatekeeping theory‟s assumption that deviance and professional routines are derived from a separation from the audience. The academic study of social media has rarely considered community journalism, local media, deviance, or news factors, however. Considerably more common are analyses of social media coverage of major events, such as the Arab Spring (Christensen, 2011; Harb, 2011; Miladi, 2011; Sayed, 2011) or American (Fernandes,

Giurcanu, Bowers, & Neely, 2010; Towner & Dulio, 2012; Woolley, et al., 2010) or overseas elections (Aparaschivei, 2011; Ifukor, 2010; Knight, 2012; Larsson & Moe, 2012).

Concerning community journalism, Greer and Yan (2010) found a surprisingly high rate of adoption of new- and social-media tools and platforms among community newspapers; to the contrary, Gilligan (2011) found that community newspaper websites are not fully utilizing or adopting electronic media, and that “an expanded understanding of the „community‟ in „community newspapers‟ … is the key to increasing sustainability for those newspapers” (Gilligan, 2011, p69). Such analyses are useful, albeit topical; they also constitute the lion‟s share of scholarship on community newspapers and social media. A few select studies analyze “local media,” however, and bear tangential relation to this dissertation (Ferguson & Greer, 2011; C. Greer & Ferguson, 2011). Analysis of citizen journalism and community journalism also offers some relevance (Brockus, 2009; S. Lewis, et al., 2009; Rossow, 2009), in that citizen journalism involves interaction between media producers and media audiences; however, as citizen journalism requires the practitioner to be a member of the audience and the professional ranks, it is a discreet phenomenon that applies tangentially here. This dissertation is concerned with interaction between journalists and audience members in a sense, not unique partnerships between them.

The closest comparison to news factor research is a look at Twitter “microblogging” practices that considered traditional news values and editorializing on the social media platform; categories such as “job talking,” “retweeting” and “personalizing” offer important insight into the characteristics of news tweets, but characteristics and factors are not identical concepts (Lasorsa, et al., 2011).

The conceptualization of community is imagined in print and online. However, community journalism - both offline and on social media platforms - offers a tangible medium with high degrees of audience interaction to test gatekeeping theory‟s assumption on audience interaction and deviance.

2.7: Summary and Literature / Hypotheses Link

This literature review has considered several concepts critical to this dissertation. The first is gatekeeping theory, which is partially predicated on an untested assumption that journalists default to deviant news due to a lack of direct interaction with media audiences (Shoemaker & Reese, 1996; Shoemaker & Vos, 2009; White, 1950). The second concept is an exploration of normative deviance, which ostensibly considers any behavior beyond societal norms but in practice has focused largely on fringe or radical behavior or ideologies (Arpan & Tuzunkan, 2011; Jong Hyuk, 2008; Shoemaker, 1984; Shoemaker & Danielian, 1991). The third concept considers the study of news factors, which have proven an effective method of distilling news into composite, analyze-able pieces (Eilders, 2006; Galtung & Ruge, 1965; Harcup & O'Neill, 2001; Kim, 2012). The fourth concept considers American community journalism, which has a long history of hyper-local orientation but has rarely been studied alongside deviance or news factors (Hansen & Hansen, 2011; Lauterer, 2006; Reader, 2006). The fifth concept considers social media and community newspapers, which have infrequently been studied in tandem (J. Greer & Yan, 2010; Hermida, 2010; Lasorsa, et al., 2011).

These sections chart a clear theoretical course through this dissertation, thus informing the hypotheses. Gatekeeping theory assumes that journalists have little practical interaction with their audiences, but a litany of past research reveals community newspapers‟ local orientation toward, and accessibility to, local audiences and communities. It is logical to assume, therefore, that community newspapers will focus more on egalitarian news factors than deviant news factors.

It is also logical to assume, given community newspapers‟ local orientation, that prominent local officials will be more accessible than elites in larger communities. A mayor or state representative would be considered a political elite in any municipality, but newspapers and elites alike are more likely to have a streamlined focus on local events and institutions, and thus more likely to interact. It seems logical, too, that perhaps such local focus might mitigate interest in outright conflict. Therefore, this study hypothesizes that community newspapers will have higher rates of prominence-based deviance than larger newspapers but will focus less on conflict-based deviance.

Social media are, by their very nature, egalitarian enterprises. They empower and enable dynamic online conversation and upend traditional journalist-to-audience dynamics; traditional websites, however, have a lower degree of interactivity. It is logical, therefore, to assume that newspapers‟ posts on social media platforms will emphasize egalitarian factors more than deviant factors while news on traditional websites emphasizes deviant factors over egalitarian factors.

H3: Regardless of circulation size, American community and daily newspapers will publish social media content with higher rates of egalitarian news than deviant news factors.

Finally, gatekeeping theory‟s assumption that journalists are out of touch with their audiences seems wholly inconsistent with the nature and business model of community journalism; its related assumption that journalists default to deviant news because of that disconnection also seems suspect. It is logical to assume, instead, that the rate of interaction between community journalists and their audiences – which is presumably very high – will influence their preference for egalitarian news.

H4: The more contact a community newspaper editor has with their community and readers, the more preference will be given to egalitarian news over deviant news.

These hypotheses will be tested using the quantitative and qualitative methodologies outlined in Chapter 3.

CHAPTER 3: METHODOLOGY

3.0: Introduction

This chapter, on the data and method of this dissertation, outlines this dissertation‟s practical steps. It summarizes the theoretical concepts and gives a detailed explanation of the principal research methods of computerized content analysis and structured, qualitative interviews. It also outlines the quantitative and qualitative sampling, data collection and coding processes.

As stated in Chapter I, this dissertation seeks to test two primary gatekeeping stipulations: (1) whether journalists rely on deviance when crafting news content, and (2) whether news routines are designed largely without input or direction from media audiences (Shoemaker & Vos, 2009; Shoemaker & Reese, 1996). To test the former theoretical relationship, news factor research and Bridges and Bridges‟ (J. Bridges, 1989; J. Bridges & L. Bridges, 1997) seven news factors are employed to measure deviant, neutral and egalitarian news factors. To test the latter theoretical relationship, data are drawn from the websites and social media pages of American community newspapers, which have a high level of interaction with their audiences, and compared to that of a wide range of less-interactive larger newspapers. This dissertation employs computerized content analysis and qualitative, structured interviews of community newspaper editors to pursue these research questions.

3.1: Hypotheses

This dissertation focuses on gatekeeping theory and Bridges and Bridges‟ (J. Bridges, 1989; J. Bridges & L. Bridges, 1997) seven news factors as they relate to American community newspapers, online and in social media; it conducts the same analysis on community daily, large daily, and national daily American newspapers. It also conducts structured, qualitative interviews of community newspaper editors to determine editorial decisions and perspective of those factors. These questions clearly point to central tenants of gatekeeping theory and the hierarchy of influences model (Shoemaker & Vos, 2009; Shoemaker & Reese, 1996). Therefore, as previously mentioned in Chapter 1.3, the hypotheses are as follows.

H3: Regardless of circulation size, American community and daily newspapers will publish social media content with higher rates of egalitarian news than deviant news factors.

H4: The more contact a community newspaper editor has with their community and readers, the more preference will be given to egalitarian news over deviant news.

H1, H2 and H3 are devoted to establishing quantitative patterns concerning the construction of deviant, neutral, and egalitarian news factors; quantitative analyses within the computerized content analysis program DICTION 6.0 will produce a composite measure which allows comparison across circulation sizes and illuminate the role circulation size plays in the use of those factors on both newspaper websites and social media platforms. H4 studies community newspaper editors directly and, through qualitative structured interviews, assesses their perspectives on audience interaction and deviance with an eye toward gatekeeping theory.

3.2: Methodology Structure

This dissertation utilizes two major methodological approaches to answer its research questions. Both methods have long academic traditions; both also are quite distinct. The first method, computerized content analysis, utilizes relatively new technology to adapt and develop traditional content analysis in very specific ways. The second method, structured qualitative interviews, applies a regimented interview format to increase the reliability of a highly subjective, qualitative methodological practice.

As a matter of organization, this dissertation explores the research paradigms, dataset and sampling procedures, data collection process, and data processing schemes for computerized content analysis and qualitative structured interviews.

3.3: Computerized Content Analysis

The computerized content analysis section of this dissertation utilizes the DICTION 6.0 software. An outline of broad computerized content analysis methodology, as well as specifics pertaining to this study, are outlined in this chapter.

3.3.A: QUANTITATIVE RESEARCH PARADIGM Content analysis is the systematic, inter-subjective study of content which is extant, and absent, within a particular text. It is typically “systematic, objective, quantitative analysis” (Nuendorf, 2002, p1) that “limits itself to the produced content alone and draws conclusions based on what is there” (Poindexter & McCombs, 2000, p188). Journalism scholars typically use content analysis to study media texts; however, similar content analysis techniques can be applied to sociology, political science, broader communication disciplines or potentially any area that considers quantitative patterns in produced material (Krippendorff, 2004; Nuendorf, 2002).

The goal of any content analysis is the valid and reliable translation of media content into useful statistical data. The method is a highly structured process with eight discreet steps: defining recording units, defining categories, testing the coding scheme, assessing coding accuracy, revision, further testing, coding the full text or data, and assessing overall reliability (Weber, 1990, p23-24). By strictly structuring the process, the content analyst removes him or herself from the analysis process, thus leaving behind an objective dataset that could (and, hopefully, will) be read, replicated and adapted by future scholars (Krippendorff, 2004; Nuendorf, 2002; Poindexter & McCombs, 2000; Weber, 1990).

To do so, coders approach a text with clear processes and research questions. The coders read, hear, or watch media content to determine the presence or absence of particular items; those presences and absences are marked in a pre-written code sheet, and the data are later processed and statistically analyzed. Content analysis is inherently a subjective and objective process; the research questions are determined subjectively, and the coding process is invariably influenced by the subjective characteristics and approaches of the individual coder; however, strict adherence to a common coding scheme and rigorous coder training reduce that subjective influence to acceptable, if not minimal, levels (Krippendorff, 2004; Nuendorf, 2002; Poindexter & McCombs, 2000; Weber, 1990).

Of critical interest in establishing that objectivity are validity and reliability. The former ensures, in a sense, that the researcher is properly measuring the objective and not something else, and that extraneous factors are not influencing the data. Too, validity refers to the appropriateness of the methodology and coding scheme to the research questions. Reliability ensures that repeated analysis by the same coder would yield similar or identical results, and that analysis by different coders also would yield similar or identical results.

Once risks have been addressed, content analysis often yields compelling statistical data that cannot be derived any other way. The original gatekeeping study required (in part) a content analysis of editorial decisions (Shoemaker & T. Vos, 2009; White, 1950); similarly, the seminal agenda-setting study utilized content analysis of election coverage in national and regional newspapers (McCombs, 2004; McCombs & Shaw, 1972). Content analysis methodologies can, and have, been applied to a diverse set of media contents across the globe (Bodle, Burriss, Farwell, Hammaker, & Joshi, 2011; Chen, 2012; Luke, Caburnay, & Cohen, 2011; Napier, 2011; ÖZkoÇAk & Tuna, 2011).

Introducing computerized content analysis software adds important simplification and intriguing depth to the research process. On the one hand, even in 2013, computerized content analysis remains limited to highly sophisticated word processing software. Studies that rely upon even simple associations or context usually are beyond content analysis software; one comparison of a traditional and computerized content analysis concerning attribute agenda setting yielded vastly different results (Conway, 2006), and content analysis scholars are quick to advocate simplicity and specificity when dealing with computers (Krippendorff, 2004; Nuendorf, 2002).

“Potential CATA (computerized content analysis software) users must be careful not to fall prey to fancy labels and abstract concepts claiming to describe what sophisticated software can do. … The use of computers is most appropriate for recurrent and repetitive tasks that can be conceptualized without uncertainty. Searching, coding, sorting, listing, and counting are obvious candidates.” (Krippendorf, 2004, p261)

Establishing validity, then, is of paramount importance for computerized content analyses. Because computerized content analysis is inherently limited, the best method of ensuring validity uses words as the units of analyses; associations and context are effectively too nuanced for computerized content analysis software (some studies, including Lowry & Xie (2007) argue that word “clustering,” or the proximity of one word to another, can appropriately tabulate conceptual associations; this researcher is unconvinced, however, given the litany of scholarship speaking to the simplicity of computer coding

and the nuanced nature of modern media). Word-based computerized content analysis is achieved primarily through the use of word dictionaries, essentially large word banks; the computer searches the text for every instance of every word in a word dictionary and groups those terms according to the researcher‟s specifications.

“Dictionaries constructed by the researcher are called custom dictionaries, as I‟ve mentioned. … For example, I have used up to 157 different custom word sets (dictionaries) for a single computer content analysis using the computer program VBPro. By using a large number or narrowly defined dictionaries, (eg, “newspaper” and all its synonyms; “television” and all its synonyms), the researcher has the option of creating a variety of flexible index combinations of the dictionaries.” (Nuendorf, 2002, p127)

For this particular dissertation, each of Bridges and Bridges‟ (J. Bridges, 1989; J. Bridges & L. Bridges, 1997) news factors, as well as factors added for this analysis, will be assigned a word dictionary; the computer will count the frequencies of those words in the text, group those frequencies accordingly, and provide one frequency for each news factor in each circulation category and online media platform. The statistics at play are relatively simple; the challenge lies in ensuring that all relevant words are included and inappropriate words are struck from the word dictionaries. This is an important step in establishing validity.

Once validity has been established, however, reliability is an extremely simplified process. Computers cannot not be reliable; as a matter of design, they approach every data point in identical manner (Krippendorff, 2004; Nuendorf, 2002). Furthermore, computerized content analysis programs are capable of almost instantaneously processing vast amounts of data – increasing the dataset well beyond the logistical limitations of most traditional content analyses.

Examples abound. Jarvis (2004) used computerized content analysis to study 14 American presidential elections between 1948 and 2000 and found three broad rhetorical trends: Democrats were more likely than Republicans to discuss campaign actors or use nouns to rhetorically build coalitions; Republicans were more likely to focus directly on ideals; and, both these trends have decreased over time. She found evidence that modern campaigns are more candidate-centric, but reliance on traditional partisan tropes remains evident (Jarvis, 2004); similarly, a study of presidential campaigns from 1952 to 1996 found an overall decline in optimism, realism and certainty (Ballotti & Kaid, 2000). A computerized content analysis of media produced by the United Church of God during its infancy identified shifting organizational identity and values over a five-year period (Aust, 2004), and television news coverage of

the September 11th terrorist attacks was consistently more emotional than newspaper coverage (Cho, et al., 2003).

“While these computerized programs are not as flexible, clever or attentive to details of context as would be a human coder, this type of coding is consistent, replicable and reliable” (Jarvis, 2004, p407).

Therefore, this dissertation adopts computerized content analysis as its first research paradigm for the following three summarized reasons:

1. Computerized content analysis offers a sound, perfectly reliable method for processing textual data; it avoids coder fatigue, inter-coder reliability issues and other human errors associated with traditional content analysis.

2. The adoption of individual words as the unit of analysis fits this dissertation perfectly, thus establishing validity for the analysis (more on this below, in the “Population and Sample” section).

3. Computerized content analysis enables the researcher to process volumes of data instantaneously, thus expanding the potential dataset beyond the logistical limits of traditional content analysis.

3.3.B: QUANTITATIVE DATASET & SAMPLING This dissertation seeks to study, through computerized content analysis, online news postings and social media feeds from American community newspapers and a wide range of daily newspapers. Establishing a nationwide random sample of community and daily newspapers is the subject of this portion of Chapter 3.

Of principle interest to this dissertation is the role that circulation size plays in the crafting of online news and social media content. Therefore, non-random criteria for circulation size, online news presence and social media presence were constructed. Initially, to ensure identical representation of circulation categories, a framework of 120 American newspapers was designed; 40 community newspapers, 40 community daily newspapers, and 40 large daily newspapers. Each circulation size was granted equal representation to ensure a balanced dataset; the balanced dataset, in turn, allowed equitable comparison between circulation categories. Community newspapers were considered weekly publications with fewer than 50,000 regular circulation; community daily newspapers were considered daily publications with fewer than 50,000 regular circulation; and, large daily newspapers were considered daily publications with more than 50,000 regular circulation. These categories are consistent with

published work on community newspapers, as well as standard industry categorization (Editor and Publisher's International Yearbook: The Encyclopedia of the Newspaper Industry, 2008; Lauterer, 2006).

Upon further consideration, categorization for national daily newspapers was introduced; these publications were considered daily newspapers with greater than 500,000 regular circulation. Although there are a handful of such newspapers in America, a number of studies have revealed the authoritative influence newspapers such as The New York Times and USA Today wield on the spectrum of American and international media (Benoit, Hemmer, & Stein, 2010; Kitch, 2007; Lee, Chin-Chuan, & Nina Luzhou, 2011; S. Lewis & Reese, 2009; McCombs, 2004; Porpora, Nikolaev, & Hagemann, 2010). As such, an additional category of five national daily newspapers brought the total number of newspapers in the quantitative dataset to 125. While comparing a group of five newspapers with three groups of 40 is potentially imbalanced, it is nonetheless proportionately reflective of the state of the American news industry – there are simply not 40 newspapers with more than half a million regular circulation, and it is better to acknowledge a discrepancy in the size of the dataset than to inflate the number of national newspapers to a comparable and artificial figure.

The next step involved ensuring geographic diversity. Although geography is not a measured variable in this dissertation, the dataset must be geographically generalizable to ensure that geographic biases do not call the quantitative results into question. Following in the footsteps of Reader (2006), who divided the United States into 14 geographic categories to derive a qualitative dataset of 28 newspapers, this study partitioned the country into eight discreet regions based on common cultural, economic, and socio-political characteristics, as well as geographic continuity.4 Eight groups were chosen so all but one group (southwest) would have at least five states, allowing equal representation among each state in each circulation category. Put another way, this ensured that no two community weekly, community daily, large daily, or national daily newspapers would be from the same state. Although this was infeasible for the southwest region, which has only four states, it provided an effective model for the remaining seven groups.

Each region was allowed 15 publications – five community newspapers, five community daily newspapers, and five large daily newspapers. Randomly selecting data is an important facet of ensuring

4 See Appendix 1 for geographic categories.

data are generalizable and not biased in any particular direction (Krippendorff, 2004; Nuendorf, 2002; Poindexter & McCombs, 2000; Weber, 1990); however, an entirely random dataset could favor particular traits, so non-random geographic controls were installed to ensure generalizability.

The random selection process utilized a set of multi-sided dice (four-sided, eight-sided and 20- sided, primarily) to determine random starting points. When possible, random selections were distributed across states in each region; put another way, each circulation category represented random newspaper selections from as many individual states as possible. The Southeast Region, for example, contains eight states and each community newspaper studied was randomly selected from a different state; the Southwest Region has only four states, however, leading to a small degree of overlap. States in each region were numbered and one was selected at random; then, listings of newspapers were located for each state primarily through current newspaper and press association websites. Dice rolls determined random selections of newspapers from those websites; first, the community newspapers were selected in this fashion, then the small dailies and large dailies.

Once newspapers were chosen at random, publication frequency and circulation size was confirmed, largely through the Ulrich Periodic Index; there is precedent for using this index in community newspaper research as well (Elliott & Greer, 2010). Circulation size was based on regular circulation, not Sunday circulation, because this dissertation focuses on routine news production rather than Sunday- specific production.

This random sample accounted for 120 American newspapers stratified by circulation size and regional geography; this sampling was based largely on a previous study of community newspapers and Benedict Anderson‟s theory of “imagined communities” (Anderson, 2006; Funk, 2012). The Funk (2012) study used identical regional categories, a 120-newspaper sample, and served as the basis for the random dataset; the researcher updated and confirmed the circulation size and categorization in October 2012. As the original study did not consider social media, the researcher ensured that each publication had Facebook and Twitter feeds updated within the last week; that test also was conducted in October 2012. Four newspapers from the Funk (2012) study, or .03 percent, lacked current social media pages and were replaced using the same random selection criteria; the lion‟s share of the dataset remained consistent, however.

Furthermore, the previous study (Funk, 2012) also did not consider national daily newspapers. There are only a small handful of daily newspapers with greater than half a million regular circulation; 60

while randomly sampling such a small field may be problematic, there are enough national newspapers to ensure geographic generalizability. As such, one national newspaper was chosen from five different regional categories; random selection was employed in regions, like the Northern Region, where more than one national newspaper is headquartered.5

This framework provides ample foundation for the valid study of community and daily newspaper online news and social media feeds. A sample of 125 publications, randomly stratified by geography and divided into comparable circulation categories, can yield a very high number of data points for computerized content analysis. The data collection process, which includes the format, schedule, and overall number of data points, is explicated in the next section.

The length of the articles themselves is irrelevant to this study. DICTION 6.0, the computerized content analysis software which this dissertation utilizes, controls for article length; its algorithms break documents into 500-word sections and then average results from each section together. Without this control, it would be impossible to compare articles of unequal length, as the unequal number of total words would invalidate analysis; however, with this control in place, relevant comparison could be achieved so long as each individual newspaper had an equal number of articles in the dataset.

3.3.C: QUANTITATIVE DATA COLLECTION The quantitative dataset for this dissertation was outlined in Chapter 3.3.a; it consists of 125 print newspapers stratified by circulation size and regional categorization, and selected at random within those categories. For the online news data, the data collection process utilized a constructed-week format; the days of those constructed weeks were selected randomly, and several such weeks were employed to ensure quantitative generalizability. For social media data, broad one-time data collection provided a great deal of data and did not risk redundant data collection, and was thus deemed sufficient.

The study of community newspapers complicated the constructed-week format, however. Research indicates that community newspapers are not fully utilizing the Internet‟s potential for instantaneous publication; instead, they remain primarily focused on the print product, and often under- utilize their online platforms (Elliott & Greer, 2010; Garfrerick, 2010; Gilligan, 2011; Hansen, 2007).

5 See Appendix 2 for full list of newspapers, circulation sizes, and regional categorizations.

There was a risk that community newspapers would update their websites potentially infrequently, or potentially only once a week; were two randomly chosen days to fall within the same calendar week, then, individual news articles could potentially be analyzed twice, providing redundant and invalid analysis. This problem is easily remedied, however, by drawing a constructed week of seven days from more than seven calendar weeks; that way, the random sample for community newspapers consists of seven discreet data points from seven different weeks, thus minimizing the chances of redundant data. Such a precaution also ensures a broader dataset for daily circulation categories.

Although a constructed month would offer more data than a constructed week, it would require a data collection over a 31-week period to ensure that every constructed data point was unique. Furthermore there is no set minimum of data to constitute a healthy quantitative analysis (Krippendorff, 2004; Nuendorf, 2002) and the constructed week format offers more than enough data for thorough computerized content analysis.

Additionally, from a logistical perspective, downloading online news content, Facebook content and Twitter content from 125 newspapers in a single day would be a daunting challenge. Therefore, the dataset of 125 newspapers was randomly sorted into five groups of 25 newspapers each. Random assignments were made regardless of circulation size; publications were numbered 1 through 125 based on region and circulation size, and a 20-sided dice was rolled to determine a random starting point. As 20 constituted roughly a sixth of the full dataset, this number was deemed a sufficient method of determining a starting point. Once a random starting point of three was determined, that newspaper was sorted into Group A, the next newspaper into Group B, and so on until Group E; the eighth newspaper was then sorted again into Group A. This method generated five random groups of 25 newspapers each.6

Once the random groups were established, five constructed weeks were designed over a nine- week period. Constructing the weeks utilized a 10-sided dice; each of the nine weeks corresponded with a side on the dice, which was then assigned to the day in question, while the tenth side elicited a re-roll. The Sunday assignment for Group A, for example, was determined by rolling the dice to see which of the Sundays (one through nine) would be assigned to Group A; then, the Monday for Group A was determined, and so on. Once a group drew a day in a particular week, that week was eliminated from the

6 See Appendix 3 for newspaper groups.

random selection process; in the event that subsequent rolls determined the same week, the next consecutive available week was assigned.

Care also was taken to ensure that no quantitative data collection days fell during qualitative data collection at the Texas Press Association‟s annual convention on Jan. 17, 18 and 19, 2013; the prospect of collecting qualitative and quantitative data in the same day simply was too daunting. As such, an initial randomly generated eight-week draft was scrapped and a new, randomly generated nine-week plan was constructed. During the data collection process, however, the data collection date for Group E on Sunday, Feb. 3 was accidentally neglected; furthermore, technical difficulties impeded the collection of two social media data points. Rather than construct an entirely new dataset, Group E‟s Sunday collection date was rescheduled for March 17, Group A‟s social media collection was rescheduled for March 18, and Group E‟s social media collection was rescheduled for March 22, leaving a final 10-week collection schedule.

This process ultimately generated five randomly constructed weeks for each of the five groups of 25 newspapers. Data were collected between Jan. 6, 2013, and March 17, 2013.7 Once these weeks were established, the three most prominent news articles on each news website were copied and pasted into Word documents during the afternoon of each day in each constructed week. The full articles, headlines and subheadlines were copied and pasted into Microsoft Word; bylines, authorship, and contact information was omitted to ensure the word-by-word analysis focused solely on the news content. The Word documents were organized by group and circulation size category, and also included the day of the week and date in the document title (ie, “A.CW.Monday.1.1” for a hypothetical data collected from community weekly newspapers in Group A on Monday, Jan. 1). Those Word documents then were saved and stored pending computerized content analysis. There were a total of 375 articles downloaded each day per group and a grand total of 2,625 articles from newspaper websites.

The constructed week format proved problematic for the study of social media content, however. Among community newspapers, there was simply no consistent updating schedule; some newspapers published Facebook and Twitter content several times a day, while others updated less frequently or even sporadically. Even a constructed week format ran the risk of redundantly coding the same data several

7 See Appendix 4 for constructed week calendars and Appendix 5 for a calendar graphic.

times, which presents a threat to the dissertation‟s validity. However, Facebook and Twitter pages store a great deal of previous posts that can serve as a rich data source.

To avoid the risk of invalid redundant coding, one-time data collection on a random day within the same 10-week period instead drew the most recent 30 social media posts from each newspaper. The frequency of Facebook and Twitter updates fluctuated based on the individual newspaper; however, collecting the most recent 30 posts ensured that the use of social media would be adequately measured without any redundant data. Each newspaper group had one day for social media data collection from both Facebook and Twitter.8 These posts were copied into Word documents (ie, A.CW.Facebook.Monday.1.1 and A.CW.Twitter.Monday.1.1) which were then saved and stored pending computerized content analysis.

3.3.D: QUANTITATIVE DATA PROCESSING Once data for computerized content analysis were copied and pasted into Word documents, the data were split into two groups. The Word documents containing data from social media posts were grouped together based on circulation size (for operationalizations of these categories, see Chapters 1.4 and 3.2). The Word documents containing data from newspaper websites, which had been stored according to group and day, were collapsed into four similar groups based on circulation size. Then, individually, each document was copied from Microsoft Word and pasted into Text Edit (a much simpler word processing program which stripped all hyperlinks from the articles) and then saved as .txt documents retaining the original save names. Hyperlinks cannot be analyzed by DICTION 6.0 software.

Data then were processed using the software DICTION 6.0, a computerized content analysis program pioneered for use in rhetorical communication studies. The software uses customizable word dictionaries to determine the presence and absence of particular words, and thus concepts, in a text; DICTION has been used abundantly in social science research (Abdelrehim, Maltby, & Toms, 2011; Crew & Lewis, 2011; Don, 2011; Gorton & Diels, 2010; Jarvis, 2004) and is consistent with the guidelines for computerized content analysis outlined in methodological texts (Krippendorff, 2004; Nuendorf, 2002).

8 See Appendix 6 for social media data collection calendar.

Using DICTION 6.0, custom word dictionaries were designed to reflect each of the seven news factors employed by this dissertation. The word dictionaries were designed to be exhaustive and exclusive, criteria considered paramount when constructing content analysis codebooks (Poindexter & McCombs, 2000). They were designed to include any word that could qualify for each of the seven news factors, as well as any potential conjugation of each word (i.e., conflict, conflicting, conflicted). Different forms of the same word, such as violence and violent, were also included.

Ultimately, four of the news factors studied here used comprehensive collections of words. Word dictionaries for prominence, conflict, oddity, and impact quickly became vast, and were designed to be inclusive; constructing the remaining three dictionaries was somewhat simpler. Proximity was split into two sections, general proximity and specific proximity. General proximity consisted of terms like “local” and “area.” Specific proximity was derived primarily through the website FreeMapTools.com, which constructed a transparent radius around each listed community in Google Maps; the map was then zoomed to a two-mile scale and scanned for any cities, towns, villages, or labeled neighborhoods located wholly or partially within the radius. For rural communities this was a relatively straightforward process, although it was admittedly more complicated for larger metropolitan cities such as New York or Chicago. The dictionary also included the name of the home community‟s county or counties, or in the case of Louisiana, parish or parishes. Each category of newspapers had its own specific proximity dictionary; so community weekly newspapers in the A group, or A.CW, had one dictionary, as did A.CD, A.LD, A.ND, B.CW, C.CW, and so on.

Similarly, timeliness was sub-divided into “recentness” and “dates.” Recentness contained words stating timeliness, such as “recent” or “current,” while dates was customized for each individual set of articles to include the date of publication as well as the two dates before and after (i.e., within the word dictionary itself, for hypothetical dataset A.CW.Monday.1.1, the dates dictionary would include “December = 30, December = 31, January = 1, January = 2, January = 3). Such programming excluded all dates not equal to the five dates of interest.

3.3.E: ABANDONING MAGNITUDE One of Bridges and Bridges‟ (J. Bridges, 1989; J. Bridges & L. Bridges, 1997) original 7 news factors was Magnitude, which was operationalized as an expression of quantitative measurement. For example, news articles about a two percent tax increase or a seventeen point victory would both qualify for Magnitude. This dissertation attempted to transfer all of Bridges and Bridges‟ news factors into the 65

current analysis; due to complications with computerized content analysis, however, this news factor was ultimately abandoned.

The dates dictionary, used here as an egalitarian factor, also provided a step in attempting to calculate Magnitude. DICTION includes a calculation for numerical figures which includes numerical digits (i.e., $14 or 14 people) as well as written expressions of quantity (i.e., fourteen dollars or fourteen people). Efforts to create a customized numerical dictionary for magnitude were unsuccessful; the program was simply unable to differentiate between numerical terms. It could not tell the difference between 2 p.m., two o‟clock, or a two percent tax increase. It also could not tell the difference between 5,125,555,555 people and the phone number (512) 555-5555; as stand-alone figures, the numerical data is too similar for a computer to distinguish. The measurement was deemed unreliable.

The next attempt hypothesized that the Dates dictionary count could be subtracted from DICTION‟s automatic calculation of numerical quantity. The goal was to remove dates from the quantitative measure; for example, to help DICTION distinguish between January 2 and 2%. This provided a rough measure of quantitative statements in articles that avoided particular days, but it unfortunately did not account for all numerical figures potentially beyond magnitude‟s intent; times and phone numbers, for example, could not be included. This also accidentally overlooked the inclusion of date words, such as “Monday” or “Tuesday,” in the same dictionary as actual numerical figures such as “January = 14.” This omission was noticed in mid-analysis, and although the error was not prohibitive and the analysis could have been repeated, it was ultimately a moot point.

Scores for numerical figures were, in some cases, nearly more than scores for all other factors combined; in every case, numerical figures were more than twice as common as the next most prevalent factor. Comparative analysis between magnitude and other factors was thus impossible, and given the remarkably high scores for magnitude, the researcher could not consider the data valid. The DICTION software was simply too sensitive to distinguish between different kinds of numbers. Unfortunately, it was abandoned as a factor.

The loss is regrettable, but not a major setback. The dictionary for impact worked flawlessly and is arguably a better indicator of social significance anyway, and the difficulty in establishing magnitude contributes to the scholastic understanding of computerized content analysis software, which will be discussed in Chapter 4.

3.3.F: CONSTRUCTING AND IMPORTING WORD DICTIONARIES

A full list of terms used in each word dictionary is available in Appendix 59; however, a sample is provided here.

Deviant News Factors:

Prominence: Mayor, governor, president, senator, executive, elite, CEO, COO, reigning, actor, musician, singer, celebrity, athlete, professional.

Conflict: War, conflict, clash, spat, difference, disagreement, disparity, confrontation, violence, violent.

Oddity: Odd, strange, bizarre, unusual, uncanny, unnerving, rare, extraordinary.

Social Significance Factors:

Impact: Impact, change, difference, meaningful, major, important, crucial, critical, altering, changing, ending, beginning, genesis, cataclysm.

Egalitarian News Factors:

Timeliness:

Recentness: Yesterday, today, tomorrow, soon, lately, (Not: next, last.)

Dates: The specific dates for the date of data collection, two days previous and two days following, as well as the corresponding days of the week, for each set of articles.

Proximity:

General Proximity: Local, area, nearby.

Specific Proximity: The name of every city, town, village, or neighborhood within 20 miles of each newspapers‟ home community, as well as the community‟s county or counties, or parish or parishes.

The construction of these dictionaries was conducted in Microsoft Excel. This enabled easy comparison and construction of dictionaries, and also allowed an expedient search for duplicate entries,

9 See Appendix 9 for full word dictionaries.

both in each individual dictionary and across dictionaries. It was important to ensure that, for every dictionary but those designed for specific proximity, every word be included in only one dictionary; for specific proximity, it was acceptable if (hypothetically) Springfield, Massachusetts and Springfield, Oregon were in two different specific proximity dictionaries as only one specific proximity dictionary would be enabled at a time. However, it was important that “Springfield” appear only once in each dictionary.

Importing the dictionaries from Excel to DICTION proved a bit tricky. DICTION requires a comma between each entry in its word dictionaries, while imports from Excel are treated as either charts or given paragraph breaks between entries, depending on the program. Neither were acceptable for DICTION, and the practice of hitting “delete, comma, space” for every entry in every word dictionary proved a nightmare.

The solution was impressive. Word dictionaries were copied from Excel, which are charts by default, into Text Edit; the overly simplistic program stripped the chart programming and considered the dictionaries a list single-word paragraphs. That list was then copied from Text Edit into Microsoft Word (which would have considered the Excel entries an un-editable chart), and a -find and -replace search was initiated which replaced every paragraph mark (¶) with a comma and space ( , ). The text then was studied to ensure there was only one comma and space between words, as superfluous commas or spaces would cause DICTION to reject the whole dictionary. After this was completed for each of the 26 dictionaries (prominence, conflict, oddity, impact, recentness, general proximity, and 20 dictionaries for specific proximity) they were individually imported into DICTION‟s custom dictionary feature. Finally, the dates dictionary was adjusted manually prior to each individual analysis.

3.3.G: QUANTITATIVE DATA UNITS Once the word dictionaries were constructed, the analysis began. The DICTION 6.0 data processing procedure is not complicated, but the units of data it produces are somewhat abstract and not very intuitive. An overview of the data units themselves, and a step-by-step review of the analysis process, is appropriate here.

The organic unit of analysis in DICTION 6.0 is individual words. Once a Word document is loaded into DICTION 6.0, the program runs a frequency analysis of every individual term in its default and custom word dictionaries and provides a count for each word dictionary. (See Chapter 3.3.f for more

on word dictionaries). The program can produce simple, raw word counts – a straight report of every word in every dictionary used in a document. In some instances or case studies, this could be effective or even ideal. However, it does not account for the length of a document, thus making comparative analyses very difficult. If DICTION 6.0 were to hypothetically compare campaign speeches from a national political convention and a whistle-stop-style campaign event, the results would be dramatically different and highly invalid. The convention speech would likely be much longer than a quick campaign pitch, thus jilting any potential quantitative comparison between them. Such a comparison could accurately compare “words” as a unit of analysis, but the comparison would be invalid. DICTION considers this type of computation a “raw score.”

Given the volume of news articles, Facebook posts and Twitter posts studied in this dissertation, a raw score would produce highly invalid data for comparison. DICTION 6.0 provides another option for analyses of texts with different lengths: An “average” or calibrated score which breaks a text into 500- word segments and runs frequency analyses for each word dictionary within each 500-word segment independently. Then, it averages the frequency scores for each 500-word segment to produce a composite frequency analysis for the entire text. If the final segment is not exactly 500-words, DICTION 6.0 computes the results and runs ratio computations to determine what the scores would be if the segment were exactly 500 words.

Put another way: If only raw word counts are considered, a study of a 1,000-word document and a 50,000-word document would be highly invalid. The average or calibrated score option would instead break the 1,000-word text into two 500-word sections, run frequency analyses of words included in the word dictionaries, and then average the results of those two analyses to produce composite word frequency scores representing the 1,000-word text document. It would follow the same procedure with the 50,000-word document, breaking it into 100 segments of 500 words each, computing frequency scores for each individual segment, and then averaging all 100 individual-segment results together. By segmenting the documents and averaging the results of the segments, calibrated scores can be compared between documents of different sizes.

To be absolutely certain of accuracy, the researcher confirmed this description of the process and terminology used in a number of publications employing DICTION. The researcher also consulted Dr. Rod Hart, one of the co-designers of the DICTION 6.0 software, to ensure accuracy.

“DICTION produces two kinds of scores: raw scores and calibrated scores. Calibrated scores, which are the default, are raw scores that have been corrected against the program‟s normative (500-word passage) standard. This „correction standard‟ lets all users of the program compare their results to one another, something that is not possible when users employ the raw score option.” (Hart, 2013)

The best terms for the units of data produced by this process, ultimately, are either “score” or “mean.” Originally, the units of data are frequency analyses of individual words; however, because a series of averages must be run to make comparison valid, the term “words” is not truly accurate. “Score” seems abstract and imprecise, although it is technically accurate. Thus, this dissertation uses the term “mean” to describe the units of quantitative data derived through DICTION 6.0 and quantitative analysis.

Because the final data points reflect a series of averages of word counts, or means, the units of data can be expressed as means. This is consistent with past research using DICTION (Ballotti & Kaid, 2000; Conway, 2006; Jarvis, 2004; Murphy, 2013).

DICTION 6.0, therefore, essentially studies frequencies of grouped words (word dictionaries) and processes text documents in 500-word segments to produce comparable data (means of word frequencies). Once DICTION 6.0 produces means, those means can in turn be compared and contrasted; ANOVA and correlation analyses also become possible.

3.3.H: DICTION 6.0 ANALYSIS With the dictionaries constructed and the data analysis process understood and articulated, the analysis was run. DICTION computed means for each word dictionary for each of the 180 .txt documents. (See Chapter 3.3.g for a description of the data units, or scores.) Each of the 180 quantitative means were then imported into Microsoft Excel individually based on circulation size – means for all the CWs were imported into the same Excel page, the CDs into the same page, and so on. Facebook and Twitter results also were imported into Excel and sub-categorized by circulation size and media type. Means then were averaged together based on circulation size and media type. Means for “recentness” and “dates” were added together within each category to form a “timeliness” composite; similarly, “general proximity” and “specific proximity” were added together to form a “proximity” composite. Results then were grouped based upon deviance (prominence, conflict, and oddity), social significance (impact), and egalitarianism (timeliness and proximity). This produced the bulk of the charts F1, F3, and F5 in Chapter 4.2.

Once the means were calculated, data were copied into a new page and an ordinal variable called “Circulation Size” replaced labels of CW, CD, LD, and ND; CW data points were given an ordinal 1 value, CD data points a 2, LD a 3, and ND a 4. The process was repeated for Facebook and Twitter data; the three collections of data were carefully kept separate, but the process for each was identical. This data page then was imported into SPSS for raw comparison, ANOVA analyses, and correlation analyses.

Raw comparisons of the means for each news factor for each circulation category were used to derive a general understanding of the data; the goal was to better illustrate differences within the means. ANOVA analysis was used to determine variance of means within each factor among circulation size categories. Put another way, the ANOVA analysis measured the standard deviations and differences between similarly categorized means to determine statistical differences between community weekly, community daily, large daily, and national daily newspapers‟ use of particular factors. This calculation produced a reliable statistical measure demonstrating variance across circulation categories. ANOVA analysis used the ordinal variables representing circulation size for website data, Facebook data, and Twitter data as factors to measure the factor data. The results were then added to charts F1, F3, and F5.

Pearson‟s correlations were used to determine relationships between circulation size and individual factors. The constructed ordinal variables representing circulation size were correlated with each factor for website and social media data. Doing so also produced correlation measurements between individual factors; although not the particular focus of this dissertation, these factor-by-factor correlations were retained in the interest of comprehensiveness.

It is worth mentioning that because the data analysis process utilized a multi-step and multi- platform process, the final “n” in the SPSS analysis is only partially representative of the full dataset. For example, concerning Facebook and Twitter analyses, SPSS processed what it considered 20 dense units of data – one unit for each .txt document representing thousands of social media posts from each group and circulation size of newspapers. One data unit represented DICTION 6.0 analyses of Facebook posts from community weekly newspapers in Group A, another for community daily newspapers in Group A, and so on. The number 20 misleads here as it represents only the final step in the data analysis process. The SPSS analysis remains valid given the density of the data and the proper procedures undertaken, however.

A useful comparison would be to consider the American population analyzed by state. A study may consider population statistics of each individual state, produce a unit of data per state, and use that data to consider national trends – instead of considering each individual American as an individual data 71

unit, which may not be feasible or practical. Subsequent statistical analysis would render an n = 50 which would vastly understate the overall analysis. The same can effectively be said for this dissertation. As such, the original n is reported here alongside an n0 which reflects the full dataset, or the number of news articles or social media posts collected and studied.

3.3.I: GENERALIZABLE VALIDITY A variety of methodological texts illustrate the potential of computerized content analysis (Krippendorff, 2004; Nuendorf, 2002; Poindexter & McCombs, 2000), and a range of studies have effectively utilized computerized content analysis techniques (Aust, 2004; Ballotti & Kaid, 2000; Conway, 2006; Jarvis, 2004; ÖZkoÇAk & Tuna, 2011).

However, establishing validity is absolutely paramount for computerized content analyses. Reliability is virtually guaranteed, but care must be taken to ensure the computer and researcher pursue identical concepts. This requires both human input and vigorous testing.

Several different ideas were explored to achieve a valid analysis. First, traditional concepts of inter-coder reliability were considered. Bridges and Bridges (J. Bridges, 1989; J. Bridges & L. Bridges, 1997) utilized a traditional, dichotomous coding scheme to determine the presence and absence of the news factors considered here; however, their unit of analysis was entire newspaper articles, not words, and dichotomous data provide an awkward comparison to the word frequency means expressed by DICTION. Next, the researcher considered having a team of human coders process data on a word-by- word basis similar to DICTION‟s programming; they would have access to the finished word dictionaries and could determine both the validity of the dictionaries while demonstrating the computer‟s efficiency at voluminous coding.

However, in practice, the latter part of this proposition proved excruciatingly daunting. Statistical methods of deriving inter-coder reliability, like Krippendorf‟s alpha, required that every word be granted an individual and consecutive line of data; if a coder accidentally neglected an “a” or “the” in the article, the entire data set would be ruined. Furthermore the practice of searching word dictionaries for individual terms was painfully tedious. Having several individuals check and reconsider the word dictionaries themselves, however, proved highly fruitful.

Indeed, the entire concept of “inter-coder reliability” is a moot point. As Krippendorf (2004) notes, computers cannot not be reliable, and it is largely guaranteed that computers will process large 72

volumes of data more reliably than humans. Validity, however, becomes the critical interest when working with computerized content analysis software.

The word dictionaries employed by computerized content analyses must be absolutely precise. They must contain appropriate words and endeavor not to include inappropriate words; they also can be vast, depending upon the measured concept. As such, it is best to crowd-source word dictionary design. One researcher will likely omit, ignore, or too eagerly include words which could bias the validity of the study; employing a team to ensure the word dictionaries are precise minimizes this risk.

The primary researcher designed a first draft of the word dictionaries used in this dissertation, and also crafted the operationalized definitions of those factors highlighted in Chapters 1:5 and 3:2. Four doctoral students then were employed to individually peruse, review, and consider these word dictionaries; each developed independent ideas on which words to include and exclude, as well as holistic thoughts on the whole analysis. Once these ideas were crafted, the group met as a whole. Ideas were combined, overlap was considered, and discussions occurred to resolve conflicting ideas. The word dictionaries then were submitted to the dissertation chair and committee for review and suggestions. By the time actual analysis began, a total of 6 people had considered, reviewed, and approved the word dictionaries.

As with any study, it is very difficult to guarantee validity. Unlike measurements for reliability, there are no statistical tests to ensure comprehensiveness and appropriateness of measurement and validity. However, employing a diverse team to consider the validity of the word dictionaries vastly strengthens the analysis. It ensures that computerized content analysis‟ biggest weakness, potential validity issues, is adjusted for and effectively neutralized.

3.4: Qualitative Structured Interviews

Chapter 3.3 outlined the research paradigm, dataset and sampling procedure, data collection process, and data processing schemes associated with the computerized quantitative content analyses employed in this dissertation. Chapter 3.4 will employ the same structure for qualitative structured interviews.

3.4.A: QUALITATIVE RESEARCH PARADIGM Even the most detailed content analysis can only explore content which is deliberately extant or absent; it cannot illuminate the logic, considerations, backgrounds, or decisions made by the individuals who shaped and crafted that extant and absent content (see Chapter 1.5 for a discussion of inferencing; this dissertation assumes that inferencing, recommended by some content analysts [Krippendorff, 2004; Weber, 1990], is inappropriate).

Qualitative methodologies must consider those inherently qualitative decisions. Gatekeeping scholars historically have relied upon interviews to document and consider editorial decisions; indeed, David Manning White (1950) learned about Mr. Gates through direct conversation, and interviews and occasional surveys are the methodological hallmark of gatekeeping theory and media sociology (Bleske, 1991; Bowman, 2008; Breed, 1955; Cassidy, 2006; Gans, 1979; Reese, 1990; Shoemaker & Vos, 2009; Tuchman, 1973). Social science interviews are direct and often highly productive, and can provide “texture and context” that go beyond surveys and content analyses (Poindexter & McCombs, 2000, p271); however, interviews are not uncomplicated.

“Understanding questions is simple – isn‟t it? Maybe not. Although we immediately recognize or understand when a question has been posed in interaction, how to ask interview questions that are comprehended by others and answered in ways that generate relevant data is more complex than initially apparent.” (Roulston, 2010, p11)

“Asking questions and getting answers is a much harder task than it may seem at first. The spoken or written word has always a residue of ambiguity, no matter how carefully we word the questions and report or code the answers.” (Fontana & Frey, 1994, p361)

Absolutely objective interviews do not exist. The characteristics of the interviewer and subject, the subject under consideration, or the interviewing procedure can bias the response and qualitative data to various degrees. For example, qualitative social science interviews concerning race and racial issues can be influenced by the racial demographics of the interviewer and subject (Fontana & Frey, 2005). 74

Furthermore, the structure and phrasing of interview questions can influence outcomes considerably. Question structures can generate preferential responses, either positive or negative, and can lead an interviewee to adopt particular answers or perspectives; additionally, poorly planned interviews can formulate an interviewee‟s responses to the point of putting words, and ideas, into his or her responses directly (Roulston, 2010). The conventional approach to minimize such biases is to design questions which are as inclusive and objective as possible.

“Many methodological texts advise qualitative interviewers to ask open, rather than closed, questions because closed questions have the possibility of generating short one-word answers corresponding with yes/no or factual information implied by the question (for example, What time is it? One fifteen). Thus, closed questions are those in which the implied response is restricted in some way.” (Roulston, 2010, p11)

However, the flexibility offered by open-ended inquiry can lead to vastly divergent responses and results; too, open-ended inquiries can fluctuate based upon the demographics of the interviewer, and are often unhelpful at producing specific, narrowly tailored data. Much of these “response biases” can be eliminated through the use of structured interviews, which also offer far greater specificity and control to the interviewer.

Through intricate planning and uniform implementation, structured interviews effectively serve as a “theatrical script” (Fontana & Frey, 1994, 2005) which ensures consistency among interviews and reduces the potential for response bias. Fontana and Frey (1994, 2005) provide key guidelines for the implementation of structured interviews:

1. Never get involved in long explanations.

2. Never deviate from the script or structured interview plan.

3. Never let anyone interrupt the interview.

4. Never suggest an answer or disagree with an answer; keep personal feelings aside.

5. Never interpret the meaning of a question.

6. Never improvise.

Assuming these guidelines are followed, structured interviews provide a controlled environment and consistent interview for social science research subjects. The goal is for total neutrality – in a sense, a kind of qualitative reliability, ensuring that the interview would proceed identically regardless of the

individual interviewer. That structured neutrality prevents spontaneity, and typically negates emotional dimensions of an interview or a subject; however, because this dissertation is not concerned with emotional dimensions, this limitation is acceptable.

Therefore, this dissertation adopts qualitative structured interviews as its second research paradigm for the following two summarized reasons:

1. Structured interviews illustrate the thoughts and motivations behind editorial decisions in a clear, concise, and scripted manner.

2. By adopting an interview structure as a script of sorts, structured interviews ensure de facto qualitative validity and reliability. They largely avoid the subjectivity inherent in other qualitative methods and keep the focus squarely on the intended research questions.

3.4.B: QUALITATIVE DATASET & SAMPLING H4, the qualitative analysis section of this dissertation, utilizes structured interviews to illuminate newspaper editors‟ perceived interaction with their audiences and their usage of deviant, neutral, and egalitarian news factors. It also explores possible relationships between the two concepts.

Of specific interest are community newspaper editors. While interviews with editors of larger publications would be illustrative, they would not speak as clearly to this dissertation‟s core questions on the relationship between deviant news and audience interaction; because previous literature has documented the high degree of local interaction between community newspapers and their audiences, community journalists are uniquely situated to consider this dynamic. As such, H4 interviews them explicitly. The explicit interest in community newspaper editors is derived from gatekeeping theory, which stipulates that professional journalism routines craft institutional preferences for news about deviance without input from audiences (Shoemaker & Reese, 1996; Shoemaker & Vos, 2009); community newspapers typically have close relationships with their communities and readers (Lauterer, 2006; Reader, 2006), and thus provide ideal ground to test these theoretical assertions. (See Chapters 1 and 2 for more on gatekeeping theory and community newspapers.)

The dataset, therefore, constitutes structured interviews with community newspaper editors. Concerning the size of the dataset, past qualitative research has indicated that repetition and 76

reinforcement of themes is more important than the raw number of interviews (Coleman, 2007; McCracken, 1993).

“Saturation was achieved when three interviews in a row yielded no substantially new information. This resulted in 18 completed interviews, which is consistent with the literature on qualitative research that notes sample size is less important than repetition of themes, and that iteration generally occurs somewhere between eight and 20 interviews.” (Coleman, 2007, p30)

A number of social science interview analyses have utilized this approach. Reader (2006) studied ethical differences at “small” and “large” American newspapers by interviewing 14 newspaper editors from each category; Coleman (2007) interviewed 18 photographers and designers to consider civic journalism and photography. Momoc (2012) used 12 semi-structued interviews of Romanian journalists and academics to compare perceptions of the media‟s role in national democratization (Momoc, 2012); Shuang and Louw (2009) interviewed 30 Chinese business owners in Brisbane, Australia, to consider intercultural challenges among ethnic businesses (Shuang & Louw, 2009); and, Richardson, Huckerby and Williams (2008) interviewed 13 journalists at Corse-Matin, the major newspaper on the semi- autonomous French island of Corsica, to study the relationships between media, language, and national identity (Richardson, et al., 2008).

Approximately 20 interviews therefore would be sufficient; this dissertation sets a goal of 30 to be abundantly thorough, but between 20 and 30 interviews is consistent with published research on, or utilizing, interview methodology.

3.4.C: QUALITATIVE DATA COLLECTION The qualitative dataset for this dissertation is highly specific. Like Reader (2006), Coleman (2007) and others, this dissertation is explicitly concerned with interviews with a particular variety of journalist; this study is focused on American community newspaper editors. As such, qualitative data collection took advantage of a natural collection of such editors to maximize efficient data collection.

The Texas Press Association (TPA) is a trade organization and occasional lobbying group for 478 paid-circulation newspapers in the Lone Star State; 75 are daily newspapers, while 403 are non-dailies. The group acts as a membership-based resource for community daily and community newspapers across Texas (Bruns, 2012). TPA also holds an annual Midwinter Conference and Trade Show; in January 2013, the show was held in Houston. The gathering offered a single venue to interview a wide variety of community newspaper editors from across Texas in relatively expedient fashion. 77

Although collecting data from a single state can be problematic, Texas offers perhaps the best candidate for such lone-state analysis; the state is vast with a great many newspapers, and a number of published studies have focused solely on Texas and Texan issues (Conway, 2006; Cope, 2011; Jensen & Uddameri, 2009; Kraeplin, 2008; S. Lewis, et al., 2009; Schweitzer & Smith, 1991; J. Stewart, 2011). The state has more than 25 million residents, 38 electoral college votes, eight of the 15 most rapidly growing large cities in the nation, three of the 10 most populous cities in the nation (Houston, Dallas, and San Antonio), and two of the five most populous metropolitan areas in the nation (Houston and Dallas/Fort Worth) ("U.S. Census Bureau," 2012) (See Chapter 1.6 for Delimitations and Limitations of this study).

The researcher designed a structured interview, which bears some resemblance to a “theatrical script” (Fontana & Frey, 1994, 2005) (See Chapter 3.4.b for more on structured interviews) prior to the TPA convention. The researcher attended the convention and intercept-interviewed 28 community weekly and daily newspaper editors. The structured interview “script” and procedure were submitted for IRB approval at the University of Texas at Austin. After review, the IRB determined the study did not meet criteria for human subjects research as it focuses on products and policies rather than people and their thoughts; subsequently, the study was granted an IRB exemption.10

The design for the structured interview‟s introduction was based on several of Stewart and Cash‟s (2000) recommended criteria for opening qualitative interviews: stating the purpose of the interview, requesting a specific amount of time, and referring to the organization the interviewer represents (C. Stewart & Cash, 2000). The introduction was designed to be direct and informative while framing and eliciting interest in the interview; it also was intended to be professional and relatively brief.11

“The few seconds or minutes you spend in the opening are often the most important portion of the interview. What you do and say, or fail to do and say – as either the interviewer or interviewee – influences how the other party perceives you and the situation. … These seconds often determine whether the interview will continue at all.” (Stewart & Cash, 2000, p57)

The structured interview focused on two core concepts: editors‟ perceived interaction with their audience and their perceived use of deviant, neutral, and egalitarian news factors (see Chapter 1 and

10 See Appendix 10 for IRB exemption notice.

11 See Appendix 8 for the full structured interview script.

Chapter 3.3.f.). The structured interview was designed to be concise and direct while promising anonymity and confidentiality. It utilized 12 short questions; five of which utilized one-to-10 Likert scales, four questions were open-ended quantitative measures, one question was open-ended and qualitative, and one question determined if editors represented a community weekly, daily, or (in uncommon instances) bi-weekly newspaper. A twelfth question was a request for a business card as a way to ensure individual editors and newspapers were not interviewed twice; providing a business card was optional, and not required.

The first question focused on circulation size. The ensuing four questions focused on perceived interaction between editors and their audience; three used Likert scales, while a fourth (Q2) was an open- ended quantitative measure. These questions were designed to measure gatekeeping theory‟s assertion that journalists, and by extension news production routines, are created largely in isolation from audience interaction (Shoemaker & Reese, 1996; Shoemaker & Vos, 2009); (See Chapter 1).

From a quantitative perspective, such a structure is potentially invalid; one editor‟s score of seven may be the equal to another editor‟s five or eight. However, from a qualitative perspective, it is an editor‟s perceptions that are of critical interest, not necessarily objective comparison between editors‟ responses. As a safeguard, an open-ended quantitative measurement (Q2) was added to serve as a check on purely qualitative (Q1). Both measure an editor‟s perception of audience interaction, but one utilizes a quantifiable number while the other relies on a Likert scale; both are analyzed independently in the results section.

Each interview was digitally recorded. A handout of the interview questions also was provided to each participant.

(Q1) Is your newspaper a community weekly, bi-weekly, or daily publication?

(Q2) On a scale of one to 10, with one being “almost none” and 10 being “a great deal,” how much interaction do you feel you have with your readership?

(Q3) On average, about how many times a week do readers visit your office to talk about news ideas or your news coverage? (Quantitative open ended)

(Q4) On a scale of one to 10, with one being “almost none” and 10 being “a great deal,” how much interaction do you feel you have with your online readership?

(Q5) What about your social media audience? On a scale of one to 10, with one being “almost none” and 10 being “a great deal,” how much interaction do you have with them? 79

The next three questions were open-ended quantitative measures about experience and staff size. Each was ultimately used to measure the relationships between professional news production routines and community interaction.

(Q6) How many years have you worked as a community newspaper editor? (Quantitative open ended)

(Q7) How large is your paid editorial staff? (Quantitative open ended)

(Q8) Do you have any unpaid or freelance reporters? If so, how many? (Quantitative open ended)

The next three questions focused on news factors, and gatekeeping theory‟s assertion that journalists are attracted to news about deviant events, ideas and individuals (Shoemaker & Reese, 1996; Shoemaker & Vos, 2009) (See Chapter 1).

(Q9) On a scale of one to 10, with one being “almost none” and 10 being “a great deal,” about how much of your news coverage is devoted to regular people and routine events?

(Q10) On a scale of one to 10, with one being “almost none” and 10 being “a great deal,” about how much of your news coverage is devoted to extraordinary events or important or distinguished individuals?

(Q11) In general, what are your thoughts on community newspaper coverage of regular and extraordinary events? Where should the priority be? (Qualitative open ended)

The final question concerned the business card request. This structured interview script provided qualitative detail to two of gatekeeping theory‟s primary assertions while speaking directly to the heart of this dissertation.

(Q12) Would it be okay if I got your business card? Your response will be completely confidential, but I‟d like to make sure I don‟t accidentally interview two people from the same newspaper.

This seemed a reasonable precaution when planning the interview; however, in practice, Q12 was largely a moot point. The majority of editors and publishers at the Texas Press Association convention were either among a very few staff members from their publications to attend the convention, or in many cases, they themselves represented the majority or sole editorial staff at their publication.

3.4.D: STRUCTURED INTERVIEW AMENDMENTS Although qualitative scholars strongly discourage impromptu changes to structured interview scripts (Fontana & Frey, 2005; McCracken, 1993; Roulston, 2010; C. Stewart & Cash, 2000), a need for two minor adjustments quickly became apparent during data collection.

The first concerned newspaper publishers. The original interview script neglected to include community newspaper publishers and instead focused exclusively on editors. While this may be a logical distinction for larger newspapers, where publishers are closer to the business staff than editorial writers, the reality is that community newspaper publishers have a very tangible understanding of their news content. Many work very closely with editors to craft news content on a weekly and daily basis; it was clear that many, too, were obviously knowledgeable enough of their newspaper‟s editorial content to answer interview questions. As such, “editor and publisher” was substituted in practice for every instance of “editor” in the script.

Secondly, and more minor, the researcher neglected to provide an opening question to ensure that participants were editors or publishers rather than advertisers. One friendly advertising manager was interviewed by mistake and the error did not become apparent until Q6; the researcher apologized for the error, erased the errant recording, and began ensuring that participants were either community newspaper editors or publishers before beginning an interview.

3.4.E: QUALITATIVE DATA PROCESSING The data analysis process for the qualitative structured interview dataset was relatively simple. First, Q8 and available business cards were checked to ensure interviews were collected from different editors and newspapers; once the researcher was confident that redundancy was avoided, the business cards were removed and destroyed and qualitative analysis began.

Five of the eight interview questions utilized a 10-point Likert scale; responses to those questions were entered into a Microsoft Excel document which then tabulated mean and median responses. Although statistically moot, given the qualitative nature of the data and the small quantitative data size, such means and medians still provided qualitative value. Similarly, the answers to the two open-ended quantitative questions were entered as a numerical value into Excel and a mean and median were computed. Responses to the one qualitative open-ended question were input into Microsoft Word and considered qualitatively. 81

3.5: Time Line

This dissertation was undertaken in the fall of 2012 and spring of 2013. The majority of the actual analysis was conducted in 2013; however, in the interest of being thorough, the entirety of the process will be considered in this section.

Comprehensive examinations were administered in August and September, 2012; the researcher entered Ph.D. candidacy in mid-September. Work on the dissertation proposal began shortly thereafter. Originally, the researcher intended to offer the full Introduction, Literature Review, and Methodology chapters as a dissertation proposal, effectively offering 60 percent of the final document as a proposal; however, because of time constraints, the proposal ultimately was limited to Chapter 1: Introduction and Chapter 3: Methodology. A draft of Chapter 1 was submitted in early October, and following revision, drafts of Chapters 1 and 3 were submitted in November. The full dissertation committee met and approved the proposal on Dec. 12, 2012.

Once the proposal was approved, the qualitative structured interview was submitted for IRB approval. IRB exemption was granted on Jan. 9, 2013.

Qualitative interviews took place during the Texas Press Association‟s 2013 Midwinter Convention and Trade Show on Jan. 17-19 in Houston, Texas. Sufficient interviews of community newspaper editors, 28, were collected during these three days.

Quantitative data collection comprised five constructed weeks over a 10-week period (See Chapter 3.3.a) in January, February, and March of 2013. Computerized content analysis was conducted once the dataset was fully collected; it was completed in March. The remaining chapters were subsequently written, edited, and approved. The full dissertation was defended on Oct. 3, 2013.

CHAPTER 4: RESULTS

4.0: Introduction

Chapter 4 exclusively examines quantitative and qualitative data produced by this dissertation. It serves as a counterpart to Chapter 3, on methodology, and presents the results of executed computerized content analyses and qualitative structured interviews substantiated and outlined in Chapter 2. Theoretical and practical meanings and implications of these findings will be explored in Chapter 5: General and Theoretical Discussion and Chapter 6: Conclusions & Opportunities for Future Research; Chapter 4 simply presents the raw data.

Quantitative results will be explored in Chapter 4.2, qualitative results will be explored in Chapter 4.3, and the accuracy of the hypotheses will be explored in Chapter 4.4. Commonalities and discrepancies between quantitative and qualitative results will be explored in Chapter 5.

4.1: Repetition of Hypotheses

It is useful here to repeat this dissertation‟s hypotheses. Although these have been described in Chapters 1.4 and 3.1, it is worth reconsidering them prior to presenting quantitative and qualitative results.

H3: Regardless of circulation size, American community and daily newspapers will publish social media content with higher rates of egalitarian news than deviant news factors.

H4: The more contact a community newspaper editor has with their community and readers, the more preference will be given to egalitarian news over deviant news.

4.2: Quantitative Results

Chapter 4.2 considers quantitative results derived from a DICTION 6.0 analysis of news content posted on 125 newspapers‟ websites and social media feeds. Analysis explored the presence of six news factors as indicators of editorial focus on deviance, social significance, and egalitarianism among community weekly newspapers, community daily newspapers, large daily newspapers, and national daily newspapers.

Scores in DICTION 6.0 were computed, essentially, by sophisticated word counting and processing. Individual statistics represent word counts that have been averaged multiple times, first by DICTION 6.0 to account for disparate document sizes (see Chapter 3.3.g for a description of DICTION‟s units of analysis and processes) and secondly by the researcher when collapsing the data into the four circulation categories operationalized in this dissertation (see Chapter 3.3.h for data processing procedures). Raw comparisons, ANOVA analyses, and Pearson‟s correlations were then conducted.

4.2.A: DEVIANCE AND EGALITARIANISM ON TRADITIONAL NEWS WEBSITES Straight comparisons of means and ANOVA analysis indicated that American newspapers of all circulation sizes demonstrated a clear preference for deviant news factors over social significance or egalitarian factors.

A straight comparison of means indicated that the overall average for deviant factors (prominence, conflict, and oddity) across all circulation categories was roughly 200 percent higher than means for social significance (impact) and egalitarian factors (timeliness and proximity). Even weekly newspapers maintained a clear preference for deviant news factors.

Furthermore, ANOVA analysis found only one statistical difference concerning circulation size and the publication of news factors. Analysis indicated highly significant variance across circulation categories concerning specific proximity (p < .001**, df = 3, n = 140, n0 = 2,625)12; smaller newspapers

12 N refers to the number of data points processed by SPSS. This number is not representative of the full dataset, however, given the multi-step process of data collection and analysis employed here (See Chapter 3.3.h for information on DICTION 6.0’s data units and processing). As such, n0 refers to the number of original data units analyzed by DICTION 6.0. These are the number of articles or social media posts downloaded and studied. 85

were significantly more likely to publish specific proximity factors than larger newspapers. However, analyses for the remaining factors and sub-factors yielded non-significant results, demonstrating a very high degree of homogeneity across circulation categories.

Put another way: Newspapers of all circulation sizes had a clear focus on deviance and published statistically consistent rates of news on prominence, conflict, oddity, impact, recentness, dates, and general proximity. Circulation size played no significant role in the publication of deviance or egalitarianism (see Figure 1 for means and ANOVA analysis).

Pearson‟s correlations were also used to determine potential relationships between circulation size and use of deviant, social significance, and egalitarian news factors. Analysis indicated only one pertinent significant correlation, between circulation size and specific proximity (r = -.342**, p < .001, n = 140, n0 = 2,625); the negative orientation indicates that the larger the circulation size, the lower the frequency of specific proximity. This is consistent with ANOVA analysis of the same data.

The remaining factors had non-significant relationships with circulation size. Analysis also indicated a significant relationship between oddity and general proximity (r = .279**, p < .001, n = 140, n0 = 2,625); although correlations between factors themselves are not of central interest to this dissertation, and this particular relationship is not particularly relevant, the full set of correlations are reported here in the interest of comprehensiveness (see Figure 2 for correlation analyses.)

F1: Word Frequency Means of News Factors on Newspaper Websites by Circulation Category

Circulation Category Deviance Social Significance Egalitarianism Timeliness Proximity

Prominence Confl ict Oddity Impact Recentness Dates General Proximity Specific Proximity

Community Weekly 0.627 1.968 0.686 5.325 11.957 7.970 0.477 6.063 Mean 2.594 6.011 20.405 6.063 8.605

Community Daily 0.748 2.121 0.548 2.565 11.410 8.563 0.369 6.558 Mean 2.869 3.113 20.342 6.558 5.982

Large Daily 0.749 1.745 0.456 3.995 10.758 8.032 0.258 7.716 Mean 2.494 4.451 19.049 7.716 6.945

National Daily 0.584 1.610 0.783 1.585 11.511 8.511 0.463 6.646 Mean 2.195 2.367 20.484 6.646 4.562

ANOVA 0.677 1.861 0.618 3.367 11.409 8.269 0.392 6.746 Overall Mean 2.538 3.986 Sum of Squares 25.712 10.155 1.073 50.889 0.748 5.434 2.2 281.637 Mean Square 8.571 3.385 0.358 16.963 0.249 1.811 0.733 93.879 F 0.191 0.169 1.014 1.715 0.373 0.67 0.929 11.113 Sig. 0.902 0.917 0.389 0.167 0.773 0.572 0.429 .000** Total Mean 20.070 6.746 6.523

F1: ANOVA analysis of word frequency means of deviant, social significance, and egalitarian news factors on newspaper websites by circulation category. Numbers in the “Community Weekly,” “Community Daily,” “Large Daily” and “National Daily” rows reflect means for all word frequency analyses per factor per circulation category; numbers for the “ANOVA” rows reflect the ANOVA analysis. For all columns, df = 3, n = 140, n0 = 2,625, and ** indicates significance at the 0.01 level. Timeliness and proximity were formed as an averaged composite for recentness and dates, and general proximity and specific proximity, respectively. 87

F2: Pearson‟s Correlations of Means Comparing Circulation Category and News Factors on Newspaper Websites

Social Deviance Egalitarianism Significance Size General Prominence Conflict Oddity Impact Recentness Dates Specific Proximity Proximity Size - -0.034 0.028 -0.029 0.103 -0.017 -0.099 0.025 -.342** Prominence - 0.015 -0.092 0.054 0.005 -0.025 0.042 -0.032 Conflict - -0.087 -0.108 0.037 0.122 -0.046 -0.055 Oddity - 0.014 -0.086 -0.022 .279** -0.149 Impact - 0.055 -0.1 0.091 -0.004 Recentness - -0.012 0.086 0.06 Dates - -0.058 0.132 General - 0.11 Proximity Specific - Proximity

F2: Pearson‟s correlation analysis of word frequency means of deviant, social significance, and egalitarian news factors on newspaper websites by circulation category. For all instances, n = 140, n0 = 2,625, and ** indicates correlation is significant at the 0.01 level (2-tailed).

4.2.B: DEVIANCE AND EGALITARIANISM ON SOCIAL MEDIA Similarly, the majority of news content on newspaper Facebook and Twitter pages favored deviant news factors over egalitarian news factors. The ratio was not quite as stark, but a straight comparison of means indicated a clear preference: On Facebook, deviant news factors were roughly 70 percent more common than egalitarian factors, and on Twitter, deviant news factors were roughly 50 percent more common than egalitarian factors.

In only one category, community weekly newspapers on Twitter, did egalitarian news factors score higher than deviant news factors.

On Facebook, ANOVA analysis indicated Prominence (p = .019*, df = 3, n = 20, n0 = 3,750) and Dates (p = .022*, df = 3, n = 20, n0 = 3,750) had slightly significant variances across circulation categories; analysis found no other significant or slightly significant variances. Newspaper circulation size had no significant relationship with conflict, oddity, impact, recentness, general proximity and specific proximity (see Figure 3 for means and ANOVA analysis for Facebook data).

Pearson‟s correlation analysis tested the relationship between circulation size and publication of individual factors on Facebook. It indicated slightly significant inverse relationships between circulation size and dates (r = -.560*, n = 20, n0 = 3,750) and specific proximity (r = -.462*, n = 20, n0 = 3,750). The remaining factors had no significant relationship to circulation size. Significant relationships between certain factors themselves also were discovered; although not of interest to this dissertation, they are presented here in the interest of comprehensiveness (see Figure 3 for Pearson‟s correlations of Facebook data.).

Similarly, on Twitter, ANOVA analysis indicated conflict (p = .013*, df = 3, n = 20, n0 = 3,750) and specific proximity (p = .04*, df = 3, n = 20, n0 = 3,750) had slightly significant variances across circulation categories. There was no other significant variance concerning newspaper circulation size and the publication of news factors (see Figure 4 for means and ANOVA analysis of Twitter data.).

Pearson‟s correlations of Twitter data indicated a slightly significant inverse relationship between circulation size and dates (r = -. 486*, n = 20, n0 = 3,750) and highly significant relationships between circulation size and conflict (r = .680**, n = 20, n0 = 3,750) and circulation size and specific proximity (r = -.583**, n = 20, n0 = 3,750). The orientation of these correlations indicates that the higher a newspaper‟s circulation size, the more likely it is to publish the conflict factor on Twitter and the less 89

likely it is to publish dates or specific proximity factors on Twitter. There was no significant relationship between circulation size and prominence, oddity, impact, recentness, or general proximity. There was only one statistically significant relationship between individual factors, concerning recentness and specific proximity; although not of interest to this dissertation, and arguably not a particularly relevant finding, full correlations are presented here in the interest of comprehensiveness (see Figure 5 for Pearson‟s correlations of Twitter data.).

F3: Word Frequency Means of News Factors on Facebook Pages by Circulation Category

Circulation Category Deviance Social Significance Egalitarianism Averages Timeliness Proximity

Prominence Conflict Oddity Impact Recentness Dates General Proximity Specific Proximity

1.19 2.606 0.686 12.076 Community Weekly 12.164 6.898 0.084 7.678 3.796 12.762 19.146 7.678 16.558

1.484 3.058 0.558 10.258 Community Daily 19.266 13.104 0.13 7.638 4.542 10.816 32.5 7.638 15.358

1.584 1.39 0.338 8.312 Large Daily 9.684 7.314 0.256 7.222 2.974 8.65 17.254 7.222 11.624

0.914 1.15 0.588 5.092 National Daily 8.784 8.268 0.324 12.636 2.064 5.68 17.376 12.636 7.744

ANOVA 1.293 2.051 0.5425 8.9345 12.4745 8.896 0.1985 8.7935 Overall Mean 3.344 9.477 Sum of Squares 338.138 122.982 0.184 99.07 1.377 12.854 0.324 133.865 Mean Square 112.713 40.994 0.061 33.023 0.459 4.285 0.108 44.622 F 4.455 2.153 0.49 1.352 0.232 4.243 0.424 1.485 Sig. .019* 0.134 0.694 0.293 0.873 .022* 0.738 0.256 Total Mean 21.569 8.7935 12.821

F3: ANOVA analysis of word frequency means of deviant, social significance, and egalitarian news factors on newspaper Facebook pages by circulation category. Numbers in the “Community Weekly,” “Community Daily,” “Large Daily” and “National Daily” rows reflect means for analyses per factor per circulation category; numbers in the “ANOVA” rows reflect the ANOVA analysis. For all columns, df = 3, n = 20, n0 = 3,750, * indicates significance at the 0.05 level, and ** indicates significant at the 0.01 level. Timeliness and proximity were formed as an averaged composite for recentness and dates, and general proximity and specific proximity, respectively.

F4: Pearson‟s Correlations Between Circulation Category and News Factors on Newspapers‟ Facebook Pages

Deviance Social Significance Egalitarianism Size Prominence Conflict Oddity Impact Recentness Dates General Proximity Specific Proximity Size - -0.362 -0.041 0.286 0.327 -0.063 -.560* -0.123 -.462* Prominence - 0.443 -0.278 -0.216 -0.056 .645** 0.166 0.347 Conflict - -0.092 0.015 0.085 0.242 0 0.049 Oddity - -0.175 -0.161 -0.271 -0.307 -0.305 Impact - 0.016 -0.259 -0.247 -0.4 Recentness - 0.011 -0.015 -0.007 Dates - .445* .542* General Proximity - .701** Specific Proximity -

F4: Pearson‟s correlations analysis of word frequency means of deviant, social significance, and egalitarian news factors on Facebook pages by circulation category. For all instances, n = 20, n0 = 3,750, * indicates correlation is significant at the 0.05 level (2-tailed), and ** indicates correlation is significant at the 0.01 level (2-tailed).

F5: Word Frequency Means of News Factors on Twitter Pages by Circulation Category

Circulation Category Deviance Social Significance Egalitarianism Timeliness Proximity

Prominence Conflict Oddity Impact Recentness Dates General Proximity Specific Proximity

Community Weekly 1.108 2.618 0.906 15.752 8.626 6.148 0.034 5.642 Mean 3.726 16.658 14.808 5.642 20.384

Community Daily 0.46 3.14 2.31 9.332 13.082 9.776 0.08 5.078 Mean 3.6 11.642 22.938 5.078 15.242

Large Daily 0.75 0.844 0.558 10.022 11.858 11.334 0.324 5.256 Mean 1.594 10.58 23.516 5.256 12.174

National Daily 0.438 0.666 0.206 4.594 7.434 13.406 0 5.646 Mean 1.104 4.8 20.84 5.646 5.904

Overall 0.689 1.817 0.995 9.925 10.25 10.166 0.1095 5.4055 Mean 2.506 10.92 Sum of Squares 105.866 140.791 0.323 1.217 1.474 23.317 12.753 313.673 Mean Square 35.289 46.93 0.108 0.406 0.491 7.772 4.251 104.558 F 2.614 4.899 2.261 0.052 0.691 2.637 1.752 3.512 Sig. 0.087 .013* 0.121 0.984 0.571 0.085 0.197 0.04* Total Mean 20.5255 5.4055 13.426

F5: ANOVA analysis of word frequency means of deviant, social significance, and egalitarian news factors on newspaper Twitter feeds by circulation category. Numbers in the “Community Weekly,” “Community Daily,” “Large Daily” and “National Daily” rows reflect means for analyses per factor per circulation category; numbers in the “ANOVA” rows reflect the ANOVA analysis. For all columns, df = 3, n = 20, n0 = 3,750, and * indicates significance at the 0.05 level. Timeliness and proximity were formed as an averaged composite of scores for recentness and dates, and general proximity and specific proximity, respectively. 93

F6: Pearson‟s Correlations Between Circulation Category and News Factors on Newspapers‟ Twitter Pages

Deviance Social Significance Egalitarianism Size Prominence Conflict Oddity Impact Recentness Dates General Proximity Specific Proximity Size - -0.134 .680** 0.068 0.008 -0.24 -.486* -0.268 -.583** Prominence - 0.143 0.161 -0.238 -0.044 0.1 0.096 -0.033 Conflict - 0.148 0.032 -0.31 -0.162 -0.293 -0.42 Oddity - -0.082 -0.114 -0.079 -0.118 0.238 Impact - 0.14 -0.145 0.104 -0.303 Recentness - 0.176 0.005 .475* Dates - -0.108 0.288 General Proximity - 0.126 Specific Proximity -

F6: Pearson‟s correlations analysis of word frequency means of deviant, social significance, and egalitarian news factors on newspaper Twitter pages by circulation category. For all instances, n = 20, n0 = 3,750, * indicates correlation is significant at the 0.05 level (2- tailed), and ** indicates correlation is significant at the 0.01 level (2-tailed).

4.2.C: EVALUATING QUANTITATIVE HYPOTHESES Quantitative data presented in chapters 4.2.a and 4.2.b reflect H1, H2, and H3.

H1 considered the rate of deviant and egalitarian news factors in community and large newspapers. Its first postulate, that community newspapers would emphasize egalitarianism over deviance, was almost completely false.

Community weekly newspapers emphasized deviance roughly 150 percent more than egalitarianism on their websites; community daily newspaper websites emphasized deviance 240 percent more than egalitarianism. ANOVA analysis indicated no significant variance between community and larger newspapers‟ use of deviant and social significance news factors, as well as timeliness and general proximity; only one sub-factor of egalitarianism, specific proximity, varied significantly across circulation categories. Furthermore, on newspaper websites, circulation size was significantly correlated only with specific proximity; there was no relationship between circulation size and any deviant news factor or the social significance factor.

Put another way: on news websites, although specific proximity was an exception influenced by circulation size, community weekly and daily newspapers gave far greater emphasis to deviant news factors. This emphasis on deviance is statistically consistent with larger newspapers‟ similar focus on, and use of, deviance.

Similarly, ccommunity weekly and daily newspapers gave priority to deviance on Facebook posts as well, and community daily newspapers favored deviance over egalitarianism on Twitter. The lone exception was community weekly newspapers‟ use of Twitter, which focused more on egalitarianism (in particular specific proximity) than deviance. On Facebook, ANOVA analysis indicated only slightly significant variance across circulation categories concerning prominence and dates; all other factors were published at statistically consistent rates. Also on Facebook, circulation size was significantly correlated only with dates and specific proximity, and not any deviant news factors. On Twitter, ANOVA analysis indicated slightly

significant variance across circulation categories concerning only conflict and specific proximity; circulation size was correlated only with conflict, dates and specific proximity.

On social media, too, community newspapers adopted a clear focus on deviance over egalitarianism.

H1‟s second claim, that larger newspapers will give priority to deviance over egalitarianism, was validated without exception. Large and national newspapers gave clear and dramatic priority to deviance, sometimes publishing deviant factors more than 300 percent more frequently than egalitarian factors, on news websites, Facebook pages, and Twitter pages. ANOVA analyses of website and social media pages indicate only slightly significant variance across circulation categories for deviant news factors. Only once (conflict factors on Twitter) is circulation size correlated with a deviant factor; the .680** correlation (N = 20) supports comparisons of means indicating larger newspapers‟ focus and priority on deviance over egalitarianism; furthermore, significant negative correlations between circulation size and egalitarian factors reinforce this conclusion.

H1 is partially accurate.

H2: American weekly and daily community newspapers will publish online news on conflict-based deviance less frequently than larger daily newspapers, but will publish prominence-based deviant news more frequently than larger daily newspapers.

H2 specifically considered prominence and deviance; both postulates are inaccurate. ANOVA analyses indicate no significant variance across circulation categories concerning prominence or conflict on newspaper websites; on social media, there are only slightly significant variances across circulation categories concerning prominence on Facebook and conflict on Twitter. Furthermore, Pearson‟s correlations indicate no significant relationships between circulation size and prominence or conflict on news websites of Facebook, although there is a highly significant positive correlation between circulation size and conflict on Twitter.

This lone exception is not enough to invalidate the general trend, however – American community and larger newspapers publish prominence and conflict factors at statistically

consistent rates, and there is no substantive relationship between circulation size and either prominence or conflict.

H2 is inaccurate.

H3: Regardless of circulation size, American community and daily newspapers will publish social media content with higher rates of egalitarian news than deviant news factors.

H3 considered newspapers‟ use of social media, Facebook and Twitter, in isolation. It postulated that, regardless of circulation size, the collaborative and discussion-based nature of social media would lead to an editorial preference for egalitarian news factors over deviant news factors.

This was largely inaccurate. Newspapers of all circulation sizes, on average, favored deviance over egalitarianism on Facebook. In some instances, the difference between the two was stark; community daily newspapers focused on deviance more than 100 percent more than egalitarianism. Other circulation sizes had a more balanced focus, but nonetheless favored deviance without exception. Similarly, community daily, large daily, and national daily newspapers all favored deviance over egalitarianism on Twitter; only community weekly newspapers gave egalitarianism, and in particular specific proximity, a priority over deviance. Overall averages, too, demonstrated clear favor for deviance over egalitarianism.

ANOVA analyses reinforced these findings. ANOVA analyses indicated only slightly significant variance across circulation categories concerning prominence and dates on Facebook, and conflict and specific proximity on Twitter; the majority of factors for both media was published at statistically consistent rates by all circulation categories, and consistently adopted a clear focus on deviance over egalitarianism and social significance.

It is worth noting that, generally, social media favored egalitarian news factors more than traditional websites. The overall averages for egalitarian factors for Facebook and Twitter were both more than 100 percent higher than the overall egalitarian average for traditional news websites. Because website and social media data are contained in independent data sets, and that lack of data overlap prevents direct statistical comparison, this is a purely cursory comparison of

means; as such, it may be statistically invalid. It is also not expressly covered by Hypothesis 3. However, it is nonetheless a noteworthy finding.

H3 is inaccurate.

4.3: Qualitative Results

Chapter 4.3 considers qualitative results derived from structured interviews outlined in Chapter 3.4. Twenty eight interviews of community newspaper editors and publishers were conducted at the Texas Press Association‟s Midwinter Convention and Trade Show on Jan. 17- 19, 2013. Qualitative data were used to analyze H4.

H4: The more contact a community newspaper editor has with their community and readers, the more preference will be given to egalitarian news over deviant news.

Twelve structured interview questions were designed and, following data collection, categorized into three sub-groups for analysis.13 The first, consisting of Q1, Q6, Q7, Q8, and Q12, primarily considers newspaper size, newsroom size and professional experience; the second, consisting of Q2, Q3, Q4, and Q5, considers perceived interaction between community newspaper editors and publishers, and their readers; the third, consisting of Q9, Q10, and Q11, considered editorial focus on deviant and egalitarian news, or news on “important people and unusual events” or “regular people and routine events.”

Some questions utilized Likert scale-style formats and some were open ended. Participants were generally very helpful and forthcoming, however, and often provided qualitative detail to essentially quantitative questions. As such, the numerical data will be presented here but also qualified with participant‟s detailed answers. The interviews were colorful, engaging, and frankly, quite a bit of fun to collect.

4.3.A: NEWSROOM SIZE AND CHARACTERISTICS Generally speaking, most participants at a state press association convention are likely to be community weekly or bi-weekly publications. Press associations mostly cater to small publications and much of the convention and trade show content was geared toward hyper-local newspapers. However, specificity is important for any qualitative analysis.

13 See Appendix 8 for the full structured interview script and Chapter 3.5.c for a full description of interview questions.

(Q1) Is your newspaper a community weekly, bi-weekly, or daily publication?

Q1 was a straightforward inquiry on circulation size. Of the 28 collected interviews, 19 editors or publishers managed weekly newspapers. Six respondents edited or published bi-weekly newspapers, one published a community daily newspaper, one respondent managed two weekly publications, and one was a publisher over a network of bi-weekly and weekly community publications. Findings were consistent with this dissertation‟s focus on community journalism.

(Q6) How many years have you worked as a community newspaper editor?

Based on their responses, participants represented at least 581 years of collective community journalism experience. This figure is an estimate because it includes some years of reporting or lower-level editor experience, and because not every participant mentioned experience outside of working as an editor or publisher. The question did not ask for industry experience, but responses were so forthcoming and the final tally was so staggering that it deserves mention here.

Concerning experience as an editor or publisher, 18 participants had more than a decade of experience. Nine had between one and nine years and only one participant had less than a year of experience. Further qualitative responses were unsolicited by the interview but nonetheless intriguing.

“I grew up in the newspaper business. So, 60.”

“Since 1989, so how many is that? 24 years?”

“I‟ve been there 33 years with that newspaper. I started as a reporter and was editor for probably … 23 years? 24? And I‟ve been publisher for I guess the last 8 or 10. Well, no, I took the publisher‟s title 20 years ago, and we had another guy that was still there and I didn‟t completely consider myself the full publisher till he left, but I basically was.”

“A publisher? Two years as publisher, but I‟ve worked there my whole life. My family owns the paper, I‟m fourth generation.”

(Q7) How large is your paid editorial staff?

Twelve participants were either the sole editorial voice in their newsrooms or had only one other journalist employed. Nine worked in newsrooms with between three and five editorial employees. Only seven participants employed more than five journalists. 100

“How large is my paid editorial staff? I‟m it! I‟m it.” (laughs)

“How large? Ooooh! Two! (laughs) Well two-ish, because we all kind of pitch in, you know. There‟s probably two or three others besides the two full-time, and that‟s not counting the sports.”

“Where I was previously, I was there for 29 years, we had a staff of anywhere from 6 to 15, it parried down over the years. Where I‟m at right now, I‟ve got 5 beside me.”

“We only have 11 people total for two papers, and on the editorial side there are 3 of us.”

(Q8) Do you have any unpaid or freelance reporters? If so, how many?

The majority of participants had a small number of unpaid or freelance contributors. Thirteen had fewer than two freelancers, 10 participants had between 3 and 5 freelancers, and five participants had more than five freelancers. Many of these freelancers were not traditional reporters, and were instead photographers, columnists or other contributors; participants volunteered the information, however, and clearly did not limit the definition of “freelance” to traditional text-based reporting. It also was clear this view was shared by all participants. As such, results are included here.

“We have freelance photographers and a columnist, and we have 4 of those, and they do that gratis, just for the exposure and so forth.”

“Basically none. We have a few columnists that submit, we get a monthly column from the school and the hospital, we had a weekly religious column but that guy figured he‟d better retire a few months ago, so we‟re blank on that.”

“Hmm, probably have about 3 or 4. Actually no, I take that back, it‟s more than that. Let‟s see … about 6 or 7, now that I think about it.”

“We have one full-time reporter, one part-time reporter, and myself. I have five student sports editors and I have two contributing writers that go to Commissioner‟s Court and city council meetings.”

(Q12) Would it be okay if I got your business card? Your response will be completely confidential, but I‟d like to make sure I don‟t accidentally interview two people from the same newspaper.

Q12 was utilized exclusively as a control variable. To ensure a valid and reliable qualitative dataset, care was taken to prevent accidentally interviewing two editors or publishers 101

from the same community newspaper. Eighteen participants volunteered their business cards. The remaining 10 did not have cards available, but provided earnest assurance that they were their newspaper‟s only visitors at the conference; in many cases, they were the only editors or publishers at their newspapers in the first place. The business cards were later checked to confirm there were no duplicates from the same publications and destroyed.

4.3.B: PERCEIVED JOURNALIST / AUDIENCE INTERACTION On the whole, community newspaper editors and publishers perceived a great deal of interaction between their newsrooms, their customers, and their community.

(Q2) On a scale of one to 10, with one being “almost none” and 10 being “a great deal,” how much interaction do you feel you have with your readership?

Community newspaper editors and publishers said they had a great deal of interaction with their audiences. The vast majority, 21 participants, rated their audience interaction as a seven or higher; six participants placed the rating between 4 and 6, and only one response gate a rating less than four. As with other questions, participants provided unexpected qualitative detail in their answers which is worth reporting.

“About 10. We see them at community events, when you go shopping at walmart, at church. We‟re a pop of about 6500, so it‟s a very small community. We see them a lot of places.”

“I‟d say 8 or 9”

“I would say probably a 5. It‟s certainly available to them, a lot of them just choose not to do anything, not to respond to anything. We don‟t have a whole lot of letters to the editor, but we put in our paper how to send a letter just about every week, so that opportunity is there for them.”

(Q3) On average, about how many times a week do readers visit your office to talk about news ideas or your news coverage?

Q3 served as a quantitative check on Q2. The nature of qualitative research accepts a certain degree of ambiguity in responses; one editor‟s seven could conceivably be another editor‟s nine. However, it seems reasonable to assume that a reader‟s visit to a newsroom has considerable external reliability.

102

Again, community newspaper editors and publishers perceived a great deal of interaction with their readers. Eight participants said readers visited more than once a day, 10 participants said readers visited about once a day, and 11 participants said readers visited less than once a day. However, even the lowest rate of perceived interaction – two or three visitors a week – reflects a high degree of accessibility to, and interaction with, the community.

“That‟s a hard question to answer. People come in every day, I‟d say about a dozen come in every day, I would say probably 2 come in for editorial things.”

“I couldn‟t count. Uh, constantly. We‟re both obviously in smaller markets surrounding the metro, and people feel more free to walk into your newspaper office and talk to you about stuff, so I bet we get 30 people a day, at least, who come in. Every bit of that, they just walk in. and that certainly doesn‟t count all the emails you get or other communication.”

“We have 2 or 3 people a week come in. They‟re not always happy, but at least they‟re reading the paper. „Thanks for reading‟ is what we send „em off with.”

(Q4) On a scale of one to 10, with one being “almost none” and 10 being “a great deal,” how much interaction do you feel you have with your online readership?

Responses concerning both Q4 and Q5, concerning interaction online and over social media, seemed highly disparate. Editors and publishers often said they had tremendous or reliable online interaction, or in other cases, they had no website or social media to speak of, and hence no interaction. Nine participants said they had a great deal of interaction with their online readership, or gave a rating of seven or more; five participants said they had some interaction or gave a rating of four to six; 13 said they had little to no interaction or gave a rating of three or less, often because they did not have a news website.

“Uhm, not a whole lot. We don‟t have a space for comments on our stories and we don‟t put a whole lot of content on the web, because it‟s a free site, so without setting up a paywall or anything we don‟t uh wanna put everything on there, so we only put probably 25 to 30% of our weekly content on the website, and everything else is in the printed product.”

“We don‟t have online readership.”

“I would say that we have a strong online readership. I would say we don‟t have as much … but a lot of the people that subscribe are also getting their news online as well. So I

103

would say probably, not as much as the regular public print dailies do. I would probably say maybe about a 5.”

“A lot. I would say we measured, well just as far as submissions, we get about 20 to 30,000 submissions a month from our community. And we have a lot, I can‟t put an exact number to it, but a bunch.”

(Q5) What about your social media audience? On a scale of one to 10, with one being “almost none” and 10 being “a great deal,” how much interaction do you have with them?

Q5 yielded similar responses. Ten participants said they had a great deal of interaction over social media or gave a rating of seven or higher; seven participants said they had some interaction or gate a rating between four and six; nine participants said they had little to no interaction or gave a rating of three or less. Many of those participants had no social media, and thus, no interaction over social media.

“Constant. With Facebook it‟s a constant thing. You post stuff and you get feedback right away, and we post to Facebook two or three times a day. Probably the average is 4 or 5 comments per post, but every now and then we post a topic that gets a lot of posts. I‟d probably say about three is the average.”

“We don‟t have Facebook. That‟s supposed to be my new job this summer.”

“Pretty much none. A little bit. We have a lot on Twitter, I‟ll say our sports guy has about 3,000 followers on Twitter, he‟s nuts. And we get maybe 700, 800 likes on Facebook, but it‟s not as good as it oughta be.”

“It‟s all about how often your editors and reporters are putting things out there to interact with. Ironically, one of my smaller communities just passed the 5000 Facebook followers point, very very small community. It has probably twice the following that some of my bigger communities get because there‟s that intimate relationship with readers. But … it‟s at least 100% growth over the last year, and it‟s all about how your reporters and your editors interact online.”

4.3.C: DEVIANT & EGALITARIAN NEWS The remaining three questions, Q9, Q10, and Q11, speak to the heart of this dissertation. Q9 asks editors and publishers about news on “regular people and routine events,” which represents a qualitative operationalization of egalitarian news. Q10 asks the same question concerning news on “important people or unusual events,” which offers a qualitative

104

operationalization of deviant news. Q11 asked editors and publishers where, ideally, priority would lie between deviant and egalitarian news.

Participants expressed a clear preference for egalitarian news on regular people and routine events. Deviant news was not marginalized, necessarily, but most participants considered egalitarian news more important, more feasible, and more profitable.

(Q9) On a scale of one to 10, with one being “almost none” and 10 being “a great deal,” about how much of your news coverage is devoted to regular people or routine events?

Of the 28 participants, 20 said a clear majority of content should focus on egalitarian news and offered a rating of seven or higher. Seven participants said focus on egalitarian and deviant news should be about even, usually rating both at five but occasionally giving other ratings between four and six. Only one community newspaper editor said egalitarian news should be in the minority and gave a rating less than four.

“Well I guess all of it, would be, wouldn‟t it? We‟re a local newspaper and we deal a lot with the schools, and other events, and then of course wrecks, and the Sheriff‟s report, right now one of our editors is doing feature stories on anybody over 90 from different parts of the county and that‟s been very interesting. Everybody‟s nosy.”

“A lot. To me the beauty of a community newspaper is, I always tell people, we‟re recording the history of this town. And, just last week we had a lot of negative things happen, we had a pick axe murder this morning when I left, but normally we love features, we love hundred-year-old ladies. We focus a lot – 1 to 10, I‟d say 8.”

“All of it. Mm-hmm.”

“I mean we deal with extraordinary, special things, but we‟ve got a city council every other week, there‟s a great deal of it that is routine. So probably 5.”

(Q10) On a scale of one to 10, with one being “almost none” and 10 being “a great deal,” about how much of your news coverage is devoted to important people or unusual events?

Conversely, participants generally said deviant news should be a lower priority. Fifteen participants said very little news coverage should focus on deviance or gave a rating less of three or less; nine participants said deviant and egalitarian news should be relatively even, or gave a rating between four and six. Only three participants said deviant news should be a majority of coverage or gave it a rating of seven or higher; furthermore, at least one of those participants gave 105

a “10” rating to both deviant and egalitarian news and seemed to imply the two should be even. Although the participant counted both as 10 and said “we do both” at that community newspaper, the expressed even composition is inconsistent with his 10 rating. The case deserves to be counted but the participant was clearly not expressing a preference for deviant news; the remaining two cases, however, clearly were.

“That‟s pretty much the focus, so I‟d say 6 or 7.”

“We have a columnist that‟s called „friends and neighbors‟ and it picks out somebody that‟s unusual. It can be a child, or a kid, or a 90-year-old man, or whatever‟s happening, and so we have that every week. And then many of our stories are about unusual or different people. Everything is community-generated, so we‟ll have Scouts that come in and we have articles about them, if somebody grows the biggest tomatah14 in the county we‟ll have an article about that, „cause that‟s what we‟re all about. So I‟d say probably about 5.”

“Hardly ever. Probably a 2.”

“Popular people are gonna get popular coverage, and that‟s not who we are. We like to cover our neighbors.”

(Q11) In general, what are your thoughts on community newspaper coverage of regular and unusual people and events? Where should the priority be?

Similarly, the same two trends were evident in responses to Q11. Only two participants said deviant news should be given a higher priority than egalitarian news. The remaining participants were split relatively evenly: 12 said egalitarian news should be favored and nine said a roughly 50/50 balance between deviant and egalitarian news should be found. Two responses were considered off topic.

Participants who clearly favored egalitarian news had several rationales. Some said routine events, such as regular city council or school board meetings, enjoyed a well-justified priority. Others said unusual events happened too infrequently in a small town, or at least too unpredictably, to be a major part of the newspaper. Many also said that community journalism must retain a hyper-local focus on ordinary happenings and individuals; their rationale was partly

14 Tomato, in Texas drawl

106

that community journalism is a niche product with a specific focus, and partly that local and egalitarian news is more profitable.

“Ideally, I think for a community newspaper, I think we‟re, that‟s a pretty good ratio. (95/5) We cover the big events when they happen and they need to be covered, but we‟re covering the people every day in our community all the time, and that‟s what people look to us for. What‟s going on in the community, and what are the people in the community doing.”

“Regular people generate most of the news. You have your school activities, your community civic groups, you have your government meetings like your school boards, city meetings and so forth, so just by osmosis it‟s going to generate most of your news. So probably 70, 80 percent of the news is generated by regular people. The unusual nuts and clowns and people who just do unusual things come in, and that‟s a little less often. So we have a solar car group that‟s traveling from here to Washington, D.C., or not from here but from our hometown, so that‟s an unusual group. And they are kinda goofy. So that‟s kind of the unusual out of the ordinary. But we cover that, I‟d just say it wouldn‟t be as often.”

“The only reason that we‟re successful is because we are covering our community. So we have a niche that no one else can cover, and we cover the students and their successes. If you came from our hometown we‟d be doing feature stories on your life about once a year keeping up with your progress. And your grandmother and your great grandmother and your aunts and uncles and the kids you went to school with, they would all be very interested in reading that. They cannot get that on Fox News, USA Today, so I let them do that stuff. I use Capital Highlights, that the TPA puts out … but I do not spend time and resource covering stuff like that. That‟s how we stay in business.”

“I would err more on the side of the, uh, routine events and regular people, because I think that‟s what makes a community newspaper thrive. Is, you know, keeping up with the little league and the high school football team an‟ all that sort of thing. I mean we also have an obligation, obviously, to cover city council and crime and all that, but, you know, we‟re the lifeblood of our community just covering everyday life.”

Participants who preferred finding a balance between egalitarian and deviant news expressed similar ideas on the role of egalitarian news. They also, however, emphasized a professional interest in “hard news” and a need to adapt to the news and events of a particular week. Some issues would be theoretically more focused on deviant news because more deviant events took place that week, so the logic goes; however, a dual focus still should be maintained on egalitarian news as well.

107

“It depends on how you define “important.” What‟s important to some of my communities also qualifies for unusual, so it could be 50 percent sometimes, (laughs) but in general terms, high-profile politicians and that kind of – that‟s not what we do. If it‟s unusual in our local community then it could be 100% of my paper that week, I don‟t care. It‟s all about what‟s important to that community, and it can be different.”

“I don‟t think I would, I think I like trying to do both of them. And we do. We have a lot of routine events, but where we live, we have a lot of special events, and we don‟t really necessarily go for somebody who‟s „important,‟ we go for somebody who is „interesting.‟”

“When you say regular people and routine events, there‟s kind of a line that comes between them. Council meetings that we cover, school board meetings, different boards and things like that, those are routine but sometimes those are timely. So we would give those a fair amount of priority. Features that are just on regular people, sometimes those aren‟t as timely. So if we had breaking news, a helicopter crash or something like that, we might bump a feature back, I dunno if that answers your question.”

“We consider ourselves a newspaper. I would say, you know, 7 or 8. We try to balance both to be honest. We have a special section we try to put a feature on every week and then the hard news is in the front, so yeah, I would say we‟re pretty balanced, seven ish, in favor of the general – although we do a lot of hard news too, so it‟s not really fair. I dunno what to tell you. I‟d say we‟re balanced, whatever a balance is.”

The two participants who expressed the minority preference for deviant news held a similar focus on news routines and a professional interest in hard news. They did not emphasize the same balance, however, between deviant and egalitarian news.

“Uh, I mean, news goes first. The other stuff that can always be held off and wait. Especially in our area, in a small newspaper, if it‟s more feature I guess on someone and been going on, that harder news will take precedent.”

“Do we do more extraordinary? I would say we do more extraordinary than we do regular people.”

4.3.D: ASSESSING THE QUALITATIVE HYPOTHESIS Generally speaking, the qualitative Hypothesis 4 was supported.

H4: Community newspaper editors‟ perceived contact with audiences positively influences reported use of, and preference for, egalitarian news over deviant news.

H4 considered qualitative data derived from qualitative, structured interviews of 27 community newspaper editors and publishers. Although it can be difficult to determine causality 108

through qualitative methods, community newspaper editors and publishers routinely reported a high level of contact and interactivity with readers, audiences, and community members. A clear majority of editors and publishers, or 21 of the 27, reported an interaction level of seven or higher (out of 10); many also expressed qualitatively that they are thoroughly immersed in the community, and try to maintain as much open communication as possible. The nature of community journalism, they argued, is very open to community input.

Much of this interaction was offline, in large part because community newspapers have web presences of varying sophistication. Online interaction was largely based upon that online presence. Many editors and publishers said they had a great deal of interaction online or over social media; many others said they had very little online interaction, often because they had underdeveloped web presences. Publications which invested in the internet and social media generally had a high degree of audience interaction, however, further demonstrating the open and interactive nature of community journalism.

Editors and publishers also expressed vast preference for egalitarian news, or news about “regular people and routine events” over deviant news, or news about “important people or unusual events.” The lion‟s share of responses indicated that egalitarian news is the lifeblood of the community newspaper industry; in no uncertain terms, editors and publishers said that egalitarian and deviant news should be split at most half in half, and most responses gave clear priority (between 70 and 100 percent of editorial content) to egalitarian news.

It seems logical to assume that the perceived interaction with audiences, which was very high, is related to the very high reported focus on egalitarian news over deviant news. This does not, however, explain the overlaps and discrepencies between quantitative and qualitative data, which will be explored in Chapter 5.

Individually, however, H4 is supported.

109

4.4: Conclusion of Results

Chapter 4 has explored the results of the computerized, quantitative content analysis of newspapers‟ websites and social media pages, as well as results of qualitative, structured interviews of community newspaper editors and publishers concerning audience interaction and preference for egalitarian and deviant news.

The remainder of this dissertation will be devoted to the discussion of these results. Chapter 5 offers a broad and practical discussion of the results and their implications for study, including a focus on implications for computerized content analysis and a discussion of the arguably contextual and circumstantial nature of localness and proximity. It also will offer theoretical development concerning deviance and gatekeeping theory; in particular, it will explore two new interpretations of deviance, as a spectrum and as a broader concept, that may strengthen the concept‟s theoretical value. It will also consider how the actual term, “deviance,” may be hindering the concept‟s theoretical development and application. Chapter 6 will conclude the dissertation and offer opportunities for future quantitative and qualitative research.

110

CHAPTER 5: GENERAL & THEORETICAL DISCUSSION

5.0 Introduction

Chapter 5 focuses on general discussion of the data presented in Chapter 4. There are noteworthy similarities between quantitative and qualitative data, particularly concerning statistical and qualitative relationships between circulation size and focus on local communities. Quantitative analysis and qualitative interviews reinforce Jock Lauterer‟s (2006) claim that community journalism is “relentlessly local,” which also speaks to the heart of community journalism research writ large (Brockus, 2009; Dill & Wu, 2009; Garfrerick, 2010; Hansen & Hansen, 2011, 2012; Reader, 2006; Smethers, et al., 2007). These important consistencies will be discussed in Chapter 5.3.

However, also noteworthy is the counter-intuitive quantitative finding that circulation size plays little to no significant role in the use of deviant or egalitarian news factors on websites or social media pages. This finding clashes, to some degree, with qualitative findings that community newspaper editors and publishers profess a clear preference for egalitarian news. There are four potential explanations for this discrepancy. The first explanation attributes the difference to an affirmation of media sociology research which argues that media institutions homogenize content to fit institutional standards and patterns; this also speaks to professional journalistic routines. The second rationale relates to the well-documented sensitivity and mechanics of computerized content analysis software and indicates a potential methodological issue or flaw. The third reason, and perhaps the simplest, relates to this dissertation‟s focus on online news and social media – not often studied traditional print media – and potential implications for the study of online news. The fourth cause points to a need for theoretical reevaluation of normative deviance, both in principle and in practice. These first three possibilities will each be discussed in Chapter 5.4; the fourth potential explanation will be explored in Chapter 5.5.

Chapter 5.6 will discuss, and largely dismiss, a central assumption of gatekeeping theory: that journalists have only “modest exposure” (Shoemaker & Vos, 2009, p52) to their audiences.

111

5.1: Quantitative Results Review

The quantitative results presented in Chapter 4.2 indicated three overarching trends about the use of deviant and egalitarian news factors on American newspapers‟ websites and social media pages.

1. American newspapers‟ online news has a clear, dominant focus on deviance. Deviant news factors were published roughly 200 percent more often than social significance or egalitarian factors.

2. American newspapers‟ social media content, posts on Facebook and Twitter, also have a clear and dominant focus on deviance. Deviant news factors were published roughly 70 percent more often on Facebook and 50 percent more often on Twitter than egalitarian factors; deviant factors were also used much more frequently than social significance factors, roughly 150 percent more on Facebook and 280 percent on Twitter. Social media posts were generally more egalitarian than news website content, however.

3. Surprisingly, circulation size played little to no significant role in newspapers‟ focus on deviance and egalitarianism. Usage of deviant, social significance and egalitarian factors was almost completely uniform across circulation categories. Highly significant variances across circulation categories existed only for specific proximity concerning news websites; slightly significant differences across circulation categories existed concerning prominence and dates on Facebook and conflict and specific proximity on Twitter. Further, Pearson‟s correlations revealed few significant relationships: between circulation size and specific proximity on news websites, circulation size, dates and specific proximity on Facebook, and circulation size, conflict, dates, and specific proximity on Twitter.

112

5.2: Qualitative Results Review

The qualitative results presented in Chapter 4.3 indicated three overarching trends among community newspaper editors and publishers.

1. By any measure, community newspaper editors and publishers claim a tremendous amount of interaction with their readers and community. Most had more face-to-face interaction than dialogue or discussion online, but this was largely due to the gulf between technically literate and illiterate newsrooms. Community newspapers with sophisticated websites and social media presences had a great deal of online interaction, but newspapers with a less substantial (or entirely absent) online presence logically had less online interaction.

2. Interviewed community newspaper editors and publishers had more than 581 years of collective experience; furthermore, there was little substantive gap between responses. Qualitative scholars recommend interviewing subjects until responses begin repeating themselves and no new information is presented (Coleman, 2007; Fontana & Frey, 2005); such qualitative consensus is clearly evident.

3. Community newspaper editors and publishers have a clear, and often dramatic, preference for egalitarian news over deviant news. Almost none of the participants reported or advocated a majority use of deviant news content. Instead, roughly three fourths reported demonstrative priority for egalitarian news, or content about “regular people and routine events,” and about a fourth reported printing about an even balance between deviant and egalitarian news content (See Chapter 4.3.c on Q9). More generally, participants were roughly evenly split between those who advocated a clear preference for egalitarian news under ideal circumstances and those who advocated a roughly even split between deviant and egalitarian news under ideal circumstances (See Chapter 4.3.c on Q11).

113

5.3: Commonalities Between Quantitative and Qualitative Results

The quantitative, computerized content analysis and qualitative, structured interviews employed by this dissertation were designed to study independent, but interconnected, concepts concerning normative deviance, gatekeeping theory, community journalism, imagined community, and online news and social media. Quantitative and qualitative analyses studied American community journalism, and a prime theme developed between the results: community journalism has a clear, tangible focus on local events, institutions, ideas, and individuals.

ANOVA analysis of quantitative data indicated that only one news factor, “specific proximity,” had significant variance in means across circulation categories (p < .001**, df = 3, N = 140). The specific proximity word dictionaries were tailored to each sub-group of newspapers, and included the name of every city, town, village, and neighborhood within 20 miles of each newspaper‟s home community. The magnitude of the significance indicates a large, demonstrable variance between community weekly, community daily, large daily, and national daily newspapers‟ use of specific proximity. Furthermore, Pearson‟s correlations indicated only one significant relationship between circulation size and other news factors – specific proximity, at - . 342** (N = 140), indicating that larger circulation categories published specific proximity at significantly lesser rates (see F1 and F2, Chapter 4.2.a).

Social media analysis yielded similar, albeit less significant statistically, results. ANOVA analysis indicated a .04* variance (df = 3, N = 20) across circulation categories for specific proximity on Twitter, and Pearson‟s correlations found a -.462* (N = 20) correlation between circulation size and specific proximity on Facebook and - .583** (N = 20) correlation between the same variables on Twitter (see F4, F5, and F6, Chapter 4.2.b).

These findings are consistent with the vast majority of comments collected during qualitative, structured interviews of community newspaper editors and publishers. The predominant priority, according to every participant, was a focus on local people, institutions, events, and governments (see Chapter 4.3).

“Popular people are gonna get popular coverage, and that‟s not who we are. We like to cover our neighbors.” 114

“The only reason that we‟re successful is because we are covering our community. So we have a niche that no one else can cover.”

“To me the beauty of a community newspaper is, I always tell people, we‟re recording the history of this town.”

Both quantitative and qualitative findings, too, are entirely consistent with past research on community journalism and its “relentlessly local” (Lauterer, 2006) focus (Bowd, 2011; Dill & Wu, 2009; Garfrerick, 2010; Hansen, 2007; Hansen & Hansen, 2011; Reader, 2006; Schweitzer & Smith, 1991; Smethers, et al., 2007). Literature, too, has indicated that niche publications function in much the same way; every article is anchored to the media‟s niche, a different kind of “local” that functions in much the same way (Cover, 2005; Gade & Krug, 2008).

Two potential eccentricities in the data merit dismissing here as well. First, although qualitative data focused only on one state and quantitative data studied a national sample, it is logical to assume the two datasets are comparable. As stated in Chapter 3.4.c, Texas offers a diverse array of newspapers and a wide range of published studies have focused exclusively on the Lone Star State (Conway, 2006; Cope, 2011; Jensen & Uddameri, 2009; Kraeplin, 2008; S. Lewis, et al., 2009; Schweitzer & Smith, 1991; J. Stewart, 2011). Quantitative data included only five newspapers from Texas, and it is very difficult to conduct statistical comparisons with such a small dataset; qualitative analysis of the newspapers and quantitative data, however, indicate very strong similarity among newspapers from Texas and other states. Second, although a number of unmeasured variables potentially weigh upon the quantitative dataset, the data themselves are valid and worthy of study. The best example is the quantitative finding concerning community daily newspapers and specific proximity, which is surprisingly lower than large daily newspapers‟ use of specific proximity (See F1). This is potentially because of a number of factors not studied here; often regional daily newspapers serve a larger regional audience than community weekly publications, and many must cover that larger area on a relatively small staff and budget. This can potentially lead to a higher use of Associated Press or wire copy (Funk, 2010). Given that ANOVA analyses indicated a highly significant variance concerning circulation size and specific proximity, it is likely that data for community daily newspapers are understated rather than a problematic anomaly.

115

Quantitative and qualitative findings here reflect past research‟s conviction that community newspapers have an institutional, and often exhaustive, focus on local news.

116

5.4: Differences Between Quantitative and Qualitative Results

Academic literature is less clear, however, on how community and large newspaper news content differs in other respects. Although “local” focus is an integral part of community journalism, however the particular “local” is defined, that focus only addresses one component of news coverage; local focus does little to predict how tension or controversy will be covered, for example. Although structural pluralism studies argue that community demographics and characteristics influence a range of coverage subjects (Donohue, et al., 1985; Donohue, Olien, & Tichenor, 1989; Jeffres, Cutietta, Sekerka, & Lee, 2000; Nah & Armstrong, 2011; Pollock, 2007), a comprehensive view of community newspaper content remains unfulfilled.

Based upon community journalism studies and structural pluralism literature, H1, H2, and H3 predicted that community weekly and daily newspapers would be generally more focused on egalitarianism than deviance on news websites and social media; they also postulated that community newspapers would be more focused on prominence, and less on conflict, than larger newspapers.

These hypotheses were largely inaccurate. There were very few statistically significant variances across circulation categories for deviant or egalitarian variables, and there were very few significant Pearson‟s correlations between circulation size and individual factors. Apart from significant differences concerning specific proximity, which were discussed in Chapter 5.3, newspaper content concerning deviant and egalitarian factors was largely homogenous.

Surprisingly, community newspapers are ultimately just as preoccupied with deviance as metropolitan and national newspapers. It may be local deviance, but it is otherwise indistinguishable from national news content. However, this obviously clashes with qualitative findings that community newspaper editors and publishers give clear and deliberate preference for egalitarian news over deviant news, or priority to news content on “regular people and routine events.”

There are four potential explanations for this dichotomy.

117

5.4.A: VALIDATION OF MEDIA SOCIOLOGY The first potential influence points to Shoemaker and Reese‟s (1996) hierarchy of influences model. The gulf between the quantitative and qualitative results fortifies media sociology and gatekeeping claims that media practitioners craft and homogenize news content to fit institutional standards and patterns (Bleske, 1991; Bowman, 2008; Cassidy, 2006; Epstein, 1973; Gans, 1979; Gieber, 1956; Hirsch, 1977; Shoemaker & Reese, 1996; Shoemaker & Vos, 2009; Sigal, 1973). Individual journalists, including particular community newspaper editors and publishers, have little practical autonomy within the hierarchy of influences (Shoemaker & Reese, 1996) and thus little incentive for creative, independent, or even simply different content.

Newspapers function, first and foremost, as newspapers – adjectives such as “community,” “niche,” “metropolitan,” or “national” only augment those institutional patterns, not override them. Arguably, too, qualitative results are reinforcing media sociology: There was very little qualitative difference between individual responses, and patterned preferences for local and egalitarian news were clearly visible across the vast majority of interviews. There was little qualitative or quantitative variance, thus possibly demonstrating the herd mentality of journalists and the power of the routines level of influences.

This interpretation points to possible cognitive dissonance and a gap between journalists‟ words and actions. Community newspaper editors‟ and publishers‟ conviction that egalitarian news overrides deviant news is inconsistent with quantitative findings indicating the exact opposite; however, given how little variance exists within quantitative or qualitative data, it seems obvious that contradictory institutional patterns are at work. Community journalists may feel obligated to profess a preference for news about “regular people and routine events,” but they feel more obliged to pursue institutional standards of professional journalism, which are more deviance-based. This, too, is consistent with gatekeeping theory and the hierarchy of influences (Shoemaker & Reese, 1996; Shoemaker & Vos, 2009). Even among very small publications, where one or two individuals operate the entire newspaper, this pressure to conform to industry standards may override the urge to craft individualized editorial philosophies.

In a sense, this interpretation argues that both quantitative and qualitative findings are accurate – community journalists reliably and accurately say one thing, and publish another.

118

Understanding the potential cognitive dissonance, either deliberate or accidental, then becomes an intriguing avenue of academic research.

5.4.B: HYPER-SENSITIVITY AND NON-CONTEXTUALIZED NATURE OF COMPUTERIZED

CONTENT ANALYSIS SOFTWARE The second potential solution argues the opposite perspective – quantitative and qualitative findings are at odds because the computerized content analysis procedure produces invalid results.

It can be difficult to establish validity for computerized content analysis. Indeed, a variety of methodological tests and academic studies caution against the temptation of over-utilizing or over-emphasizing computerized content analysis (Ballotti & Kaid, 2000; Conway, 2006; Jarvis, 2004; Krippendorff, 2004; McCombs, 2004; Nuendorf, 2002))). The word dictionaries utilized in this dissertation were designed to be broad and inclusive; there is a possibility that dictionaries for deviant factors, in particular conflict and prominence, were simply too comprehensive.

Constructing voluminous word dictionaries runs the risk of finding results everywhere, effectively weighing down every scale until no differences or variances can be found. This possibility is highly similar to Conway‟s (2006) findings when testing conventional second-level agenda setting results against computerized content analysis. The computerized content analysis software was too sensitive for proper analysis, counted too many words because of an inability to account for context or circumstance, and ultimately rendered a high volume of inapplicable data. This is effectively exactly what happened with this study concerning magnitude, and forced the abandonment of the news factor; it is a possibility that the same applies to prominence, conflict, oddity, and magnitude, all of which used large customizable word dictionaries.

For similar reasons, it is possible that significant variances and relationships between circulation size and specific proximity are actually understated. The word dictionaries for specific

119

proximity were tailored to individual newspaper groups15 and included all cities, towns, villages, and neighborhoods within 20 miles of each newspapers‟ home community; the dictionaries were meticulously defined, but they were not as vast as dictionaries for prominence, conflict, oddity, and impact. The dictionaries also considered only clear and extant articulations of local community names; however, as Benedict Anderson (2006) notes, community identity can be articulated in a wide variety of imagined and circumstantial ways. The name of the community alone is only one type of local identity; indeed, scholars have furthered Anderson‟s “Imagined Community” concept into a wide range of fields with individualized and circumstantial expressions of broadly defined identity and community (Arduser, 2011; Brennen & dela Cerna, 2010; Cover, 2005; Funk, 2012; Kennedy, 2010; S. C. Lewis, 2008; Pentzold, 2011; Reader, 2007; Richardson, et al., 2008). DICTION 6.0 likely would have missed most of those unique expressions of local identity, thus understating proximity.

For example, for a media audience in San Antonio, Texas, the word “Spurs” would clearly qualify as both prominence and specific proximity. The San Antonio Spurs are a prominent professional basketball team with a major following in the Alamo City. This reflects Anderson‟s (2006) theory of imagined community; through media coverage, a sports team and its name is now a clearly articulated aspect of local identity. DICTION 6.0, however, would likely be unaware of the Spurs. It would register “San Antonio” as specific proximity under this dissertation‟s methodology because “Antonio” would be included in the word dictionary for specific proximity; “Spurs,” however, would likely be absent. Including the name of every prominent team or institution in every city studied would quickly become overwhelming; furthermore, even with meticulous effort, the researcher could neglect a prominent institution. The Spurs are a national brand; the budding San Antonio Scorpions FC, a minor league soccer team, are not well known. Even if the researcher using DICTION included “Spurs” in the appropriate dictionaries, “Scorpions” would likely be omitted. Broader articulations of

15 ie, community weekly newspapers in A group (A.CW), community daily newspapers in B group (B.CD), and so on. See Chapter 3.4.d for more information.

120

Prominence, like “basketball,” “coach,” or “quarterback” would likely be included, however, thus potentially mitigating the loss and elevating the word dictionary‟s validity.

It is telling that quantitative analysis indicated significant variances and correlations concerning circulation size and specific proximity; however, given the circumstantial nature of identity and the volume of literature indicating the “relentlessly local” (Lauterer, 2006) nature of community journalism, it is possible that DICTION 6.0 understated the real relationships between circulation size and local content. This reinforces the claim that perhaps quantitative findings for prominence, conflict, oddity, and impact also are suspect. If one word dictionary is flawed and lacks contextual validity, could the other word dictionaries suffer the same shortcoming?

The researcher is reluctant to endorse this perspective. A great deal of time and effort was devoted to the construction of these word dictionaries, much of it quite tedious, and the introduced concept of generalizable validity (see Chapter 3.3.h) was designed to minimize potential validity issues within the word dictionaries. However, given the complications surrounding the news factor magnitude, the potential understatement of proximity, and the gulf between quantitative and qualitative results, it remains a possibility worth mentioning.

The computerized content analysis process may be too hyper-sensitive, and too non- contextualized, to properly study news factors. Another possibility is that the program‟s lack of context makes it more difficult, or even impossible, to study anything concerning proper nouns, like prominence or specific proximity; hyper-sensitivity is an advantage, but only concerning rhetoric and language, not concepts requiring contextual proper nouns.

5.4.C: GAP BETWEEN ONLINE AND OFFLINE COMMUNITY JOURNALISM The third potential influence is arguably the simplest. This dissertation studied online media exclusively, both newspapers and social media posts on Facebook and Twitter; however, the bulk of research on community journalism has focused on traditional print products (Dill & Wu, 2009; Gade & Krug, 2008; Garfrerick, 2010; Givens, 2012; Hansen & Hansen, 2011; Lauterer, 2006; S. Lewis, 2008; Rossow, 2009; Smethers, et al., 2007). Studies of online community journalism, or of community newspapers‟ web presence, are relatively uncommon (Funk, 2012; Gilligan, 2011; J. Greer & Yan, 2010; Hansen, 2007).

121

Several interviewed community newspaper editors and publishers had no web presence at all; more had a limited online presence, and social media in particular seemed under-utilized. Most also expressed a desire to develop their online and social media presences, but to date, many editors and publishers said they had a limited presence on the World Wide Web. Their print product remains a top editorial priority and the top avenue for advertising revenue.

It seems plausible, then, that editors and publishers were primarily referring to their print publications when discussing broad preferences for egalitarian news, either consciously or sub- consciously. The structured interview did not address potential content differences between community newspapers‟ print versions, websites and social media platforms; perhaps their news selections, or even emphases within selected news selections, would vary between print and the web. Furthermore quantitative analysis focused exclusively on online content, on newspaper websites and social media platforms, creating a potential disconnect between the quantitative and qualitative datasets.

Furthermore, although community newspapers were meticulously categorized by circulation size, and the size of community newspapers sampled in quantitative and qualitative data are likely highly similar, the quantitative data gave obvious priority to community weekly and daily newspapers with functioning websites and social media presences. This reduces the overlap between qualitative and quantitative datasets, potentially to a considerable degree.

Given the nature of computerized content analysis, studying traditional print newspapers would have been extremely difficult. However, in retrospect, it seems an oversight to not either focus more exclusively on interviews of editors and publishers with strong web presences or to simply ask how print, online, and social media content differ concerning deviance or egalitarianism. Perhaps community newspapers prioritize their edgier, or more deviant, news content on their websites while granting egalitarianism higher priority in print? Perhaps media sociological effects are more profound online? Or perhaps the act of establishing a strong web and social media presence, at a small community newspaper, somehow indicates broader loyalty to industry standards and media sociological patterns – standards and patterns which might be ignored at publications which also are largely ignoring the Internet, either by choice or because of a lack of resources?

122

This perspective argues that quantitative and qualitative data are accurate; however, comparisons between them are awkward because the datasets are largely dissimilar. Quantitative and qualitative data may simply be studying different media.

123

5.5: Reconsidering Normative Deviance

The fourth potential influence points to theoretical development and reconsideration for gatekeeping theory and normative deviance. It argues that quantitative and qualitative data are accurate and referring, essentially, to the same concepts. The discrepancy between quantitative focus on deviant news and qualitative focus on news on “regular people and routine events” relates to complexities within that phraseology. Although “regular people” can easily be considered egalitarian, “routine events” potentially trips both deviance and egalitarianism. A city council meeting might be considered “routine” by journalists, but would arguably constitute both prominence and conflict from an academic perspective.

Quantitative and qualitative results suggest theoretical development and reconsideration of normative deviance in two possible directions: deviance and egalitarianism as a spectrum, or deviance as a broadened concept. Visualizing the dynamic within a two-by-two typology is also helpful.

5.5.A: THEORETICAL DISCUSSION & DEVELOPMENT OF NORMATIVE DEVIANCE: A TWO

OPTION APPROACH The concept of normative deviance deserves reconsideration and reevaluation in light of quantitative and qualitative findings presented here. Gatekeeping theory (Shoemaker & Vos, 2009; White, 1950) and studies of deviance (Arpan & Tuzunkan, 2011; Boyle & Armstrong, 2009; Daley & James, 1988; Jong Hyuk, 2008; Paletz & Entman, 1981; Shoemaker & Danielian, 1991; Shoemaker, 1996) argue that journalists rely on news about deviant events, institutions, groups, and ideologies because of a lack of exposure to media audiences. The news, these theorists argue, is focused on physical and ideological threats to the status quo because of a basic sociological need to identify and avoid potential threats and predators.

To follow this logic, journalists are not responsible civil servants, but rather sensitive alarm systems blaring warnings whenever a potential risk or threat appears. The bigger the risk, the louder the barking. Quantitative and qualitative results presented here complicate this theoretical framework, however.

124

On the one hand, quantitative findings indicate that circulation size plays little to no significant role in American newspapers‟ clear preference of deviant news over egalitarian news. Although community newspapers are reliably, and statistically significantly, more focused on local communities on their websites and social media pages than larger publications, that local content was constructed with deviance very much in mind. This arguably speaks to the omnipresence of deviance a validation of gatekeeping theory (Shoemaker & Vos, 2009) and media sociology (Shoemaker & Reese, 1996). (See Chapter 4.2 for full quantitative results.)

On the other hand, these quantitative findings are at odds with the majority of qualitative interviews of community newspaper editors and publishers, who said that deviant news about “important people or unusual events” should occupy between 0 and 30 percent of community newspaper content. Instead, egalitarian news about “regular people and routine events” should occupy the lion‟s share of community news. (See Chapter 4.3 for full qualitative results.) These qualitative results arguably challenge both this dissertation‟s quantitative findings and previous assertions of a broad journalistic focus on deviance and deviant news.

Chapter 5.4 explores potential practical explanations for this discrepancy. However, practical explanations may not offer a valid response to the question of normative deviance; they also do little to enhance or improve gatekeeping theory or media sociology. Instead, perhaps a more fundamental theoretical exploration is in order. Shoemaker and Vos (2009) make two claims of normative deviance that seem at odds with findings presented here. The first:

“Laws and norms define the boundaries of the civilized world. Inside the boundaries is civilization, society as it is supposed to exist. The outside is deviance, a world full of norm and rule breaking, some minor and some fully evil. Events happening outside of the boundaries are more likely to become news items (Shoemaker, Chang, & Brendlinger, 1987; Shoemaker & Cohen, 2006). For example, the normal efficient and ethical work of government officials gets little media coverage, whereas a questionable or inefficient action can generate protracted public debate and many news items.” (Shoemaker & Vos, 2009, p25)

Concerning community journalism, this assertion is problematic and somewhat paradoxical. The first part of the first quote seems validated by quantitative findings (including this study) indicating newspapers‟ strong focus on deviant news, regardless of circulation size, on their websties and social media platforms; the second part of the first quote, however, is clearly

125

rebuked by this study‟s qualitative data demonstrating that community journalists do, very much, focus on “normal efficient and ethical work of government officials” (Shoemaker & Vos, 2009, p25). Interviewed editors and publishers clearly interpreted “news about regular people and routine events” to include routine government functions; local city council and school board meetings were mentioned particularly often. Community journalists said they would cover those routine events regardless of their level of excitement; a 20-minute meeting with three agenda items would get the same priority as a week-long budget debate or controversy.

This presents a theoretical paradox: How can community newspapers be reliable publishers of news on deviance and dedicated to giving covering routine events?

The second Shoemaker and Vos (2009) passage is similarly complicated:

“There is also a tendency for news items to be predominantly about celebrities and other prominent people. In the context of most people‟s lives, the lives of celebrities are deviant and therefore interesting to gatekeepers. Whereas information about your neighbor‟s cancer surgery may be important to you, it is probably not important to news gatekeepers. If people are prominent enough, however, even routine activities can leap tall gates and result in an astonishing number of news items.” (p25)

Community newspapers would clearly be interested in a neighbor‟s cancer treatments – arguably, such news is the bread and butter of the community newspaper industry. It exemplifies the kind of news on “regular people” that community newspaper editors and publishers said should constitute the majority of community newspaper content. However, once again, this argument runs aground against quantitative findings arguing that community newspapers are just as focused on deviance as larger newspapers.

This, too, presents a theoretical paradox: How can community newspapers keep a loyal, institutional focus on regular people while maintaining a published emphasis on deviant news?

On a theoretical level, these questions must be answered by re-evaluating normative deviance as a concept. On its face, routine city council meetings do not qualify as “deviance” if the meetings themselves are not somehow conflicted or controversial; however, as a concentration of political power, even small and routine city councils qualify as collections of

126

prominent individuals. Why would a routine meeting not, then, qualify as deviant? If prominence is an indicator of deviance, should not even local and predictable prominence qualify?

Similarly, why would cancer not qualify as a predatory threat to the status quo? The American Cancer Society estimates that 580,350 people will die of various cancers in the United States in 2013, with as many as 1,660,290 new cases developing the same year. It is the second most common cause of death in the United States; only heart disease is more prevalent (Society, 2013). Why should its frequency discount its potential threat? And why, theoretically, should its effect on a “regular person” reduce its deviant potential?

Thus, conceptual normative deviance needs to be re-evaluated or redefined. This dissertation proposes two theoretical adjustments. The first views deviance and egalitarianism as part of a spectrum rather than mutually exclusive constructs. The second broadens and expands normative deviance.

5.5.B: THE SPECTRUM OPTION: DEVELOPING AN INVERSE RELATIONSHIP BETWEEN

DEVIANCE & EGALITARIANISM One option for theoretical development of normative deviance is to reconsider the relationship between deviance and egalitarianism. On the one hand, quantitative results indicate that community newspapers, on their websites and social media platforms, are equally focused on deviant news as large newspapers; on the other hand, qualitative results indicate a clear preference for egalitarian news over deviant news among community newspaper editors and publishers. Perhaps, theoretically, both are valid – community newspapers do focus on deviance, and do give priority to “regular people and routine events,” but the relationship between deviance and egalitarianism is not mutually exclusive.

Perhaps community newspapers can effectively focus on both because they are effectively opposite sides of the same concept. Perhaps a better operationalization of normative deviance includes the proposed concept of normative egalitarianism:

Normative Deviance: Behavior, ideas, groups, or events are deviant when they break social rules or norms.

127

Normative Egalitarianism: Behavior, ideas, groups, or events are egalitarian when they confirm to social rules or norms.

If the concept of normative deviance is to remain static, then a complimentary and inverse concept of normative egalitarianism is an appropriate conclusion. Normative deviance has a long scholarly history and has perhaps earned its place in communication theory; however, at least according to quantitative and qualitative findings here, normative deviance offers only partial predictive power concerning the news agendas of community newspapers.

A new concept of normative egalitarianism could serve as a related predictor of the remaining news agenda, and a spectrum-based relationship between inverse concepts is arguably more comprehensive than a dichotomous conceptualization of normative deviance. News is not simply deviant or non-deviant; instead, news content is driven by individual editorial choices and philosophies falling between the two concepts.

Community journalism philosophies, at least according to these data, typically give clear focus to local communities and local individuals, but also clearly emphasize deviance in news about those local communities and individuals. Quantitative and qualitative data indicate a significant preference for news about local communities; although qualitative and quantitative data differ on editorial preference for deviant and egalitarian news, perhaps those distinctions are related. If normative deviance and normative egalitarianism are on a spectrum, and relate to a news item‟s subject matter and its coverage, then perhaps community newspaper editors simply prefer deviant news about egalitarian subjects. They fall at a particular midway point on the Deviance-Egalitarian spectrum that is customized for the community journalism business model. Perhaps editors and publishers at larger media, other media platforms, or other nations would utilize a different combination of normative deviance and egalitarianism; indeed, quantitative data suggest that metropolitan and national newspapers in America are much less concerned with covering local communities, thus suggesting a position on the Deviance-Egalitarian spectrum much closer to the deviant end. Deviant news about deviant subjects is obviously in a different point on a spectrum than deviant news about egalitarian subjects.

The Deviance-Egalitarian Spectrum: A spectrum, visualized as a quantitative range between one and 10, which considers normative deviance and normative egalitarianism as related and inverse concepts. Journalistic choices between deviant and egalitarian news 128

topics, and deviant and egalitarian coverage of those topics, reflect a particular news outlet‟s position on the spectrum.

And, indeed, Shoemaker (1984) posits deviance as a spectrum more than binary variable and also notes that newsworthiness assessments, including deviance and potentially including egalitarianism, are only one facet of dynamic news production cycles (Shoemaker & Reese, 1996; Shoemaker & Vos, 2009). The introduction of normative egalitarianism arguably strengthens the conceptualization of normative deviance, and offers the full theory more predictive power.

This may also support Shoemaker‟s (1996) assertion that news audiences are attracted to deviance on a sociological level – media consumers living in relatively safe, isolated communities are sociologically unaccustomed and un-attuned to patrolling for threats or predators. Community journalists are adapting to community interests and to practical reality. This also is consistent with research on structural pluralism and the community structure model, which argues that journalists are influenced by community demographics (Donohue, et al., 1985; Donohue, et al., 1989; L. Jeffres, et al., 2011; Nah & Armstrong, 2011; Pollock, 2007), and could theoretically be expanded to consider the presence or lack of threats or triggers for deviant news.

5.5.C: THE EXPANSION OPTION: BROADENING OPERATIONALIZATIONS OF NORMATIVE

DEVIANCE A simpler explanation, conversely, is that the concept of normative deviance may be structurally and theoretically sound and simply poorly executed, with too narrow a focus on fringe groups and violent conflict.

Operationally, normative deviance is structured fairly broadly. News events that “break social rules or norms” cover a wide swath of theoretical ground, given how broadly social rules and norms can be considered. But in practice, the bulk of research on deviance implies that deviance is contingent either upon a dramatic or violent break from society writ large, or upon massive or nationwide celebrity. Shoemaker and Vos (2009) stipulate that routine functions of government and a neighbor‟s cancer treatments are simply not deviant enough to qualify as deviant. Instead, aggressive or violent protest rhetoric or imagery qualifies as deviant (Arpan & Tuzunkan, 2011; Boyle & Armstrong, 2009), just as news on the Ku Klux Klan or Nazi Party (Shoemaker & Danielian, 1991) and papal molestation trials (Breen, 1997) clearly threaten social 129

rules and norms. If this is true, or if this is theoretically sound, then either community newspapers are simply not deviant enough to be considered deviant, or the theoretical spectrum between deviance and egalitarianism is necessary to strengthen the theory.

The question is not if angry mobs or military invasions are deviant. Obviously, unquestionably, such news fits the definition of normative deviance. The question is if such content comprises the entirety of deviance, or simply the realm of extreme examples.

There remains the possibility that normative deviance is broader in theory than in practice. It depends on how “social rules and norms” are operationalized. Even in a small town, the average resident is not likely to visit many city council meetings or school board functions. Officials such as mayors and local business leaders qualify as political elites just as minor league pitchers and high school quarterbacks qualify as athletic elites. The difference between a small- town mayor and the mayor of New York City, or the gap between the pitcher for the minor league Round Rock Express and the major league New York Yankees, is a question of magnitude rather than philosophy.

If deviant news is concerned with any news content which feels foreign to the standard, routine, nine-to-five lives of the average resident, then for most Americans the local city council chambers are almost as unfamiliar as the halls of the Kremlin or the United Nations. Exposure to local institutions, while theoretically highly tangible, is simply not taken advantage of on a massive scale. Furthermore, as far as potential threats to an individual‟s status quo are concerned, the financial and staffing decisions made by local school boards have arguably far more influence over local residents than congressional actions.

It does not take a colossally deviant event to intrude on daily life. Potentially, any routine act of government could disrupt social rules and norms to one degree or another.

Community newspaper editors and publishers, as such, cover those routine political meetings and local political institutions on a regular basis – so much so that interview participants on the whole considered city council meetings routine events, not unusual ones. The focus on government routines can be considered highly regular and egalitarian; indeed, this study‟s qualitative analysis utilized such a rubric in the interest of consistency with current notions of

130

normative deviance. However, it seems plausible that community newspapers focus on routine government meetings and officials precisely because they are out of the ordinary for regular citizens, and because local government decisions carry so much weight – because, in effect, they are deviant despite being highly routine. If egalitarianism therefore cannot be simplified to “routine events,” the notion of deviance becomes considerably murkier.

The same can be said for news on regular people, which was considered a clear priority by all qualitative interview participants and the top priority by a great many. Regular lives are filled with challenges to personal safety and a person‟s individual status quo. Articles written about “regular people” are not simply factual depictions of personal routines, they are stories focused on unique residents facing relatable challenges. There is no logical connection between a lack of political or celebrity stature, on the one hand, and a dedication to bland or typical existence on the other.

Returning to the bear-watching caveman metaphor, an expanded understanding of normative deviance seems logical. If journalists know the authority vested in a government institution, they should be more likely – not less – to cover judiciously their every meeting and action to be on the lookout for threats to the status quo. Similarly, if journalists are not preoccupied with political or celebrity status, they can cover any number of threats to individuals that can be related to the whole community. As far as frightening predators are concerned, frankly, cancer can be among the scariest – regardless of the political stature of the victim.

Perhaps the metaphoric, vigilant caveman simply has a more comprehensive understanding of deviance, threats and predators. Instead of looking simply for the dramatic threats, the journalist is cognizant of the routine and personal nature of many deviant events. This seems to fit with the conceptual definition of normative deviance, and offers a plausible alternative to the fringe-centric focus of past deviance studies.

5.5.D: VISUALIZING DEVIANCE, PROMINENCE, AND CONFLICT In many ways, deviance can be considered a combination of prominence and conflict. The scale of that prominence includes a wide range of local executives and celebrities important to a

131

particular community, but in theory, news focuses on the overlap between the two concepts (See F7: Deviance Typology of Prominence and Conflict).

132

F7: Deviance Typology of Prominence and Conflict

F7: Visualization of deviance using a two-by-two typology. The X axis refers to conflict in events and the Y axis refers to the level of prominence. News within quadrant 1, of prominent people and conflicting events, is the most deviant; news falling in quadrants 2 and 3 are also deviant based upon either conflict or celebrity. Only quadrant 4, regular people and non- conflicting events, has low deviance and is rarely considered news.

133

This visualization expresses deviance in terms of conflict (the X axis) and prominence (the Y axis). News which falls into the first quadrant, concerning prominent people and conflicting events, is the most deviant and would certainly grab headlines. The types of prominence and conflict vary, from elections to athletes to celebrity awards, but the central principles of prominence and conflict apply in each case. The second quadrant, concerning prominent people and non-conflicting events, focuses more on prominence and celebrity than conflict. These are events which are cordial and friendly, or even dull, but revolve around prominent individuals. In a small town, a local Rotary Club lunch is an excellent example; small community elites are often members of civic groups, and even if the agenda for a particular meeting is short or uneventful, any gathering of mayors and judges and business leaders constitutes deviance and often earns news coverage.

The third quadrant, concerning regular (non-prominent) people and conflicting events, also constitutes deviance. This refers to conflict which occurs or envelops ordinary Americans; local teachers going on strike would qualify, as would a charity drive for local cancer patients. As always, conflict can manifest itself in a variety of ways, including sports or achievements; a high school graduation would fall into this category, for example, given the difficulty and achievement inherent in passing high school.

In each of these quadrants and cases, prominence and conflict are broadly defined by design. Obviously, international warfare is not the same kind of conflict as a student struggling to graduate; however, from a theoretical perspective, both represent deviance and both offer threats (of varying severity) to the status quo of the average reader. Most media consumers can relate to the conflicts surrounding academic achievement and can identify the threats to their personal status quos should a child, friend, or relative not graduate high school; similarly, media consumers can imagine the personal and political threats presented by an invading army.

Only in the fourth quadrant, concerning ordinary people and non-conflicting events, does deviance not apply. These events fall purely within the realm of the average nine-to-five and do not constitute threats to any status quo, personal or political; they are routine and ordinary. In this example, Jane Doe mowing her lawn is in no way an expression of deviance. Jane Doe has a lawn; for most media consumers, this is not unusual. She must mow it, often frequently in the

134

summers; this, too, is relatable and routine. Mowing a lawn usually is an easy chore with nothing special about it, and the hypothetical Jane Doe is not a prominent individual. The entire image is wholly pedestrian, and because it does not relate to prominence or conflict, it is not news.

It could potentially become news should additional events occur, of course. If, while mowing her lawn, Jane Doe discovers a Cold War-era bomb shelter buried beneath her flower bed, the news event would migrate to the third quadrant – a regular person encountering a highly unusual event rooted in conflict: in this case, worries over nuclear fallout. If American soccer star Landon Donovan or Oscar-award winner Anne Hathaway were to drive up and offer to help Jane mow her lawn, the news event would shift to the second quadrant of prominent individuals encountering non-conflicting events. And if Jane lived in a drought-prone area and became frustrated with the state of her lawn, she may chose to run for city council with hopes of changing the municipal water rationing schedule or procedures; this decision would catapult the event into the first quadrant, with Jane shifting from a non-prominent individual to a prominent electoral candidate and the event changing from a casual Saturday lawn mowing to a conflict-drenched election.

This typology is a useful visualization of deviance. It emphasizes the importance of a broad understanding of deviance while reiterating the centrality of deviance in news construction.

135

5.6: Reconsideration of Journalist-Audience Interaction and Gatekeeping Theory

Although similarities and differences between quantitative and qualitative results presented here need some reconciliation and consideration, and the concept of normative deviance needs theoretical reconsideration, there is one theoretical stipulation of gatekeeping theory that can be accurately addressed and largely dismissed.

Qualitative data presented here clearly rebuke the stipulation that “Individual journalists have modest exposure to their audience” (Shoemaker & Vos, 2009, p52). The overwhelming majority of structured interviews indicate that community newspaper editors and publishers have a great deal of contact with their readers and community. Editors and publishers reported an open door policy that led to, in most cases, visits every day from readers interested in discussing news and news coverage. Levels of online and social media interaction were typically less, but this is understandable since many community newspapers‟ web presences are underdeveloped or underutilized.

Furthermore, editors and publishers said they perceived a remarkably high level of interaction with readers and community members. They craft editorial content with that perceived interaction in mind, not in spite of it. This is entirely consistent with a variety of previous studies of community newspapers (Bowd, 2011; Brockus, 2009; Cover, 2005; Garfrerick, 2010; Hansen, 2007; Hansen & Hansen, 2011; Lauterer, 2006; Reader, 2006; Smethers, et al., 2007). Furthermore, community newspapers occupy a statistical majority of American newspapers (Lauterer, 2006). Only a handful of media have nationwide or international reach. Those media may potentially craft news content in isolation from media audiences, but these data compliment previous research findings that the majority of American newspapers have incredible interaction and dialogue with readers.

Here, gatekeeping theory makes an invalid assumption. Community journalists, at least, have a highly interactive relationship with readers. This does not invalidate the heart of gatekeeping theory or the hierarchy of influences model by any means, but it does indicate one clear assumption which needs revision. 136

Journalists, as a rule or standard practice, simply do not craft editorial content in isolation from their audiences. This may vary from media to media, but the broad generalization is clearly invalid. At least community journalists have far more than “modest” exposure to their audiences.

137

5.7: Implications for Journalists and Publishers

These findings also merit consideration from a journalist‟s and publisher‟s eye-view. Data presented here speak fundamentally to the core question of academic research into journalism: What is news? News is deviance, content about ideas, people, or events which fall outside the typical realm or paradigm of the average reader. What might the omnipresence of deviance in American journalism mean for the average journalist? And, concerning community media in particular, how does the spurious relationship between circulation size and deviance impact the small newspaper‟s editorial mission and business model?

Clearly, community media remain “relentlessly local” (Lauterer, 2006); however, quantitative and qualitative data presented here demonstrate that localness is effectively the only alteration in community journalism‟s mission. Put another way, community newspapers seem to be simply newspapers in every other capacity –focused on deviance, on prominence and conflict, even if their definitions of prominence and conflict are potentially broader than conceptualizations of the same ideas at larger publications. Community newspapers are not reinventing journalism; in a sense, they are simply adapting institutional journalistic standards in one key dimension. Focusing exhaustively on local news gives community newspapers a marketable identity, but the focus does not change the nature of the news content itself; once these papers establish localness in news, the remaining characteristics of community newspaper content are largely consistent with mainstream journalistic practices.

Even in a small community, then, journalists resemble threat-watching cavemen more than responsible civil servants; publishers are in turn encouraging that caveman approach even as new media technologies offer new opportunities for media businesses. The primary goal remains the preservation of the community‟s physical and ideological status quo. Events, ideas, or individuals who present a prominence- or conflict-based threat of change to that status quo, to one degree or another, become news. The community journalist utilizes the same news factors, the same news judgment, and the same news instincts as news producers in larger markets; the community newspaper publisher builds a business around that traditional focus on deviance, and seems committed to staying the course on developing online and social media platforms.

138

This affirms what Burd (1979) called the inseparability of community and media. If the association of physical geography and community among small weekly and daily newspapers does little to influence editorial focus on deviance, then it seems logical to assume that media catering to other communities would also merge an overall focus on deviance with a niche focus on their community – be those communities physical or ideological. It is clear from this data that newspapers serving small and large communities share an equal focus on deviance; online or niche media may well adopt the same perspective. A magazine catering to college students, for example, may have an editorial mission dedicated to a particular 18-22-year-old demographic; beyond that, however, data imply that news content would be devoted to persons, ideas, and events which deviate beyond the typical routines of those college students. Articles concerning watching college football games at home would not be common, but a guide to obtaining 50- yard-line seats could be prominent. Similarly, online news catering to computer programmers would be devoted to deviations from typical computer programming routines – an article about an advanced new processor, for example, or news about new cyber-security initiatives or emboldened cyber-attacks.

Data imply that the kind of community, and indeed the variety of journalist and publisher, is irrelevant. If a focus on deviance does not threaten or conflict with the editorial mission of traditional community journalism, and does not infringe on community newspaper‟s exhaustively local focus, then it seems logical to assume that deviance-centric content would fit with journalism focused on other kinds of community.

That, in turn, indicates that all media are primarily focused on deviance – that all media content are devoted to exceptions to daily routines, be they physical or ideological. The nature of journalism itself is devoted first to deviance and second to whatever communities or subject matters that media cover. This offers an answer to the critical question of journalism studies. What is news? News is deviance.

This can be a benefit for the average journalist, and the average publisher. It demonstrates that adopting a niche focus compliments and reinforces journalistic focus on deviance; in the case of community newspapers, that niche focus on local geography has proven reliable and profitable. A niche focus distinguishes community newspapers from other media without challenging

139

industry-wide journalistic practice, and a variety of media potentially can adopt that dynamic between niche focus and deviance-centric news coverage and apply it to a number of diverse communities.

These are deliberate choices by publishers who are identifying new target audiences and new media platforms, potentially to great effect; however, the decision to retain that traditional editorial focus on deviance is very much by design, and seems independent of dialogue with media audiences or the interactive nature of new media. In an era where anybody can engage with media publishers and journalists, or even take up the pen online themselves, the innate journalistic focus on deviance remains a clear and unflinching element of media business models and journalistic practice.

This is in some ways a very empowering finding for journalists and publishers struggling with a rapidly evolving media ecosystem. Newspapers face a variety of challenges, many of which were not measured by this dissertation; obstacles like market competition, fluctuations in the national and local economy, and finding the right balance between local and non-local editorial content all influence a newspaper‟s content and business model. These issues are compounded as media continue to migrate toward online and social media, and as those are eventually replaced by the next technological innovation.

Ultimately, journalists and publishers may find comfort in the knowledge that the core principle of news – deviance – is unchallenged by divergent and evolving media platforms, and can in fact be complimented with a potentially profitable niche focus on a community of one kind or another.

140

CHAPTER 6: CONCLUSIONS & OPPORTUNITIES FOR FUTURE RESEARCH

6.0: Introduction

This dissertation has made major contributions to the study of deviance, gatekeeping theory, community journalism, and online and social media. Data indicating that circulation size plays no significant role in the use of deviance on newspaper websites, and little significant role on social media feeds, provide statistical evidence that the effectiveness and predictive power of deviance, as a theory, is vastly understated.

The concept is simple: Journalists are cavemen, worried and staring out onto the veldt for potential threats, dedicated to guarding the physical and ideological status quo of their tribe or society. As creatures of habit, cavemen learned that anything beyond the most ordinary of routines registered as a deviation from safety and security – and, thus, a potential threat to be broadcast to the tribe. Certainly, not all deviations are created equally; a “yard of the month” award is not the same kind of deviance from routine as a violent crime spree or government shutdown, but the basic principle is the same. Media are about exceptions to daily life, not a reflection of reality. Journalists are sociological safeguards for security and preservation, not agents of civil service or democracy. This is consistent with past journalism studies of deviance (Boyle & Armstrong, 2009; Daley & James, 1988; Gans, 1979; Shoemaker, 1984; Shoemaker & Vos, 2009; White, 1950). Shoemaker (1984) older studies of “social control” (Lauderdale & Estep, 1980; Miliband, 1969; Paletz & Entman, 1981), and quantitative and qualitative findings presented here. This also resonates with popular psychological understanding about anxiety. Returning to Wilson and Dufrene‟s (2010) anxious caveman is helpful here:

141

Arguably, mankind remains perched at the edge of the cave, staring off at vague shapes that may or may not be threats. The caveman has evolved, however, and developed powerful tools for mass communication (the Internet, Facebook, and Twitter) which can almost instantaneously distribute breaking news about the blurry shape on the horizon. The entire tribe – today an entire nation, or potentially the whole world – can be transfixed by real-time updates and reports on the blurry shape, or #Beargate2013. This human behavior has changed little over the centuries, it seems, even as our technology has advanced dramatically.

Quantitative data presented here indicate that a wide range of American media, across diverse dimensions of circulation size and concerning news websites and the most popular social media on Earth, remain predominately preoccupied with news about deviance. Prominence and conflict remain the driving factors in news construction, even among hyper-local community newspapers which have a well-documented reputation for local news (Bowd, 2011; Funk, 2013; Givens, 2012; Hansen & Hansen, 2011; Hume, 2005; Lauterer, 2006; Martin, 1988). It is not surprising that quantitative data indicated that circulation size is related to geographic focus; in fact, this is consistent with the editorial mission and business model of community journalism. It is surprising, however, that news content in hyper-local community newspapers as focused on deviance as national media like The New York Times.

The weekly, hyper-local WeeklyObserver is certainly concerned with local news about its home in Hemingway, South Carolina; The Los Angeles Times, conversely, is less concerned with news exclusively about Los Angeles and more devoted to major national and international news, patterns, trends, and events. Once that local focus is accounted for, however, there are no significant differences between the two concerning the remainder of the editorial content.

Put another way: If quantitative analysis had not measured for specific proximity and had instead been solely concerned with the use of deviant news factors, then the news content in The Weekly Observer and The Los Angeles Times would yield statistically identical results. Thus, the only important difference between community journalism news content and national journalism news content is a focus on localness – the “community” may generate differences in newspaper business models, but a pervasive industry standard on deviance clearly defines the “journalism” part, regardless of a publication‟s circulation size.

142

The size of the metaphorical cave, in a sense, makes no difference. A caveman-journalist in a smaller tribe may concentrate on bear-watching in a smaller physical area than a caveman- journalist in a larger tribe, but the instinctive bear-watching is identical in both cases. As mentioned in Chapter 5.7, the nature of the cave is also irrelevant – a newspaper‟s “community” may be a small geographic region, a particular ideological niche, or a specific audience demographic, but in every case a predominant focus on deviance within that community is almost certain to dominate the publication‟s editorial content.

Similarly, qualitative data illuminate the pervasiveness of deviance. At first glance, the qualitative interviews clash with the quantitatie analysis; community newspaper editors and publishers repeatedly and reliably expressed a priority for news about “regular people and routine events.” Closer analysis, however, reveals that even these news items are focused on deviance. Editors and publishers routinely mentioned “routine events” like City Council meetings and bond elections which are clearly deviant from the daily routines of average readers, even in a small town; furthermore, particularly in today‟s polarized political climate, most Americans would agree that political elites and government actions or inactions represent structural challenges to the status quo. Focusing on those “routine” events reveals a tacit reliance upon deviance, on prominence and conflict. Too, even the “regular people” highlighted by community newspaper editors and publishers are not truly average. High school quarterbacks are not trule “average,” for example, and ordinary neighbors who make the news do so because of a noteworthy event or action.

One editor mentioned that “we love hundred-year-old ladies” – a population which clearly constitutes celebrity in a small town, let alone prominence and cultural clout within the community. News about centenniels is clearly deviant, as is news about any and all prominence and conflict.

Exploring these findings in isolation has clear merit, and both have profound theoretical value to the academic study of journalism. The remaining chapters in Chapter 6.0 will explore findings using a more compartmentalized framework, effectively looking at this dissertation‟s major contributions more individually. However, questions raised by academic research are often

143

more important than the answers themselves. As such, two broad sets of questions are implicated and posed here which speak fundamentally to human communication and behavior.

The first series of questions concerns human behavior. As Chapter 5.7 and others outline, all media share a focus on deviance. Circulation size, surprisingly, plays little significant role in the formulation of deviant news content. Newspapers‟ websites and social media are excellent examples of mass communication. They can potentially reach an audience of thousands in a matter of seconds. If such iconic mass media are fixated on deviance, then, there is a clear implication that all human communication – if not all human behavior – is focused on deviance, on deviations from regular routines, and on pathological threat identification and assesment. This possibility is underscored by the ineractivity of both community journalism and social media. Community newspaper editors have a great deal of interaction with audiences, readers, and community members; Facebook and Twitter are intrinsically highly participatory. Deviance remains the driving force behind both community journalism and newspapers‟ social media feeds, however.

This begs the question, “Is all human communication rooted in deviance?” Or even, “Is all human behavior rooted in deviance?” If all media are a sophisticated method of monitoring deviant behavior and potential threats, then would lower-tech human behavior not be devoted to the same caveman instincts? And, what might constant and consistent threat-watching indicate about human behavior as technology and societies continue to evolve? As a journalism-centric analysis, this dissertation cannot answer these questions. Doing so would require rigorous study rooted in psychology, sociology, and potentially even neuroscience. But this dissertation can, and has, identify the clear potential for meritorious study.

The second line of inquiry concerns language and identity construction. As mentioned in Chapter 2.3, a certain degree of ambiguity surrounding the term “deviance” complicates potential analysis; it has a popular connotation of sex offenders, radical ideologies, and fringe groups. These associations are not accidental, nor are they impertinent. As studies of linguistics (Candel, 2013; Enghels & Jansegers, 2013; Goddard, 2012; Regier, Khetarpal, & Majid, 2013; Shariq, 2013) indicate, words are effectively the building blocks of identity and communication. Associations between words and ideas are powerful and relate directly to how humans percieve

144

and judge the world. Most of those words are interpreted and applied subjectively to help formulate, define, and reinforce personal and political ideologies.

“Often, it seems, we are less interested in „facts‟ and more interested in what we say about them, and consequently, it is the talking about them that becomes important. We refine upon the so-called facts, modify and color them, put them on condition, and make them provisional.” (Embler, 1962, p223)

On a practical level, this can be easily identified and overcome. As mentioned in Chapter 2.3, although the word “deviance” has narrow connotations in popular culture and everyday usage, the journalism studies‟ operationalization of normative deviance is tangible, easily explained, and potentially omnipresent in American media. This is useful for social science research, but ultimately says little about the linguistic or cultural implications of the term “deviance.”

It is telling that a word which speaks to the heart of American media – and potentially all human behavior – is laden with connotation, alternative definitions, and largely negative implications. Using popular definitions of “deviance,” deviant behavior is considered truly fringe or threatening. It is not a broad behavioral construct, but instead a true aberration from normal routines. This speaks to Embler‟s (1962) perspectives on linguistics and identity construction. If “facts” are subjective and unimportant, and humans use words and language to reinforce their ideologies and world views, then the omnipresence of normative deviance in media may underscore its prevalence in human communication. Humans‟ instinctive bear-watching, that innate compulsion to monitor surroundings for threats and change, may itself be so pervasive that the word itself becomes understated or exaggerated.

Put another way: Clear, extant threats may be so toxic to this threat-watching instinct that they are categorized as “deviant,” a label reserved for vile acts or fringe groups despite the fact that it may speak to the whole of human behavior. Preserving the status quo is so critical that terms for breaking it become associated with profound negativity. If facts are relative and perception is everything, as linguistics studies argue, then associations between an instinctive behavior (watching for deviance) and the embellishment of the word “deviance” to clear and present dangers underscores the prevalence of threat-watching and broad monitoring of deviance

145

in human behavior. How, and why, the term “deviance” became so narrowly focused then becomes an intriguing question for scholars of linguistics and language.

If all media are devoted to deviance, and potentially all human behavior is devoted to monitoring deviance, then perhaps the fullness of human languages are constructed to reinforce these behaviors? Perhaps the metaphoric caveman is not simply watching the veldt for bears – perhaps every word in his vocabulary is devoted to either threat-watching or preserving the status quo. If so, why? And has human language truly evolved beyond troglodytic grunts and gestures of warning? This media-centric dissertation cannot answer these questions, which require rigorous analysis in rhetoric, communication studies, psychology, anthropology, and sociology. But the implications from data presented here are clear and intriguing.

The remainder of Chapter 6 is broken into three parts. Chapter 6.1 returns to points made in Chapter 1.3: Importance and Significance of the Study and evaluates each of 1.3‟s major claims in light of the data analysis and results. Chapter 6.2 further expounds on the methodological and theoretical implications of this study, particularly concerning items developed or discovered during the analyses.

Chapter 6.3 explores opportunities for future research. It returns to ideas and issues highlighted in Chapter 5.4 and 5.5. The concepts identified as possible explanations for discrepancies between quantitative and qualitative results offer plain direction for future research; indeed, each potential explanation is deserving of independent study.

146

6.1: Evaluating the Importance & Significance of the Study

Chapter 1.3 identified three major avenues of exploration for this dissertation. Subsequent data analysis addressed those points.

This dissertation very much broadened and deepened the study of deviance. Computerized quantitative analysis indicated that circulation size plays no significant role in a newspaper‟s editorial focus on deviance (prominence, conflict, and oddity) concerning web news and little significant role on a newspaper‟s Facebook and Twitter feeds. This is a strong affirmation of gatekeeping theory and the work of Pamela Shoemaker and others who argue that mediated deviance is a psychological device used to identify real and imagined threats to the status quo (Arpan & Tuzunkan, 2011; Boyle & Armstrong, 2009; Jong Hyuk, 2008; Pontell, 2007; Shoemaker & Danielian, 1991; Shoemaker, 1996).

Based on these findings, it seems clear that media are less as a watchdog and more as a caveman. The preservation of the tribe, so to speak – the physical and ideological status quo of American newspaper readers – is the primary objective of newspapers of any size. Quantitative data clearly reflect a focus on deviance. Qualitative data also indicate a professional focus on prominence and conflict; the majority of interviews said that important and deviant events, such as government meetings or violent crime, would always be covered. Furthermore, this institutional focus on deviance is in no way circumstantial – newspapers of every circulation size, from tiny hyper-local weeklies to newspapers with national and international circulation, plainly prioritize being a sentry over a watchdog. This is a major finding.

That social media, Facebook and Twitter, also adopt a clear focus on deviance also is significant.

147

It is worth noting that circulation size does influence egalitarianism. Quantitative data identified a significant inverse relationship between circulation size and proximity, or expressions of local place; as mentioned in Chapter 5.4.b, that significance may even be understated. Qualitative data also indicated a clear focus on local news among community newspaper editors and publishers. This is highly consistent with vast the majority of literature on community journalism, which emphasizes the “relentlessly local” (Lauterer, 2006) focus of community journalism (Funk, 2013; Garfrerick, 2010; Hansen, 2007; Hansen & Hansen, 2011; Reader, 2006; Smethers, et al., 2007). Most community journalism literature, however, does not directly address other aspects of the news content itself. A few studies have explored how community newspapers address crisis (Dill & Wu, 2009; Hansen & Hansen, 2012); these studies argue that community newspapers address crisis seriously and professionally but retain their predominately local focus. Community newspapers cover local crisis in much the same way that larger newspapers cover crises. This, too, is consistent with data presented here. Quantitative and qualitative data indicate an immutable emphasis on local communities among community weekly and daily newspapers; however, once that local criteria are met, the remainder of the news adopts a major focus on sentry journalism and deviant events.

As mentioned in Chapter 5.4 and 5.5, there is some discrepancy between quantitative and qualitative data. This researcher argues that the theoretical development explored in Chapter 5.5 accounts for this discrepancy.

All media are focused on deviance – differences are simply a matter of degree. For some communities, violent crime or high-profile sex scandals constitute deviant events; for others, an exemplary Yard of the Month award qualifies as an out-of-the-ordinary occurrence. Although it sounds strange to equate violent crime with horticultural awards, both events are in the news because they are unique. They are unusual, they are noteworthy, they stand out from the daily routines of readers in one way or another and are thus newsworthy. The scale of that uniqueness is irrelevant, but the nature of that uniqueness is of critical interest to media studies.

148

applying news factors to divergent American media validates the news factor approach generally while also expanding it into less-examined American media.

This study affirmed the use of news factors as a barometer for deviance (Shoemaker & Vos, 2009); it also introduced news factors as a barometer for egalitarianism. Deviance itself can be an abstract concept; identifying tangible operationalizations always benefits research. In particular this study develops the Bridges and Bridges (J. Bridges, 1989; J. Bridges & L. Bridges, 1997) model into new territory online, in social media, and concerning computerized content analysis.

The use of DICTION 6.0 also is noteworthy here. The use of news factors fit well with the evolving use of computerized content analysis; the framework provided salient objectives to measure and clear ideas to conceptualize and operationalize. This is a major asset for computerized content analysis research, for which specific research goals and clear operationalizations of terms are absolutely critical (Conway, 2006; Krippendorff, 2004; Nuendorf, 2002). News factors could easily be used for future computerized content analysis studies. As addressed in chapter 5.4.b, news factors which are not predicated on proper nouns (e.g., conflict or oddity) are potentially more valid than news factors requiring specific names (such as specific proximity); news factors without proper nouns also are certainly easier to program. However, with or without proper nouns, the methodology is effective and a prime tool for future analyses.

The cornerstone of the community newspaper industry is a dedicated, in some cases almost fanatical, focus on local events, places, ideas, and individuals. This is consistent with the lion‟s share of literature on community journalism (Bowd, 2011; Burroughs, 2006; Hansen, 2007; Hansen & Hansen, 2011; Lauterer, 2006; Rossow, 2009; Schweitzer & Smith, 1991; Smethers, et al., 2007); it also is consistent with quantitative and qualitative data presented here.

149

The ideal adjective to describe community journalism is local. If the business model were to be condensed into one concept, it would be local.

Business models rarely can be boiled down to simply one word, however, and it is telling that beyond that localness, community newspapers differ little from metropolitan and national newspapers as to deviance. Although the scale of the deviance itself might vary from publication to publication or issue to issue, it seems clear that an expanded concept of deviance remains the driving force in large and small newspapers. Sentry journalism is the primary focus of editorial content.

This acknowledgment of community journalism deviance, too, furthers the theoretical expansion of community journalism study. Traditionally, the study of community journalism has focused on suburban and rural American newspapers; certainly, these small publications meet the community journalism criteria. They do not have dominion over the term, however; particularly in the internet age, “community” is a highly malleable term. As Benedict Anderson (2006) noted, all communities are “imagined;” thus, the study of any niche media effectively qualifies as community journalism. A number of studies explore niche media (Arduser, 2011; Brennen & dela Cerna, 2010; Cover, 2005; Kennedy, 2010; Pentzold, 2011); this study broadens the definition of community journalism, and in so doing supports the inclusion and evolution of niche media studies within community journalism studies.

150

6.2: Summary of Major Contributions

With the importance and significance of this study, the interest of clarity suggests the major practical and theoretical contributions of this study be directly stated.

1. Newspapers craft news content based on deviance. Behavior, ideas, groups, or events which break social rules or norms are the predominant focus of American newspapers across circulation categories (from hyper-local weeklies to national dailies) and new media platforms (websites, Facebook, and Twitter). This is a major affirmation of gatekeeping theory (see chapters 4 and 5).

2. The theoretical concept of deviance, and its practical study, requires expansion.

a. A spectrum approach considers normative deviance and normative egalitarianism as opposite ends of a range. News content varies somewhat and falls between the two (see Chapter 5.5.b).

b. A broadening approach expands “social rules and norms” in the normative deviance operating definition to include any broad rule or norm, from the dramatic to the routine (see Chapter 5.5.c).

c. A two-by-two typology explores the roots of deviance, prominence and conflict, and the combinations which lead to news coverage (see Chapter 5.5.d).

3. American community journalism, online and on Facebook and Twitter, remains relentlessly local (see Chapter 5.3).

4. The news factor approach is a strong operationalization of deviance and egalitarianism. It also works well with computerized content analysis software (see chapters 3 and 4).

5. Computerized content analysis is a powerful tool which can effectively analyze theoretical concepts in media sociology (see chapters 3 and 4).

151

6.3: Opportunities for Future Research

Chapter 6.0 identifies a pair of meta-level opportunities for future research related to human behavior and communication. Chapter 6.3 explores more journalism-centric avenues for upcoming scholarship. This dissertation makes several major contributions to the study of deviance, gatekeeping theory, community journalism, and online and social media. It also highlights important opportunities for future research.

This study‟s quantitative data clearly indicate that newspapers of all circulation sizes focus on deviance and deviant news; circulation size wields no significant influence over the use of Prominence, Conflict, or Oddity on news websites and little significant influence over the same news factors on Facebook or Twitter feeds (see Chapter 4.2). This study‟s qualitative data, however, indicate that community newspaper editors and publishers adopt a clear focus on news about “regular people and routine events” (egalitarian news) over news about “important people or unusual events” (deviant news) (see Chapter 4.3).

Chapters 5.4 and 5.5 highlighted four possible explanations for these discrepancies. Two theoretical explanations discounted the differences between datasets, and two practical explanations expounded upon the dataset differences. All four deserve independent consideration and dedicated research, and all four offer opportunities for important future studies.

6.3.A: MEDIA SOCIOLOGY & COGNITIVE DISSONANCE The first research opportunity returns to Chapter 5.4.a: Validation of Media Sociology. This opportunity argues that the gap between the quantitative and qualitative datasets can be equated to a gap between journalists‟ words and actions. It is not at all a flattering explanation for journalists; however, it is worth considering.

Media sociology and gatekeeping theory both argue that media practitioners homogenize media content to fit institutional standards (Shoemaker & Reese, 1996; Shoemaker & Vos, 2009). Individual journalists‟ attitudes, beliefs, perspectives and demographics ultimately wield little influence over media content; instead, media conform to standards crafted by the industry as a whole and socialized over time (Bleske, 1991; Bowman, 2008; Cassidy, 2006; Gans, 1979; Sigal,

152

1973; White, 1950). The theory argues that journalists adopt a herd mentality and behave, first and foremost, as journalists.

Data presented here certainly validate this to some degree: The exhaustive local focus of community journalism qualifies as a professional norm and institutional standard. No community journalist in the country, in this researcher‟s opinion, would argue that community journalism as an industry is not predominately locally focused. It‟s also not a stretch to consider a focus on deviance as an institutional norm; ANOVA analyses of quantitative data indicated no significant variance concerning deviance among a spectrum of large and small newspapers.

But are those institutional norms so entrenched they would dictate a journalist‟s behavior in an interview? Are they so universal that they would instruct 28 diverse individuals to toe an institutional party line when answering a question, but then pursue contradictory policies when crafting news content?

Interviews offer the best avenue for illumination. Speaking to community journalists about institutional pressures would shed light on distinctions between working as a professional journalist and working as a journalist for their individual paper. How much pressure do community newspaper editors and publishers feel to adhere to professional journalistic standards? Where is that pressure coming from, or not coming from? Apart from the well-established focus on local news, how do community journalists decide what is “news?” Where do they look for inspiration or guidance?

If community journalists indicate broad pressure to meet certain industry standards, then media sociology and gatekeeping research may be supported (Gans, 1979; Shoemaker & Reese, 1996; Shoemaker & Vos, 2009). If, instead, community journalists demonstrate individual editorial autonomy or greater sensitivity to local interests or issues, then results may point more firmly in the direction of structural pluralism (Donohue, et al., 1985; Olien, et al., 1978; Pollock, 2007; Tichenor, et al., 1973). Either way, data would be intriguing.

153

6.3.B: TILT, SENSITIVITY, AND PROPER NOUN USAGE IN COMPUTERIZED CONTENT

ANALYSIS The second research opportunity returns to chapter 5.4.b: Hyper-Sensitivity and Non- Contextualized Nature of Computerized Content Analysis Software. It argues that qualitative data are accurate while quantitative data are invalid. Computerized content analysis requires a specific research goal and finely tuned word dictionaries; programs such as DICTION 6.0 are very precise, but also very dull, instruments (Conway, 2006; Jarvis, 2004; Krippendorff, 2004; Nuendorf, 2002). If a researcher is too worried about accidentally omitting terms in a word dictionary, he or she may go overboard by including too many words and compromise the dataset. A useful metaphor involves pinball machines.

Normally, pinball machines give points based on where a pinball is hit. Hitting the ball onto specific targets or specific ramps awards points; hitting other locations with the ball, like traps or blank walls, award no points. The game is challenging, but the score is accurate. Some players may attempt pinball wizardry by cheating, however. Shaking the machine in a certain direction, or lifting it off the ground entirely, allows the ball to roll in the desired direction more deliberately and easily – however it risks “pushing tilt,” whereupon a gravity sensor detects the motion, the machine shuts down, and the score is erased. The player can knock the ball in any direction he or she chooses, but there can be no victory and no proof of accomplishment.

In a similar way, an overly zealous computerized content analyst could reach tilt with software like DICTION 6.0. By including too many terms in too many custom dictionaries, the software could find a high volume of invalid results in any number of texts – the metaphoric ball could be lifted and shaken into the desired position, but the difference between valid terms and invalid terms would be completely missed. Everything would be counted and the score, so to speak, would be meaningless.

This is a potential explanation for this study‟s small number of significant ANOVAs and Pearson‟s correlations. If too many words were included in the word dictionaries, then false positives could have been too plentiful in every news article or social media feed. It also could explain why the most significant statistical relationships occurred concerning Specific Proximity, which relied on specific proper nouns that could not be misinterpreted.

154

Conversely, it also is possible that the word dictionaries relying upon proper nouns were invalid. As mentioned in Chapter 5.4.b, the number of local proper nouns in any community could have been vastly understated by dictionaries used here. Dictionaries which do not use proper nouns would be more valid; indeed, the bulk of computerized content analyses have focused on broader rhetorical studies.

The researcher is inclined to believe the latter possibility. It is possible that proper noun- centric dictionaries like Specific Proximity were understated; however, significant relationships were more common with proper noun-centric word dictionaries than broader word dictionaries. This implies that while computerized content analysis is a dull instrument, it remains precise enough to record significant relationships.

A comprehensive test of computerized content analysis software would be fruitful. By designating a few particular texts for analysis and manipulating word dictionaries, formulae and best practices could be developed to determine the ideal size and comprehensiveness of word dictionaries. The distinction between proper noun-centric and non-proper noun-centric dictionaries could be explored and a method could be developed for determining validity in computerized content analysis results. The research would be nuanced and potentially tedious, but very worthwhile.

It also is worth mentioning that qualitative safeguards may be an effective method of detecting invalidity in computerized content analysis data. While validity may be difficult to determine, invalidity can be flagrantly obvious with a little investigation. This was the case with the news factor Magnitude, ultimately abandoned in this study (see Chapter 3.3.e). The results simply did not seem valid. While it sounds simple, topical qualitative analysis of the results and the data could shed a great deal of light on the quantitative content analysis process.

6.3.C: ONLINE AND OFFLINE COMMUNITY JOURNALISM The third opportunity for future research returns to Chapter 5.4.c. It explores a distinction between online and offline community journalism which was accidentally overlooked during the course of this research.

155

Quantitative data looked exclusively at online data. Community newspapers were featured, but only those with active websites, Facebook pages and Twitter accounts. This potentially omits a great number of community newspapers which are less technologically advanced; indeed, a number of interviewed community newspaper editors and publishers said they had underdeveloped or even non-existent web presences. Editors and publishers said the Internet either was or should be a top priority, but the low-tech, small-town newsrooms were admittedly behind the curve. Thus, interviewing editors and publishers without major web presences and comparing those interviews to online-only data may have provided a mismatch.

If Chapter 6.3.a was unkind to journalists, Chapter 6.3.c is unkind to this researcher. Steps should have been taken to ensure that interviews were conducted with editors and publishers with stronger online commitments; or, alternatively, questions should have addressed the difference between planning online and offline news. The researcher also should have pretested the interview script more vigorously.

This omission presents an obvious opportunity for further qualitative interviews and future research. Do community journalists craft or plan their online, offline, and social media content similarly or differently? How have their media production routines changed in the Internet era? And what pressures, and from whom, are they facing concerning online migration – are readers encouraging them to utilize the Web, or are advertisers and market pressures keeping them grounded in print? Questions such as these could be posed directly to community newspaper editors and publishers with great effect.

Furthermore, the potential for convention or conference interviews should not be discounted. The researcher had a welcoming experience at the Texas Press Association‟s Midwinter Conference and Trade Show. Organizers were easy to work with and interview subjects were plentiful and talkative. It‟s true that community journalists are hardly bombarded with academic interview requests, and perhaps the situation would be different for a more frequently studied population, but the potential for interviews at conferences and conventions should be explored and encouraged.

156

6.3.D: INQUIRING ABOUT DEVIANCE: SPECTRUM AND BROADENING The fourth avenue for future research returns to Chapter 5.4. The distinctions between quantitative and qualitative data may center around the phrase “regular people and routine events.” The researcher intended this as an indicator of egalitarianism, and it fits that role to a degree; the phrase “regular people” is clearly egalitarian, and also quite specific. The phrase “routine events,” though, ultimately highlighted several areas of theoretical development for the study of deviance.

The researcher originally intended the phrase “routine events” to refer to mundane routine, thus solidifying the phrase within the realm of egalitarianism. Most interviewed community newspaper editors and publishers, however, equated “routine events” with routine government events – city council meetings, school board hearings, and other chronologically regular events. These events are “routine” in the sense that they happen frequently, and they are “routine” to journalists who cover them frequently. They are not “routine” to the average American, however, who has likely rarely set foot in a local city council or representatives‟ office. “Routine” to journalists, “routine” to academic researchers, and “routine” to newspaper readers may ultimately constitute three very different ideas.

The first subsequent research opportunity lies in properly operationalizing deviance. Chapter 5.4 argues that academic studies of deviance should not be focused solely on fringe groups, violent crime, or dramatic protest movements (Arpan & Tuzunkan, 2011; Boyle & Armstrong, 2009; Breen, 1997; Chan & Lee, 1984; Jong Hyuk, 2008; Pritchard & Hughes, 1997); such news clearly qualifies as deviant, but it does not constitute the entirety of deviance. Instead, deviance is a much broader concept which includes any instance where social rules or norms are broken. This study explores two potential avenues for theoretical development: deviance as a spectrum between normative deviance and normative egalitarianism (see Chapter 5.5.b) and deviance as a broader concept in general (see Chapter 5.5.c). Dedicated exploration and theoretical reconsideration is necessary to further explore these ideas, and to further the study of deviance as a whole. This particular researcher is partial to the broadening option, although theoretical development could and should lead in diverse directions.

157

6.4: Final Thoughts

Chapter 6.0 has identified broad and specific conclusions and opportunities for future research. It seems clear, based on these findings, that news is about deviance – deviations from the regular routines of media consumers, based particularly on prominent individuals and conflict, and rooted in a pathological human need for safety and security. This is one strong answer to the central question of journalism research, “What is news?” While this dissertation‟s strongest contribution relates to human communication and behavior, and identified in Chapter 6.0, perhaps a simpler conclusion is also valuable to journalism and journalism studies.

We should be talking about deviance. We should be asking journalists, media audiences, and academics about deviance. We should be exploring the concept in the field to properly conceptualize it, identify it, and understand it. Given connotations and popular confusion surrounding the term “deviance” identified in Chapter 2.5, a tangible discussion about the term and theoretical concept of normative deviance seems appropriate and useful.

How might journalists, academics, and media audiences define “deviance?” What events would the average news reporter consider “deviant,” and how deviant would a news item need to be to merit coverage? How would a reporter define deviance, and are there different kinds of deviance? How would those perspectives differ from an academic‟s or a reader‟s view of deviance, in the media or in daily life? For a term with such clear theoretical potential, it seems odd that this study has focused primarily on theoretical and content analysis.

There is clear value in conducting interviews with paraphrases and content analyses with news factor approaches to deviance, as this dissertation has done. However, there may be even more value in simply tackling the subject head-on. Summarize media sociology and the study of gatekeeping for journalists, academics, and media audiences and frankly ask each group for their thoughts. The results would clearly speak to the practical and theoretical development of deviance and expand academic understanding of media industries.

They also would do what social science does best – explore and articulate the broad patterns and deep meanings behind day-to-day images and routine media. In an era of rapid media evolution, such exploration and articulation has never been more valuable. 158

APPENDICES

Appendix 1: Map of Regional Categorizations

A1: Map of Regional Categorizations. The dataset was randomly collected from eight geographic American regions, as illustrated here. Regional categories were based upon common cultural, political and economic features, as well as geographic contiguousness. Within each region, five small weekly newspapers, five community daily newspapers, five large daily newspapers and (when applicable) one national daily newspaper were randomly selected.

159

Appendix 2: List of Selected Newspapers by Circulation Category

All circulation sizes taken from the Ulrich Periodical Index; however, five newspapers were unlisted, and data from the 2008 Editor and Publisher‟s Yearbook was substituted and marked *. In one case, The New Haven Leader, circulation figures were unavailable in both databases, so the publication was contacted and self-reported circulation was used and marked **.

PACIFIC:

Weekly:

1. Port Townsend Leader. Port Townsend, Washington. Circulation: 9,300

2. The Tillamook Headlight-Herald. Tillamook, Oregon. Circulation: 9000

3. Half Moon Bay Review. Half Moon Bay, California. Circulation: 7,200

4. St. Helena Star. St. Helena, California. Circulation: 6,000

5. Capital City Weekly. Juneau, Alaska. Circulation: 20,000

Community Daily:

1. The Ellensburg Daily Record. Ellensburg,, Washington. Circulation: 6,000

2. The Grant‟s Pass Daily Courier. Grant‟s Pass, Oregon. Circulation: 18,000

3. The Santa Cruz Sentinel. Scott‟s Valley, California. Circulation: 28,000.

4. West Hawaii Today. Kailuna Koa, Hawaii. Circulation: 12,308

5. Juneau Empire. Juneau, Alaska. Circulation: 7,000

Large Daily:

1. The Seattle Times. Seattle, Washington. Circulation: 215,000

2. The Oregonian. Portland, Oregon. Circulation: 239,071

3. The Torrance Daily Breeze. Torrance, California. Circulation: 83,800

160

4. The Honolulu Star-Advertiser. Honolulu, Hawaii. Circulation: 146,252

5. The Anchorage Daily News. Anchorage, Alaska. Circulation: 64,947

ROCKY MOUNTAIN:

Weekly:

1. Reno News and Review. Reno, Nevada. Circulation: 27,000

2. The Vernal Express. Vernal, Utah. Circulation: 4,883

3. Missoula Independent. Missoula, Montana. Circulation: 22,541

4. Colorado Springs Independent. Colorado Springs, Colorado. Circulation: 36,000

5. Idaho Business Review. Boise, Idaho. Circulation: 2,550

Community Daily:

1. Idaho Press-Tribune. Caldwell, Idaho. Circulation: 18,691.

2. Bozeman Daily Chronicle. Bozeman, Montana. Circulation: 18,000

3. Casper Star Tribune. Casper, Wyoming. Circulation: 30,502

4. Lahontan Valley News & Fallon Eagle Standard. Fallon, Nevada. Circulation: 4,700

5. Steamboat Today. Steamboat Springs, Colorado. Circulation: 9,500

Large Daily:

1. Las Vegas Review-Journal. Las Vegas, Nevada. Circulation: 190,100

2. Pueblo Chieftain. Pueblo, Colorado. Circulation: 53,250

3. Ogden Standard-Examiner. Ogden, Utah. Circulation: 60,768

4. The Deseret News. Salt Lake City, Utah. Circulation: 71,741

5. The Idaho Statesman. Boise, Idaho. Circulation: 143,098

161

SOUTHWEST:

Weekly:

1. San Pedro Valley News-Sun. Benson, Arizona. Circulation: 3,200

2. The Las Cruces Bulletin. Las Cruces, New Mexico. Circulation: 22,000*

3. The Santa Fe Reporter. Santa Fe, New Mexico. Circulation: 23,000

4. Mannford Eagle. Mannford, Oklahoma. Circulation: 2,000

5. The Katy Times. Katy, Texas. Circulation: 6000**

Community Daily:

1. Sierra Vista Herald. Sierra Vista, Arizona. Circulation: 12,000

2. The Alamogordo Daily News. Alamogordo, New Mexico. Circulation: 8,710

3. The Claremore Daily Progress. Claremore, Oklahoma. Circulation: 6,100

4. The Lawton Constitution. Lawton, Oklahoma. Circulation: 23,824

5. The McAllen Monitor. McAllen, Texas. Circulation: 42,068

Large Daily:

1. The Arizona Daily Star. Tuscon, Arizona. Circulation: 165,029

2. Albuquerque Journal. Albuquerque, New Mexico. Circulation: 107,947

3. The Tulsa World. Tulsa, Oklahoma. Circulation: 142,000

4. The El Paso Times. El Paso, Texas. Circulation: 87,500

5. The San Antonio Express News. San Antonio, Texas. Circulation: 356,367

162

GREAT PLAINS:

Weekly:

1. New Haven Leader. New Haven, Missouri. Circulation: 1,380**

2. Buffalo Center Tribune. Buffalo Center, Iowa. Circulation: 1,700

3. Marion County Record. Marion, Kansas. Circulation: 3023

4. North Bend Eagle. North Bend, Nebraska. Circulation: 1,467*

5. Caledonia Argus. Caledonia, Minnesota. Circulation: 2,150

Community Daily:

1. The Emporia Gazette. Emporia, Kansas. Circulation: 10,000

2. The Southeast Missourian. Cape Girardeau, Missouri. Circulation: 17,215

3. The Pierre Capital Journal. Pierre, South Dakota. Circulation: 10,000

4. The Minot Daily News. Minot, North Dakota. Circulation: 26,000

5. The Duluth News Tribune. Duluth, Minnesota. Circulation: 40,000

Large Daily:

1. The Sioux Falls Argus Leader. Sioux Falls, South Dakota. Circulation: 54,000

2. The Omaha World Herald. Omaha, Nebraska. Circulation: 192,607

3. The Cedar Rapids Gazette. Cedar Rapids, Iowa. Circulation: 68,700

4. The St. Louis Post Dispatch. St. Louis, Missouri. Circulation: 286,310

5. The St. Paul Pioneer Press. St. Paul, Minnesota. Circulation: 197,477

163

GREAT LAKES:

Small Weekly

1. The Salyersville Independent. Salyersville, Kentucky. Circulation: 5,500

2. The Wetzel Chronicle. Wetzel, West Virginia. Circulation: 7,000

3. The Hudson Star-Observer. Hudson, Wisconsin. Circulation: 7,500

4. The Wednesday Journal. Oak Park, Illinois. Circulation: 10,000

5. The Brown County Democrat. Nashville, Indiana. Circulation: 4,700

Community Daily

1. The Owensboro Messenger-Inquirer. Owensboro, Kentucky. Circulation: 31,900.

2. The Parkersburg News and Sentinel. Parkersburg, West Virginia. Circulation: 21,473

3. The Greenville Daily Advocate. Greenville, Oho. Circulation: 8,000

4. The Terre Haute Tribune-Star. Terre Haute, Indiana. Circulation: 36,000

5. The Grand Haven Tribune. Grand Haven, Michigan. Circulation: 10,500

Large Daily

1. The Green Bay Press-Gazette. Green Bay, Wisconsin. Circulation: 61,000

2. The Rockford Register Star. Rockford, Illinois. Circulation: 65,968

3. The Detroit News. Detroit, Michigan. Circulation: 232,434

4. The Dayton Daily News. Dayton, Ohio. Circulation: 139,462

5. The Northwest Indiana Times. Munster, Indiana. Circulation: 79,393*

164

SOUTHEAST:

Small Weekly:

1. The Fairfield Bay News. Fairfield Bay, Arkansas. Circulation: 3,000

2. The Calhoun County Journal. Bruce, Mississippi. Circulation: 4,700

3. Weekly Observer. Hemingway, South Carolina. Circulation: 2,100

4. The Alachua County Today. Alachua, Florida. Circulation: 5,000

5. The Mississippi Business Journal. Jackson, Mississippi. Circulation: 10,500

Community Daily:

1. The Valdosta Daily Times. Valdosta, Georgia. Circulation: 21,150

2. The Houma Courier. Houma, Louisiana. Circulation: 19,000

3. The Russellville Courier. Russellville, Arkansas. Circulation: 10,500

4. The Cullman Times. Cullman, Alabama. Circulation: 20,000

5. The Morristown Citizen Tribune. Morristown, Tennessee. Circulation: 21,322

Large Daily:

1. The Shreveport Times. Shreveport, Louisiana. Circulation: 63,211

2. The St. Petersburg Times. St. Petersburg, Florida. Circulation: 316,007

3. The Chattanooga Times Free Press. Chattanooga, Tennessee. Circulation: 80,000

4. The Charleston Post and Courier. Charleston, South Carolina. Circulation: 101,288

5. The Macon Telegraph. Macon, Georgia. Circulation: 70,000

MID-ATLANTIC

Small Weekly:

1. The Cranbury Press. Cranbury, New Jersey. Circulation: 3,088 165

2. The Cherokee Scout. Murphy, North Carolina. Circulation: 9,500

3. The Westmoreland News. Westmoreland, Virginia. Circulation: 8,200

4. The Dover Post. Dover, Delaware. Circulation: 29,400

5. The Central Penn Business Journal. Harrisburg, Pennsylvania. Circulation: 12,000

Community Daily:

1. The Scranton Times-Tribune. Scranton, Pennsylvania. Circulation: 40,500

2. The Burlington County Times. Willingboro, New Jersey. Circulation: 40,067

3. The Hagerstown Herald-Mail. Hagerstown, Maryland. Circulation: 13,332

4. The Salisbury Post. Salisbury, North Carolina. Circulation: 26,500

5. The News Virginian. Waynesboro, Virginia. Circulation: 8,070

Large Daily:

1. The Charlotte Observer. Charlotte, North Carolina. Circulation: 235,469

2. The Baltimore Sun. Baltimore, Maryland. Circulation: 321,165

3. The Wilmington News-Journal. Wilmington, Delaware. Circulation: 111,459*

4. The Roanoke Times. Roanoke, Virginia. Circulation: 100,249

5. The Press of Atlantic City. Atlantic City, New Jersey. Circulation: 78,288

NORTHERN REGION

Small Weekly:

1. The Greenwich Post. Darien, Connecticut. Circulation: 24,000

2. The Springville Journal. Springville, New York. Circulation: 4,368

3. The Arlington Advocate, Arlington, Massachusetts. Circulation: 8,547

4. The York County Coast Star. Portsmouth, New Hampshire. Circulation: 9,724

166

5. The Stowe Reporter. Stowe, Vermont. Circulation: 5,600

Community daily:

1. The Westerly Sun. Westerly, Rhode Island. Circulation: 10,100

2. The Concord Monitor. Concord, New Hampshire. Circulation: 21,145

3. The Brattleboro Reformer. Brattleboro, Vermont. Circulation: 11,235

4. The Middletown Press. Middletown, Connecticut. Circulation: 12,360

5. The Waterville Morning Sentinel. Waterville, Maine. Circulation: 22,400

Large Daily:

1. The Providence Journal. Providence, Rhode Island. Circulation: 173,000

2. The Bangor Daily News. Bangor, Maine. Circulation: 63,244

3. The Quincy Patriot Ledger. Quincy, Massachusetts. Circulation: 74,500

4. The Staten Island Advance. Staten Island, New York. Circulation: 72,000

5. New Hampshire Union Leader. Manchester, New Hampshire. Circulation: 60,000

NATIONAL NEWSPAPERS:

1. The Los Angeles Times. Los Angeles, California. Circulation: 955,211

2. The Washington Post. Washington, DC. Circulation: 732,872

3. The Chicago Tribune. Chicago, Illinois. Circulation: 600,988

4. The New York Daily News. New York, New York. Circulation: 763,975

5. The Houston Chronicle. Houston, Texas. Circulation: 553,462

167

Appendix 3: Newspaper Groups

GROUP A: The Arlington Advocate, Arlington, Massachusetts. Circulation: 8,547

The Brattleboro Reformer. Brattleboro, Vermont. Circulation: 11,235

The Casper Star Tribune. Casper, Wyoming. Circulation: 30,502

The Cedar Rapids Gazette. Cedar Rapids, Iowa. Circulation: 68,700

The Chattanooga Times Free Press. Chattanooga, Tennessee. Circulation: 80,000

The Chicago Tribune. Chicago, Illinois. Circulation: 600,988

The Claremore Daily Progress. Claremore, Oklahoma. Circulation: 6,100

The Detroit News. Detroit, Michigan. Circulation: 232,434

The Greenville Daily Advocate. Greenville, Oho. Circulation: 8,000

The Hagerstown Herald-Mail. Hagerstown, Maryland. Circulation: 13,332

The Half Moon Bay Review. Half Moon Bay, California. Circulation: 7,200

The Hudson Star-Observer. Hudson, Wisconsin. Circulation: 7,500

The Marion County Record. Marion, Kansas. Circulation: 3023

The Missoula Independent. Missoula, Montana. Circulation: 22,541

The Ogden Standard-Examiner. Ogden, Utah. Circulation: 60,768

The Pierre Capital Journal. Pierre, South Dakota. Circulation: 10,000

The Quincy Patriot Ledger. Quincy, Massachusetts. Circulation: 74,500

The Russellville Courier. Russellville, Arkansas. Circulation: 10,500

168

The Santa Cruz Sentinel. Scott‟s Valley, California. Circulation: 28,000

The Santa Fe Reporter. Santa Fe, New Mexico. Circulation: 23,000

The Torrance Daily Breeze. Torrance, California. Circulation: 83,800

The Tulsa World. Tulsa, Oklahoma. Circulation: 142,000

The Weekly Observer. Hemingway, South Carolina. Circulation: 2,100

The Westmoreland News. Montross, Virginia. Circulation: 8,200

The Wilmington News-Journal. Wilmington, Delaware. Circulation: 111,459

GROUP B: The Alachua County Today. Alachua, Florida. Circulation: 5,000

The Charleston Post and Courier. Charleston, South Carolina. Circulation: 101,288

The Colorado Springs Independent. Colorado Springs, Colorado. Circulation: 36,000

The Cullman Times. Cullman, Alabama. Circulation: 20,000

The Dayton Daily News. Dayton, Ohio. Circulation: 139,462

The Deseret News. Salt Lake City, Utah. Circulation: 71,741

The Dover Post. Dover, Delaware. Circulation: 29,400

The El Paso Times. El Paso, Texas. Circulation: 87,500

The Honolulu Star-Advertiser. Honolulu, Hawaii. Circulation: 146,252

The Lahontan Valley News & Fallon Eagle Standard. Fallon, Nevada. Circulation: 4,700

The Lawton Constitution. Lawton, Oklahoma. Circulation: 23,824

169

The Mannford Eagle. Mannford, Oklahoma. Circulation: 2,000

The Middletown Press. Middletown, Connecticut. Circulation: 12,360

The Minot Daily News. Minot, North Dakota. Circulation: 26,000

The New York Daily News. New York, New York. Circulation: 763,975

The North Bend Eagle. North Bend, Nebraska. Circulation: 1,467

The Roanoke Times. Roanoke, Virginia. Circulation: 100,249

The Salisbury Post. Salisbury, North Carolina. Circulation: 26,500

The St. Helena Star. St. Helena, California. Circulation: 6,000

The St. Louis Post Dispatch. St. Louis, Missouri. Circulation: 286,310

The Staten Island Advance. Staten Island, New York. Circulation: 72,000

The Terre Haute Tribune-Star. Terre Haute, Indiana. Circulation: 36,000

The Wednesday Journal. Oak Park, Illinois. Circulation: 10,000

The York County Coast Star. Portsmouth, New Hampshire. Circulation: 9,724

West Hawaii Today. Kailuna Kona, Hawaii. Circulation: 12,308

GROUP C: Steamboat Today. Steamboat Springs, Colorado. Circulation: 9,500

The Anchorage Daily News. Anchorage, Alaska. Circulation: 64,947.

The Brown County Democrat. Nashville, Indiana. Circulation: 4,700

The Caledonia Argus. Caledonia, Minnesota. Circulation: 2,150

170

The Capital City Weekly. Juneau, Alaska. Circulation: 20,000

The Central Penn Business Journal. Harrisburg, Pennsylvania. Circulation: 12,000

The Duluth News Tribune. Duluth, Minnesota. Circulation: 40,000

The Grand Haven Tribune. Grand Haven, Michigan. Circulation: 10,500

The Houston Chronicle. Houston, Texas. Circulation: 553,462

The Idaho Business Review. Boise, Idaho. Circulation: 2,550

The Idaho Statesman. Boise, Idaho. Circulation: 143,098

The Juneau Empire. Juneau, Alaska. Circulation: 7,000.

The Katy Times. Katy, Texas. Circulation: 6,000

The Macon Telegraph. Macon, Georgia. Circulation: 70,000

The McAllen Monitor. McAllen, Texas. Circulation: 42,068

The Mississippi Business Journal. Jackson, Mississippi. Circulation: 10,500

The Morristown Citizen Tribune. Morristown, Tennessee. Circulation: 21,322

The New Hampshire Union Leader. Manchester, New Hampshire. Circulation: 60,000

The News Virginian. Waynesboro, Virginia. Circulation: 8,070

The Northwest Indiana Times. Munster, Indiana. Circulation: 79,393

The Press of Atlantic City. Atlantic City, New Jersey. Circulation: 78,288

The San Antonio Express News. San Antonio, Texas. Circulation: 356,367

The St. Paul Pioneer Press. St. Paul, Minnesota. Circulation: 197,477

The Stowe Reporter. Stowe, Vermont. Circulation: 5,600

171

The Waterville Morning Sentinel. Waterville, Maine. Circulation: 22,400

GROUP D: The Arizona Daily Star. Tuscon, Arizona. Circulation: 165,029

The Charlotte Observer. Charlotte, North Carolina. Circulation: 235,469

The Cranbury Press. Cranbury, New Jersey. Circulation: 3,088

The Ellensburg Daily Record. Ellensburg, Washington. Circulation: 6,000

The Emporia Gazette. Emporia, Kansas. Circulation: 10,000

The Fairfield Bay News. Fairfield Bay, Arkansas. Circulation: 3,000

The Green Bay Press-Gazette. Green Bay, Wisconsin. Circulation: 61,000

The Greenwich Post. Darien, Connecticut. Circulation: 24,000

The Idaho Press-Tribune. Caldwell, Idaho. Circulation: 18,691.

The Las Vegas Review-Journal. Las Vegas, Nevada. Circulation: 190,100

The Los Angeles Times. Los Angeles, California. Circulation: 955,211

The Owensboro Messenger-Inquirer. Owensboro, Kentucky. Circulation: 31,900.

The New Haven Leader. New Haven, Missouri. Circulation: 1,380

The Port Townsend Leader. Port Townsend, Washington. Circulation: .9,300

The Providence Journal. Providence, Rhode Island. Circulation: 173,000

The Reno News and Review. Reno, Nevada. Circulation: 27,000

The Salyersville Independent. Salyersville, Kentucky. Circulation: 5,500

172

The San Pedro Valley News-Sun. Benson, Arizona. Circulation: 3,200

The Scranton Times-Tribune. Scranton, Pennsylvania. Circulation: 40,500

The Seattle Times. Seattle, Washington. Circulation: 215,000

The Shreveport Times. Shreveport, Louisiana. Circulation: 63,211

The Sierra Vista Herald. Sierra Vista, Arizona. Circulation: 12,000

The Sioux Falls Argus Leader. Sioux Falls, South Dakota. Circulation: 54,000

The Valdosta Daily Times. Valdosta, Georgia. Circulation: 21,150

The Westerly Sun. Westerly, Rhode Island. Circulation: 10,100

GROUP E: The Alamogordo Daily News. Alamogordo, New Mexico. Circulation: 8,710

The Albuquerque Journal. Albuquerque, New Mexico. Circulation: 107,947

The Baltimore Sun. Baltimore, Maryland. Circulation: 321,165

The Bangor Daily News. Bangor, Maine. Circulation: 63,244

The Bozeman Daily Chronicle. Bozeman, Montana. Circulation: 18,000

The Buffalo Center Tribune. Buffalo Center, Iowa. Circulation: 1,700

The Burlington County Times. Willingboro, New Jersey. Circulation: 40,067

The Calhoun County Journal. Bruce, Mississippi. Circulation: 4,700

The Cherokee Scout. Murphy, North Carolina. Circulation: 9,500

The Concord Monitor. Concord, New Hampshire. Circulation: 21,145

173

The Grants Pass Daily Courier. Grants Pass, Oregon. Circulation: 18,000

The Houma Courier. Houma, Louisiana. Circulation: 19,000

The Las Cruces Bulletin. Las Cruces, New Mexico. Circulation: 22,000

The Omaha World Herald. Omaha, Nebraska. Circulation: 192,607

The Oregonian. Portland, Oregon. Circulation: 239,071

The Parkersburg News and Sentinel. Parkersburg, West Virginia. Circulation: 21,473

The Pueblo Chieftain. Pueblo, Colorado. Circulation: 53,250

The Rockford Register Star. Rockford, Illinois. Circulation: 65,968

The Southeast Missourian. Cape Girardeau, Missouri. Circulation: 17,215

The Springville Journal. Springville, New York. Circulation: 4,368

The St. Petersburg Times. St. Petersburg, Florida. Circulation: 316,007

The Tillamook Headlight-Herald. Tillamook, Oregon. Circulation: 9,000

The Vernal Express. Vernal, Utah. Circulation: 4,883

The Washington Post. Washington, DC. Circulation: 732,872

The Wetzel Chronicle. New Martinsville, West Virginia. Circulation: 7,000

174

Appendix 4: Constructed Week Schedules

GROUP A: Monday: January 14

Tuesday: February 12

Wednesday: February 6

Thursday: February 28

Friday: February 22

Saturday: March 9

Sunday: January 27

GROUP B: Monday: February 25

Tuesday: February 19

Wednesday: March 6

Thursday: March 14

Friday: January 25

Saturday: February 2

Sunday: February 10

175

GROUP C: Monday: February 18

Tuesday: February 26

Wednesday: March 13

Thursday: January 31

Friday: February 8

Saturday: January 26

Sunday: March 3

GROUP D: Monday: January 28

Tuesday: January 22

Wednesday: February 13

Thursday: March 7

Friday: March 1

Saturday: February 9

Sunday: February 17

GROUP E: Monday: February 11

Tuesday: March 12

Wednesday: February 20

Thursday: January 24 176

Friday: March 8

Saturday: March 2

Sunday: March 17

177

Appendix 5: Constructed Week Calendar Graphic

A5: Calendar graphic expressing the constructed week schedule for quantitative data collection. Data was collected over a ten week period between January and March, 2013. The five groups of 25 newspapers each are expressed here through color coordination. Red is Group A, orange is Group B, green is Group C, blue is Group D, and purple is Group E. Dates were selected at random using multi- sided dice; however, those random selections were structured to ensure that dates for each group were at least approximately a week apart to reduce the likelihood of redundant data. 178

Appendix 6: Social Media Data Collection Schedule

GROUP A: Monday, March 18

GROUP B: Wednesday, February 27

GROUP C: Friday, February 22

GROUP D: Monday, February 11

GROUP E: Friday, March 22

179

Appendix 7: Social Media Data Collection Graphic

A7: Color coordinated graphic representation of social media data collection.

180

Appendix 8: Structured Interview Script

Hello! My name is Marcus, and I‟m a Ph.D. candidate at the University of Texas School of Journalism. I‟m conducting brief structured interviews for my dissertation on community journalism. Do you have five minutes to answer a few quick questions, and is it okay if I use a tape recorder? Your answers will be kept totally anonymous and in confidence, and the interview has been approved by the Institutional Review Board at UT-Austin.

If no: Thank you for your time.

If yes: Thanks! My dissertation is looking at two main ideas: community newspaper editor‟s interaction with their audiences, and topics covered in the news. Let‟s talk about audience interaction first.

(Q1) Is your newspaper a community weekly, bi-weekly, or daily publication?

(Q2) On a scale of one to 10, with one being “almost none” and 10 being “a great deal,” how much interaction do you feel you have with your readership?

(Q3) On average, about how many times a week do readers visit your office to talk about news ideas or your news coverage? (Quantitative open ended)

(Q4) On a scale of one to 10, with one being “almost none” and 10 being “a great deal,” how much interaction do you feel you have with your online readership?

(Q5) What about your social media audience? On a scale of one to 10, with one being “almost none” and 10 being “a great deal,” how much interaction do you have with them?

Let‟s talk briefly about your office.

(Q6) How many years have you worked as a community newspaper editor? (Quantitative open ended)

(Q7) How large is your paid editorial staff? (Quantitative open ended)

(Q8) Do you have any unpaid or freelance reporters? If so, how many? (Quantitative open ended)

Now, just three questions on news coverage.

181

(Q9) On a scale of one to 10, with one being “almost none” and 10 being “a great deal,” about how much of your news coverage is devoted to regular people and routine events?

(Q10) On a scale of one to 10, with one being “almost none” and 10 being “a great deal,” about how much of your news coverage is devoted to important people or unusual events?

(Q11) In general, what are your thoughts on community newspaper coverage of regular and extraordinary events? Where should the priority be? (Qualitative open ended)

Thanks! I really appreciate it.

(Q12) Would it be okay if I got your business card? Your response will be completely confidential, but I‟d like to make sure I don‟t accidentally interview two people from the same newspaper.

If no or not available: That‟s okay, thanks!

If yes: Thanks!

182

Appendix 9: Full Word Dictionaries

All 27 full word dictionaries are presented here as they appeared in the DICTION 6.0 word dictionary analysis.

1: PROMINENCE academic, academics, academies, academy, actor, actors, actress, actresses, administrator, administrators, adviser, advisers, AG, airport, airports, Ambassador, ambassadors, Attorney, attorneys, authorities, authority, bank, banker, bankers, banks, baseball, baseballs, basketball, basketballs, bible, bibles, bishop, bishops, blizzard, Boss, bosses, broker, brokers, budget, budgets, bureaucrat, bureaucrats, Captain, captains, cardinal, cardinals, CEO, CFO, Chair, Chairman, chairmen, chairpeople, Chairperson, chairs, Chairwoman, chairwomen, chief, chiefs, coach, coaches, college, colleges, commander, commanders, Commissioner, commissioners, congress, congressman, congressmen, congresswoman, congresswomen, Consul, consulate, consulates, consuls, Controller, controllers, COO, cop, cops, corporal, corporals, council, councilman, councilmen, councils, councilwoman, councilwomen, court, courthouse, courthouses, courts, CTO, dean, deans, Delegate, delegates, department, departments, deputies, Deputy, Director, directors, dispatcher, dispatchers, drought, droughts, duke, dukes, earl, earls, educator, educators, enforcer, enforcers, Exec, Executive, executives, flood, floods, football, footballs, friar, friars, goalie, goalies, golf, Gov, governor, governors, hail, hockey, hurricane, hurricanes, ice, instructor, instructors, Judge, judges, juries, jury, keeper, keepers, koran, korans, law, lawmaker, lawmakers, laws, lawyer, lawyers, LB, Leader, leaders, lecturer, lecturers, legend, legends, legislature, legislatures, lieutenant, lieutenants, linebacker, linebackers, lt, Manager, managers, marine, marines, Mayor, mayors, mentor, mentors, minister, ministers, Officer, officers, Official, officials, Olympics, Overseer, overseers, Parliament, pastor, pastors, playoff, playoffs, police, pope, popes, Premier, premiers, President, presidents, Prez, priest, priests, principal, principals, producer, producers, professor, professors, prominence, prominent, quarterback, quarterbacks, rain, rains, RB, rector, rectors, rep, representative, representatives, reps, researcher, researchers, sale, sales, salesman, salesmen, salespeople, salesperson, saleswoman, saleswomen, sergeant, sergeants, scholar, scholars, school, schools, schoolteacher, schoolteachers, scientist, scientists, scout, scouts, sen, senator, senators, sens, sleet, snow, 183

snowstorm, snowstorms, soccer, sold, soldier, soldiers, storm, storms, superintendent, superintendents, Supervisor, supervisors, teacher, teachers, team, teams, torah, torahs, tornado, tornados, tournament, tournaments, trainer, trainers, trial, trials, tutor, tutors, universities, university, varsity, vicar, vicars

2: CONFLICT absence, absences, absent, acquires, acquit, acquits, acquittal, acquitted, act, acted, action, actionable, acts, affray, affrayed, affrays, agitate, agitated, agitates, agitation, agreement, agreements, alarm, alarmed, alarming, alarms, alert, alerted, alerting, alerts, allegation, allege, alleged, alleges, altercate, altercated, altercates, altercation, animosity, animus, annex, annexed, annexes, annexing, approval, approve, approved, approves, acquire, acquired, acquiring, argue, argued, argues, argument, argumentative, argumentatively, arrest, arrested, arresting, arrests, arson, attain, attained, attainment, attains, ban, banned, banning, bans, Battle, battled, battles, battling, bicker, bickered, block, blocked, blocker, blocking, blocks, blowout, blowouts, brawl, brawled, brawler, brawling, brawls, breach, breached, breaches, break, breaking, breaks, broke, campaign, campaigned, campaigner, campaigning, campaigns, champion, champions, Clash, clashed, clashes, clashing, clashingly, close, closed, closer, closes, closing, collide, collided, collides, colliding, collision, Combat, combated, combative, combats, compete, competition, competitive, competitively, completed, completes, Conflict, conflicted, conflicting, conflicts, contended, contends, contention, contentious, contentiously, contest, contestable, contested, contestedly, contests, controversial, controversially, controversies, controversy, convert, converted, converts, convicted, convicts, counter, countered, counters, critic, criticize, criticized, criticizes, crusade, crusaded, crusader, crusades, crusading, crusadingly, cut, cuts, cutting, damn, damnably, damned, damning, damns, danger, dangerous, dangerously, deal, deals, dealt, debate, debated, debates, debating, debater, decide, decided, decides, decision, declination, decline, declined, declines, declining, derivative, derive, derived, derives, destroy, destroyed, destroyer, destroys, destructible, disaccord, disaccorded, fighter, Fight, disaccording, disaccordingly, disaccords, disagree, disagreeable, disagreeably, disagreed, disagreement, disagrees, discord, discords, dispute, disputed, disputes, disputing, disputingly, dissension, dissent, dissented, dissenting, dissentingly, dissents, dissidence, disturb, disturbance, disturbed, disturbing, disturbingly, disturbs, divide, divided, divides, division, divisive, divisively, dominant, dominate, 184

dominated, dominates, domination, elect, elected, electing, election, elections, elector, eliminate, eliminated, eliminates, eliminating, elimination, eliminator, encounter, encountered, encountering, encounters, enraged, escape, escaped, escapes, evil, evilly, faction, factional, factionalism, factions, fail, failed, failing, failingly, fails, failure, bickering, bickers, championed, championing, fighting, fightingly, fights, fire, fireable, fired, fires, fissure, fissured, fissures, fought, foul, fouled, fouler, fouling, fouls, fracas, fracture, fractured, fractures, fray, friction, fuss, fussed, fusses, fussily, fussy, gain, gained, gaining, gainingly, gains, goal, goals, graffiti, hang, hanging, hangs, hassle, hassled, hassles, hassling, hire, hireable, hired, hires, hit, hits, holdup, holdups, hung, hurt, hurting, hurts, indestructible, infect, infected, infection, infections, injure, injured, injures, injuring, injury, interference, interfere, interfered, interferes, interfering, interferingly, interrogate, interrogated, interrogates, interrogator, kill, killed, killer, kills, lawsuit, lose, loser, losing, loss, , losses, missing, murder, murdered, murders, opponent, oppose, opposed, opposes, opposing, opposingly, outrage, outrageous, outrageously, overrule, overrules, overruling, point, pointed, pointer, pointing, points, politic, political, politician, politics, poll, polled, polling, polls, problem, problematic, problematically, problems, procure, procured, procures, procuring, procuringly, proposal, propose, proposed, proposes, quarrel, quarreled, quarreling, quarrelous, quarrels, rage, rages, reason, reasonable, reasonably, reasons, reckless, recklessly, refusal, refuse, refused, refuses, replaceable, replace, replaced, replaces, resign, resigned, resigns, rift, rifted, rifts, rivaled, rivalries, rivalrous, rivalry, rival, runoff, runoffs, schism, schisms, score, , scored, scorer, scores, scoring, separate, separated, separately, separates, separation, sever, severance, severed, severs, shatter, shattered, shatters, shock, shocked, shocking, shocks, shoot, shooting, , shootings, shoots, , shot, shots, skirmish, skirmished, skirmishes, slay, slayed, slayer, slays, smear, smeared, smearing, smears, spat, spats, spatted, spatter, split, splits, splitting, splittingly, squabble, squabbled, squabbles, squabbling, squabbly, stab, stabbed, stabbing, stabs, strafe, strafed, strafes, strafing, strand, stranded, strands, strife, strifed, strifes, strive, strived, strives, striving, struggle, struggled, struggles, struggling, strugglingly, stun, stunned, stunning, stuns, success, successes, succeeded, successful, successfully, successor, sue, sued, sues, suing, suit, suits, tangle, tangled, tangles, tense, tensed, tensely, tension, touchdown, touchdowns, trifle, trifled, trifles, trifling, triflingly, triumphant, triumphantly, triumphed, triumphs, triumph, tug-of-war, unreasonable, unreasonably, update, updated, updates, , verdict, verdicts, victor, victorious, victoriously, victors, victory, violate, 185

violated, violating, violation, violations, violator, war, warfare, warring, wars, win, winner, winning, wins, won, convict, defiantly, defiance, defy, defied

3: ODDITY aberrant, aberrate, aberated, aberating, aberration, aberrations, abnormal, abnormalities, abnormality, abnormally, astonish, astonished, astonishing, astonishingly, astonishment, astonishments, atypical, atypicalities, atypicality, atypically, avant-garde, bizarre, bizarrely, conspicuous, conspicuously, craze, crazed, crazies, crazy, curious, curiosities, curiosity, curiously, defiantly, deviate, deviated, deviates, deviation, deviations, differ, differed, difference, differences, different, differently, eccentric, eccentrically, eccentricities, eccentricity, erratic, erratically, estrange, estranged, exotic, exotics, irregular, irregularities, irregularity, irregularly, kook, kooks, kooky, odd, oddities, oddity, oddly, offbeat, off-the-wall, outcast, outcasts, outlandish, outlandishly, peculiar, peculiarities, peculiarity, peculiarly, perplex, perplexed, perplexes, perplexing, perplexingly, psycho, psychos, queer, queered, queerly, queers, random, randomly, rare, rarely, spacey, strange, strangely, stranger, strangers, uncannily, uncanny, uncommon, uncommonalities, uncommonality, uncommonly, unusual, unusually, weird, weirdly, whim, whimsical, whimsically, whimsies, whimsy, wierdo, wierdos

4: IMPACT acute, acutely, acutes, afflict, afflicted, affliction, afflictions, afflictive, baleful, balefully, befoul, befouled, befouls, big, bigger, biggest, casual, casually, consequence, consequences, consequential, consequentially, consider, considerable, consideration, considerations, considered, core, cores, critical, critically, crucial, crucially, decisive, decisively, depend, depended, dependent, dependently, dependents, determinately, determine, determined, develop, developed, developing, development, developments, dinky, dire, direly, essential, essentially, esteem, estimably, esteemed, exception, exceptional, exceptionally, exceptions, extensive, extensively, extent, fall, fallen, falling, falls, first, firstly, first-string, foremost, fragile, frail, grave, gravely, graves, huge, icy, immaterial, immaterially, impact, impacted, impactful, impacting, impacts, imperative, imperatively, imperatives, inconsequential, inconsequentially, inconsiderable, inconsiderably, inconsiderate, inferior, inferiors, infinitesimal, infinitesimally, insignificance, insignificant, insignificantly, irrelevancies, irrelevant, irrelevantly, key, keys, large, largely, least, 186

legendary, less, lesser, lessers, main, major, majorly, majors, maximal, maximally, maximum, maximums, meager, meagerly, meaning, meaningful, meaningfully, meaningless, meaninglessly, meanings, minimal, minimally, minimum, minimums, miniscule, minor, minorly, minors, minute, minutely, momentous, momentously, more, most, necessarily, necessary, necessities, necessity, negligible, negligibly, new, newer, newest, nightmare, nightmares, paltry, paramount, pointless, pointlessly, premiere, press, pressed, pressingly, primaries, primarily, primary, prime, primed, relevancies, relevancy, relevant, relevantly, repeat, repeated, repeater, repeats, rise, risen, rises, rising, salient, saliently, scant, scantly, scanty, second-string, serious, seriously, severe, severely, significance, significances, significant, significantly, slight, slighted, slightly, slights, small, smaller, smallest, spurious, spuriously, standout, standouts, subordinate, subordinated, subordinately, subordinates, subsidiaries, subsidiary, substantial, substantially, sudden, suddenly, superior, superiorly, superiors, supreme, supremely, Supremes, third-string, tinily, tiny, trend, trended, trending, trends, trendy, trivial, trivialities, triviality, trivially, unconsidered, unimportant, unimportantly, urgent, urgently, valuable, valuables, valuably, vital, vitally, vitals, wane, waning, warned, warnings

5: TIMELINESS

5.1: Recentness current, currently, recent, recently, today, tomorrow, yesterday

5.2: Dates This dictionary fluctuated based upon the individual data point. It included five days of the week: the day in question, two days previous, and two days prior. It also included the date of data collection, two dates previous, and two days prior.

A hypothetical example for Wednesday, March 13: Monday, Tuesday, Wednesday, Thursday, Friday, March = 11, 12, 13, 14, 15

187

6: PROXIMITY

6.1: General Proximity area, civic, civically, local, locally, neighborhood

6.2: Specific Proximity

6.2.a: Proximity A.CW Acton, Afton, Agua, Altos, Andover, Arlington, Ashland, Auburndale, Baldwin, Bayport, Bayview, Bedford, Belmont, Bernal, Beverly, Bienerman, Billerica, Boston, Boxborough, Boxford, Braintree, Brisbane, Brookline, Bruno, Burlingame, Burlington, Cambridge, Canton, Carlisle, Carlos, Chelmsford, Chelsea, Colonial, Como, Concord, Corner, Cottage, Croix, Daly, Danvers, Dayton's, Dedham, Dover, Elmo, Emerald, End, Everett, Excelsior, Fairview, Farnham, Florence, Foster, Framingham, Francisco, Granada, Grant, Gregorio, Gresham, Hastings, Hathome, Hemingway, Highwood, Hillsboro, Hillsborough, Hingham, Honda, Hudson, Hugo, Hyde, Inver, Johnsonville, Kendal, Kinsale, Lake, Lakeland, Lawrence, Leonardtown, Lincoln, Linda, Lolo, Lowell, Loyola, Lynn, Lynnfield, Mahtomedi, Malden, Maplewood, Marblehead, Marine, Marion, Marlborough, Mateo, Maynard, Medfield, Medford, Melrose, Menlo, Middlesex, Middleton, Midtown, Millbrae, Milton, Missoula, Montara, Montross, Moon, Moss, Mountain, Natick, Needham, Nesmith, Newton, Norwood, Oak, Oakdale, Oaks, Pacifica, Pajarito, Palo, Pattee, Paul, Payne-Phalen, Peabody, Peaks, Pecos, Prescott, Quincy, Randolph, Reading, Redwood, revere, Richmond, River, Roberts, Rock, Roseville, Roxbury, Salem, Santa, Saugus, Scandia, Serramonte, Sherborn, Somerset, Somerville, Southborough, Stanford, Stillwater, Stoneham, Stoughton, Sudbury, Sunset, Swampscott, Tappahannock, Tewksbury, Topsfield, Vadnais, Wakefield, Walpole, Waltham, Warsaw, Watertown, Wayland, Wellesley, Wenham, Westborough, westford, Westlake, Westmoreland, Weston, Westside, Westwood, Weymouth, White, Williamsburg, Wilmington, Winchester, Winthrop, Woburn, Woodbury, Woodside

188

6.2.b: Proximity A.CD Amesti, Aptos, Aqua, Arcanum, Atkins, Bellows, Bernardston, Boonsboro, Bowers, Braddock, Bradford, Brattleboro, Bulder, Cabrian, Casper, Catoosa, Chambersburg, Chelsea, Chesterfield, Chouteau, Claremore, Clover, Collinsville, Corralitos, Covington, Cress, Cruz, Dardanelle, Darke, Deerfield, Dover, Dummerston, Evansville, Falling, Farm, Felton, Fountainhead-Orchard, Fox, Greencastle, Greenfield, Greenville, Guilford, Hagerstown, Halfway, Harmonyville, Hattieville, Hinsdale, Hughes, Inola, Interlaken, Jamaica, James, Jerusalem, Keene, Lewisburg, Lomond, Mar, Marion, Marlborough, Martinsburg, McKinley, Mercersburg, Middletown, Milton, Myersville, Natrona, Newfane, Northfield, Oak, Owasso, Paris, Pierre, Piqua, Pryor, Putney, Richmond, Robinwood, Rogers, Russellville, Saratoga, Scotts, Sereno, Shelburne, Sheperdstown, Smithsburg, Stoneville, Summit, Swanzey, Thurmont, Townshend, Tulsa, Turley, Union, Verdigris, Vernon, Versailles, Vista, Wagon, Walpole, Washington, Watsonville, Waynesboro, Westminster, Westmoreland, Williamsport, Wilson-Conochocheague, Winchester, Windham

6.2.c: Proximity A.LD Abington, Alamitos, Alloway, Altos, Amana, Amherstburg, Angeles, Arlington, Auburndale, Avon, Bagley, Barton-McFarland, Baynard, Beach, Bear, Bell, Bella, Belle, Bellflower, Belmont, Beverly, Birmkingham, Bixby, Bloomfield, Bolsa, Boothwn, Boston, Boynton, Brainerd, Braintree, Bridgewater, Brigham, Brightmoor, Brockton, Broken, Brook, Brookhaven, Brookline, Brookside, Brookwood, Browntown, Bryantville, Buena, Cabridge, Canton, Carneys, Castle, Catoosa, Cedar, Center, Cerritos, Chattanooga, Chelsea, Chester, Chickamuga, Clair, Clawson, Claymont, Clearfield, Clinton, Cohasset, Collinsville, Commerce, Compton, Cornerstone, Cou, Culver, Cypress, Darby, Dearborn, Dedham, Deepwater, Delisle's, Delray, Deptford, Detroit, Dover, Downey, Drexel, Easton, Eastpointe, Ecorse, Eden, Edgars, Edgemoor, Elkton, Elsmere, Essex, Exton, Eye, Farmington, Farr, Ferndale, Figueroa, Fitzgerald, Folsom, Forest, Foxboro, Franklin, Fraser, Garden, Gardena, Garnet, Gesto, Glasgow, Glen, Glenpool, Green, Greenville, Greenwich, Grosse, Grove, Halifax, Hamilton, Hamtramck, Hanover, Hanson, Harrison, Harrisville, Havertown, Hawaiian, Hawthorne, Hiawatha, Highlands, Hingham, Hixson, Hockessin, Holbrook, Hollywood, Homestead, Hooper, Hull, Huntington, Hyde, Inglewood, Inkster, Jamaica, Jasper, Jefferson, Jenks, Kaysville, Kendal, Kennett, Kenwood, Lakewood, Landenberg, Lawndale, Layton, Leimert, Lennox, Lexington, Liberty, Lincolnway, Linn, Lisbon, 189

Livnoia, Logan, Lomita, Lynn, Lynnfield, Lynwood, Madison, Malden, Malvern, Manhattan, Mannington, Mansfield, Mantua, Marblehead, Marcus, Marina, Marion, Marriott-Slaterville, Marshfield, McGregor, McKinley, Medfield, Medford, Melrose, Mexicantown, Mid, Middle, Milton, Mirada, Monica, Monponsett, Montebello, Morgan, Morningside, Mountain, Nahant, Nantasket, Natick, Needham, Neward, Newark, Newport, Newton, Noelridge, Norfolk, Norwalk, Norwell, Norwood, Oakhurst, Ogden, Oglethorpe, Oldmans, O'Main, Ooltewah, Owasso, Painters, Palos, Paquette, Paulsboro, Pedricktown, Pedro, Pembroke, Penn, Penns, Peoria, Perry, Philadelphia, Piety, Pilesgrove, Plain, Pleasant, Points, Quincy, Randolph, Reading, Red, Redondo, Regent, Revere, Rey, Richmond, Ridge, Ridley, Ringgold, Rivera, riverdale, Riversdale, Riverview, Rock, Rockland, Romulus, Rosewille, Roslindale, Rossmoor, Rossville, Rouge, Roy, Royal, Salem, Sand, Sapulpa, Saugus, Scituate, Segundo, Sharon, Sherborn, Signal, Silver, Sinasac, Skiatook, Soddy-Daisy, Solon, Somerville, Southfield, Southgate, Springfield, Springwells, Stanton, Sterling, Stoneham, Stoughton, Sunset, Swampscott, Syracuse, Talleyville, Taylor, Tecumseh, Toddville, Torrance, Tower, Trenton, Troy, Tulsa, Tunnel, Turley, Turner, Twin, Uintah, Venice, Verdigris, Vernon, Victoria, Wagon, Wakefield, Walker, Walpole, Waltham, Warren, Washington, Watertown, Wayland, Wayne, Weber, Wellesley, Westland, Westminster, Westmont, Weston, Westtown, Wetmouth, Whitman, Whittier, Whitwell, Wilmington, Winchester, Windsor, Winthrop, Woburn, Woodhaven, Woodlands, Woolwich, Wyandotte, Yorklyn

6.2.d: Proximity A.ND Addison, Alsip, Austin, Bellwood, Bensenville, Berwyn, Bridgeview, Broadview, Bronzeville, Brookfield, Burbank, Burnham, Burr, Calumet, Chicago, Cicero, Clearing, Cook, Dolton, Edgewater, Elmhurst, Evanston, Evergreen, Forest, Glencoe, Grange, Harvey, Hegewisch, Hickory, Hinsdale, Holland, Hubbard, Island, Kenilworth, Lawndale, Lincolnwood, Logan, Lombard, Lyons, Maywood, Midlothian, Midway, Morgan, Morton, Niles, Norridge, Northbrook, Northfield, Northlake, Oak, Palos, Plaines, Portage, Posen, Prospect, Pullman, Robbins, Roseland, Schiller, Side, Skokie, South, Summit, Villa, Village, Westchester, Westmont, Whitnig, Willowbrook, Wilmette, Winnetka

190

6.2.e: Proximity B.CW Acres, Addison, Agua, Alachua, Amesbury, Appletree, Archer, Arlington, Auburn, Austin, Barrington, Bellwood, Belmont, Bend, Bensenville, Berwick, Berwyn, Bloomingdale, Boyes, Brentwood, Briar, Bridgeview, Broadview, Bronzeville, Burbank, Burr, Butler, Butterfield, Calistoga, Camden, Capri, Carol, Carson, Chicago, Cicero, Cimarron, Clayton, Clearing, Cleveland, Cook, Creek, Darien, Deer, Deerfield, Dodge, Dolton, Dover, Downers, Dunning, Durham, Eliot, Elk, Ellyn, Elmhurst, Elmwood, Epping, Evanston, Evergreen, Exeter, Felton, flowerfield, Forest, Fountain, Franklin, Fremont, Gainesville, Glencoe, Glendale, Gleneagle, Glenview, Grange, Greenland, Hampton, Harrington, Helena, Hickory, Highland, Hinsdale, Hominy, Hubbard, Itasca, Jenkins, Kenilworth, Kennebunk, Kensington, Kent, Keystone, Kittery, Lawndale, Lebanon, Lee, Leipsic, Logan, Lombard, Lyons, Manitou, Mannford, Maywood, Middletown, Midlothian, Milford, Monument, Morgan, Morton, Napa, Naperville, Neddick, Newburyport, Newington, Newmarket, Niles, Norridge, Northbrook, Northfield, Northgate, Northlake, Nottingham, Oak, Oakhurst, Ogunquit, O'Main, Orland, Palatine, Palos, Paso, Pawnee, Petaluma, Plaines, Portage, Portsmouth, Posen, Prague, Prospect, Pullman, Ridge, Riverside, Robbins, Rockingham, Rohnert, Rollinsford, Rosa, Roseland, Roselle, Rye, Salisbury, Sand, Sapulpa, Schaumburg, Schiller, Schuyler, Sebastopol, Security-Widefield, Skokie, Smyrna, Sonoma, Springs, Sticknet, Stickney, Stone, Stonegate, Stratham, Stratmoor, Suburban, Townsend, Tulsa, Verano, Villa, Wahoo, Westchester, Westmont, Wheaton, Willowbrook, Wilmette, Windsor, Winfield, Winnetka, Woodmoor, Woodside, Yards, York, Yountville

6.2.f: Proximity B.CD Advance, Andover, Apache, Barry, Berlin, Bloomfield, Blountsville, Bolton, Branford, Brazil, Bridgeton, Bristol, Britain, Buckley, Cache, Cheshire, Chester, Churchill, Clinton, Colchester, Columbia, Comanche, Concord, Cook, Crompton, Cromwell, Cullman, Dennison, Denton, Durham, Elgin, Enochville, Essex, Falkville, Fallon, Farmersburg, Farmington, Forestville, Glastonbury, Grove, Guilford, Haddam, Hamden, Hanceville, Harmony, Hartford, Hartselle, Haute, Haven, Hawai'i, Hebron, Hills, Holualoa, Honalo, Honaunau, Honaunau-Napoopoo, Kahaluu-Keauhou, Kalaoa, Kannapolis, Kealakekua, Keney, Kensington, Killingworth, Knightsville, Kona, Landis, Lawton, Lexington, Lyme, Manchester, Marlborough, Marshall, Meriden, Middlefield, Middlesex, Middletown, Minot, Misenheimer, Mocksville, Mooresville, 191

Newington, Northford, Oaks, Paris, Plainville, Pleasant, Portland, Prospect, Quinnipiac, Robertson, Rockwell, Rowan, Salisbury, Southington, Spencer, Vernon, Vigo, Wallingford, Walters, Ward, Waterbury, Welcome, Wethersfield, Windsor, Wolcott, Woodleaf

6.2.g: Proximity B.LD Affton, Ahuimanu, Aiea, Album, Alexandria, Alimanu, Alto, Alton, Amboy, Americas, Ampliacion, Annadale, Anthony, Arden, Arlington, Arnold, Ashley, Baden, Ballwin, Barrancas, Beavercreek, Bellbrook, Bellefontaine, Belleville, Belmont, Berkeley, Bloomfield, Bluffdale, Borderland, Botetourt, Bountiful, Brentwood, Bridgeton, Brighton, Brooklyn, Brookville, Brunswick, Cahokia, Canutillo, Carbon, Carlisle, Carlton, Carrollton, Caseyville, Castleton, Castner, Cave, Cedarville, Centerville, Centro, Chaparral, Charleston, Clair, Clayton, Clifton, Clint, Cloverdale, Colinsville, College, Columbia, Concord, Coney, Cottonwood, Country, Craig, Crestwood, Creve, Daleville, Daybreak, Dayton, Dewees, Dorado, Draper, Drexel, Dupo, Dutchtown, Eastern, Eastview, Edison, Edwardsville, Elizabeth, Elizario, Elliston, Emerson, Englewood, Enon, Ewa, Fairmont, Fairview, Farmington, Fenton, Ferguson, Florissant, Folly, Franklin, Futureland, Garden, Glassgow, Goose, Granite, Halawa, Hanahan, Hartford, Hazelwood, Hazlet, Heroes, Herriman, Hoboken, Holladay, Hollins, Hollywood, Holmdel, Honolulu, Horizon, Huber, Hunter, Imperial, Isidro, James, Jennings, Johns, Jordan, Juarez, Kailua, Kaimuki, Kalihi-Palama, Kaneahe, Kapahulu, Kapolei, Kaysville, Keansburg, Kearns, Kettering, Keyport, Kiawah, Kirkwood, Kohlberg, Ladson, Lafayette, Lake, Lakehurst, Lebanon, Lewisburg, Liliha-Kapalama, Linden, Livingston, Louis, Madison, Magna, Makakilo, Manchester, Manhattan, Manoa, Mariners, Marlborough, Maryland, Maryville, Maunawili, McKinley, Mehlville, Mendez, Miamisburg, Midvale, Mililani, Military, Millcreek, Millstadt, Milton, Moanalua, Mokapu, Moneta, Montana, Montanas, Montclair, Montgomery, Montvale, Moraine, Morningside, Murphy, Murray, Newark, Norflour, Normandy, Norte, Nuuanu, Oakville, Oakwood, O'Fallon, Olivette, Oquirrh, Orange, Overland, Palms, Palolo, Palomino, Parkland, Paso, Pearl, Pecan, Pershing, Pine, Pleasant, Polk, Pontoon, Praderas, Prince's, Pueblos, Raleigh, Ravenel, Richmond, Ridge, Riverside, Riverton, Roanoke, Robbins, Rocky, Roxana, Rushfair, Salem, Sandy, Sauzal, Sayreville, Seabrook, Shiloh, Smithton, Socorro, Southview, Spanish, Springboro, Springfield, Staten, Sullivan's, Sunrise, Sunset, Taylorsville, Tipp, Tobin, Trotwood, Troutville, Troy, ueens, Union, Unoin, Valley, Vandalia, Venice, Village, Ville, 192

Vinton, Waco, Wadmalaw, Wagener, Wahiawa, Waikiki, Waimalu, Waimanalo, Waipahu, Walnut, Washington, Waterloo, Webster, Westwood, Willberforce, Woodbourne-Hyde, Woodbridge, Woodrow, Xenia, Yellow, York, Zaragoza

6.2.h: Proximity B.ND Amboy, Bronx, Brooklyn, Clifton, Elizabeth, Freeport, Hempstead, Kings, Manhattan, Orange, Paterson, Queens, Richmond, Staten, Teaneck, Yonkers, York

6.2.i: Proximity C.CW Ada, Alief, Bainbridge, Bend, Berlin, Bloomfield, Bloomington, Boiling, Boise, Bolton, Brandon, Brookshire, Brown, Brownsville, Byram, Calais, Caledonia, Camp, Carlisle, Cinco, Clinton, Columbus, Crescent, Crosse, Cy, Cypress, Dauphin, Dillsburg, Dorchester, Dover, Duncannon, Eagle, Edinburgh, Eitzen, Elizabethtown, Elizabethville, Ellettsville, Emigsville, Fairfax, Felipe, Fletcher, Flora, Florence, Flowood, Goldsboro, Grantville, Greatwood, Grove, Hardwick, Harris, Harrisburg, Hershey, Highspire, Hinds, Hockley, Houston, Hummelstown, Jackson, Jeffersonville, Jericho, Johnson, Juneau, Katy, Kuda, Lamoille, Lawton, Linglestown, Madison, Manchester, Martinsville, Marysville, Mechanicsburg, Meridian, Middlesex, Middletown, Millersburg, Mission, Montpelier, Morgantown, Morristown, Morrisville, Nampa, Nashville, Newport, Palmyra, Pattison, Paxtonia, Pearl, Plainfield, Progress, Rankin, Raymond, Richmond, Ridgeland, Rosenberg, Rushford, Rutherford, Schlusser, Sealy, Sharpstown, Simonton, Skyline, Smithville, Star, Steelton, Stowe, Sugar, Terry, Trafalgar, Underhill, Valley, Waller, Wallis, Waterbury, Westside, Williston, Wormleysburg

6.2.j: Proximity C.CD Afton, Alamo, Albion, Allendale, Alton, Augusta, Belgrade, Bellair, Benton, Blanca, Bravo, Brinnington, Bulls, Burnham, Canaan, Candlewyck, Carlos, Carlton, Charlottesville, Chavez, Churchville, Citrus, Clinton, Cloquet, Conklin, Coopersville, Cornville, Covesville, Crozet, Dalton, Dandridge, Defiance, Doce, Doffing, Donna, Doolittle, Duluth, Eagleton, Edcouch, Ednam, Elsa, Esko, Faber, Fairfield, Farmington, Fisherville, Flordon, Freedom, Fruitport, Garden, Glenaire, Greenville, Greenwood, Grottoes, Hamblen, Haven, Hidalgo, Hills, Holland, Homa, Inglecress, Jefferson, Jolivue, Joya, Juan, Juneau, Kennebec, Leigh, Liberty, Lopezville, 193

Louis, Lyndhurst, Manchester, Market, McAllen, Mercedes, Mercer, Meriwether, Mirian, Missoin, Monte, Montville, Morristown, Mosheim, Muniz, Muskegon, Nellysford, Newport, Norridgewock, Norton, Nunica, Nurillo, Olivarez, Olive, Ottawa, Palermo, Palhurst, Palmview, Park, Peacock, Penitas, Pharr, Pike, Pine, Pittsfield, Proctor, Progreso, Range, Ravenna, Readfield, Reynosa, Rome, Routt, Russellville, Rutledge, Saginaw, Scissors, Sharon, Sharyland, Sidney, Skowhegan, Smithfield, Spring, Station, Staunton, Steamboat, Stuarts, Superior, Talbott, Thorndike, Thurston, Union, Unity, Vassalboro, Verde, Verona, Vienna, Villa, Waterville, Waverly, Waynesboro, Weslaco, Westover, Weyers, Windsor, Winslow, Zealand

6.2.k: Proximity C.LD Abbott, Absecon, Ada, Afton, Alamo, Allenstown, Alsip, Altgeld, Amherst, Anchorage, Anthony, Antonio, Apple, Armatage, Ashburn, Atkinson, Atlantic, Auburn, Balcones, Bayport, Bayshore, Bear, Beecher, Bexar, Bibb, Blaine, Bloomington, Boise, Boston, Bow, Brentwood, Bridgeview, Brigantine, Brooklyn, Bryn, Burbank, Burnham, Burnsville, Byron, Calhoun, Calumet, Canada, Candia, Castle, Centerville, Chester, Chicester, Chugaik, Cibolo, Columbia, Como, Concord, Converse, Cooper, Cottage, Crown, Crystal, Danville, Deerfield, Derry, Dolton, Dyer, Eagan, Eagle, Eden, Edgewood, Edina, Elmo, Epping, Epsom, Flossmoor, Folwell, Frankfort, Fremont, Fridley, Fulton, Gage, Galloway, Gary, Glenwood, Goffstown, Golden, Gordon, Gray, Gresham, Griffith, Hamilton, Hammond, Harbor, Harlandale, Harvey, Hastings, Haven, Hegewisch, Hiawatha, Highland, Hillsborough, Hobart, Holland, Homewood, Hooksett, Hopkins, Hopkinton, Howe, Hudson, Hugo, Hyde, Island, Jeffersonville, Jordan, Judson, Kirby, Kuda, Lake, Lakeland, Lauderdale, Leon, Linwood, Litchfield, Lizella, Londonderry, Loudon, Lowell, Lynnhurst, Lynwood, Macon, Mahtomedi, Manchester, Maplewood, Margate, Markham, Matanuska-Susitna, Mat-Su, Matteson, Mays, Mendota, Meridian, Merrillville, Merrimack, Midlothian, Milford, Minneapolis, Minnetonka, Mokena, Morgan, Morthfield, Munster, Mystic, Nampa, Newport, Nokomis, Northcliff, Northside, Northwood, Nottingham, Oak, Oakdale, Oakland, Ocean, Olmos, Olympia, Orland, Palos, Paul, Pembroke, Penacook, Phillips, Pinardville, Pleasantville, Plymouth, Pomona, Portage, Posen, Powderhorn, Prescott, Ramsey, Rapids, Raymond, Richfield, Robbins, Roseland, Rosemount, Roseville, Salem, Sandown, Sauk, Savage, Schererville, Schertz, Seward, Shavano, Shingle, Somers, Star, Station, Stone, Suncook,

194

Tangletown, Tinley, Vadanis, Ventnor, Victory, Warner, Weare, Whiting, Whittier, Willard-Hay, Wilton, Windcrest, Windham, Winfield, Woodbury

6.2.l: Proximity C.ND Aldine, Alief, Atascocita, Bellaire, Channelview, Clear, Crosby, Cy, Cypress, Deer, Fresno, Friendswood, Galena, Harris, Highlands, Houston, Humble, Jacinto, Jersey, Meadows, Mission, Pasadena, Pearland, Porte, Southbelt, Spring, Sugar, Uptown, Webster, Westside

6.2.m: Proximity D.CW Allentown, Amboy, Armonk, Bayville, Bedford, Belle, Benson, Bordentown, Branchburg, Bridgeport, Brunswick, Buren, Camano, Canaan, Carlsborg, Chappaqua, Charleston, Chesterfield, Chimacum, Cleburne, Clinton, Cochise, Colts, Commack, Concordia, Coupeville, Cranbury, Darien, David, Dunellen, Eastchester, Easton, Edison, Egypt, Elmwood, Elwood, Ewing, Fairfield, Fairview, Farmingdale, Fords, Franklin, Freehold, Glen, Greenbank, Greenlawn, Greenville, Greenwich, Hadlock-Irondale, Halesite, Hamilton, Hanover, Hansville, Harrison, Hartsdale, Haven, Hawthorne, Heathcote, Heber, Hermann, Holmdel, Hopewell, Howell, Huntington, Jackson, Jamesburg, Jefferson, Katonah, Keasbey, Kendell, Keyport, Kings, Kingston, Kisco, Langley, Larchmont, Laurence, Lawrenceville, Levittown, Liberty, Locust, Ludlow, Magoffin, Mamaroneck, Manalapan, Mansfield, Martinsville, Matawan, Metuchen, Middlesex, Millstone, Millwood, Muttontown, Neshanic, Northport, Norwalk, Norwich, Oak, Paintsville, Piscataway, Plainfield, Pleasantville, Prestonburg, Princeton, Purchase, Redding, Reno, Ridgefield, Rivers, Robbinsville, Rochelle, Rossmoor, Rye, Salem, Salonga, Salyersville, Sayreville, Scarsdale, Sequim, Skillman, Somerville, Spanish, Sparks, Spotswood, Springs, Stamford, Stanwood, Stony, Stratford, Syosset, Thornwood, Titusville, Townsend, township, Trenton, Trumbull, Union, Valhalla, Valleys, Warrenton, Washington, Washoe, Weston, Westport, Whetstone, White Plains, Wilton, Windsor, Woodbury, Woodside, Wright, Yardley

6.2.n: Proximity D.CD Ada, Alton, Archbald, Ariel, Ashley, Avoca, Blakely, Boise, Carbondale, Charlestown, Clarks, Cochise, Dallas, Daviess, Dickson, Dunmore, Dupont, Eagle, Easton, Ellensburg, Elum, Emporia, Exeter, Factoryville, Forest, Forge, Forty, Fry, Gouldsboro, Greentown, Greenwich, Griswold, 195

Hahira, Hatfield, Hereford, Hopeville, Huachua, Jennings, Jermyn, Jessup, Kingston, Kingstown, Kittitas, Kuda, Lackawanna, Lakeland, Ledyard, Lisbon, Livermore, London, Lowndes, Lyon, Maceo, Meridian, Montville, Mosaic, Moscow, Mystic, Nampa, Narragansett, Newfoundland, Noank, Norwich, Owensboro, Pittston, Plymouth, Preston, Quitman, Ray, Remerton, Richmond, Rockport, Roslyn, Scranton, Shavertown, Sierra, Snoqualmie, Star, Statenville, Stockton, Stonington, Suncadia, Swoyersville, Taylor, Thorp, Throop, Tobyhanna, Tombstone, Tunkhannock, Uncasville, Valdosta, Vantage, Voluntown, Washington, Waterford, Waymart, Westerly, Wilkes-Barre, Wyoming

6.2.o: Proximity D.LD Allouez, Anne, Ashwaubenon, Assonet, Attleborough, Balalrd, Ballantyne, Barrington, Bellevue, Bellingham, Belmont, Benton, Berkley, Bethel, Bitter, Blackstone, Blanchard, Bossier, Bothell, Boulder, Brandon, Bremerton, Brier, Bristol, Brown, Burien, Burrillville, Caddo, Canton, Carnation, Casas, Cascade-Fairwood, Cashon, Catalina, Cathcart, Charlotte, Chase, Chico, Chute, Clark, Clyde, Columbia, Concord, Cornelius, Cortaro, Cottage, Coventry, Covington, Cramerton, Cranston, Creek, Cumberland, Davidson, Dell, Denmark, Dighton, Downtown, Doyline, Eastgate, Eastwood, Edmonds, Elm, Enterprise, Esperance, Exeter, Fall, Flowing, Foster, Franken, Franklin, Freetown, Fremont, Frierson, Garreston, Gastonia, Glochester, Gorst, Green, Greenleaf, Greenville, Greenwood, Harrisburg, Hartford, Haughton, Henderson, Hobart, Holly, Houghton, Howard, Huntersville, Indian, Issaquah, Johnston, Keithville, Kenmore, Kent, Keyport, King, Kingston, Kingstown, Kirkland, Klahanie, Lennox, Lincoln, Luxemburg, Lynbrook, Maltby, Manchester, Mansfield, Martha, Marvin, Matthews, Mecklenburg, Medina, Mercer, Midland, Mill, Millville, Minnehaha, Mint, Mirrormont, Moines, Mooringsport, Mountlake, Newcastle, Newell, Norton, Olalla, Orchard, Oro, Paradise, Pere, Pima, Pineville, Plainville, Portsmouth, Poulsbo, Princeton, Providence, Pulaski, Rainier, Rancho, Redmond, Renton, Rumford, Sahuarita, Sammamish, Scituate, SeaTac, Seattle, Seekonk, Seymour, Shoreline, Shreveport, Silverado, Silverdale, Sioux, Sloan, Smithfield, Sobieski, Somerset, Spring, Stanley, Stonewall, Suamico, Summerlin, Summit, Sunrise, Suquamish, Swansea, Tanque, Taunton, Tea, Tega, Tiverton, Tortolita, Tracyton, Tukwila, Tuscon, Union, Vail, Valencia, Vegas, Ward, Warren, Warwick, Waskom, Weddington, Wedgewood, Wesley, Whitney, Winchester, Woodinville, Woonsocket, Wrentham, Wrightstown, Wylie 196

6.2.p: Proximity D.ND Altadena, Angeles, Arcadia, Bel, Burbank, Compton, Covina, Downey, Gardena, Glendale, Habra, Hawthorne, Lakewood, Lynwood, Manhattan, Monte, Montebello, Monterey, Norwalk, Nuys, Pasadena, Pick, Rowland, Sherman, Torrance, Venice

6.2.q: Proximity E.CW Andrews, Arcade, Ashford, Aurora, Bancroft, Blairsville, Blue, Boston, Bricelyn, Bruce, Buffalo, Burt, Calhoun, Cameron, Chaffee, Cherokee, Coffeeville, Collins, Cruces, Derby, Dona, Erie, Evans, Falls, Forest, Gowanda, Hamburg, Hayesville, Hiawassee, Maeser, Manzanita, Martinsville, Mesilla, Moundsville, Murphy, Nehalem, Orchard, Pacific, Paden, Radium, Sistersville, Springville, Swea, Tillamook, Uintah, Valley, Vernal, Wetzel, Wheeler, Winnebago, Woodsfield

6.2.r: Proximity E.CD Abington, Alamogordo, Allenstown, Allentown, Amsterdam-Churchill, Anna, Ashland, Atco, Auburn, Bala, Barnstead, Barrington, Bayou, Bedford, Belgrade, Belmont, Bensalem, Berlin, Blennerhassett, Bordentown, Boscawen, Boston, Bourg, Bow, Bozeman, Bridesburg, Bristol, Burlington, Camden, Candia, Canterbury, Chaffee, Chauvin, Cherry, Chesterfield, Chichester, Churchville, Columbus, Concord, Contoocook, Corners, Croydon, Cutler, Debtford, Deerfield, Delanco, Devola, Dulac, Echelon, Edgewater, Egypt, Elizabeth, Ephraim, Epsom, Erdenheim, Ewing, Florence, Flourtown, Frankford, Franklin, Germantown, Gibbsboro, Gibson, Gilmanton, Girardeau, Glendora, Glenside, Glouchester, Goffstown, Gold, Grants, Gray, Haddon, Haddonfield, Hainesport, Henniker, Hillsborough, Holly, Hooksett, Hopkinton, Horse, Horsham, Houma, Huntingdon, Jackson, Jacksonville, Jonesboro, Josephine, Kensington, Langhorne, Larose, Laurel, Lawrence, Levittown, Lindenwold, Lockport, Loudon, Lumberton, Luz, Manchester, Manhattan, Mansfield, Marietta, Marlton, Matthews, Medford, Mellmawr, Mercerville-Hamilton, Merlin, Merrimack, Mills, Mineral, Montegut, Moorestown, Newton, Northfield, Northwood, Olney, Oreland, Otero, Parkersburg, Pemberton, Pembroke, Penacook, Pennsauken, Philadelphia, Pinardville, Pine, Pittsfield, Raceland, Ramblewood, Redwood, Reedsville, Richboro, Robbinsville, Rogue, Roslyn, Runnemede, Sanbornton, Schriever, Scott, Severly, Shamong, Somerton, Southampton, Stratford, Suncook, Tabernacle, Terrebonne, 197

Thibodaux, Titusville, Trenton, Tularosa, Tullytown, Turnersville, Vienna, Voorhees, Warminster, Warner, Warrington, Waterford, Weare, Webster, Westampton, Westville, Williams, Willingboro, Wood, Woodbourne, Woodlynne, Woodside, Yardley

6.2.s: Proximity E.LD Albuquerque, Arbutus, Baltimore, Bangor, Beaverton, Belleair, Bellevue, Beloit, Belvidere, Bernalillo, Boring, Bosque, Bucksport, Burnie, Byron, Carmel, Carter, Cedar, Chalco, Clackamas, Clearwater, Cockeysville, Columbia, Corinth, Corrales, Council, Cully, Damascus, Dell, Dixmont, Douglas, Dundalk, Dunedin, Edgewood, Egypt, Eldersburg, elkhorn, Elkridge, Ellenton, Ellicott, Essex, Etna, Fallston, Felida, Ferndale, Finksburg, Frankfort, Gandy, Garden, Gate, Gibson, Gibsonton, Gladstone, Glen, Glenburn, Glenwood, Gretna, Hermon, Highpoint, Hillsboro, Hudson, Kingsville, Largo, Laurel, Lealman, Levant, Loves, Lutherville-Timonium, Machesney, Madeira, Marsh, Meade, Memphis, Metzger, Milford, Millard, Millersville, Monkton, Multnomah, Oatfield, Odenton, Oldsmar, Omaha, Orchards, Oregon, Orland, Orono, Owings, Palmetto, Papillion, Parrish, Pecatonica, Penobscot, Peralta, Petersburg, Pikesville, Pinellas, Placitas, Plain, Plattsmouth, Portland, Pueblo, Pumphrey, Rancho, Reisterstown, Richmond, Ridgecrest, Ridgefield, Riverview, Riviera, Rock, Rockford, Rockton, Roscoe, Rossville, Ruskin, Salmon, Sandia, Scappoose, Severn, Severna, Sherwood, Sparrows, Sunnyside, Tampa, Tierra, Tigard, Tijeras, Towson, Troutdale, Tualatin, Valley, Vancouver, Vista, Washington, Washougal, Westchase, Westgate, Wilkes, Wilsonville, Winnebago, Winterport, Woodlawn

6.2.t: Proximity E.ND Accokeek, Annandale, Arbor, Aspen, Beltsville, Bethesda, Bowie, Brandywine, Bryans, Burke, Calverton, Clinton, College, Crofton, Croom, Fairfax, Fairland, Galthersburg, Groveton, Highland, Hyattsville, Hybla, Kettering, Lakes, Lanham, Laurel, Lorton, Marlboro, Marlton, McLean, Mitchellville, Newington, Oakton, Olney, Petworth, Potomac, Reston, Rockville, Spring, Springfield, Springs, Suitland-Silver, Travilah, Vienna, Washington

198

Appendix 10: IRB Exemption Notice

The IRB exempted the qualitative structured interview script in January, 2013, claiming it did not constitute human subjects research. Their exemption is attached. The title of the dissertation has since changed, but the interview script itself was identical.

199

200

REFERENCES

Abdelrehim, N., Maltby, J., & Toms, S. (2011). Corporate Social Responsibility and Corporate Control: The Anglo-Iranian Oil Company, 1933–1951. Enterprise and Society, 12(4), 824-862. doi: 10.1093/es/khr032 Ahmad, A. N. (2010). Is Twitter a useful tool for journalists? [Article]. Journal of Media Practice, 11(2), 145-155. doi: 10.1386/jmpr.11.2.145_1 Anderson, B. (2006). Imagined Communities: Reflections on the Origin and Spread of Nationalism (Revised Edition ed.). London: Verso. Aparaschivei, P. A. (2011). The Use of New Media in Electoral Campaigns: Analysis on the Use of Blogs, Facebook, Twitter and YouTube in the 2009 Romanian Presidential Campaign. [Article]. Journal of Media Research, 4(2), 39-60. Arduser, L. (2011). Warp and Weft: Weaving the Discussion Threads of an Online Community. [Article]. Journal of Technical Writing & Communication, 41(1), 5-31. doi: 10.2190/TW.41.1.b Arpan, L. M., & Tuzunkan, F. (2011). Photographic Depiction of Normative Deviance and Informational Utility as Predictors of Protest News Exposure, Related Perceptions, and Story Comprehension. [Article]. Mass Communication & Society, 14(2), 178-195. doi: 10.1080/15205431003650486 Aust, P. J. (2004). Communicated values as indicators of organizational identity: A method for organizational assessment and its application in a case study. Communication Studies, 55(4), 515-534. doi: 10.1080/10510970409388636 Badii, N., & Ward, W. J. (1980). The Nature of News in Four Dimensions. Journalism Quarterly, 57, 243-248. Ballotti, J., & Kaid, L. L. (2000). Examining verbal style in presidential campaign spots. Communication Studies, 51(3), 258-273. doi: 10.1080/10510970009388523 Benoit, W. L., Hemmer, K., & Stein, K. (2010). New York Times' Coverage of American Presidential Primary Campaigns, 1952-2004. [Article]. Human Communication, 13(4), 259-280. Bhattacharyya, A. (2011). IS NATION'S WATCHDOG REALLY PEEPING INTO THE NATION? [Article]. Global Media Journal: Indian Edition, 1-10. Bishop, R. (2009). "Little More than Minutes": How Two Wyoming Community Newspapers Covered the Construction of the Heart Mountain Internment Camp. [Article]. American Journalism, 26(3), 7-32. Bleske, G. L. (1991). Ms. Gates Takes Over. [Article]. Newspaper Research Journal, 12(4), 88- 97. Bodle, J. T., Burriss, L., Farwell, T., Hammaker, S., & Joshi, J. (2011). Gender and Journal Scholarship in Mass Communication: How Well Are Women Doing? A Twenty-year Content Analysis. [Article]. Journalism & Mass Communication Educator, 66(2), 118- 133. Bowd, K. (2011). 4. Reflecting regional life: Localness and social capital in Australian country newspapers. [Article]. Pacific Journalism Review, 17(2), 72-91. Bowman, L. (2008). RE-EXAMINING "GATEKEEPING". [Article]. Journalism Practice, 2(1), 99-112. doi: 10.1080/17512780701768550

201

Boyle, M. P., & Armstrong, C. L. (2009). Measuring Level of Deviance: Considering the Distinct Influence of Goals and Tactics on News Treatment of Abortion Protests. [Article]. Atlantic Journal of Communication, 17(4), 166-183. doi: 10.1080/15456870903156134 Breed, W. (1955). Social Control in the Newsroom: A Functional Analysis. Social Forces, 33, 326-355. Breen, M. J. (1997). A cook, a cardinal, his priests, and the press: Deviance as a trigger for intermedia agenda setting. [Article]. Journalism & Mass Communication Quarterly, 74(2), 348-356. Breen, M. J. (1997). A cook, a cardinal, his priests, and the press: Deviance as a trigger for intermedia agenda setting. [Article]. Journalism & Mass Communication Quarterly, 74(2), 348-356. Brennen, B., & dela Cerna, E. (2010). JOURNALISM IN SECOND LIFE. [Article]. Journalism Studies, 11(4), 546-554. doi: 10.1080/14616701003638418 Bridges, J. A. (1989). News Use on the Front Pages of the American Daily. Journalism Quarterly, 66, 332-337. Bridges, J. A., & Bridges, L. W. (1997). Changes in News Use on the Front Pages of the American Daily Newspaper, 1986-1993. Journalism & Mass Communication Quarterly, 74(4), 826-838. Briones, R., Catona, D., Keefe, B., Zimbaldi, C., & Pollock, J. (2008). Nationwide Newspaper Coverage of Childhood Obesity: A Community Structure Approach. [Article]. Conference Papers -- National Communication Association, 1. Brockus, S. (2009). Rethinking the Reader's Role in Community Journalism. Paper presented at the Newspapers and Community-Building Symposium XV, Mobile, Alabama. Bruns, A. (2012). JOURNALISTS AND TWITTER: HOW AUSTRALIAN NEWS ORGANISATIONS ADAPT TO A NEW MEDIUM. [Article]. Media International Australia (8/1/07-current)(144), 97-107. Burd, G. (1979). What is community? Grassroots Editor. 20(1), 3-14. Burd, G. (2008). THE MEDIATED METROPOLIS AS MEDIUM AND MESSAGE. [Article]. International Communication Gazette, 70(3/4), 209-222. doi: 10.1177/1748048508089948 Burroughs, A. (2006). Civic Journalism and the Community Newspaper: Opportunities for Civic and Social Connections. Master of Mass Communication, Louisiana State University. Candel, D. (2013). Advanced literacy and the place of literary semantics in secondary education: A tool of fictional analysis. [Article]. Semiotica, 2013(195), 305-329. doi: 10.1515/sem- 2013-0031 Cassidy, W. P. (2006). Gatekeeping Similar For Online, Print Journalists. [Article]. Newspaper Research Journal, 27(2), 6-23. Chan, J. M., & Lee, C. C. (1984). The journalistic paradigm on civil protests: A case study of Hong Kong. In A. Arno & W. Dissanayake (Eds.), The news media in national and international conflict (pp. 183-202). Boulder, CO: Westview. Cho, J., Boyle, M. P., Keum, H., Shevy, M. D., McLeod, D. M., Shah, D. V., & Pan, Z. (2003). Media, Terrorism, and Emotionality: Emotional Differences in Media Content and Public Reactions to the September 11th Terrorist Attacks. Journal of Broadcasting & Electronic Media, 47(3), 309-327. doi: 10.1207/s15506878jobem4703_1

202

Christensen, C. (2011). Discourses of Technology and Liberation: State Aid to Net Activists in an Era of 'Twitter Revolutions'. [Article]. Communication Review, 14(3), 233-253. doi: 10.1080/10714421.2011.597263 Chung, D. S., & Nah, S. (2009). The Effects of Interactive News Presentation on Perceived User Satisfaction of Online Community Newspapers. [Article]. Journal of Computer-Mediated Communication, 14(4), 855-874. doi: 10.1111/j.1083-6101.2009.01473.x Coleman, R. (2007). Picturing civic journalism: How photographers and graphic designers visually communicate the principles of civic journalism. Journalism, 8(1). Connelly, L. (2012, September 30, 2012). Surfing dogs brave the waves in Huntington Beach, The Orange County Register. Retrieved from http://www.ocregister.com/news/waves- 373188-dogs-record.html Conway, M. (2006). THE SUBJECTIVE PRECISION OF COMPUTERS: A METHODOLOGICAL COMPARISON WITH HUMAN CODING IN CONTENT ANALYSIS. [Article]. Journalism & Mass Communication Quarterly, 83(1), 186-200. Cope, L. (2011). From ethnocultural pride to promoting the Texas Czech vernacular: current maintenance efforts and unexplored possibilities. [Article]. Language & Education: An International Journal, 25(4), 361-383. doi: 10.1080/09500782.2011.578003 Corrigan, D. M. (1990). Value Coding Consensus in Front Page News Leads. Journalism Quarterly, 67(4), 653-662. Cover, R. (2005). Engaging sexualities: Lesbian/gay print journalism, community belonging, social space and physical place. [Article]. Pacific Journalism Review, 11(1), 113-132. Crew, J. R. E., & Lewis, C. (2011). Verbal Style, Gubernatorial Strategies, and Legislative Success. Political Psychology, 32(4), 623-642. doi: 10.1111/j.1467-9221.2011.00832.x Daley, P. J., & James, B. (1988). Framing the News: Socialism as Deviance. [Article]. Journal of Mass Media Ethics, 3(2), 37-46. David, P. (1996). ROLE OF IMAGERY IN RECALL OF DEVIANT NEWS. [Article]. Journalism & Mass Communication Quarterly, 73(4), 804-820. De Vreese, C. H., Peter, J., & Semetko, H. A. (2001). Framing Politics at the Launch of the Euro: A Cross-National Comparative Study of Frames in the News. [Article]. Political Communication, 18(2), 107-122. doi: 10.1080/105846001750322934 Dill, R. K., & Wu, H. D. (2009). Coverage of Katrina in Local, Regional, National Newspapers. Newspaper Research Journal, 30(1). Don, W. (2011). Jokes inviting more than laughter Joan Rivers political-rhetorical world view. Comedy Studies, 2(2), 139-150. doi: 10.1386/cost.2.2.139_1 Donohue, G. A., Olien, C. N., & Tichenor, P. J. (1985). Reporting Conflict by Pluralism, Newspaper Type and Ownership. [Article]. Journalism Quarterly, 62(3), 489-507. Donohue, G. A., Olien, C. N., & Tichenor, P. J. (1989). Struture and constraints on community newspaper gatekeepers. Journalism Quarterly, 66(4), 807-845. Donohue, G. A., Tichenor, P. J., & Olien, C. N. (1995). A guard dog perspective on the role of media. Journal of Communication, 45, 115-132. Editor and Publisher's International Yearbook: The Encyclopedia of the Newspaper Industry. (2008). (88 ed.). New York, NY: Editor and Publisher. Eilders, C. (2006). News factors and news decisions. Theoretical and methodological advances in Germany. [Article]. Communications: The European Journal of Communication Research, 31(1), 5-24. doi: 10.1515/commun.2006.002

203

Elliott, C. W., & Greer, C. F. (2010). Newsworthiness and Islam: An Analysis of Values in the Muslim Online Press. [Article]. Communication Quarterly, 58(4), 414-430. doi: 10.1080/01463373.2010.525153 Embler, W. (1962). Language and Truth. In S. I. Hayakawa (Ed.), The Use and Misuse of Language (pp. 219-234). New York, NY: Premier Books. Enghels, R., & Jansegers, M. (2013). On the crosslinguistic equivalence of sentir(e) in Romance languages: A contrastive study in semantics. [Article]. Linguistics, 51(5), 957-991. doi: 10.1515/ling-2013-0034 Entman, R. M. (2003). Cascading Activation: Contesting the White House's Frame After 9/11. [Article]. Political Communication, 20(4), 415. Epstein, E. J. (1973). News from nowhere: Television and the news. New York: Random House. Ferguson, D. A., & Greer, C. F. (2011). Local Radio and Microblogging: How Radio Stations in the U.S. are Using Twitter. [Article]. Journal of Radio & Audio Media, 18(1), 33-46. doi: 10.1080/19376529.2011.558867 Fernandes, J., Giurcanu, M., Bowers, K. W., & Neely, J. C. (2010). The Writing on the Wall: A Content Analysis of College Students' Facebook Groups for the 2008 Presidential Election. [Article]. Mass Communication & Society, 13(5), 653-675. doi: 10.1080/15205436.2010.516865 Firmstone, J. (2008). THE EDITORIAL PRODUCTION PROCESS AND EDITORIAL VALUES AS INFLUENCES ON THE OPINIONS OF THE BRITISH PRESS TOWARDS EUROPE. [Article]. Journalism Practice, 2(2), 212-229. doi: 10.1080/17512780801999378 Fontana, A., & Frey, J. H. (1994). Interviewing: The Art of Science. In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of Qualitative Research. Thousand Oaks, CA: Sage Publications. Fontana, A., & Frey, J. H. (2005). The Interview: From Neutral Stance to Political Involvement. In N. K. Denzin & Y. S. Lincoln (Eds.), The Sage Handbook of Qualitative Research (3 ed., pp. 695-727). Thousand Oaks, CA: Sage Publications. Fuller, J. (1996). News Values: Ideas for an Informative Information Age. Chicago: University of Chicago Press. Funk, M. (2010). The 30-50-60 Curve: Locally Generated News Content in Small and Medium- Sized American Dailies. Grassroots Editor, 51(4). Funk, M. (2012). Imagined commodities? Analyzing local identity and place in American community newspaper website banners. New Media & Society. doi: 10.1177/1461444812458433 Funk, M. (2013). The Next President is at the Front Door Again: An Analysis of Local Media Coverage of the 2012 Republican Iowa Caucus, New Hampshire Primary and South Carolina Primary. Community Journalism, 2(1), 26-48. Gade, P., & Krug, C. (2008). What Community Newspapers Can Learn from the Fast-Growing Free U.S. Daily Industry. Paper presented at the Newspapers and Community Building Symposium XIV, St. Paul, Minnesota. Galtung, J., & Ruge, M. (1965). The structure of foreign news: the presentation of the Congo, Cuba and Cyprus crises in four Norweigan newspapers. Journal of International Peace Research, 1, 64-91. Gans, H. J. (1979). Deciding what's news. New York, NY: Pantheon.

204

Garcia, E. P. (2007). Interactivity in Argentinean Online Newsrooms. [Article]. Zer: Revista de Estudios de Comunicacion, 12(23), 7-25. Garfrerick, B. H. (2010). The Community Weekly Newspaper: Telling America's Stories. American Journalism, Summer 2010. Gieber, W. (1956). Across the desk: A study of 16 telegraph editors. Journalism Quarterly, 33(4), 423-432. Gilligan, E. (2011). Online Publication Expands Reach of Community Journalism. [Article]. Newspaper Research Journal, 32(1), 63-70. Givens, D. T. (2012). Mr. and Mrs. John Doe announce the engagement of...: A study of the decline in reader-submitted content in four Eastern Kentucky community newspapers. Grassroots Editor, 53(3-4), 19-24. Gladney, G. A. (1996). How Editors and Readers Rank and Rate the Importance of Eighteen Traditional Standards of Newspaper Excellence. Journalism & Mass Communication Quarterly, 73(2), 285-303. Goddard, C. (2012). Semantic primes, semantic molecules, semantic templates: Key concepts in the NSM approach to lexical typology. [Article]. Linguistics, 50(3), 711-743. doi: 10.1515/ling-2012-0022 Gorton, W., & Diels, J. (2010). Is political talk getting smarter? An analysis of presidential debates and the Flynn effect. Public Understanding of Science. doi: 10.1177/0963662509357010 Greer, C. F., & Ferguson, D. A. (2011). Using Twitter for Promotion and Branding: A Content Analysis of Local Television Twitter Sites. [Article]. Journal of Broadcasting & Electronic Media, 55(2), 198-214. doi: 10.1080/08838151.2011.570824 Greer, J. D., & Yan, Y. (2010). New ways of connecting with readers: How community newspapers are using Facebook, Twitter and other tools to deliver the news. [Article]. Grassroots Editor, 51(4), 1-7. Gupta, P. (2012, October 23, 2012). Weird news: New species of ferns named after Lady Gaga. Salon.com. Hanitzsch, T. (2005). Journalists in Indonesia: educated but timid watchdogs. [Article]. Journalism Studies, 6(4), 493-508. doi: 10.1080/14616700500250396 Hansen, E. K. (2007). Community insights on newspaper Web sites: What readers want. Paper presented at the Newspapers and Community-Building Symposium XIII, Norfolk, Virginia. Hansen, E. K., & Hansen, G. L. (2011). Newspaper Improves Reader Satisfaction By Refocusing on Local Issues. Newspaper Research Journal, 32(1). Hansen, E. K., & Hansen, G. L. (2012). Community crisis and community newspapers: A case study of The Licking Valley Courier. Grassroots Editor, 53(3-4), 8-12. Harb, Z. (2011). Arab Revolutions and the Social Media Effect. [Article]. M/C Journal, 14(2), 2- 2. Harcup, T., & O'Neill, D. (2001). What Is News? Galtung and Ruge revisited. [Article]. Journalism Studies, 2(2), 261-280. doi: 10.1080/14616700120042114 Harry, J. C. (2001). Covering Conflict: A Structural-Pluralist Analysis of How a Small-Town and a Big-City Newspaper Reported an Environmental Controversy. Journalism & Mass Communication Quarterly, 78(3), 419-436. Hart, R. P. (2013, Augut 21, 2013). [Diction 6.0 question from Journalism Ph.D.?].

205

Heckert, D. M., & Heckert, D. A. (2007). Positive Deviance. In G. Ritzer (Ed.), Blackwell Encyclopedia of Sociology: Blackwell Publishing. Hermida, A. (2010). Twittering the News. [Article]. Journalism Practice, 4(3), 297-308. doi: 10.1080/17512781003640703 Hirsch, P. M. (1977). Occupational, Organizational and Institutional Models in Mass MEdia Research. In P. M. Hirsch, P. M. Miller & F. G. Kline (Eds.), Strategies for Communication Research. Beverly Hills, California: Sage. Hoffer, T. W., & Butterfield, G. A. (1976). The Right to Reply: A Florida First Amendment Abberation. [Article]. Journalism Quarterly, 53(1), 111-116. Holt, L. F., & Major, L. H. (2010). FRAME AND BLAME: AN ANALYSIS OF HOW NATIONAL AND LOCAL NEWSPAPERS FRAMED THE JENA SIX CONTROVERSY. [Article]. Journalism & Mass Communication Quarterly, 87(3/4), 582-597. Hume, J. (2005). Life and Death in a Small Town: Cultural Values and Memory in Community Newspaper Obituaries. Grassroots Editor, 46(4), 24-29. Ifukor, P. (2010). “Elections” or “Selections”? Blogging and Twittering the Nigerian 2007 General Elections (Vol. 30, pp. 398-414). Jarvis, S. E. (2004). Partisan patterns in presidential campaign speeches, 1948–2000. Communication Quarterly, 52(4), 403-419. doi: 10.1080/01463370409370209 Jeffres, L. W., & Dobos, J. (1983). Neighborhood Newspaper Audiences. [Article]. Newspaper Research Journal, 4(2), 31-42. Jeffres, L., Cutietta, C., Sekerka, L., & Lee, J.-W. (2000). Newspapers, pluralism, & diversity in an urban context. Mass Communication & Society, 3(2/3), 157-184. Jeffres, L., Horowitz, E., Bracken, C. C., Jian, G., Neuendorf, K. A., & Yoon, S. (2011). Structural Pluralism and the Community Context: How and When Does the Environment Matter? [Article]. Mass Communication & Society, 14(6), 787-815. doi: 10.1080/15205436.2010.515371 Jensen, R., & Uddameri, V. (2009). Using communication research to gather stakeholder preferences to improve groundwater management models: a South Texas case study. [Article]. JCOM: Journal of Science Communication, 8(1), 1-8. Johnson, T. J., & Kelly, J. D. (2003). Have New Media Editors Abandoned the Old Media Ideals? The Journalistic Values of Online Newspaper Editors. [Article]. New Jersey Journal of Communication, 11(2), 115-134. Johnson, T., & Perlmutter, D. (2009). “The Facebook Election: New Media and the 2008 Election Campaign” Special Symposium. Mass Communication and Society, 12(3), 375-376. doi: 10.1080/15205430903021525 Jong Hyuk, L. (2008). EFFECTS OF NEWS DEVIANCE AND PERSONAL INVOLVEMENT ON AUDIENCE STORY SELECTION: A WEB-TRACKING ANALYSIS. [Article]. Journalism & Mass Communication Quarterly, 85(1), 41-60. Jong Hyuk, L. (2009). NEWS VALUES, MEDIA COVERAGE, AND AUDIENCE ATTENTION: AN ANALYSIS OF DIRECT AND MEDIATED CAUSAL RELATIONSHIPS. [Article]. Journalism & Mass Communication Quarterly, 86(1), 175- 190. Jong Hyuk, L., & Yun Jung, C. (2009). News values of sports events: an application of a newsworthiness model on the World Cup coverage of US and Korean media. [Article]. Asian Journal of Communication, 19(3), 302-318. doi: 10.1080/01292980903039012 206

Joye, S. (2010). News media and the (de)construction of risk: How Flemish newspapers select and cover international disasters. [Article]. Catalan Journal of Communication & Cultural Studies, 2(2), 253-266. doi: 10.1386/cjcs.2.2.253_1 Kennedy, J. (2010). DON'T YOU FORGET ABOUT ME. [Article]. Journalism Studies, 11(2), 225-242. doi: 10.1080/14616700903290635 Kepplinger, H. M., & Ehmig, S. C. (2006). Predicting news decisions. An empirical test of the two-component theory of news selection. [Article]. Communications: The European Journal of Communication Research, 31(1), 25-43. doi: 10.1515/commun.2006.003 Kim, H. S. (2012). War journalists and forces of gatekeeping during the escalation and the de- escalation periods of the Iraq War. [Article]. International Communication Gazette, 74(5), 323-341. doi: 10.1177/1748048512439815 Kitch, C. (2007). Selling the 'Authentic Past': The New York Times and the Branding of History. [Article]. Westminster Papers in Communication & Culture, 4(4), 24-41. Knight, M. (2012). Journalism as usual: The use of social media as a newsgathering tool in the coverage of the Iranian elections in 2009. [Article]. Journal of Media Practice, 13(1), 61- 74. doi: 10.1386/jmpr.13.1.61_1 Kong, T. S. K., Mahoney, D., & Plummer, K. (2002). Queering the interview. In J. Gubrium & J. Holstein (Eds.), Handbook of qualitative research: Context and method (pp. 239-258). Thousand Oaks, CA: Sage. Kraeplin, C. (2008). Two Tales of One City: How Cultural Perspective Influenced the Framing of a Pre-Civil Rights Story in Dallas. [Article]. American Journalism, 25(1), 73-97. Krippendorff, K. (2004). Content Analysis: An Introduction to its Methodology (Second Edition ed.). Thousand Oaks, CA: Sage Publications. Larsson, A. O., & Moe, H. (2012). Studying political microblogging: Twitter users in the 2010 Swedish election campaign. [Article]. New Media & Society, 14(5), 729-747. doi: 10.1177/1461444811422894 Lasorsa, D. L., Lewis, S. C., & Holton, A. E. (2011). Normalizing Twitter. Journalism Studies, 13(1), 19-36. Lauderdale, P., & Estep, R. E. (1980). The bicentennial protest: An examination of hegemony in the definition of deviant political activity. In P. Lauderdale (Ed.), A Political Analysis of Deviance. Minneapolis: University of Minnesota Press. Lauterer, J. (2006). Community Journalism: Relentlessly Local. Chapel Hill, North Carolina: University of North Carolina Press. Lauterer, J. (2012). Toto, I don't think we're (just) in Kansas anymore: How U.S. community newspapers are serving as a model for China. Grassroots Editor, 53(3-4), 1-7. Lee, F. L. F., Chin-Chuan, L., & Nina Luzhou, L. (2011). Chinese Peasants in the Process of Economic Reform: An Analysis of New York Times's and Washington Post's Opinion Discourses, 1981-2008. [Article]. Communication, Culture & Critique, 4(2), 164-183. doi: 10.1111/j.1753-9137.2011.01097.x Lewin, K. (1947). Frontiers in Group Dynamics II. Channels of group life: Social Planning and Action Research. Human Relations, 1, 143-145. Lewis, J., & Cushion, S. (2009). THE THIRST TO BE FIRST. [Article]. Journalism Practice, 3(3), 304-318. doi: 10.1080/17512780902798737 Lewis, S. C. (2008). News, Nationalism, and the Imagined Community. [Article]. Journalism Studies, 9(3), 409-428. doi: 10.1080/14616700801999212

207

Lewis, S. C., & Reese, S. D. (2009). WHAT IS THE WAR ON TERROR? FRAMING THROUGH THE EYES OF JOURNALISTS. [Article]. Journalism & Mass Communication Quarterly, 86(1), 85-102. Lewis, S. C., Kaufhold, K., & Lasorsa, D. L. (2009). Thinking About Citizen Journalism. Journalism Practice, 4(2), 163-179. Little, C. B. (2007). Deviance, Absolutist Definitions of. In G. Ritzer (Ed.), Blackwell Encyclopedia of Sociology: Blackwell Publishing. Lowrey, W., Brozana, A., & Mackay, J. B. (2008). Toward a Measure of Community Journalism. [Article]. Mass Communication & Society, 11(3), 275-299. doi: 10.1080/15205430701668105 Luke, D. A., Caburnay, C. A., & Cohen, E. L. (2011). How Much Is Enough? New Recommendations for Using Constructed Week Sampling in Newspaper Content Analysis of Health Stories. [Article]. Communication Methods & Measures, 5(1), 76-91. doi: 10.1080/19312458.2010.547823 Martin, S. R. (1988). Proximity of Event as Factor In Selection of News Sources. [Article]. Journalism Quarterly, 65(4), 986-1043. Matthes, J. (2009). Framing Responsibility for Political Issues: The Preference for Dispositional Attributions and the Effects of News Frames. [Article]. Communication Research Reports, 26(1), 82-86. doi: 10.1080/08824090802637114 McCombs, M. (2004). Setting the Agenda: The Mass Media and Public Opinion. Cambridge, UK: Polity Press. McCombs, M., & Funk, M. (2011). Shaping the Agenda of Local Daily Newspapers: A Methodology Merging the Agenda Setting and Community Structure Perspectives. [Article]. Mass Communication & Society, 14(6), 905-919. doi: 10.1080/15205436.2011.615447 McCombs, M., & Shaw, D. (1972). The Agenda-Setting Function of Mass Media. [Article]. Public Opinion Quarterly, 36(2), 176-187. McCracken, G. (1993). The Long Interview. Newbury Park, CA: Sage. McKinney, B. C., Kelly, L., & Duran, R. L. (2012). Narcissism or Openness?: College Students‟ Use of Facebook and Twitter. [Article]. Communication Research Reports, 29(2), 108- 118. doi: 10.1080/08824096.2012.666919 Miladi, N. (2011). Tunisia: A media led revolution? [Article]. Media Development, 58(2), 8-12. Miliband, R. (1969). The State in Capitalist Society. London: Wiedenfeld & Nicholson. Miller, C., Rainie, L., Purcell, K., Mitchell, A., & Rosenstiel, T. (2012). A Closer Look: How people get local news and information in different communities: Pew Research Center. Mitchell, E., Pollock, J., Schumacher, S., & de Zutter, L. (2006, 2006 Annual Meeting). Gay Rights: Nationwide Newspaper Coverage of Gay Adoption, Gay Marriage, and Gays in the Boy Scouts: A Community Structure Approach. Mitchell, H. (2012, October 25, 2012). Live leopard shark drops from sky onto golf course, The Los Angeles Times. Retrieved from http://www.latimes.com/sports/sportsnow/la-sp-sn- shark-golf-course-20121025,0,7711527.story Momoc, A. (2012). Perception of Romanian Journalists on Mass-media and Democracy. [Article]. Romanian Journal of Journalism & Communication / Revista Romana de Jurnalism si Comunicare- RRJC, 12-19. Mowlana, H. (1996). Global communication in transition: The end of diversity? London: Sage.

208

Murphy, A. C. (2013). On "true" portraits of Letters to Shareholders --- and the importance of phraseological analysis. [Article]. International Journal of Corpus Linguistics, 18(1), 57- 81. doi: 10.1075/ijcl.18.1.06mur Nah, S., & Armstrong, C. L. (2011). Structural Pluralism in Journalism and Media Studies: A Concept Explication and Theory Construction. [Article]. Mass Communication & Society, 14(6), 857-878. doi: 10.1080/15205436.2011.615446 Napier, J. (2011). ''It's not what they say but the way they say it''. A content analysis of interpreter and consumer perceptions towards signed language interpreting in Australia. [Article]. International Journal of the Sociology of Language, 2011(207), 59-87. doi: 10.1515/ijsl.2011.003 Nuendorf, K. A. (2002). The Content Analysis Guidebook. Thousand Oaks, CA: Sage Publications. Olien, C. N., Donohue, G. A., & Tichenor, P. J. (1978). Community Structure and Media Use. [Article]. Journalism Quarterly, 55(3), 445-455. Östgaard, E. (1965). Factors Influencing the Flow of News. Journal of Peace Research, 2(1), 39- 63. doi: 10.1177/002234336500200103 ÖZkoÇAk, L., & Tuna, Y. (2011). A Content Analysis: Environment Themes and Tools in Newspapers Advertisements. [Article]. Online Journal of Communication & Media Technologies, 1(3), 1-13. Paletz, D. L., & Entman, R. M. (1981). Media, Power, Politics. New York: Free Press. Papacharissi, Z. (2009). The virtual geographies of social networks: a comparative analysis of Facebook, LinkedIn and ASmallWorld. New Media & Society, 11(1-2), 199-220. doi: 10.1177/1461444808099577 Pentzold, C. (2011). Imagining the Wikipedia community: What do Wikipedia authors mean when they write about their „community‟? [Article]. New Media & Society, 13(5), 704- 721. doi: 10.1177/1461444810378364 Pinto, J. G. (2009). DIFFUSING AND TRANSLATING WATCHDOG JOURNALISM. [Article]. Media History, 15(1), 1-16. doi: 10.1080/13688800802583240 Poindexter, P. M., & McCombs, M. E. (2000). Research in Mass Communication: A Practical Guide. Boston & New York: Bedford / St. Martin's. Pollock, J. C. (2007). Tilted Mirrors: Media Alignment with Political and Social Change - A Community Structure Approach. Cresskill, N.J.: Hampton Press, Inc. Pollock, J., Piccillo, C., Leopardi, D., Gratale, S., & Cabot, K. (2005). Nationwide Newspaper Coverage of Islam Post-September 11: A Community Structure Approach. [Article]. Communication Research Reports, 22(1), 15-27. doi: 10.1080/0882409052000343480 Pontell, H. N. (2007). Deviance, Reactivist Definitions of. In G. Ritzer (Ed.), Blackwell Encyclopedia of Sociology: Blackwell Publishing. Porpora, D. V., Nikolaev, A., & Hagemann, J. (2010). Abuse, Torture, Frames, and the Washington Post. [Article]. Journal of Communication, 60(2), 254-270. doi: 10.1111/j.1460-2466.2010.01481.x Pritchard, D., & Hughes, K. D. (1997). Patterns of deviance in crime news. [Article]. Journal of Communication, 47(3), 49. Protess, D. L., Leff, D. R., Brooks, S. C., & Gordon, M. T. (1985). Uncovering Rape: The Watchdog Press and the Limits of Agenda Setting. [Article]. Public Opinion Quarterly, 49(1), 19-37.

209

Reader, B. (2006). Distinctions that Matter: Ethical Differences at Large and Small Newspapers. Journalism and Mass Communication Quarterly, 83(4), 860. Reader, B. (2007). Air Mail: NPR Sees "Community" in Letters From Listeners. [Article]. Journal of Broadcasting & Electronic Media, 51(4), 651-669. Reese, S. D. (1990). The News Paradigm and the Ideology of Objectivity: A Socialist at the Wall Street Journal. Critical Studies in Mass Communication, 7, 390-409. Reese, S. D., & Lewis, S. C. (2009). Framing the War on Terror: The internalization of policy in the US press. [Article]. Journalism, 10(6), 777-797. doi: 10.1177/1464884909344480 Regier, T., Khetarpal, N., & Majid, A. (2013). Inferring semantic maps. [Article]. Linguistic Typology, 17(1), 89-105. doi: 10.1515/lingty-2013-0003 Reinemann, C., & Schulz, W. (2006). Introduction to the special issue: News decisions and news values. [Article]. Communications: The European Journal of Communication Research, 31(1), 1-4. doi: 10.1515/commun.2006.001 Richardson, J. E., Huckerby, S., & Williams, M. (2008). NATION, CULTURE AND IDENTITY IN A FRENCH-LANGUAGE CORSICAN NEWSPAPER. Journalism Studies, 9(4), 561-577. doi: 10.1080/14616700802114324 Rossow, M. D. (2009). Connecting with the community: Hiring the right "face" for the newspaper. Paper presented at the Newspapers and Community Building Symposium XV, Mobile, Alabama. Roulston, K. (2010). Reflective Interviewing: A Guide to Theory and Practice. London: SAGE Publications Ltd. Sayed, N. (2011). Towards the Egyptian Revolution: Activists' perceptions of social media for mobilization. [Article]. Journal of Arab & Muslim Media Research, 4(2&3), 273-298. doi: 10.1386/jammr.4.2-3.273_1 Schulz, W. (1976). Die Konstruktion von Realitat in den Nachrichtenmedien: Analyse der aktuellen Berichterstattung [The construction of reality in the news media. An analysis of the news coverage]. Freiburg: Alber. Schweitzer, J. C., & Smith, B. L. (1991). Community Pressures on Agenda-Setting. [Article]. Newspaper Research Journal, 12(3), 46-62. Semetko, H. A., & Valkenburg, P. M. (2000). Framing European politics: A content analysis of press and television news. [Article]. Journal of Communication, 50(2), 93. Shariq. (2013). Exploring the Semantic Changes in Persian Loan Words Used in Brass Industry of Moradabad. [Article]. Language in India, 13(3), 371-380. Shoemaker, P. (1983). Bias and Source Attribution. [Article]. Newspaper Research Journal, 5(1), 25-31. Shoemaker, P. (1984). Media Treatment of Deviant Political Groups. [Article]. Journalism Quarterly, 61(1), 66-82. Shoemaker, P. (2006). News and newsworthiness: A commentary. [Article]. Communications: The European Journal of Communication Research, 31(1), 105-111. doi: 10.1515/commun.2006.007 Shoemaker, P. (1996). Hardwired for News: Using Biological and Cultural Evolution to Explain the Surveillance Function. Journal of Communication, 46(3), 32-47. doi: 10.1111/j.1460- 2466.1996.tb01487.x Shoemaker, P., & Cohen, A. A. (2006). News around the world: Content, practitioners, and the public. New York: Routledge.

210

Shoemaker, P., & Reese, S. (1996). Mediating the Message: Theories of Influence on Media Content (2 ed.). White Plains, N.Y.: Longman. Shoemaker, P., & Vos, T. (2009). Gatekeeping Theory: Routledge. Shoemaker, P., Chang, T. K., & Brendlinger, N. (1987). Deviance as a predictor of newsworthiness: Coverage of international events in the U.S. media. In M. McLaughlin (Ed.), Communication yearbook (Vol. 10, pp. 348-365). Newbury Park, CA: Sage. Shoemaker, P., & Danielian, L. (1991). Deviant Acts, Risky Business and U.S. Interests: The Newsworthiness of World Events. [Article]. Journalism Quarterly, 68(4), 781-795. Shuang, L., & Louw, E. (2009). Cultural Translation and Identity Performance: A Case of Chinese Business People in Australia. [Article]. China Media Research, 5(1), 1-9. Sigal, L. (1973). Reporters and officials: The organization and politics of news-making. Lexington, MA: D.C. Heath. Smethers, S., Bressers, B., Willard, A., Harvey, L., & Freeland, G. (2007). Kansas Readers Feel Loss When Town's Paper Closes. Newspaper Research Journal, 28(4), 6-21. Society, A. C. (2013). Cancer Facts & Figures 2013. Atlanta: American Cancer Society. Stassen, W. (2010). Your news in 140 characters: exploring the role of social media in journalism. [Article]. Global Media Journal: African Edition, 4(1), 1-16. Stewart, C. J., & William B. Cash, J. (2000). Interviewing: Principles and Practices (9th ed.). Boston: The McGraw-Hill Companies, Inc. Stewart, J. (2011). Discovering Black Film History: Tracing the Tyler, Texas Black Film Collection (Vol. 23, pp. 147-173): Indiana University Press. Stone, G. C., & Banning, S. (1997). Public perceptions of newspaper's watchdog role. [Article]. Newspaper Research Journal, 18(1/2), 86-102. Tichenor, P. J., Donohue, G. A., & Olien, C. N. (1973). Mass communication research: Evolution of a structural model. Journalism Quarterly, 50, 419-425. Tichenor, P. J., Donohue, G. A., & Olien, C. N. (1980). Community conflict and the press. Beverly Hills, CA: Sage. Towner, T. L., & Dulio, D. A. (2012). New Media and Political Marketing in the United States: 2012 and Beyond (Vol. 11, pp. 95-119). Tuchman, G. (1973). Making News by Doing Work: Routinizing the Unexpected. American Journal of Sociology, 79(1), 110-131. Tuchman, G. (1978). Making News. New York: Free Press. U.S. Census Bureau. (2012). from U.S. Department of Commerce http://www.census.gov/# Watson, B. R., & Riffe, D. (2011). Structural Determinants of Local Public Affairs Place Blogging: Structural Pluralism and Community Stress. [Article]. Mass Communication & Society, 14(6), 879-904. doi: 10.1080/15205436.2011.611922 Wilson, K. G., & Dufrene, T. (2010). things might go terribly, horribly wrong: a guide to life liberated from anxiety. Oakland, CA: New Harbinger Publications, Inc. Weber, R. P. (1990). Basic Content Analysis (Second Edition ed.). Newbury Park, CA: SAGE Publications. White, D. M. (1950). The 'Gate Keeper': A Case Study in the Selection of News. Journalism Quarterly, 27(3). Woolley, J. K., Limperos, A. M., & Oliver, M. B. (2010). The 2008 Presidential Election, 2.0: A Content Analysis of User-Generated Political Facebook Groups. Mass Communication and Society, 13(5), 631-652. doi: 10.1080/15205436.2010.516864

211

Wu, G., Hu, X., & Wu, Y. (2010). Effects of Perceived Interactivity, Perceived Web Assurance and Disposition to Trust on Initial Online Trust. [Article]. Journal of Computer-Mediated Communication, 16(1), 1-26. doi: 10.1111/j.1083-6101.2010.01528.x Yamamoto, M. (2011). Community Newspaper Use Promotes Social Cohesion. [Article]. Newspaper Research Journal, 32(1), 19-33.

Yoo, C. Y. (2011). Modeling Audience Interactivity as the Gratification-Seeking Process in Online Newspapers. [Article]. Communication Theory (10503293), 21(1), 67-89. doi: 10.1111/j.1468-2885.2010.01376.x

212