Citation
Corporate crises in cyberspace

Material Information

Title:
Corporate crises in cyberspace extending public relations media monitoring to the public dialogues on the usenet
Creator:
Dzwo, Tzong-Horng, 1964-
Publication Date:
Language:
English
Physical Description:
x, 134 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Content analysis ( jstor )
Corporations ( jstor )
Internet ( jstor )
Journalism ( jstor )
News content ( jstor )
News media ( jstor )
Newsgroups ( jstor )
Newspapers ( jstor )
Public opinion ( jstor )
Public relations ( jstor )
Dissertations, Academic -- Mass Communication -- UF ( lcsh )
Mass Communication thesis, Ph.D ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1998.
Bibliography:
Includes bibliographical references (leaves 125-133).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Tzong-Horng Dzwo.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
029539621 ( ALEPH )
40074479 ( OCLC )

Downloads

This item has the following downloads:


Full Text










CORPORATE CRISES IN CYBERSPACE:
EXTENDING PUBLIC RELATIONS MEDIA MONITORING TO THE PUBLIC DIALOGUES ON THE USENET















By

TZONG-HORNG DZWO














A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1998













ACKNOWLEDGMENTS

Earning a Ph.D. from a university in a foreign country is not easy. The academic demands are challenging enough without the added pressures of an unfamiliar language and culture. While eager to return to my homeland, I am none-the-less proud of my achievement at the University of Florida and thankful for all the support I have received during my academic journey in America.

First and foremost, I wish to offer a tribute to Dr. John Wright who restored my confidence and encouraged me to remain for the fourth year. He absolutely altered my view of American teachers. Without him, I would have struggled continuously in prospectus writing and sobbed in a dark comer of the past. He deserves all the warm sunshine emanating from my heart.

To say that Dr. John Sutherland is my major advisor is an understatement, because he not only advised me on my prospectus and dissertation writing, but guided me to a new world without hatred and revenge. He shared his brilliant insights with me on mass communication theories and philosophies of life. He also showed me how to behave like a responsible teacher and researcher in the future. Dedicated to him for his strict but friendly guidance in my final year is this Chinese motto: Be a teacher one day, be a Fatherforever!

Never is it enough to say "thank you" to Dr. Marilyn Roberts. She makes me feel that she has more confidence in me than I do myself. All my knowledge, if any, about








political communication and American presidential campaigns is attributed to her caring and warm teaching. She constantly draws a bright rainbow at the other side of the tunnel whenever I feel downcast. The more I look back at the past, the greater debt I owe to her.

The original idea for this dissertation about crisis management came from Dr. Baker, a public relations expert who provided useful tips on crisis communication and literature. The theoretical framework of this dissertation was based upon her lectures and ideas about crisis management.

Dr. Michael Martinez deserves special thanks for his insightful advice on

prospectus writing and instructions in understanding the dimensions of public opinion in terms of political communication. Leading about political behavior became enjoyable with his discipline.

My sincere kudos are paid to three doctoral colleagues: Walt McDowell, for sparing no time in proofreading my prospectus and most of my dissertation text; Tom Kelleher and Michelle O'Malley, for their kindness in coding messages for the reliability check. I also feel very lucky to have two friends in the Graduate Division, Jody Hedge and Margi Hatch, who helped me with all the paper work with which I was unfamiliar during my four years at the college.

I owe a lot to my lovely wife, who had to take care of my cute, but naughty, children from the first day 1 arrived at this school. Without her, the pursuit of this doctoral degree would have been out of the question. Finally, I also thank my parents, who offer me all the support I need, financially and spiritually. They also exemplify why they are the greatest parents of all in my mind. Hence, this doctoral degree is solely dedicated to them.


iii














TABLE OF CONTENTS
page


A CKN OW LED GM EN TS .................................................................................................. ii

LIST OF TABLES ............................................................................................................. vi

LIST OF FIGURE S .......................................................................................................... viii

A BSTRA CT ....................................................................................................................... ix

CHAPTERS

1. INTRODU CTION .............................................................................................. I

Overview ............................................................................................................. 1
Problem D efined ................................................................................................ 2
Research Purpose ............................................................................................... 3
D issertation Outline ........................................................................................... 5

H. LITERA TURE REVIEW ................................................................................... 7

Review of A genda-setting Theory ...................................................................... 8
The Internet: A Brief History, Functions, and Current Situation ...................... 16
The Challenge of M onitoring the Internet ........................................................ 20
The Internet Research ....................................................................................... 23

111. CRISIS COMMUNICATION LITERATURE ............................................... 28

Overview .......................................................................................................... 28
Crisis M anagem ent Research and Public Opinion ........................................... 31
Sturges' Public Opinion Model of Crisis Management ................................... 38
M acK uen's Strategy M odel of Public D ialogue .............................................. 40
Research Questions .......................................................................................... 44
Research Hypotheses ....................................................................................... 46

IV .M ETH OD OLO GY .......................................................................................... 50

Overview .......................................................................................................... 50
Corporate Crisis Selection ............................................................................... 51


iv








N ew s Coverage Analysis ................................................................................. 52
M edia Selection for Analysis of N ew s Stories ................................................ 53
On-line D iscussion M essage Analysis ............................................................. 54
Definitions of Coding Categories .................................................................... 57
Conceptual D efinitions .................................................................................... 59
Operational D efinitions .................................................................................... 61
Independent and Dependent Variables ............................................................. 63
Introduction of Com puter Software: DICTION ............................................... 65

V RESULTS ......................................................................................................... 67

Descriptive Analysis ......................................................................................... 67
M attel Cabbage Patch D oll D efect Case ........................................................... 70
Test of Hypotheses ............................................................................................ 72
Hudson Foods M eat Contam ination Case ......................................................... 79
Test of Hypotheses ............................................................................................ 83
Post Hoc Analysis ............................................................................................. 89

VI. DISCU SSION ................................................................................................. 101

Sum m ation of Findings ................................................................................... 102
Research Questions ......................................................................................... 103
Proposed Crisis Com m unication M odel ......................................................... 106
Lim itations ...................................................................................................... 108
Im plications and Suggestions for Future Study .............................................. 110



APPENDICES

A HOME PAGE OF DEJA NEWS SEARCH ENGINE ................................ 116
B POWER SEARCH PAGE OF DEJA NEWS SEARCH ENGINE ............. 117
C CODING SCHEM E FOR NEW S STORIES .............................................. 118
D CODING SCHEME FOR ON-LINE MESSAGES ..................................... 120
E SAMPLE OF REPORT FILE IN DICTION SOFTWARE ........................ 122


REFEREN CES ................................................................................................................ 125

BIOGRAPHICAL SKETCH ........................................................................................... 134









v













LIST OF TABLES

Table pqge

3-1. Payoff matrix for player encountering friend ............................................................ 42

3-2. Payoff matrix for player encountering friend ............................................................. 42

5-1. Frequencies of news stories on each news media in the Mattel case ....................... 71

5-2. Frequencies of messages appearing on each sampling day in the Mattel case .......... 73

5-3. Results of Hypothesis 1 test in the Mattel case ........................................................ 74

5-4. Results of Hypothesis 2 test in the Mattel case .......................................................... 74

5-5. Results of first Hypothesis 4 test in the Mattel case .................................................. 75

5-6. Results of second Hypothesis 4 test in the Mattel case ............................................. 76

5-7. Results of Hypothesis 5 test in the Mattel case ........................................................ 76

5-8. Results of Hypothesis 6 test in the Mattel case .......................................................... 77

5-9. Results of Hypothesis 7 test in the Mattel case ........................................................ 77

5 10. Frequencies of news stories on each news media in Hudson Foods case ................. 81

5-11. Frequencies of messages appearing on each sampling day in Hudson Foods case. .82 5-12. Results of Hypothesis 1 test in the Hudson Foods case ............................................ 83

5-13. Results of Hypothesis 2 test in the Hudson Foods case ............................................ 84

5 -14. Results of comparison of three stages in Hypothesis 4 in the Hudson Foods Case .85 5-15. Results of comparison of two stages in Hypothesis 4 in the Hudson Foods Case. ..85 5-16. Results of comparison of two stages in Hypothesis 4 in the Hudson Foods Case ... 86



vi








5-17. Results of Hypothesis 5 test in the Hudson Foods case ............................................ 86

5-18. Results of comparison of two stages in Hypothesis 6 in the Hudson Foods Case ... 87 5 -19. Results of comparison of two stages in Hypothesis 6 in the Hudson Foods Case... 8 8 5-20. Results of Hypothesis 7 test in the Hudson Foods case ............................................ 88

5-21. Results of Chi-square test with combined data ........................................................ 99









































vii













LIST OF FIGURES

Figure pM,e

3-1. Group opinion form ation process ............................................................................. 39

3-2. Sturges'model of public opinion with crisis management plan ................................ 40

3-3. Causal relationship between the media agenda and the public agenda ..................... 47

5-1. Bivariate correlation between the number of negative on-line messages and the
Optim ism Score in the M attel case ............................................................................ 78

5-2. Bivariate correlation between the number of negative on-line messages and the
Optimism Score in the Hudson Foods case .............................................................. 89

5-3. Relationship between the media agenda and the public agenda in the Mattel case ... 90 5-4. Relationship between the media agenda and the public agenda in the Hudson
F ood s case .................................................................................................................. 90

5-5. Cubic relationship between the media agenda and the public agenda in the Hudson
F o o d s case ................................................................................................................. 9 1

5-6. Distribution of numbers of on-line messages by types in the Mattel case ................ 95

5-7. Distribution of numbers of on-line messages by types in the Hudson Foods case .... 95 5-8. Frequencies of combined on-line messages and media stories ................................. 98

5-9. Frequencies of combined negative on-line messages and the Optimism Scores in
m edia stories ............................................................................................................. 99

6-1. An optimal crisis communication model dealing with public agenda and media
agen d a ..................................................................................................................... 10 8







viii














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy CORPORATE CRISES IN CYBERSPACE: EXTENDING PUBLIC RELATIONS MEDIA MONITORING TO THE PUBLIC DIALOGUES ON THE USENET By

Tzong-Horng Dzwo

August 1998

Chairman: John C. Sutherland
Ma or Department: Mass Communication

Managing unexpected crises is a primary concern of the public relations

profession. With the remarkable growth of on-line communication using the Internet, companies can no longer deal exclusively with conventional media such as newspapers, radio, and television. Keeping a crisis situation under control has become an even greater challenge in the age of cyberspace. This dissertation applies basic agenda-setting theory to two crisis cases involving the interaction of conventional media coverage and the USENET dialogues by on-line users. The on-line dialogues were sampled from a search engine on the Internet. The conventional media stories were gathered from the Lexis/Nexis data base. Two corporate crises selected for testing the hypotheses were the Mattel Doll Defect case and the Hudson Foods Meat Contamination case.

The crisis communication model predicted that after a crisis happens, public

opinion about the company would be more negative than before its occurrence. As the


ix








company adopted interventions to ameliorate the detrimental impact caused by the crisis, the number of negative on-line dialogues would decrease. Unlike the traditional method of examining the public agenda in agenda-setting studies, content analysis of the public dialogues on the USENET, a dependent variable, was employed in comparison to the conventional media agenda, an independent variable. The number of on-line dialogues was assumed to have a corresponding relationship with the media coverage of the crisis.

While the number of negative on-line messages about the company increased

after the crisis and were reduced due to intervention strategies adopted by each company, the total number of on-line messages did not decrease after the intervention in the Hudson Foods case.

Results in the Mattel case revealed a significant correlation between the number of media coverage and that of on-line messages. The significant correlation between the tone of media stories and the number of negative on-line messages was also found in the same case. However, this relationship that derived from agenda-setting theory was not detected in the Hudson Foods case.

Despite the specific findings in each case, further examination of the combined

data supported the basic components of agenda-setting theory. There were correspondent relationships between the media reports and the public agenda as well as between the tone of media stories and the number of negatively toned on-line messages.












x













CHAPTER I
INTRODUCTION


Overview

The topic of media effects has been a primary research domain for over three decades. Among the many theoretical approaches that have been proposed, one of the most popular has been agenda-setting theory. Hundreds of studies have utilized aspects of the agenda-setting framework to reveal positive correlations between media coverage of issues and their distinctive importance in people's minds.

The development of agenda-setting theory can be understood from two levels of research. The first level study can be traced back to 1972 when McCombs and Shaw wanted to explain how and why people think about different social issues and rank their importance. The findings showed a significant correlation between people's ranking of issues' importance in their minds and the amount of media coverage. In other words, the salience of issues in people's minds is greatly influenced by and transmitted from the media coverage.

Besides the original focus on the issue salience transmitted from media to the

public's mind, media effects in view of agenda-setting theory have also been found from another perspective. The selection of objects and attributes of an issue in the news media becomes a major concern. How news frames and agenda attributes impact the public agenda plays a central role in the second level study of agenda-setting studies (McCombs,



1





2

1997). The research question in the second level has moved from "Does the amount of media coverage transmit the issue salience to people's minds?" to "To what extent is the public's view of an issue shaped by the media, that frame objects or attributes of the issue in different ways?" To flifther understand the media effects on people's interpretation of social issue(s), researchers look into the content of media coverage, instead of the number of media stories on particular issues, and then examine how the news content has been framed in ways that most affect people. The key difference between the first level and the second level approaches to agenda-setting theory is in the way researchers study the media coverage.

The central concept of agenda-setting theory is an explanation of the causal

influence of the media agenda on the public agenda. Regardless of the level of agendasetting research that measures the media agenda from different aspects, the most common methodology is the survey. For example, researchers will ask people "What is the most important issue facing the society or yourself" or "How do you think about a particular issue or person?" These responses are compared with media content analyses to determine if the media content influences people's thinking about different issues.

Problem Defined

Whether the survey is conducted by telephone or mail, this method can be

somewhat obtrusive and may not capture the true opinions of subjects. What people really think or say about the issue(s) in the interest of the researchers may not be completely detected by asking conventional survey questions.





3

In contrast to surveying, content analysis has been available as a less obtrusive method for analyzing messages for many years. Researchers have used content analysis to evaluate the intentions or opinions of the professional communicators such as newspapers, magazine, radio, television, and so on. However, no studies have used content analysis to probe the expressed public opinions of the receivers of these mediated messages about some particular issues within agenda-setting theory. With the introduction of new interactive media technologies, particularly the Internet, ordinary people can now express their views to their fellow citizens without the use of the conventional media. Based upon the public relations crisis management concept combined with new communications technology, this research will employ content analysis to measure the public agenda for the purpose of extending the dimensions of agenda-setting theory.

Research PMose

Given the Jnternet's dynamic characteristic of transforming information senders to receivers, this new electronic delivery system provides both possibilities and problems for the public relations professional. In particular, the Internet offers special challenges when a company experiences some type of crisis.

The public relations field has long been criticized for its lack of a formal and

consistent theoretical framework to help its empirical practice; neither is there a mutual understanding between public relations theorists and practitioners (Terry, 1989). This tension is exacerbated when discussing crisis management communication. Because public relations crisis communication demands an accurate depiction of public opinion





4

during the crisis, the Internet needs to become an integral part of contemporary agendasetting theory.

By modifying the concepts of agenda-setting theory founded by McCombs and Shaw in 1972, the current study will examine the relationship between the traditional media coverage of a corporate crisis and public opinion expressed on the USENET, one of the popular functions on the Internet. While previous agenda-setting studies examined public opinion by either conducting cross-sectional telephone surveys or using secondary public opinion surveys, this study will break new ground by abandoning survey methodologies in favor of content analysis of the on-line public messages on the USENET. This new type of analysis will permit public relations practitioners to "eaves drop" on people's conversations about the crisis event without disturbing their communication. This method also will consolidate the theoretical framework of agendasetting theory in the domain of the public agenda. In addition, the time frame effects of media coverage of corporate crises on the on-line public discussions that occur daily on the Internet are examined.

Another goal of this study is to understand the flow of on-line discussions when a corporation is in the process of a crisis and how corporate crisis management strategy impacts the flow of on-line discussions.

The examination of on-line messages as opposed to traditional media news

coverage will provide preliminary insights into understanding the formation of on-line public opinion. Comparison of the media agenda and the public on-line agenda is needed to have a better understanding of how these two factors interact.





5

After analysis of two case studies involving both the media agenda and the public agenda concerning a corporate crisis, this research outlines a dynamic crisis communication model that is intended to help companies prevent further damage as a result of an organizational crisis. Based on a public opinion model of crisis management proposed by Sturges (1994), this restructured model will provide greater insights into how to effectively manage public opinion and control the crisis to the advantage of the corporation.

Dissertation Outline

The following chapter describes the essential concepts and development of

agenda-setting theory that forms the basis of this study. Studies related to the Internet are presented as well. In the third chapter, crisis management literature is reviewed to understand the points of view of both scholars and public relations practitioners about how to manage the corporate crises. To apply a general public opinion model to the flow of on-line discussion messages, Sturges' (1994) model along with MacKuen's strategy model of public dialogue is introduced. The research questions and hypotheses will conclude the third chapter.

Chapter four discusses the methodology used in this study: content analyses of both traditional media coverage and on-line discussion messages. It also explains the interceder reliability test, independent and dependent variables, and conceptual definitions. The fifth chapter presents the findings of hypothesis testing for two corporate crises and post-hoc analysis. In the final chapter, a summation of the results, the





6

discussion of findings, limitations of this study, and the practical implications and suggestions for future research are discussed.













CHAPTER H
LITERATURE REVIEW


When a crisis occurs, the media can shape the general public's perception about the targeted organization and how it deals with a particular crisis. To better understand how mass media affect people's perceptions about the organization, a sound theoretical background is necessary. A viable theory will help public relations professionals understand not only how their company is portrayed in the media, but to what extent the general public has been affected by the media coverage of the crisis.

Having been tested for nearly three decades in the mass communication field, agenda-setting theory emphasizes a unidirectional impact from the media agenda to the public agenda. Applying this theory to the crisis management communication, public relations professionals are better able to detect the trend of public opinion and how long-term media coverage influences people's opinions as a crisis event evolves. After understanding how the public thinks about an issue, the corporation can then adopt appropriate strategies to manage and ultimately resolve the crisis.

The following chapter will first review the agenda-setting theory in detail, followed by a discussion of pertinent studies of the Internet.










7





8


Review of Agenda-Setting Theo

First Level Research

Agenda-setting theory was originally proposed by McCombs and Shaw in 1972

when they conducted their pioneering study in Chapel Hill, NC. In that study, the amount of media coverage of various issues within a period of time was compared with the public's ranking of their importance. The results revealed the powerful effects of media on transmitting the salience of issues to people's minds.

Three types of agenda are involved in the agenda-setting research: media agenda, public agenda, and policy agenda. The hierarchical effects derived from this theory are that the media agenda would influence the public agenda, which in turn may influence policy agenda (Dearing & Rogers, 1996). Sometimes a fourth variable is studied in agenda-setting studies: a real-world indicator, that Dearing & Rogers (1996) defines as "a variable that measures more or less objectively the degree ofseverity or risk ofsocial problem" (p. 23). However, since the real-world indicator varies depending on the issue topic, this variable to test agenda-setting theory is seldom used. Measurement of Media and Public Agenda

Media agenda are usually measured by content analysis of the news media,

initially employed by McCombs and Shaw (1972) and Funkhouser (1973), to determine the number of news items about an issue or issues of study. The more news coverage is devoted to an issue, the more salient that issue is considered to be for the media. In this case, number counting is a very important way to measure media agenda. With growing popularity of computer usage in recent years, computerized content analysis has become





9


an efficient alternative to counting news stories or identifying words in a text when they appear together. It also facilitates qualitative analysis in some cases.

Public agenda were originally measured by asking a question in public opinion surveys: "What is the most important problem facing this country today?" This type of question has been used for decades in national public opinion surveys. It can indicate the relative importance of an issue in the public's mind. In contrast to the measurement of media agenda, the researchers will determine whether people take cues from the amount of media coverage to answer the WIP" (most important problem)-type question. Intervening Variables in Affecting the Public Agenda

Although the original agenda-setting study has specified the direct relationship between the media agenda and the public agenda, many other factors intervening in the causal relationship have been found in various studies including intensity and variation in news coverage. Brosius and Kepplinger (1990) found that agenda-setting effects were most likely to occur when coverage was intense and when there was significant variation in the coverage from month to month. In addition, the type of issue is another focus in the agenda-setting study. Wanta and Hu (1993) found that press coverage, besides increasing public concern with certain issues, can also decrease concerns. It mainly depends upon recipients' self-involvement and interest in the issues.

Media credibility is another determining factor. Rogers and Dearing (1988) concluded that the Herald-Tribune, a local newspaper, transmitted more salience of a news item to the subjects than the National Enquirer, a national tabloid magazine. Wanta and Hu (19 94) also indicated that individuals who perceive the media as more credible





10


would rely more on the media for information about issues and hence be more susceptible to the media influencing their understanding of issues.

The extent of media exposure is another intervening variable. The more media exposure, the more possible the public agenda will be influenced by the mass media (Weaver et al., 1981; Mullins, 1977; Wanta & Hu, 1994). People's need for orientation also transfers the media agenda to the public agenda. Zucker (1978) pointed out that if people have fewer direct experiences with an issue, the news media's influence on public opinion about that issue will be greater than otherwise. By analyzing television and newspaper issue coverage for four weeks, Wanta and Wu (1992) found that interpersonal communication can reinforce the media agenda-setting effects on the public agenda when the conversation deals with the same issue that media have emphasized. Studies of Intermedia Influence

Not only do newspapers and television play a central role in setting media agenda, but the wire services set the agenda for newspapers as well. The coefficient of concordance between the contents of 24 Iowa daily newspapers across 13 categories and those of the Associated Press amounted to 0.915 in Gold and Simmons' (1965) study. After analyzing 20 afternoon newspapers around the nation, Stempel (1964) found that average use of the Associated Press news items by these papers was 22 percent, ranging from a low of 11 percent by a New York paper to 34 percent by the Rochester TimesUnion.

Whitney and Becker (19 82) examined the gatekeeping effects of wire service

news among 46 editorial managers from newspapers and commercial television stations. They concluded that at the least local media are greatly influenced by "the decisions of a





11


relatively few editors operating at the regional, national and international bureaus of the wire services" (p. 65). Reese and Danielian (1989) studied the media coverage of the drug issue in 1986 and provided evidence of strong intermedia agenda-setting effects from the New York Times to other television and newspaper organizations. Time Frame Issues in Agenda-Setting Research

How long will an issue remain salient in people's minds? One of the most

essential considerations in testing agenda-setting hypotheses is the time frame utilized by the researchers in their studies (Eaton, 1989; Funkhouser, 1973; Mullins, 1977; Sohn, 1978; Stone & McCombs, 1981; Wanta & Hu, 1994). In general, time-lag selection is important because it demonstrates the time-varying causal effects. The length of time taking effects from the media agenda to the public agenda depends upon factors such as nature of the issue, total amount of media coverage, personal relevance, and so on (Eyal, 1979; Mazur, 1987). Salwen (1988) urged that any time discrepancies in the measurement of the public agenda may affect the public's evaluations of issue salience.

Agenda-setting studies have to be concerned with the time frame over which media coverage has the most impact on public opinion. Winter and Eyal (198 1) suggested the "optimal effect span" is between 4 and 6 weeks, whereas Stone and McCombs (19 8 1) thought that it takes two to six months for changes in the media agenda to be fully "translated" to the public agenda. Shoemaker et al. (1989) used two months as a basis to analyze the relationship between drug coverage and public opinion. Shoemaker et al. (1989) proposed that coverage which recurs in emphasis on a three- or four-month schedule may have the most influence on public opinion. Consequently, the causal relationship between media and public opinion will be detected within a period of time





12


when media repeatedly cover one particular issue. As the above findings dealt with the traditional media agenda-setting effects, no research exists in previous agenda-setting studies regarding the time frame of on-line newsgroups and issue salience. Direction of Causal Relationship between Media Agenda and Public Agenda:

With respect to the relationship between the media agenda and the public agenda, Ito (1993) believed that "mass media effects will not take place unless media stand on the majority side or the mainstream in the tripolar relationship among the mass media, the government and the masses" (p. 123). According to Behr and Iyengar's (1985) finding, the direction of causality between media coverage and public awareness of the issue was unidirectional, that is, media coverage on the issues had quite an impact on people's awareness of them. Beniger (1978) maintained that the public's attitudes and opinions may be more closely correlated with media coverage than with more objective social conditions. Indeed, Zhu et al. (1993) found that the public's issue priorities are much more influenced by media coverage than by social interaction.

Willnat and Zhu (1996) also provided evidence of one-sided newspaper coverage influencing public opinion about a governor's overall performance when they compared time-series data of public opinion polls with three leading newspapers in Hong Kong. To investigate the impact of press coverage on the general public's belief about HIV transmission, Hertog and Fan (1995) discovered a significant causal relationship from news contents to the public opinion, but not vice versa. From these findings, agendasetting theory suggests to public relations professionals the importance of monitoring media coverage of issues that have grabbed public attention.





13


On the first level of agenda-setting theory, the researchers put much emphasis on what media say about different issues and then correlate it with how people rank the importance of those issues. The content of media is not examined in particular. Most studies focus on the amount of news coverage in the media versus public opinion about the examined issues. The public agenda is measured by surveying people's opinion, instead of actually analyzing how and what people say about the issues. The core concept of the first level agenda-setting theory is "where public opinion about an issue comes from" (Whitney & Becker, 1982; McCombs & Shaw, 1972; Wanta & Hu, 1994; Zucker, 1978).

Second Level Research

In addition to the original focus on issue salience transmitted from the media to the public's mind, agenda-setting theory studies have moved to the second level. How news stories are framed and issue attributes are presented play a central role in impacting the public's understanding of the issues (McCombs, 1997). On the second level, the researchers attempt to answer the question: "To what extent is the public's view of an issue shaped by the media that frame objects or attributes of the issue in different ways?"

To frame, as Entman (1993) explains, "is to select some aspects of a perceived

reality and make them more salient in a communicating text, in such a way as to promote a particular problem definition, causal interpretation, moral evaluation andlor treatment recommendation for the item described" (p. 52). Because all issues have different dimensions, they can be interpreted from all kinds of perspectives. Through selection and use of some attributes or objects of an issue, media can make some news prominent and transmit that prominent aspect to the audience's mind.





14


Given that the agenda-setting theory has progressed from counting numbers

(media agenda) and ranking issues (public agenda) to analyzing and explaining the textual meanings of the media coverage, it can be assumed that the media are not only telling us what to think about (Cohen, 1975), but also tell us how to think about some objects or attributes of particular issues.

To understand the conceptualization of media frames and attributes, Ghanem (1997) breaks down four dimensions of media frames (p. 10): 1. The topic of a news item (what is included in the frame);

2. Presentation (size and placement);

3. Cognitive attributes (details of what is included in the frame);

4. Affective attributes (tone of the picture).

Although the idea of news framing is quite novel in recent years, McCombs

(1997) provided some research studies confirming the effects of agenda-setting on the second level. In regard to issue framing, Takeshita and Mikami (1995) found a highly positive correlation between the salience of system-related aspects of reform on the public agenda and people's attentiveness to political news on both newspapers and television news. Maher (1995) showed a high correspondence in the relative salience of the attributes that defined the pictures of local environmental issues in the newspaper and among the public.

In earlier studies, Weaver, Graber, McCombs, and Eyal (1981) found a striking degree of correspondence between the agenda of attributes in the Chicago Tribune and the agenda of attributes in Illinois voters' descriptions of Jimmy Carter and Jerry Ford. The correspondence between the agenda of attributes in Newsweek and the agenda of





15


attributes in New York Democrats' descriptions of the contenders for the presidential nomination has also been reported by Becker and McCombs (1978).

Concerning candidate attributes framed in the news stories, King (1997) compared newspapers agenda on candidates for Taipei mayor with voters' images of three candidates and revealed six significant correlations ranged from 0.59 to 0.75. McCombs et al. (1998) studied the second-level effects of agenda setting in the Spanish election and found significant correspondence between various news and political advertising agendas and the pictures of the parliamentary and mayoral candidates in voters' minds. In that study, the strongest effects resulted from the affective correlations between news coverage and voters' affective description of candidates.

We have already seen how the media agenda directly affect the public agenda, but there is another aspect of the media that must also be considered: paid advertising. In addition to news stories and commentary contained within programs, the general public is often bombarded with advertising messages. For instance, to examine the effects of political advertising, Roberts (1997) found that political advertising in the 1990's Texas gubernatorial election campaign had great impact on media news items. Rank orders of the media agenda were correlated with the political advertising agenda at two time frames. The salience of candidates' image and political issues were also transmitted to the voters' mind and further influenced their voting decision. This two-step process of agenda-setting, as Roberts contended, reveals that candidates'paid political advertising can shift the priority of the media agenda as well as shape the public agenda. The agenda-setting process has gone beyond the influence on people's cognitive level to the behavioral outcomes.





16


The idea of the second level agenda-setting research sheds much light on how and what to examine in media effects on the direction of public opinion. The findings also set a new path to delve into the relationship between the media and the public agenda. Scholars focusing on the latter begin to look into the contexts of media in order to probe why public opinion is formed. To understand people's cognitive knowledge and affective reactions to the issues has broadened the explanatory domain of agenda-setting theory.

Although the contexts of the media agenda have been thoroughly examined in

terms of the second level of agenda-setting theory, measurement of the public agenda still relies on the survey to understand how people's understanding of distinctive issues was influenced by the media news framing. The growing development and adoption of new communication technologies, which allow people to talk to each other interactively, provide another useful tool for media researchers or professionals to analyze the public agenda without directly surveying people.

The Internet: A Brief History, Functions, and Current Situation

The origin of the Internet can be traced back to 1969 when the first node was

installed at UCLA in California for research purposes. At that time, the Internet was used only by scientists and government organizations. By December 1969, four nodes were connected on the early computer network, namely ARPANET. With the establishment of ARPANET, scientists and researchers could share one another's computer facilities at different places. Ironically, the main traffic on ARPANET was not long-distance computing. Instead, it was news and personal messages. Many researchers through ARPANET were collaborating on projects, trading work experiences, and even





17


exchanging some gossip. Throughout the 1970s, ARPANET was expanding at a fast pace.

Three decades later, the Internet has moved beyond its original functions for

military and research institutions. and entered the world of public education as well as the commercial sectors. The Internet's pace of growth in the early 1990s is so rapid that currently there are tens of thousands of nodes on the Internet that extend to over 40 countries in the world, with more coming on-line every day. The primary attraction that the Internet has is that no official censorship has been implemented at least up until now. Nobody controls this so-called "Information Highway."

Generally, people make use of the Internet in the following areas: World Wide

Web' (WWW) surfing, electronic mail (e-mail), the USENET, Telnet, Gopher searching, and file transfer.

To exchange information or ideas on the Internet, electronic bulletin board

systems (BBS), mailing lists (Listserv), and moderated newsgroups (USENET), have been among the most popular electronic discussion groups for more than a decade.

Electronic bulletin boards function as a specialized medium and can be set up by inter-connecting personal computers for the purpose of serving the debate, association, and exchange function of that community (Thomsen, 1996). Thousands of people could contribute to an open discussion or an informal dialogue at a specific time on this system.

A Listserv is a program that maintains one or more of mailing lists. A listserv automatically distributes an e-mail message from one member of a list to all other


I The World Wide Web is a protocol that allows users to easily make hypertext information available to other users.





18


members on that list. Listservs maintain thousands of lists in the form of digests, electronic journals, discussion groups and the like. In general, the topics discussed in the Listservs are more serious than those in the newsgroups and tend to be more focused on particular issues.

The USENET, also known as newsgroups, is a system where messages about any subject can be posted, and other people on the Internet can reply to them. These on-line discussion groups are a world of their own. Vast amounts of news information, debates and arguments involving miscellaneous topics are transmitted and posted among more than 5,000 discussion groups every day. The USENET also distributes various free electronic j journals and publications.

Another unique characteristic of the USENET is that senders and receivers do not have to communicate with each other at the same time. Once posted on a newsgroup, the messages are distributed to other readers immediately and can be responded to at any time. Unlike the Listserv networks, the USENET has no central authority and each contribution is passed throughout the system of interconnected hosts-- systems that receive and pass along each contribution they receive.

Although the above three electronic discussion groups provide a relatively free

environment for opinion or ideas exchange, informal but implicit rules of behavior called etiquettet" have always been required when one person attempts to post his or her message on the discussion groups. It is a user's duty to post the messages relevant to the topic issues of the appropriate discussion group without intentionally offending other participants. Nevertheless, due to the user's ability to conceal his or her identity in the computer-mediated communication, a user might post the messages with no factual





19


information or provide excessively critical comments. This form of message is called '!flamingI'2 which often generates a series of public posts in which people flame one another rather than contribute useful information. Hill and Hughes (1997) defined flaming messages as personal attacks, usually accompanied by profanity. In his article, Groper (1996) thought that flaming words or phrases are extreme criticism, sarcasm, cursing, exclamations, multiple question marks, emotions like :-(, which connote anger, and block capital words. These socially undesirable acts mainly come from lack of standard rules in the electronic communication world, namely, the Internet (McLaughlin et al., 1995).

While radio took 30 years to reach an audience of 50 million, and television took 13, the Internet took just four years (Weise, 1998). In a recent report on its first major study of the economic impact of the Internet, the Commerce Department of the United States disclosed that "Net traffic is doubling every hundred days and electronic commerce VI
should reach $300 billion by 2002. More than 100 million people are now on line.... (Weise, 1998).

How many users are surfing on the Internet today? According to a survey

conducted by the RelevantKnowledge company in late February of 1998, the Web users in the United States have grown to 57,037,000, a growth of over 1.6 million users since the company's last study in January 1998. Jeff Levy, CEO of the RelevantKnowledge




2 A public post or email message that expresses a strong opinion or criticism. Flames can be fan when they allow people to vent their feelings, then return to the topic at hand. Others are simply insulting and can lead to flame wars.





20


company, attributes the growth to an increase in awareness of the Internet among the public in the United States.

The Challenge of Monitoring the Internet

As new technologies have changed life patterns in the past decades, the services provided by these new technologies have grown in number. Providing the users more control over the communication process, the interactive characteristics of the Internet have changed the conventional communication paradigm that only sources have ability and power to disseminate information. The communication capabilities of the Internet are too enormous to be ignored. On the Internet, the devices such as electronic mail, newsletters, and on-line discussion groups allow corporations around the world to promote products, get feedback or comments from customers, and respond to questions from a variety of publics. Similarly, the Internet has transfigured the arena in which the public relations professionals operate.

For instance, when an unexpected corporate crisis event occurs, a story containing inaccuracies about the crisis published on the electronic newspapers will stay on the Net and be retrieved and used by otherjournalists as well as most on-line users. The journalists can also obtain the first or longest story, correct or not, on any particular issue and re-distribute it to other news sources or readers. Given the potential of instant and interactive worldwide communication, the Internet can no longer be considered a secondary communication medium.

In Kalish's article appearing in the Reuters North American Wire (Feb. 13, 1997) on public relations firms handling the Internet, he remarked that public relations veterans





21


agreed that about the worst thing a company could do is ignore persons who badmouth them in cyberspace. Speaking of the advantages of the Internet for public relations, Strenski (1995) pointed out that "professional public relations communicators can choose from a blossoming array of cyberspace resources through which to channel their messages" (p. 33).

Issue tracking is an integral part of public relations work. Whether it is called issues management or crisis management, scanning the environment to seek public opinion trends should be a top priority. There are many ways to scan and monitor the evolution of issues. The most widely used method is the scanning of trade publications, books, scholarly journals and, most importantly, the news media. In addition, themes from popular entertainment such as movies, plays, novels and television shows should be evaluated as well since they are consumed by millions of persons.

Besides traditional ways of monitoring the external environment, the Internet offers another dynamic channel for public relations practitioners. One of the most popular functions utilized by on-line users is called newsgroups. Thousands of electronic forums have been established to exchange messages on countless issues or topics, The subscribers to each newsgroup can remain anonymous, if desired, when they express their own opinion. To explain the popularity of the newsgroups on the Internet, Berger (1995) believes that the popularity stems from the loss of credibility of traditional media by the general public. However, it is also recognized by most users that "much of the discussion is trivial, and that, more often than not, debate rapidly disintegrates into verbal assaults and idle banter" (Revah, 1995, p. 10). The dark side of the Internet is that "anyone can say anything to anybody at anytime," commented Don Middleberg, president of a public





22


relations firm (Kalish, 1997). As a result, some people are participating in a variety of newsgroups that damage companies' reputations.

To emphasize the importance of monitoring newsgroups discussion for references to the business corporations, Kalish cited the suggestion of Adam Cooper, creative manager at the Interactive Solutions Group of Edelman Public Relations Worldwide, stating that, "We fmd that the newsgroups are much more interactive than the web, so the newsgroups are really what we would consider the key place to listen to people" (Reuters North American Wire, Feb. 13, 1997). Two examples involving corporate crises on the Internet-related discussion bolster Cooper's viewpoint.

One example was the crisis pertinent to the introduction of Intel's new Pentium computer processor. The Intel organization discovered the Pentium flaw in the summer of 1994 but declined to acknowledge the problem publicly. By failing to recognize the importance of the message posted on a CornpuServe3 forum, Intel paid out over $450 million to replace the flawed chip. This is one example of ignoring the public's concerns about its own products, as well as its own image.

Another rumor-like incident occurred in November 1996 when people on the

Internet labeled fashion designer Tommy I-Elfiger a racist and called for a boycott of his products. In contrast to the Intel's reaction, the fashion designer Tommy Hilfiger paid attention to the offending Internet sites and responded by launching an e-mail campaign that assured minorities on the Internet that they were valued customers and that the racist rumors were totally groundless.


3 CompuServe has been one of major commercial on-line services in the United States for over fifteen years.





23


Learning firom the above examples, it is obvious that an Internet-caused crisis can have profound influences on public opinion. On-line users can discuss any company in different newsgroups and detractors can design a web page full of factitious information about a specific company. These examples of business disruptions as a result of the Internet discussions and the resulting media coverage can be influenced by an organized group or a single individual operating in cyberspace. For whatever the reasons the situation arises, the "on-line crisis" has the potential to damage a corporate image and creates a corporate management problem for public relations practitioners.

The Internet Research

Is the Internet truly a mass medium? The answer is both yes and no. A properly constructed website can address thousands of audiences while still dealing one-on-one as an interactive medium. The more important issue is that these new websites must be monitored in additional to the traditional media such as television, radio, and newspapers. To what extent can the Internet be regarded as a mass medium? The Internet encompasses interpersonal conversation and mass distribution of messages and each message can reach a large audience. Morris and Ogan (1996) cited Valente's idea about the critical mass: "the critical mass is achieved when about 10 to 20 percent of the population has adopted the innovation. When this level has been reached, the innovation can be spread to the rest of the social system" (p. 45).

Before delving more into the impact of new communication technologies, it may be worthwhile to reevaluate the definition of mass media. Does the Internet play the role of an interactive medium? Rafaeli (1986) regarded this new type of communication





24


medium as similar to other media; however, it indeed integrates two interests in communication with the technological innovation of computer networking: "interpersonal conversation and mass propagation of messages" (p. 123). In Bucy's (1995) article discussing an objective measurement of interactivity, he suggested that the interactive event should not only activate an interpersonal conversation, but it has to be "a form of mass communication in which the sender broadcasts a message to a mass audience in the process of responding to an individual" (p. 6). Mebert (1992) considered message transmission through a mass medium as an effective communication that elicits feedback from the audience.

Rafaeli's (1988) definition of interactivity based on the issue of responsiveness recognizes three pertinent levels: noninteractive communication sequences, quasiinteractive (reactive) communication, and fully interactive communication. The distinction between quasi- and full interactivity depends on the nature of the communication responses. In the fully interactive communication, sequential messages "depend upon the reaction in earlier transactions, as well as on the content exchanged" (p. 118-9). Deriving from Rafaeli's model, interactivity is "feedback that relates both to previous messages and to the way previous messages related to those preceding them" (p. 120).

Rafaeli and LaRose's (1993) study on electronic bulletin boards treated electronic bulletin boards as a kind of collaborative mass medium, "a new type of mass communications medium in which the audience acts both as the source and the receiver of the message" (p. 277). An interactive medium is a "many-to-many" communication channel unlike conventional mass media such as television, radio, and static print media,





25


which are a "one-to-mass" channel. Also, electronic bulletin boards represent an interactive medium that demands users' contribution as well as consumption.

Garramone, Harris, and Anderson (1986) explored the motivations for

participation on a political bulletin board. They observed no statistically significant relationships between the use of BBS features and demographics or political partisan predispositions. Also in their study, surveillance was ranked as the number one motivation for BBS use, followed by curiosity, and knowledge of others' opinions. In his article defining electronic bulletin board users, Rafaeli (19 86) concluded that the users of boards could be characterized as both a faithful and active audience.

To further probe the success of electronic bulletin board systems, Rafaeli et al. (1993) found that the diversity of board contents had positive impact on board users' contribution levels and small group size contributed to adoption level of the BBSs. After content analyzing public relations practitioners' use of an electronic newsgroup called PRForum4, Thomsen (1996) distinguished three main functions in using the PRForum: exchange of information and advice, debate over issues affecting the profession, and cultivation of a sense of self-validation and enhanced efficacy both at a personal and professional level.

Among tools on the Internet that are influential in both public relations and advertising fields, Bobbitt (1995) recommended the USENET newsgroup as a public relations tool for both issues and audience research. Because study of issues and


4 As noted in Thomsen's article, PRForum, founded in 1993, is an online discussion group for public relations professionals. Updated in List search website (http://www.Isoft.com/lists/list-q.html), there were 1,224 people subscribing to this group as of May 13, 1998.





26


audience analysis is dependent on the credibility of the respondents, the researcher must assume that expressed opinions or comments are indeed truthful. Treating the USENET as a political discussion forum, Groper (1996) concluded that the political leadership did not exist in the three sampled newsgroups. No single participant in the newsgroups discussion was able to dominate the dialogues and provoke passion among his or her followers. Hill and Hughes (1997) analyzed messages in 22 political newsgroups and found that most political USENET groups exhibited the traditional characteristics of a socially cohesive group. The authors further argued that although these groups became a new form of political community for the purpose of political dialogue, it did not signalize any paradigm shift. Instead, "people are merely moving their age-old patterns of interaction into a new realm" (p. 25).

To what extent should the public relations people evaluate this new interactive medium, the Internet? According to a comprehensive demographic survey released in March of 1997 by CommerceNet and Nielsen Media Research, a large majority of Web users--73 percent--spent some portion of their time online searching for a specific product or service information. Given the growing popularity of the Internet, it is inevitable for public relations professionals to add the Internet to the list of communication channels.

As two-way communication turns the monologue into an interactive dialogue through the USENET discussion forums, the Internet offers a great opportunity to examine the public agenda which has become more interactive than ever before. Since the Internet discussion groups have the potential to galvanize public opinion in a matter of hours due to asynchronous and rapid communication, the dynamic attitudes of on-line





27


users in the discussion groups toward the corporation during the crisis time need to be investigated.

Even with many scholars and public relations practitioners emphasizing the

importance of monitoring media coverage of and public reactions to the corporate crises, no formal study has yet been done to examine how people react to a corporate crisis on the USENET in contrast to the conventional news media coverage. After analysis of both media coverage and public discussion messages on the USENET, it would be of value to track how the interpretation of the crisis varies among the news media and on-line users over time.

From theoretical considerations, since the flow of public opinion is such an

essential component of crisis communication management, the USENET on the Internet will fiction as an opinion forum to analyze the public agenda for not only public relations practitioners, but also the agenda-setting researchers. The contexts of the public conversations on the Internet can be fin-ther examined to test agenda-setting theory.

The next chapter will review the literature of crisis communication management and introduce a crisis public opinion model proposed by Sturges (19 94) and MacKuen's (1990) model of public dialogue. At the end of the third chapter, research questions and hypotheses are presented.













CHAPTER Ill
CRISIS CONEWNICATION LITERATURE Overview

In recent decades in American business, there has often been a tension between corporate public relations professionals and media journalists. When things are going well, public relations managers desperately want media coverage, but "good news" is seldom a high priority for people assigning reporters and writing headlines. However, during a business crisis, the news media appear far more inspired to investigate the issue. In these "damage control" situations, public relations managers must be aware totally of what is being printed and said in all media. Often this information is inaccurate or biased and the corporation must somehow correct the misconception and rehabilitate the corporation's public reputation (Birch, 1994; Traverso, 1992).

As with treating a spreading disease, early intervention is essential to successful treatment of a corporate crisis. Prevention of something detrimental to the organization remains a vital part of corporate public relations work. As no organization is exempt from a crisis, the art and science of crisis management is no longer an unusual specialty but an imperative job for public relations practitioners (Maggart, 1994). Rather than coping with a crisis after it has reached overwhelming proportions in the media, intelligent public relations professionals must learn to detect potential crises at the embryonic stage. Dealing with an emerging crisis before it reaches the mass public is the key to success.


28





29


Media play an essential role in helping people understand how the corporation

reacts to the crisis. Media also provide public relations professionals with information as to how the general public feels about the crisis event, so the corporation can adopt appropriate management tactics to ameliorate the crisis situation.

Effective communication with key publics in the society is the purpose of public relations practice in a business corporation, especially when it encounters an unexpected crisis. Among all key publics, the media stand out as the most significant in the public relations field, because they disseminate information and serve as interpreters of social phenomena. However, the media do not operate in isolation. Public relations professionals can often affect the behavior of the mass media.

Not only should an organization pay attention to media coverage, which

sometimes generates negative effects on the public agenda if no prompt action has been implemented by the corporations, but public perception of and opinion about the crisis must be evaluated. Because most people perceive truth to be whatever public opinion expresses at the crisis time, the essential purpose of crisis communications is to affect the atmosphere of public opinion, hoping that anything good about the organization will persist and bad connotations will soon disappear. Frequently asked questions in a crisis by a professional public relations person are: "Does the general public feel the same way as the media j journalists and in terms of the perception of a crisis, "To what degree are people affected by media coverage of the crisis and the organization?" The assumption behind these questions posits that the general public is often influenced by the media coverage.





30


In recent years a number of new media technologies have made the public relations' goal of effectively reaching the general public more challenging. New technologies affect and change the way people communicate with each other. The most conspicuous distinction between old and new media is the control that new media venues offer users over the communication process. But this new control also raises new challenges and concerns for public relations practitioners, for they must broaden the boundaries of surveying media contents. The new challenges go beyond conventional media such as newspapers, radio and television and include new technologies where information receivers can also become information senders.

In view of the explosive development of communication technology, new

message delivery systems such as cable and satellites offer dozens if not hundreds of programming options, each of which has the potential of carrying a message about a corporate crisis. While these message channels are almost always managed by established companies, a new media technology that does not require a large company infrastructure is the Internet. Given dollars and technical know-how, anyone can set up an Internet web page for people all over the world to read. Furthermore, these web pages can be interactive, meaning that individual users of a website can respond immediately to the message senders. Breaking the prevalent concept of source-oriented media, the Internet's greatest advantage is its interactivity between information providers and receivers. An audience member may become an active message producer, notjust a passive receiver.

From the crisis management perspective, the Internet also offers an excellent opportunity for public relations professionals to gauge the impact of the crisis on the





31


corporation covered by the media and to monitor the public's reaction to it. The exchange between the media, public relations professionals, and the public is imperative for the modem corporation, especially in a crisis situation. By understanding these interactions, the Intemet can function as a more useful tool for public relations practitioners.

When a corporate crisis occurs, the news media are tempted often to rely on

information not only from conventional j oumalistic sources such as wire services but on information derived from the Internet websites (Birch, 1994). As a result, a company faces a new source of pressure from both on-line users and the j oumalists who might seek information related to the crisis through the Internet. As the Internet becomes another important area in the public relations field, crisis managers must take a much broader view of it as a potential problematic source.

Since monitoring of traditional media coverage about the corporate crisis has always been a fundamental part of crisis management in the public relations field, the advent of the Internet sets another path to understand the general public's opinion about the crisis.

Crisis Management Research and Public Ovinio

Nothing is more challenging and unnerving to public relations professionals than the unpredictable crisis. A crises can cause a temporary disruption of activity with no long-lasting consequences to the bottom line or it can permanently damage a corporation's reputation, resulting in reduced profitability. Understanding properly how a crisis is contained and managed may in extreme cases determine the very survival of a business organization.





32


Definition of Crisis

What is a crisis? No universally accepted definition has been adopted, but in general, a crisis is a situation that threatens the normal activity of an organization. A crisis can proceed from a mere disruption of activity to damaging the corporate reputation and reducing profitability. Hayes (1985) believes that a crisis results from a major incongruence between the expectations of a corporation and what happens in the environment. Weick (1988) thinks that "crises are characterized by low probability/high consequence events that threaten the most fundamental goals of an organization" (p. 305). Dutton (1986) identifies three essential dimensions involved in a crisis in the study of the decision-making processes between crisis and non-crisis issues in an organization: importance, immediacy and uncertainty. Lerbinger (1997) treats a crisis as an event that endangers an organization's future profitability, growth, and even its survival. FeamBanks (1996) defines a crisis broadly as "... a major occurrence with a potentially negative outcome affecting an organization, company, or industry, as well as its publics.. ." (p. 1). The Institute for Crisis Management refers to crisis as a si gn ificant business disruption, which results in extensive news media coverage and public scrutiny

(Irvine & Millar, 1996).

To sum up, a crisis can be something unexpected that occurs by surprise in any type of corporation (Woods, 1996); stems from the interaction of failures between the corporation and the external environment (D'Aveni et al., 1990); requires fast and accurate reaction to neutralize high threat to corporate values (Gonz~lez-Herrero & Pratt, 1995); creates uncertainty (Mitchell, 1986); threatens the reputation and assets of the





33


organization (Barton, 1993); and causes publics, especially media, to scrutinize the organization (Irvine & Millar, 1996).

Types of Crises

Although corporate crisis can occur suddenly without forewarning, it is wise to identify the types of crises when they take place in order to adopt the proper strategic plan. The Institute for Crisis Management puts the crisis into 16 categories and identifies three origins of crises, management, employees, and other (Irvine & Millar, 1996). The 16 categories are business catastrophe, environmental damage, consumer action, discrimination, financial damages, labor disputes, sexual harassment, white collar crime, casualty accident, class action suits, defects/recalls, executive dismissal, hostile takeover, mismanagement, whistle blowing, and workplace violence. Mitroff et al. (1996) categorizes 11 types of crisis which include: criminal attacks; economic attacks; loss of proprietary information; industrial disaster; natural disaster; equipment/plant malfunction; legal problem; perceptual/reputational; human resources/occupational; environmental/health; and regulatory. In her case study book, Kathleen Fearn-Banks (1996) lists the following five types of crisis: product tampering, environmental, natural disasters, violence, and celebrities and crises. According to Lerbinger (1997), there are seven crisis types: natural, technological, confrontation, malevolence, skewed management values, deception, and management misconduct. Environmental Monitoring in Public Relations

Business crises put corporate reputations and economic survival to the test.

Effective crisis communication planning not only results in good issue management but in favorable public perception of the corporation. One of Hearit's (1994) suggestions to





34


deal with crises was to diffuse hostility directed toward the corporation suffering from it. But as a potential crisis is evolving, how can companies keep track of how they are being portrayed and how can this information be intercepted before it reaches important stakeholders? In the public relations field, environment monitoring involves a long-term examination of media coverage before any issues related to the organization can transform into a crisis that will detrimentally impact the business. Needless to say, environment monitoring becomes an indispensable component of crisis management. From a synthesis of the literature emphasizing the importance of environmental monitoring in crisis management (Birch, 1994; Camey, 1993; Fombrun, 1996; Hayes, 1985; Umansky, 1993; Wilcox et al., 1992), several reasons for monitoring media coverage are identified:

1. The effectiveness of general issue management can be determined by analyzing the

tone of media coverage;

2. The information obtained from the media content also helps public relations people

examine public perception of an organization's response to a crisis;

3. Monitoring and analyzing media coverage help public relations practitioners audit the

reputation of the organization;

4. Scrutinizing media coverage on issues provides public relations practitioners ideas

about how to position an organization in front of publics and how to modify the

strategies or tactics when handling a crisis.

5. Understanding media coverage is one part of formative research in the

implementation of public relations strategies.





35


Public Opinion in Public Relations

Public opinion research has always been a vital topic across a diversity of social science fields such as political science, psychology, and mass communication. To understand public attitudes, polls are often employed as an efficient way of measuring the public agenda on major issues. Whether public opinion represents an aggregation of individual views or a collective movement remains debatable (Price, 1992).

Public opinion, more often than not, can determine what governments and

corporations do when encountering the unexpected crisis. In a typical crisis situation, the public perceives "the truth" to be whatever public opinion says that it is. Public relations practitioners are responsible for surveying questions of public opinion. Therefore, it is vital for an organization to prove to its stakeholders that the prevailing negative public opinion is not accurate.

Public opinion can and must be shifted to the corporation's advantage if an

organization desires to emerge from a crisis situation with a minimum of damage to its reputation. It is difficult enough to deal with a situation in which the corporation is indeed guilty, but even more frustrating is coping with bad information and false accusations. Looking back to past crisis events under various circumstances, many companies--such as Exxon, NASA, Dennys, and Intel--failed to communicate effectively with the general public, including their loyal customers, when they faced a crisis or precrisis situation. Conventional press conferences or general press statements may not subdue the anxieties of people who are directly influenced by a crisis. Corporations need to go directly to the public to gauge how well public relations efforts are working; this is where proactive tactics become essential.





36


Damage control is the key objective of crisis management. The more a public relations practitioner knows about how a crisis is being interpreted by the media and the general public, the more effective the damage control will be. The communication challenge is to convince various audiences of the company's ability to alleviate the damage, eliminate public doubts and position itself for future growth. The success of crisis management is evaluated by how effective the company affects the public opinion process.

Media and Crisis Management

The first lesson for a public relations practitioner to learn is that it is always better to deal with a potential crisis before it blooms into a real crisis. Without tracing and monitoring the shifts in public opinion, public relations programs become meaningless. Public relations practitioners cannot afford to operate in a vacuum. Accurate information concerning media messages and resultant public opinions is crucial for continued success. Among the research devices used to accomplish this task are personal contacts, content analysis of media coverage, field reports, and polling (Wilcox et al., 1992). Media play a crucial role in understanding public opinion trends on issues. Although the impact of mass media on attitude formation and change has proven to be less influential than some expected (Klapper, 1960), mass media still function as a primary resource for information about many topics. Therefore, it becomes impossible to ignore what mass media cover or say about an issue pertinent to an organization while simply probing public opinion.

Media impact on spreading a crisis is often time-dependent. It is better to control an issue in the early stage than wait until it becomes a crisis management issue. As recommended by Birch (1994), the use of tracking research, even daily in the early





37


stages, allows corporate management to know what the general public or key stakeholders are really thinking and saying. Permanent monitoring of active publics, getting continuous feedback through different channels from these publics, also facilitates issue tracking (Umansky, 1993; Vendrell, 1993). In summary, when dealing with an unexpected crisis, the two-way research, monitoring both the media and public agenda, is pivotal.

During a corporate crisis, communicating with the general public is very

important. Crisis communication tends to evolve through three stages: mass media publicity, public opinion arousal, and then public policy makers' response (Mayer, 199 1). Public opinion stems from the media issue coverage and then impacts government policy makers. At the initiate phase of a crisis, an organization may have the power to influence events. After the crisis has received large amounts of attention from both media and the public, the ability of the company to control its public reputation diminishes. In her book on crisis management, Fearn-Banks (1996) terms "containment" as efforts exerted by the organization to limit the duration of the crisis or keep it from spreading to other areas affecting the organization.

An effective crisis communications plan must involve a sound media monitoring program to ensure accuracy of media coverage. Traverse (1992) urged that a media monitoring program is necessary when facing the crisis as media provide a fast channel to deliver information to the public. Situated in such a dynamic information era, Strenski (1995) pointed out that a key area in which public relations professionals must excel in crisis control is to take responsibility for monitoring the on-line services such as BBSs and all kinds of newsgroups that exist in cyberspace. He warned that ignoring these





38


groups is missing a major portion of the potential audience. Kotcher(1992) offered reasoning why new communication media must not be neglected by public relations practitioners. He states,

Communicators can no longer operate under the old assumption
that mass communication is passive in nature and that interactive, personal communication can only really occur in small groups or face-to-face. In addition, communicators should not assume their present understanding of technology and its applications would be
sufficient at a time of crisis (p. 20).

Crises can start with rumors, which can be absolutely false or partly false without substantial verification and credible sources. In her ten guidelines for reducing legal risks in crisis management, Fitzpatrick (1995) stressed the urgent need to respond to rumors spreading all over the variety of information channels that may turn out to cause a longterm, sustained crisis for a company.

Sturges' Public Opinion Model of Crisis Manageme

Communication is an important tool in handling the impression of key

stakeholders at the time of crisis (D'Aveni et al., 1990). Sturges (1994) emphasizes the value of communication during the crisis to the publics. The crisis does not happen in isolation so the communication plan should consider how public opinion, interacting with other social events and environment, is formed before, during and after the crisis. He illustrates the evolution of group opinion through an eight-step process as shown in Figure 3 1.





39



8. Social Norm 10 1. = Latent Issue





2. Event Occurs
7. Social




3. Pro and Con
orm
Factions Form
6. Public's
Opinion Forms




4. DebEaate
6. Time Lapses .4 Occu rs




FIGURE 3-1. Group opinion formation process


In the first stage, an issue must be salient but latent for the public members. As

public awareness is the necessary condition in the formulation of public opinion, an event occurs and comes into people's mind in the next step after they become aware of that event. Then the pro and con factions among the members solidify in the third stage, followed by public debate among the members (step 4). After a time lapse occurs (step 5), when enough has been said and done, public opinion thus emerges in step 6. In step 7, some action will be used to reinforce the public opinion formed in the previous step. Finally, due to the performance of action, a social norm is going to be established.





40


Sturges applies this concept of opinion formation to the direction of public opinion in the crisis situation as illustrated in Figure 3-2. The public opinion, either negative or positive, expands as the crisis erupts and fades away while the crisis abates and ends. Therefore, the objective of crisis communication is "to influence the public opinion development to the point that opinions held in the post-crisis period are at the same level or greater in positive opinions or lower in negative opinion" (p. 303) among the members of the public.







-Negative Opinion
-Positive Opinion
Pre-crisis Build Up Break Out Abatement Termination





FIGURE 3-2. Sturges'model of public opinion with crisis management plan


MacKuen's Strategy Model of Public Dialogue

Regardless of the form (face-to-face, electronic bulletin boards, newsgroups, etc.) one chooses to take part in the political conversation, MacKuen's (1990) discussion about personal strategies in the public dialogue to promote deliberative democracy is employed to examine how on-line users get involved in the discussion groups. The basic idea behind MacKuen's strategy model for public dialogue is that deliberative democracy is a function of political participants who interact with one another. These interactions set the political and intellectual contexts for democratic decisions and may take two forms. One





41


occurs between those who hold similar views and another stems from individuals of different views. However, only the latter allows genuine debate and an exchange of ideas. As a result, the conversation which each individual creates becomes crucial in shaping public opinion and then develops into deliberative democracy.

Regarding engagement in political discourse, MacKuen holds that one's

willingness to engage in public discourse depends "not only on how much one enjoys discussion or hates controversy, but also on the likelihood of running into sympathetic and contrary viewpoints" (p. 61). Individuals can actively choose when and where to participate without passively getting involved. Four strategies will affect the outcome of interactive political conversations: TALK, CLAM, REACT, and SIGNAL. Either social consensus or disagreement comes from different strategies that people employ when encountering the political discussion. Similarly, the outward environment would play a role in determining what strategy individuals may use while becoming involved in the political dialogue.

MacKuen uses Expressivity (p. 64), the ratio of pleasure from rewarding conversation to pain of disagreements, to depict how an individual feels about the conversation and his or her own tolerance of opposing viewpoints. It also represents a positive incentive to engage in the political conversation. Low levels of Expressivity require friendly environments to make conversation worthwhile. By contrast, high levels of Expressivity render political conversation attractive even when facing opposing majorities.





42


Applications of Four Types of Strategies

The outcome of use of TALK/CLAM strategies in the political dialogue is shown in Table 3-1 and 3-2.

TABLE 3-1.

Payoff matrix for player encountering friend Friend's Strategy


TALK CLAM

Player's Strategy TALK Reinforcement Bluster

CLAM Music Pass






TABLE 3-2.

Payoff matrix for player encountering friend Friend's Strategy


TALK REACT CLAM

TALK Reinforcement Reinforcement Bluster


Player's REACT Reinforcement Pass Pass
Strategy

CLAM Music Pass Pass





43


As seen in Table 3 1, the conversation that reinforces democratic dialogue

between two parties who reward self-expression, willingness to speak out, happens only if both individuals choose TALK strategy. By the same token, if both individuals choose CLAM strategy, a nonpolitical conversation (Pass) will follow. The theme in this matrix of employment of both strategies is that all outcomes but Reinforcement and Disagreement result in little or temporary value through the interactive conversation.

Other than the aforementioned two strategies, an individual might opt for the third one--- REACT.

As shown in Table 3-2, the unique advantage of this strategy is to avoid

disagreements when both player and friend adopt either TALK or REACT strategy in the conversation. However, it also limits the attractiveness of the TALK option if an opponent is involved. By using this strategy, a person can cheerfully respond to friendly TALKERS and remaining silent before opposing viewpoints. The REACTOR never initiates a political conversation, but simply follows the partners, with whom she or he agrees. As the number of TALKing friends increases, the attractiveness of REACT will also increase likewise.

The last strategy MacKuen mentions is called SIGNAL. The consequence of

signaling for individual choice produces a community of isolated groups and eliminates meaningful public dialogue. So to speak, signaling not only prevents a consideration of alternative views, but also generates a consistency of opinions around the symbols embodied in the signal itself (p. 82).

Because the level of Expressivity in the public environment determines an individual's willingness to engage in the political dialogue, MacKuen's integrated





44


TALK/CLAMIREACT/SJGNAL models provide an insightful understanding about how a person would contribute to the meaningful dialogues in any public discussion forum. For instance, employment of the TALK/CLAM strategies is most likely to generate a democratic dialogue; on the other hand, use of REACT/SIGNAL strategies will help to frame a friendly interactive discussion environment without disagreement.

Research Questions

Studies on crisis or issue management have created stages or life cycles from the emergence to the termination of a crisis (Feamn-Banks, 1996; Gaunt et al., 1995; Gonzdlez-Herrero et al., 1995; Hainsworth, 1990; Meng, 1992). With the opinions identified as positive or negative or combination of both, people tend to react positively to good news and negatively to bad news (Sigelman et al., 1993). Based upon Sturges' (1994) public opinion formation model during crisis, negative public opinion about the corporation at the breakout of the crisis reaches its highest point and then ebbs to the precrisis stage after the crisis.

As Sturges demonstrates earlier in Figure 3-2, it is anticipated that the highest number of negative messages about an organization will be posted as the crisis erupts, following the initial media coverage.

1 How will public opinion on an interactive communication. channel, such as the Internet, differ from Sturges' model on public opinion during a crisis?

As commercial providers increase on the Internet and more political information is provided, the problem of who sets the agenda for the new medium also becomes a concern. Behr et al. (1985) maintained that public concerns about energy and inflation were determined by television coverage, which in turn was dependent upon world





45


conditions. By comparing the contents of three local and regional daily newspapers with the results surveying three waves of respondents drawn from the same study population across eight months, Salwen (1988) detected a strong influence of media coverage on environmental issues on the public agenda in the first and second wave. Despite timevarying effects for agenda-setting, the direction of influence has traditionally been from the media to the public.

2. Will this unidirectional influence from the media to the public hold when the

interactive public agenda is examined?

Famous public relations crisis cases such as the Intel Pentium flaw, the Pepsi

syringes tampering, and the Nike discrimination against Islam suggested the importance of a quick and proper reaction to the potential crisis issue, whether it is only a rumor or not. Communication with the target publics during the crisis becomes a top priority task for public relations professionals; hence, preparation of complete crisis management planning should be highly stressed (Barton, 1993; Mitroff, 1996).

3. To what degree is the frequency of on-line dialogues on the USENET affected by

an organization's strategic crisis intervention?

MacKuen (1990) demonstrates that people who feel high levels of Expressivity will more likely adopt a TALK strategy than those who do not. Although CLAM is one of the strategies used by individuals to engage in the public dialogue, it is impossible to detect who remains silent by adopting this strategy in the newsgroups discussion. CLAM users are regarded as those who read the messages without posting their own opinions on the newsgroups.





46


4. How will the frequency and tone of on-line dialogues on the USENET posted by

TALKERS and REACTORS vary in the light of MacKuen's strategy model?

Research Hypotheses

Given that the Internet provides public relations professionals access to listening to what people really think and say about the crisis or the targeted corporation on the USENET, hypotheses in this study stress the change of number of on-line discussion messages before, during and after the corporate crisis. Analysis of these messages will also help advance the measurement of the public agenda in agenda-setting theory.

Sturges (1994) divides the time span during the crisis into several stages, that is, pre-crisis, build-up, break out, abatement, and termination. He asserts that both negative and positive opinions about the organization from the public will increase when the crisis breaks out. To compare the difference in the number of on-line messages before and after the crisis erupts, ten days will be used as a time frame to test the following hypothesis:

Hypothesis 1: The average number of daily USENET dialogues regarding the corporation will be greater during the ten days following the eruption of the crisis than the average number of daily on-line dialogues within ten days before the crisis.

Also illustrated in Sturges' model is that public negative opinion will

comparatively increase more than public positive opinion since the crisis can have negative impact on the society as well as the general public. Therefore, by using all the sampling days after the crisis eruption as a time frame, a second hypothesis is formed.





47


Hypothesis 2: The average number of daily negative USENET dialogues will be greater than the average number of daily positive USENET dialogues after the eruption of the crisis.

Numerous agenda-setting studies lead to a common conclusion that news media coverage of an issue has considerable impact on the public agenda. Concerning the first level of agenda-setting theory, the more media coverage of an issue, the more salient that issue will become in people's minds. The next hypothesis is hence made based upon the agenda-setting concept as displayed in Figure 3-3.



Media News 2:* Public Public Opinion
Reports Stories Agenda on the USENET


FIGURE 3-3. Causal relationship between the media agenda and the public agenda


Hypothesis 3: The daily number of the USENET dialogues will be related to the daily number of news stories about a crisis after its eruption.

In Sturges' model, to move to the abatement stage where the amount of public

negative opinion sharply decreases from the breakout stage, the corporation must employ proper strategies to contain the crisis spreading. To test if change in the amount of public negative on-line dialogues does occur after the company adopts the intervention strategies, the following hypothesis is formed.

Hypothesis 4: The average number of daily negative USENET dialogues

following a crisis intervention strategy will be less than the negative USENET dialogues prior to the intervention.





48


Derived from the above hypothetical assumption, it is expected that the

intervention strategies will have positive influences on people's opinion about the corporation. In Sturges' model, the amount of public positive opinion about the company remains similar at both pre-crisis and abatement stages, so the next hypothesis will be as follows:

hypothesis 5: Following a crisis intervention strategy, the average umber of

daily positive on-line dialogues on the USENET concerning the company will be equal to the average number of daily positive on-line dialogues before the crisis eruption.

After the intervention strategies, amount of both negative and positive opinion from the public decreases in comparison with those when the crisis erupts according to Sturges'model. Thus, the hypothesis is developed as follows:

Hypothesis 6: The number of the USENET dialogues following a crisis intervention will be less than the number of the USENET dialogues before the intervention and after the crisis.

No study has applied MacKuen's strategy model of public dialogue to the on-line users who express their opinion on the USENET. As TALKERS are inclined to express their own opinion to initiate the public dialogue or disagree with other people, the next hypothesis is suggested to test how both TALKERS and REACTORS voice their opinion about the company after a crisis happens.

Hypothesis 7: The number of daily negative USENET dialogues posted by TALKERS will be greater than that by REACTORS.

The concept and findings of the second-level agenda-setting research indicate that content of media coverage of the corporate crisis can greatly affect the public opinion.





49


By using computer software to analyze the tone of media coverage, the hypothesis is formed to test the relationship between the tone of the media agenda and the public online dialogues.

Hypothesis 8: The Optimism Score of media messages will be positively related to the number of negatively toned USENET messages posted after the media stories were released.

The next chapter will describe in detail the method used to test the hypotheses and to answer the research questions.













CHAPTER IV
METHODOLOGY


Overview

To test the agenda-setting theory, most researchers have traditionally used content analysis to examine the media agenda and public opinion surveys to examine the public agenda. All studies investigating the media content have focused on the conventional media such as television, magazines, newspapers, and radio. However, the Internet has quickly evolved into another important communication medium not only as a corporate communications tool but as a forum for public opinion. People have become more active and willing to get on the Net to express their attitudes toward or opinion about particular issues or persons. The Internet hence provides public relations professionals another opportunity to understand the trend of public opinion about their own organization at times of crisis. To better explore public opinion on the USENET, this study will employ content analysis to analyze the on-line public dialogues as well as the media coverage of the selected corporate crises.

This chapter will first briefly state the definition and functions of content analysis, followed by detailed descriptions of how to select crisis cases, news media, and public dialogues on the USENET. Following the listing of coding categories, both conceptual and operational definitions of key terms stated in the hypotheses are provided. Then, all independent and dependent variables are identified with an intercoder reliability check in




50





51


the next section. The final section introduces of a computer software called DICTION for the purpose of testing the last hypothesis. Description of Content Analysis

The most popular and frequently cited definition of content analysis is offered by Berelson (1952), "Content analysis is a research technique for the objective, systematic, and quantitative description of the manifest content of communication" (p. 18). Krippendorff (1980) also defines content analysis as a research technique for making replicable and valid inferences from data to their context. The inferences involve the sender, receiver of the message, the message itself, as well as the communication channels. This unobtrusive quantitative measure of texts serves many purposes such as identification of intentions of the communicators, understanding of attitudinal and behavioral responses to communications, and description of trends in communication content (Webber, 1990).

Corporate Crisis Selection

This study is intended to predict the trend of public dialogues on the USENIET about corporate crises. Tfhe impact of news media coverage and the company's intervention strategy will be integrated into the model. Two examples of the same type of crisis will be randomly selected. The first crisis will be analyzed and used to modify the coding categories and to improve the ways of retrieving a sufficient number of media stories and on-line discussion messages.

For the purposes of generalizability of fi-ndings and to understand the influences of the crisis in the on-line USENET dialogues, product tampering was selected as the target crisis. Feamn-Banks (1996) considers product tampering crises as a result of claims





52 0


made against manufacturers' products. After analyzing news stories about 16 types of business crisis, Irvine and Millar (1996) found that crises of defects & recalls of products has increased 41 percent between 1994 and 1995. Because product tampering has the public safety and health at high risk, Lerbinger (1997) classifies it as a crisis of malevolence, which often necessitates product recalls. Past examples include Johnson & Johnson's Tylenol murders in 1982, Gerber's product tampering in 1986, Perrier's recalling of water bottles in 1990, and Pepsi's syringe threat in 1993.

The time period from which the two crisis cases were selected was from March, 1995 to December, 1997. The crisis event was chosen for study because it met the conditions that are pre-defined in the latter part of this chapter.

News Coverage AnalyaLs

A corporate crisis has emerged when the news media report an issue in a negative or adversary tone. Content analysis of crisis coverage by the media has usually been used to guide an organization's reaction or responses to handling the crisis and to measure how the general public perceives the organization. Continuing coverage by the mass media may serve to help comprehend the agenda-setting effects on the crisis and delineate a crisis development model. Research on testing agenda-setting hypotheses has traditionally focused on impact over time, which required at least two weeks to six months to discover the power of media setting the public agenda (Eaton, 1989; Funkhouser, 1973; Mullins, 1977; Salwen, 1988; Shoemaker et al., 1989; Sohn, 1978; Stone & McCombs, 1981; Wanta & Hu, 1994; Winter et al., 1981).

With the considerable number of public dialogues on the USENET, the time

frame for analysis of media coverage is six weeks, 10 days before occurrence of a crisis





53


and 32 days following the crisis eruption. Although on-line discussion may or may not end one month after occurrence of the crisis, the random error becomes larger as more days of messages are retrieved when applying statistical analysis to test the effects of news media coverage in relation to the number of on-line dialogues.

The on-line LexisNexis data base served as a foundation of media news

coverage. The keywords for searching news articles in the LexisNexis data base varied depending on each crisis.

Media Selection for Analysis of News Stories

Television provides political information to most citizens in Western industrial

societies and has become the most credible source of information (Brosius & Kepplinger, 1990). Zucker (1978) maintained that at the national level the public maybe more influenced by the three networks' newscasts than by newspapers because of television's accessibility. Shaw and McCombs (1977) argued that television news might have a stronger short-term impact, however, newspaper content may have a more consistent effect across longer periods of time.

The LexisNexis data base was selected for capturing news media coverage. The Lexis/Nexis contains both newspaper and television news transcripts that have national circulation and audiences. The media chosen as a part of sampling frame included three television news networks, and three elite nationally known newspapers. They were: the CNN news, the ABC news, the NBC news, the New York Times, the Wall Street Journal, the Washington Post. Because the CBS news transcripts are currently not in the LexisNexis data base, the CNN news, the ABC news, and the NBC news were selected for the analysis of television network news analysis. The New York Times was chosen for





54


the study because, besides having a national circulation, it is thought to have an agendasetting effect on smaller media that could be seen in survey data of a cross section of the population. The New York Times was shown to set the agenda for the television networks when the cocaine issue coverage was examined (Reese & Danielian, 1989). The Wall Street Journal, a business-oriented national as well as international newspaper, devotes considerable coverage to news about major corporations. Because news magazines were not found to contribute as strong agenda-setting effects as newspapers and television (Shoemaker, 1989), this type of media was not included in the analysis.

On-line Discussion Message Analysis

In addition to crisis coverage by the six news sources, on-line dialogues will also be content analyzed. There are three advantages in using the unmodified USENET discussion groups to study political behavior in cyberspace (Groper, 1996): these newsgroups are simple to access, posts to them are free of charge, and messages in the USENET groups can be downloaded instantly from the Internet without subscribing to the groups. In contrast, both Listservs and BBSs make it difficult to download data without subscribing to the group itself.

The time frame for downloading the messages from on-line newsgroups was the same as sampling media news coverage, that is, 10 days before and 32 days after the crisis eruption. A search engine called De/a News' on the Internet (See its homepage in




I The Deja News Inc., founded in May, 1995, is the first company to organize and provide easy access to discussion groups. As of now, the company's (http://www.dejanews.com) archives are extended back to March, 1995 and has about 138 million articles appearing in over 20,000 newsgroups, accounting for more than 180 gigabytes of disk space. Similar to the use of Lexis/Nexis data base, input of keywords





55


Appendix A.) was adopted to facilitate the message downloading. To have enough but not too many redundant on-line discussion messages about the targeted corporation, the keywords used the Deja News search engine were slightly different from those used in the Lexis/Nexis data base.

Unit of Analysis

The unit of analysis is the single message by one reader regardless of its length within each thread pooled from the Deja News search engine. Every sampled message was coded for the manifest content with the assistance of pre-defmned categories. Although a thread has to include at least two response messages besides the original subject message, based upon the concept of "responsive interactivity" proposed by Rafaeli (198 8), this study still retrieved and analyzed any single message under a thread as long as it was related to the crisis event. In addition, the same message that was posted to different newsgroups or on different days by the same on-line user was singled out and counted once only.

Threads across newsgroups from the Deja News data base were retrieved for analysis. Not all news stories were retrieved. For instance, a newsgroup called "biz.clarinet.webnews" often posts the news stories directly from other media sources, so this type of message was ignored. Nevertheless, if the news stories were cited or posted by a self-identified on-line user who did not express any personal opinion, those messages would be sampled. Some messages containing several segments were read and treated as one single message.



will automatically display messages related to the topic interesting to the users. The articles on the USENET can go back as far as 1979.





56


Steps to Download On-line Messages

Because on-line messages of 42-day duration would be downloaded for analysis, the efficient way is to use the power search" method in the Deja News search engine (See Appendix B), which enables a user to specify a time frame without worrying about having too many messages beyond the particular time span of 42 days. The following describes how to search for the messages in the Deja News search engine step by step:

1. Type in the keywords in the "Search for:" blank row;

2. Use "threaded" as the results format and sorted by "Date.";

3. Leave "Group(s)", "Author(s)", and "Subject(s)" blank;

4. Specify the beginning and ending date in "Date from:" and "To:" blank rows;

5. Click on "Find" button.

After the search results showed on the screen, every message was downloaded for further analysis. While downloading all pertinent messages based on abovementioned criteria, the researcher has also clicked on "view thread" within the listed message to see exactly how many messages had been posted under that threaded topic. Then, every single message under that threaded topic was examined to ensure that all messages relevant to the crisis had been downloaded.

The reason for examining each single message under one thread is because the

content of on-line discussion is often diverted to other topics that can be totally irrelevant to the crisis event or the involved company. Without careful inspection of each message, the number of on-line messages can be erroneously expanded, thus threatening the content validity of the study. Another rationale is that some on-line users may mention or





57


discuss the crisis as responding to other users' messages without mentioning any keywords that are used to search for crisis-related messages.

Definitions of Coding Categories

The categories of discussion messages needed to be defined in order to compare the progress of media coverage on the corporate crisis. Stempel 111 (1989) suggested that it is efficient to use an established category system that had been found functional and manageable by other researchers. Although a large number of articles emphasize how crucial evaluation of public reaction to the crisis is, specific definitions of positive or negative public reaction are unavailable. Since no preset categories related to this study were available for the on-line message analysis, the researcher synthesized various articles dealing with either the public reaction to a crisis or corporate planning for crisis cases (Birch, 1994; Carney et al., 1993; Fearn-Banks, 1996; Fitzpatrick, 1995; Gonzd1ezHerrero et al., 1995; Hearit, 1994; Lerbinger, 1997; Mitroff, 1996; Stanton, 1995; Thomsen, 1996; Traverso, 1992; Wilcox et al., 1992; Woods, 1996).

As noted in the second chapter, it is not unusual to receive so-called "flaming" messages from the newsgroups discussions (Bucy, 1995); therefore, the flaming messages were also coded. Coding categories were tested and re-tested for reliability. When training the coders for reliability, the researcher modified the coding items, if necessary, to make them mutually exclusive and exhaustive and create a common frame of reference between coders and the researcher. When a negative or positive message involved more than one item listed below, it was coded as one single number for that category. For example, if one negative message mentioned the adoption of legal action





58


and boycott of the product or service, this message would be coded as 1 ", which indicates a negative message. This rule was also applied to positive messages.

The coding schemes for news stories and on-line messages are shown in Appendix C and Appendix D.

1. Negative posted messages will be coded as those that a) request legal action against the company, b) request compensation from the company, c) mention of personal or others' experiences suffering from the crisis event,

d) urge the boycott of the company's product or service, e) give sarcastic jokes or harsh comments about the company, or f) blame or accuse the company for its wrongdoing or its defective product(s).

2. Positive posted messages will be coded as those that a) urge end of a boycott,

b) compliment the company, c) mention the laws as unfair to the organization, d) praise company's intervening actions, or e) deny that the crisis event is solely due to the company.

3. Neutral posted messages will be coded as those including a) forwarded media reports without expressing opinion, b) forwarded website information without expressing opinion,





59


c) personal questions or responses to others' messages without expressing any opinions,

or

d) statements that take the crisis issue for granted without blaming or praising the

company.


Conceptual Definitions

In the following two sections, some special terms stated in the hypotheses are defined both conceptually and operationally along with concise reviews of previous studies germane to the terms.

On-line Dialogujes on the USENET

The idea of dialogue has been extended to apply to the messages on the

computer-mediated communication channels, such as the USENET, BBS, and e-mail system. As far as messages on the USENET are concerned, MacKinnon (1992) points out that these newsgroup postings have the quality of spontaneity and are uncensored, frnctioning more like true dialogues. Unlike traditional conversation in the real world, the feelings expressed in the postings are not mirrored either physically or symbolically, and the dialogue is "absent from physical proximity, face-to-face interaction, and nonverbal cues" (p. 114). Morris and Ogan (1996) regard the USENET as a many-tomany asynchronous communication. So to speak, two or more on-line users communicating with each other do not have to be on-line at the same time. Each could leave messages to others and retrieve replies at his or her convenience (Groper, 1996). To contrast with dialogic communication mainly applied to interpersonal communication, Thomnsen (1996) cited Ball-Rokeach and Reardon' s concept of "telelogic communication" for the analysis of bulletin board systems. According to Ball-Rokeach





60


and Reardon (1988), telelogic communication "involves alternating dialogue between people at a distance who use both conventional and unconventional language and electronic or optical channels" (p. 135). This type of communication is often geographically and temporally unbounded. Bucy (1995) considers mediated exchanges of messages as a form of mass communication in which the sender broadcasts a message to a mass audience in the process of responding to an individual.

All messages on the USENIET have a subject heading and those which carry the same subject line are called a "thread." Hill and Hughes (1997) define thread as responses to the original message, or to something said in someone else's reply. Once a thread has been started though, it is no longer under the sender's control.

For this study, on-line dialogues on the USENET are defined as--Multiple messages about the same subject matter, with a single sender to a group not in the same physical space which was formed because of its interest in the subject matter, who can respond immediately to the message
and who can direct that response to the entire group.

A message is considered to be the total written materials contained in the text. This includes all words, sentences, phrases and paragraphs. The subject matter in the multiple messages refers to the discussion topic of one single thread. A thread is defined as messages responding to the original message or to other messages under the same discussion topic.

Crisis

The media coverage of a corporate crisis provides people most information about the crisis event itself and, thus, shapes the public's perception of the organization. Aiming to examine the relationship between the media coverage of the crisis and on-line





61


public opinion about it, a crisis in this study is defined as similar to that by Irvine and Millar (1996)--A significant business disruption, which negatively impacts its stakeholders, such as employees, stockholders, customers, or community members and results in extensive news media coverage and public
scrutiny.

Crisis Intervention

Crisis communication results from a need to offset potential negative

consequences of not communicating (Sturges, 1994). Heart (1994) suggests that to intervene in the crisis is to diffuse hostility directed toward the corporation suffering from it. Resembling the idea of crisis intervention, Fearn-Banks (1996) uses "containment" to refer to "the effort to limit the duration of the crisis or keep it from spreading to other areas affecting the corporation" (p. 7). Lerbinger (1997) regards crisis intervention as an action that helps remove ambiguity during a crisis and improves the understanding of the public and the media.

For this study, crisis intervention is defined as--When the organization acts strategically to reduce the negative impact of
the crisis on stakeholders after the occurrence of the crisis.

Operational Definitions

The operational definitions of key terms are listed as follows:

On-line Dialogues on the USENET: "number of messages from different individuals within one subject thread." Here the thread is defined as messages responding to the

original message or to other messages under the same discussion topic or subject.





62


Crisis: "the events that

negatively impact the target subjects, e.g.: the general public, customers, and

employees,

are covered by both a national elite newspaper and a national television

network,

continue to be reported by major national news media for at least a one-week

duration and covered by some news sources for a total of seven days during

the month, and

are discussed by an Internet newsgroup periodically for at least two weeks

since the initial news media coverage."

The crisis is considered to have erupted on the first day that any of the selected news media reported the crisis event.

Corporate Crisis Interventions: "the actions that the corporation adopts after the

media coverage of the negative event and which are intended to reduce the damage

caused by the media coverage."

The following would be coded as crisis interventions:

I evacuation of employees or particular publics involved in the crisis for the sake of

safety,

2. hosting a news conference to admit or disclose happening of the crisis,

3. setting up hot lines to answer requests from media and publics,

4. provision of medical services to injured or dead people,

5. preparation for negotiation with interest groups or specific publics concerning the

crisis,





63


6. product recalls or destruction,

7. decision to enforce laws to combat the crisis,

8. offer of monetary or product compensation,

9. making a public apology,

10. acknowledgement of wrongdoing,

11. replacement of persons who caused the crisis, or 12. promise to prevent a similar crisis in the future. TALKER: "the on-line users who either respond to without expressing agreement, or

express different opinions, or initiate the subject message within one thread."

" REACTOR: "the on-line users who express their agreement to the initial message or

merely reply to other users without voicing their opinions about others' statements.

Independent and Dependent Variables

Four independent variables and two dependent variables are included in this

study. The number of news stories covered by the media about the crisis event is the first independent variable. The second independent variable is considered the eruption of a crisis. The contents of selected media will be analyzed to determine when the crisis occurs. The first day that any one of the media reports the crisis event is going to be the day of the crisis eruption. As the primary purpose of the crisis intervention is to alleviate the damage done by that crisis to the corporation, how a corporation manages the crisis intervention becomes indispensable in changing or reducing negative public opinion about the company. The third independent variable is the direct action(s) taken by the corporation during the crisis period. The last independent variable is the number of online messages posted by both TALKERS and REACTORS.





64


Since the number of negative as well as positive on-line dialogues on the

USENET is the main item examined in the hypotheses, which presumably varies with the other variables, the number of negative and positive on-line dialogues on the USENET become the two dependent variables.

Reliability

In regard to content analysis, reliability is the fundamental requirement for yielding similar results with the same instrument on a given data sample. Intercoder reliability represents the consistency of shared understandings or meanings held by coders. Weber (1990) urged that this type of reliability be a minimum standard for content analysis.

To test the interceder reliability, 20 messages randomly selected from the first

case were given to two coders, who are native English speakers. They were instructed in detail on how to code a nominal variable--tone of message. New items suggested by the coders were added to the coding scheme.

Although there is no agreement on a minimum interceder reliability coefficient, most content analysts suggest that the researcher should look at whether or not the extent of agreement exceeds chance (Stempel HI, 1989). Krippendorff (1980) found that an agreement correlation of less than 0.7 tended to be statistically insignificant and that 90 percent of agreement is not plausible in most cases. For this study, the pre-test interceder reliability coefficient was expected to reach at least 80 percent before the researcher started to code messages in the second case.

According to Scott's (1955) formula, reliability reached 0.76; that is, both coders had consistent agreements on 16 out of 20 messages in the same category. Because they





65


added some reasons why they coded the negative tone of messages, two more items were added to the negative tone category. As a result, both coders were given the modified coding scheme to re-code another 20 messages from the second case. The reliability this time improved to 0.82.

Introduction of Computer Software: DICTION

In order to test the eighth hypothesis that the Optimism Score of media messages will be positively related to the number of negatively toned USENET messages posted after the media stories were released, a computer software, DICTION, designed by Hart (1997) for textual analysis was employed to calculate the number of negative words appearing in each news story.

After selected articles are analyzed, this textual analysis program will calculate scores of five "Master Variables": Certainty, Optimism, Activity, Realism, and Commonality. Every "Master Variable" score has its own formula. Take "Optimism" as an example. The definition of the Optimism Score is "language endorsing or highlighting the positive entailments of some person, group, concept, or event" (Hart, 1997, p. 46). The formula for this variable is (Praise + Satisfaction + Inspiration) (Blame + Hardship + Denial). Terms of Praise, Satisfaction, Inspiration, Blame, Hardship, and Denial are consisted of different words stored in the dictionaries in the software.

To see to what degree a news story covered the crisis event in a negative tone, the Optimism Score was used. However, to simplify the formula, only numbers of words representing above six terms were put into the formula, instead of the standardized score displayed in the report file (See Appendix E). Once the Optimism Score for each news





66


story was obtained, it was compared with the number of daily negative USENET messages to test hypothesis 8.

The next chapter will first present descriptive analysis. Then, description of each crisis case along with results of downloading media stories and USENET messages is provided, followed by the discussion of outcomes of hypotheses testing. Post-hoc analysis will end the chapter.













CHAPTER V
RESULTS



Two corporate crises were selected for testing the hypotheses. The first case is "Mattel Cabbage Patch Doll Defect," which occurred at the end of 1996. The second crisis case, "Hudson Foods Meat Contamination," happening in the mid-1997 was analyzed on the basis of modified coding categories derived firom. the Mattel case.

This chapter begins with descriptive analysis. Then, the results of downloading media stories and on-line discussion messages and test of hypotheses are presented for each case. Post-hoe analysis concludes this chapter.

Descriptive Analysis

Media Stories

The media stories covering both crises were retrieved solely from the Lexis/Nexis data base. Except for a few news stories with the same content, all the news articles appearing in the search results had been gathered and served as a sample of the media agenda. Consequently, 46 media stories covering the Mattel case and 172 stories about the Hudson Foods company were analyzed. On-line Messages on the USENET

The Dej a News search engine ftinctions as the only source for downloading the on-line messages concerning each crisis event. With different keywords used in each case, 488 messages of the Mattel case and 250 of the Hudson Foods case appeared on the


67





68


results pages. However, because not all the messages displaying on the results pages were related to the crisis, many messages were eliminated. Finally, 139 and 154 on-line messages from the Mattel and the Hudson Foods cases were sampled through detailed examination. They were used as the basis for the public agenda to test the hypotheses. Tone of On-line Messages

Tone of on-line messages was coded as either positive, negative, neutral or

unidentified by coders. In the Mattel case, 59 negative (42%) and 5 positive messages

(4%) were posted discussing the crisis. In the Hudson Foods case, 66 messages (43%) were considered negative and 11 messages positive (7%). The remaining messages in both cases were either neutral or unidentified. Interestingly, percentages of negative and positive messages in both cases were very close. Frequency Appearance of User Names in the Discussion Messages

According to Groper's (1996) findings, no single participant in the newsgroups discussion was able to dominate the dialogues and provoke passion among his or her followers. After counting the user name in the e-mail address, it was found that only two persons posted more than three times in all 139 messages in the Mattel case. In the Hudson Foods case, six on-line users posted more than three messages and some even only forwarded media reports without expressing opinions. The majority of on-line users involved in the crisis discussion posted one or two messages. Even though some people might post one single message in different newsgroups, this phenomenon may imply that the corporate crisis did not arouse much self-interest from on-line users who often post messages to the newsgroups.





69


Users' Strateizies

Two types of users' strategy were categorized. One is TALKER strategy applied to users who initiate the discussion thread or express disagreement with other users. Another is REACTOR strategy adopted by users who tend to agree with other people's opinion or do not express their own opinions.

In the Mattel case, 38 users adopted TALKER strategy, while REACTOR

strategy was employed by 78 users. In the Hudson Foods case, 46 users were identified to have a TALKER strategy, and 83 employed REACTOR strategy. Because some online users had posted more than one message, they could adopt both TALKER and REACTOR strategies in various postings. In total, 83 percent of users, 154 for Mattel and 139 for Hudson Foods, were classified as either TALKERS or REACTORS. Frequencies of Flaming Messages

As discussed earlier in Chapter two, flaming messages are often posted and seen in the newsgroups, a public discussion forum without a central authority to regulate users' behavior. Contrary to expectation, the analysis of the on-line messages in the two cases revealed that very few messages could be coded as flaming.

In the Mattel case, 14 flaming messages (10%) were analyzed: 9 targeted on the Mattel company, 4 on other users, and I on both the company and users. As to the Hudson Foods case, flaming messages were even fewer. Only 7 were retrieved: the government was the flaming target in 3, 3 messages targeted on other users, and 1 targeted on both the government and users.





70


Mattel Cabbage Patch Doll Defect Case

Case Description

The Mattel toy company encountered a product defect crisis in the United States at the end of 1996. The following story is based on the media stories retrieved from the Lexis/Nexis data base.

On December 28, 1996, the New York Times reported that a battery-operated

cabbage patch doll manufactured by Mattel, an international toy company headquartered in HongKong, munched a seven-year-old girl's hair and would not let go until her family worked to dissemble the components of the doll. After the news was disclosed in the conventional media, a company spokesman defended the toy, saying that the toy was safe and that no other complaints had been presented. However, in the next few days, more than nine accidents had been reported to the company and the news media as well. The Consumer Product Safety Commission then started to investigate the doll accidents to determine if legal action would be needed.

After January 1, 1997, Mattel decided to take action to resolve the crisis. First, they stuck warning labels on the toy and then recalled all the defective cabbage patch dolls from the retailers, offering 40 dollars as a reftirid to the customers who had bought this type of doll.

After these interventions, no further serious damage case was discovered or reported by mass media after January 10, 1997. News stories concerning the Mattel company after January 10 had nothing to do with this crisis.

The eruption day of the crisis was December 28 and adoption of the intervention strategies was first reported on January 1, 1997.





71


Retrieval of Media Stories from the Lexis/Nexis Data Base

As this crisis was solely related to the Mattel company, the keyword for searching the news stories about the crisis was "mattel." After identifying the date when any of six major media first reported the crisis event, the time frame was set up to retrieve news stories. In this case, the date was December 28, 1996 when the New York Times disclosed this crisis event on Section 1. According to the preset criteria, news stories that appeared on the selected media 10 days before and 32 days after the crisis had to be retrieved from the Lexis/Nexis data base. The time frame hence was from December 18, 1996 to January 28, 1997. The search string was "mattel and date (aft 12/18/1996 & bef 01/28/199 7)."v

Forty-six news stories were downloaded and served as the basis of the media

agenda. Frequencies of news stories appearing in these media sources are shown in Table 5-1.


TABLE 5-1.

Frequencies of news stories on each news media in the Mattel case Valid Cumulative
Frequency Percent Percent Percent Valid ABC News 4 8.7 8.7 8.7
CNN 8 17.4 17.4 26.1
NBC News 4 8.7 8.7 34.8
New York11 2. 23957
Times11 2.23957
Wall Street 14 30.4 30.4 89.1
Journal
Washington 5 10.9 10.9 100.0
Post
Total 46 100.0 100;0
Total 46 100.0





72


In this case, most media stories came from two major news sources, the New York Times and the Wall Street Journal; 5 4 % of total 46 news articles were gathered from these two media.

Downloading of On-line Discussion Messages from Deia News Search Engine

The Deja News search engine provides different ways of retrieving on-line messages that interest the readers. To simplify the search process, the method called "Power Search" was employed, which allows on-line users to specify a time frame and type in the keywords in order to get limited as well as desirable messages related to users' topics of interest.

The time frame for the on-line search was identical to the one established for the conventional media: from December 18, 1996 to January 28, 1997. By using the keywords, "mattel AND (cabbage OR doll)," 488 messages displayed on the results pages. To further examine each message, 139 were actually related to the crisis discussion and therefore were used for testing the hypotheses. Frequencies of messages appearing on each sampling day are shown in Table 5-2.

Test of Hypotheses

Hypothesis 1: The average number of daily USENET dialogues regarding the
corporation will be greater during the ten days following the eruption of the crisis
than the average number of daily on-line dialogues within ten days before the
crisis.

As people did not discuss the crisis event until it erupted, all 488 messages downloaded by using the keywords were used to test this hypothesis. Consequently, there were 66 and 13 6 discussion messages 10 days before and after the crisis eruption,





73


respectively. Table 5-3 shows a significant difference between the two groups.

Therefore, this hypothesis was supported.


TABLE 5-2.

Frequencies of messages appearing on each sampling day in the Mattel case


Valid Cumulative
Frequency Percent Percent Percent Valid 12/28/96 4 2.9 2.9 2.9
12/29/96 8 5.8 5.8 8.6
12/30/96 17 12.2 12.2 20.9
12/31/96 17 12.2 12.2 33.1
01/01/97 5 3.6 3.6 36.7
01/02/97 5 3.6 3.6 40.3
01/03/97 11 7.9 7.9 48.2
01/04/97 10 7.2 7.2 55.4
01/05197 9 6.5 6.5 61.9
01/06/97 1 .7 .7 62.6
01/07/97 10 7.2 7.2 69.8
01/08/97 3 2.2 2.2 71.9
01/09/97 6 4.3 4.3 76.3
01/10/97 1 .7 .7 77.0
01/11/97 11 7.9 7.9 84.9
01/12/97 2 1.4 1.4 86.3
01/13/97 5 3.6 3.6 89.9
01/14/97 8 5.8 5.8 95.7
01/16/97 5 3.6 3.6 99.3
01/17/97 1 .7 .7 100.0
Total 139 100.0 100.0
Total 139 100.0





74


TABLE 5-3.

Results of Hypothesis 1 test in the Mattel case N Mean Std. Std. Error t value
Deviation
10 days 66 6.6 5.1 1.61 -2.756*
before
crisis
10 days 136 13.6 6.2 1.96
after
crisis I

*p:! 0.05 level

Hypothesis 2: The average number of daily negative USENET dialogues will be
greater than the average number of daily positive USENET dialogues after the
eruption of the crisis.

The numbers of positive and negative messages during 32 days after the crisis eruption were compared. Table 5-4 shows a significant difference between the average numbers of daily negative and positive USENET dialogues. It thus supported the hypothesis.

TABLE 5-4.

Results of Hypothesis 2 test in the Mattel case N Mean Std. Std. Error t value
Deviation
Positive 5 0.16 0.51 0 3.59*
Messages
Negative 58 1.81 2.76 0.49
Messages

*p:! 0.05 level

Hypothesis 3: The daily number of the USENET dialogues will be related to the
daily number of news stories about a crisis after its eruption.

Consistent with the concept of agenda-setting theory, the more news stories are about a crisis, the more on-line discussion messages should be posted on the USENET.





75


Simple regression analysis was utilized to test this hypothesis. The number of news stories on each day after the crisis eruption was the independent variable, whereas that of the on-line dialogues was the dependent variable. As r was 0.635 (r2 = 0.403) and significant (F (1, 3 1) = 20.24, p < 0.000), this supported the hypothesis. While the correlation was moderately strong, the relationship was significant and in the expected direction.

Hypothesis 4: The average number of daily negative USENET dialogues
following a crisis intervention strategy will be less than the negative USENET
dialogues prior to the intervention.

Twenty-seven negative on-line dialogues in four days had been posted before the adoption of a crisis intervention and 31 messages had been posted in 28 days after its intervention. Table 5-5 reveals a significant difference that supports the hypothesis. Nevertheless, to compare the average number of negative on-line dialogues four days before and after the intervention, the results shown in Table 5-6 differed and the hypothesis was not supported.

TABLE 5-5.

Results of first Hypothesis 4 test in the Mattel case N Mean Std. Std. Error t value
Deviation
4 days before 27 6.75 2.75 1.38 3.96*
intervention
28 days after 31 1.11 1.95 0.37
intervention

*p:! 0.05 level





76


TABLE 5-6.

Results of second Hypothesis 4 test in the Mattel case

N Mean Std. Std. Error t value
Deviation
4 days before 27 6.75 2.75 1.38 1.075N'
intervention
4 days after 19 4.75 2.5 1.25
intervention I


Hypothesis 5: Following a crisis intervention strategy, the average number of
daily positive on-line dialogues on the USENET concerning the company will be
equal to the average number of daily positive on-line dialogues before the crisis
eruption.

As no positive message was posted before the crisis eruption and three were

posted after the adoption of intervention, no statistical significance was found in Table

5-7. The hypothesis was supported, although the positive messages were very few.


TABLE 5-7.

Results of Hypothesis 5 test in the Mattel case

N Mean Std. Std. Error t value
Deviation
10 days 0 0 0 0 -1.362 NS
before crisis
28 days after 3 0.11 0.42 0
intervention I I


Hypothesis 6: The number of the USENET dialogues following a crisis
intervention will be less than the number of the USENET dialogues before the
intervention and after the crisis.

Comparing the average number of on-line dialogues on each day, four days before

(n, = 46) versus 28 days after the intervention adoption (n2 = 93), the hypothesis was

supported as demonstrated in Table 5-8.





77


TABLE 5-8.

Results of Hypothesis 6 test in the Mattel case N Mean Std. Std. Error t.- value
Deviation
4 days after crisis 46 11.5 6.56 3.28 3.544*
and before intervention
28 days after 93 3.32 3.99 0.75
intervention

*p < 0.05 level

Hypothesis 7: The number of daily negative USENET dialogues posted by TALKERS will be greater than that by REACTORS.

TALKERS posted 22 negative on-line messages, whereas 34 negative on-line

messages were posted by REACTORS. By using a Chi-square test, this hypothesis was not supported as shown in Table 5-9, X2 (1, N = 61) 0.727, p < 0.394.


TABLE 5-9.

Results of Hypothesis 7 test in the Mattel case Negative Positive Total df Chi-square
value
TALKER 22 1 23 1 0.727 NS

REACTOR 34 4 38
1 1 1 1 1 1
Total 56 5 61




Hypothesis 8: The Optimism Score of media messages will be positively related
to the number of negatively toned USENET messages posted after the media
stories were released.






78


To test this hypothesis, the correlation was examined between the average number of negative on-line messages each day and the Optimism Score, which was calculated by subtracting the number of negative words from the number of positive words found in each media story. The Optimism Score is a summed score for all media messages on a particular day.

Simple regression analysis revealed that the Optimism Scores in news stories

were positively related to the number of negative on-line messages posted each day as the correlation coefficient was 0.467, r2 =0.22, p < 0.02. Although the variance explained was small, this hypothesis was supported.


>~12

ID 0
U)
(D
CD .3



M 0 001
z
4-
0 03 001
CD
-0 03 3
E
z 0 D 03 C 0O 0 0 0
z
-2 _____________________________________40 -30 -20 -10 '0 10 2

Optimism Score for All News Stories Each Day


FIGURE 5-1. Bivariate correlation between the number of negative on-line messages
and the Optimism Score in the Mattel case





79


Hudson Foods Meat Contamination Case Case Description

The Hudson Foods company encountered a serious meat contamination crisis

across the United States after July of 1997. The following story is based upon the media stories retrieved from the Lexis/Nexis data base.

On August 12, 1997, the ABC World News Tonight had the following coverage:

an Arkansas meat processor is recalling 20,000 pounds of frozen hamburger patties which may be contaminated with the E. coli bacteria. Hudson Foods distributed the hamburger nationwide. The problem was first spotted in Colorado when several people who had eaten the hamburger became ill." This news article is the earliest one that showed "Hudson Foods" in all news documents of the Lexis/Nexis data base in both July and August of 1997. Therefore, the beginning day of this crisis was considered to be August 12, 1997.

After the government urged more recall of the contaminated meat, Hudson Foods recalled about 1.2 million beef products from the market on August 16, 1997, closed its plant in Nebraska and recalled 25 million beef patties on August 21, 1997, according to the ABC news reports.

In the beginning of September, 1997, Hudson Foods agreed to merge with its rival company, Tyson Foods.

As to this crisis case, the eruption day was August 12, 1997 and the adoption of the intervention strategies occurred twice. The first was on August 16, 1997, when the company recalled one million pounds of beef patties and the second was on August 2 1,





80


1997, when about 25 million beef patties were recalled and a food plant in Nebraska was

shut down.

The intensity of this crisis extended beyond the 42 days framed in this study. The

U.S. News & World Report and the Newsweek in their September editions had stories

featuring the food contamination cases in the past decade to demonstrate the importance

of food safety. The New York Times had the following news story in Section A of its late

edition dated on October 2, 1997:

A Federal grand jury is investigating whether a
meatpacking company tried to hide the extent of potential E. coli
bacteria contamination in what became the largest meat recall.
Tom Monaghan, the United States Attorney for Nebraska,
issued a statement saying the investigation had arisen from
information received last month from the Agriculture Department's inspector general regarding the August recall of 25 million pounds
of hamburger produced at the Hudson Foods plant in Columbus,
Neb.
Mr. Monaghan would provide no other details, but
company officials confirmed today that they had received a
subpoena for documents related to the recall and expected some
employees to be called.
Hudson Foods, based in Rogers, Ark., said in a statement
today, "As it has in the past, Hudson Foods will continue to
cooperate fully."
Hudson shut down the Columbus plant at Agriculture
Department insistence after the amount of meat recalled due to
possible E. coli contamination rose from 20,000 pounds initially to
25 million pounds.

Retrieval of Media Stories from the Lexis/Nexis Data Base

To search for news articles about this crisis event in the Lexis/Nexis data base, the

company's name "Hudsonfoods" did not appear until August 12, 1997 by using the

keywords "Hudson foods and date (bef 081311199 7 and aft 0 71011199 7). Following the

preset criteria, news stories that appear on the selected media 10 days before and 3 2 days

after the crisis need to be retrieved from the Lexis/Nexis data base. The time frame





81


hence was from August 2, 1997 to September 12, 1997 and 172 news stories were downloaded and served as the basis of the media agenda.


TABLE 5-10.

Frequencies of news stories on each news media in Hudson Foods case Valid Cumulative
Frequency Percent Percent Percent Valid ABC News 20 11.6 11.6 11.6
CNN 51 29.7 29.7 41.3
NBC News 29 16.9 16.9 58.1
New York 32 18.6 18.6 76.7
Times
Wall Street 20 11.6 11.6 88.4
Journal
Washington 20 11.6 11.6 100.0
Post
Total 172 100.0 100.0
Total 172 100.0



As Table 5-10 shown above indicates, CNN had 51 stories related to the Hudson Foods company and the New York Times had the second most, 32 news stories covering this company. Interestingly, the ABC did not have any news stories in September containing the keywords of "Hudson foods," so only 20 news stories were analyzed from the ABC news media.

Downloading of On-line Discussion Messages from Deja News Search Engine

As similar to the Mattel case, the search function of "Power Search" in the Deja News search engine was used to obtain on-line discussion messages pertaining to the crisis. Because the last news article was retrieved on September 12, 1997, discussion messages of one more day were downloaded for time-lag analysis. The time frame then started from August 2, 1997 to September 13, 1997. Keywords were "hudson ANDfoods







82




OR beef OR meat." That is, any messages containing hudson foods or hudson and beef,



or hudson and meat would have been retrieved and examined to determine if they were



relevant to the crisis discussion. As a result, 154 messages were regarded as related to



the crisis and became the basis of the public agenda. Frequencies of messages appearing


on each sampling day are displayed in Table 5-11.





TABLE 5-11.


Frequencies of messages appearing on each sampling day in Hudson Foods case



Valid Cumulative
Frequency Percent Percent Percent
Valid 08/08/97 1 .6 .6 .6
08/13/97 1 .6 .6 1.3
08/15/97 2 1.3 1.3 2.6
08/16/97 3 1.9 1.9 4.5
08/17/97 3 1.9 1.9 6.5
08/18/97 5 3.2 3.2 9.7
08/19/97 7 4.5 4.5 14.3
08/20/97 2 1.3 1.3 15.6
08/21/97 2 1.3 1.3 16.9
08/22/97 11 7.1 7.1 24.0
08/23/97 11 7.1 7.1 31.2
08/24/97 8 5.2 5.2 36.4
08/25/97 1 .6 .6 37.0
08/28/97 1 .6 .6 37.7
08/27/97 3 1.9 1.9 39.6
08/28/97 8 5.2 5.2 44.8
08/29/97 6 3.9 3.9 48.7
08/30/97 4 2.6 2.6 51.3
08/31/97 3 1.9 1.9 53.2
09/01/97 5 3.2 3.2 56.5
09/02/97 3 1.9 1.9 58.4
09/03/97 1 .6 .6 59.1
09/04/97 5 3.2 3.2 82.3
09/05/97 3 1.9 1.9 84.3
09/06/97 3 1.9 1.9 66.2
09/07/97 3 1.9 1.9 68.2
09/08/97 9 5.8 5.8 74.0
09/09/97 6 3.9 3.9 77.9
09/10/97 6 3.9 3.9 81.8
09/11/97 11 7.1 7.1 89.0
09/12/97 14 9.1 9.1 98.1
09/13/97 3 1.9 1.9 100.0
Total 154 100.0 100.0
Total 154 100.0





83


Test of Hypotheses

HUothesi-s1: The average number of daily USENET dialogues regarding the
corporation will be greater during the ten days following the eruption of the crisis
than the average number of daily on-line dialogues within ten days before the
crisis.

One message regarding the company was posted ten days before the crisis eruption and 25 were posted ten days after the crisis. The t-test shows a significant difference in Table 5-12. Therefore, this hypothesis was supported.


TABLE 5-12.

Results of Hypothesis I test in the Hudson Foods case

N Mean Std. Std. Error t value
Deviation
10 days 1 0.1 0.32 0 -3.456*
before
crisis
10 days 25 2.5 2.17 0.69
after crisis

*p:! 0.05 level


Hypothesis 2: The average number of daily negative USENET dialogues will be
greater than the average number of daily positive USENET dialogues after the
eruption of the crisis.

The numbers of positive and negative messages during 32 days after the crisis

eruption were compared. The t-test reveals a significant difference between the average numbers of daily negative and positive on-line dialogues in Table 5-13. Consequently, this hypothesis was supported.





84


TABLE 5-13.

Results of Hypothesis 2 test in the Hudson Foods case

N Mean Std. Std. Error t value
Deviation
Positive 11 0.33 0.54 0 5.71
Messages
Negative 68 2.06 1.85 0.32
Messages

*p < 0.05 level


Hypothesis 3: The daily number of the USENET dialogues will be related to the
daily number of news stories about a crisis after its eruption.

Different from the finding in the Mattel case, this result, r = 0.23 1, p < 0. 197,

(1, 32) = 1.742, ? = 0.053, does not support this hypothesis. The relationship between number of news stories and that of on-line messages was not significant.

Hypothesis 4: The average number of daily negative USENET dialogues
following a crisis intervention strategy will be less than the negative USENET
dialogues prior to the intervention.

To test this hypothesis, a one-way ANOVA was used. Unlike Mattel, which had only one intervention, Hudson Foods adopted intervention strategies at two different times. In the Hudson Foods case, hypothesis 4 was tested in three ways.

The first comparison was the average number of daily negative USENET

dialogues among three stages, before the first intervention, after the first intervention but before the second intervention, and after the second intervention. Table 5-14 shows no significant relationship among three stages, F (2, 32) = 2.966, p < 0.067.






85



TABLE 5-14.

Results of comparison of three stages in Hypothesis 4 in the Hudson Foods Case


Sum of Mean
Squares df Square F Sig.
Number of Between 18.140 2 9.070 2.966 .067
negative Groups
message Within 91.739 30 3.058
Groups
Total 109.879 32
a. Non-Significant


The second test was to compare the number of negative USENET messages

before and after the first intervention stages. Table 5-15 reveals a significant

relationship, F (1, 32) 5.38, p < 0.027.



TABLE 5-15.

Results of comparison of two stages in Hypothesis 4 in the Hudson Foods Case


Sum of Mean
Squares df Square F Sig.
Number of Between 16.250 1 16.250 5.380 .027
negative Groups
message Within 93.629 31 3.020
Groups
Total 109.879 32
a. p < 0.05 level



The final test was to compare the number of negative USENET messages before

and after the second intervention stages. A significant relationship was found in Table 516, F (1, 32) = 4.247, p < 0.048.






86



TABLE 5-16.

Results of comparison of two stages in Hypothesis 4 in the Hudson Foods Case


Sum of Mean
Squares df Square F Sig.
Number of Between 13.240 1 13.240 4.247 .048 a
negative Groups
message Within 96.639 31 3.117
Groups
Total 109.879 32
a. p < 0.05 level



Hypothesis 5: Following a crisis intervention strategy, the average number of
daily positive on-line dialogues on the USENET concerning the company will be
equal to the average number of daily positive on-line dialogues before the crisis
eruption.

The ANOVA results in Table 5-17 supported this hypothesis, nj = 1, n2 = 11,

(1, 37) = 2.375, p < 0.132. That is, whether the company adopted the intervention or

not, no significant difference was found in the average number of daily positive on-line

dialogues after the intervention and before the crisis regardless of the few positive on-line

messages.



TABLE 5-17.

Results of Hypothesis 5 test in the Hudson Foods case


Sum of Mean
Squares df Square F Sig.
Average # of Positive Between .632 1 .632 2.375 .132
On-line Messages Groups
Each Day Within 9.579 36 .266
Groups
Total 10.211 37
a. Non-Significant





87


Hypothesis 6: The number of the USENET dialogues following a crisis
intervention will be less than the number of the USENET dialogues before the
intervention and after the crisis.

As Hudson Foods adopted interventions at two different times, this hypothesis was tested in two ways. The first test was to see if the total number of on-line dialogues decreases after the first intervention. By the t-test, Table 5-18 did not support the hypothesis. In other words, the total number of on-line dialogues did not decrease but actually increased with adoption of the first crisis intervention.


TABLE 5-18.

Results of comparison of two stages in Hypothesis 6 in the Hudson Foods Case N Mean Std. Std. Error t value
Deviation
5 days after crisis 6 1.2 1.3 0.58 -4.607*
and before 1 st intervention
28 days after I st 147 5.25 3.48 0.66
intervention

*p:! 0.05 level


The other time frame to test this hypothesis was before and after the second

intervention. Similarly, the t-test results in Table 5-19 showed no evidence of a decrease in the total number of USENET dialogues after the second crisis intervention. The number of USENET discussion dialogues increased after the second intervention.

Hypothesis 7: The number of daily negative USENET dialogues posted by
TALKERS will be greater than that by REACTORS.

TALKERS posted 23 negative USENET dialogues and REACTORS posted 42. By using a Chi-square test, this hypothesis was not supported in Table 5-20, X 2 (1, N 76) = 0.004, p < 0.95.





88


TABLE 5-19.

Results of comparison of two stages in Hypothesis 6 in the Hudson Foods Case N Mean Std. Std. Error t_- value
Deviation
10 days after 25 2.5 2.17 0.69 -2.982*
crisis and before
2nd intervention
23 days after 2nd 128 5.57 3.67 0.76
intervention __ _____ ______ _____ _____*p 0.05 level


TABLE 5-20.

Results of Hypothesis 7 test in the Hudson Foods case

Negative Positive Total df Chi-square
____ ____ ____ ___value
TALKER 23 4 27 1 0.004 NS

REACTOR 42 7 49

Total 65 11 76



Hypothesis 8: The Optimism Score of media messages will be positively related
to the number of negatively toned USENET messages posted after the media
stories were released.

To test this hypothesis, bivariate correlation between the average number of

negative on-line messages each day and the Optimism Score was examined as shown in Figure 5-2.





89





CU
6 13




13 03 0 0 C3 L3



2 n 0 3 11 0 0C
E0000 00
z
-2 _ _ _
-120 -100 -80 -60 -40 -20 0 20

Optimrism Score for All News Stories Each Day FIGURE_ 5-2. Bivariate correlation between the number of negative on-line messages
and the Optimism Score in the Hudson Foods case


The above figure suggested that the Optimism Scores in news stories were not

related to the number of negative messages posted each day as r = 0. 112 (r2 = 0.012) with p < 0.536. As a result, this hypothesis was not supported.


Post Hoc Analysis

According to the agenda-setting theory, there should be a positive correspondent relationship between the media agenda and the public agenda. In the Mattel case, this relationship has been found. The relationship was not found in the Hudson Foods case. To examine why the results of Hypothesis 3 differed between the two cases, additional analyses were conducted. Figures 5-3 and 5-4 show the patterns of USENET messages and the number of news stories.







90




Is 1 6 1 4


12


N no s n s to/i
ay












"0

.2 -..50 Oy
S.mnplin g D ay.




FIGURE 5-3. Relationship between the media agenda and the public agenda in the

Mattel case







25



# of On-line 20 Messages

A 9 of News Stories


15



10











\1) \'4\ \4') 4' \'4\ 4'q 41 411 1\4 Iq\ 0\~\ 4 41 Sampling Days FIGURE 5-4. Relationship between the media agenda and the public agenda in the

Hudson Foods case





The lack of a significant relationship between the media agenda and the public agenda in the Hudson Foods case is evident (r = 0.012). Using SPSS curve-fitting, a cubic relationship was found to improve the explanatory power of the media agenda




Full Text
CORPORATE CRISES IN CYBERSPACE:
EXTENDING PUBLIC RELATIONS MEDIA MONITORING TO
THE PUBLIC DIALOGUES ON THE USENET
By
TZONG-HORNG DZWO
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1998

ACKNOWLEDGMENTS
Earning a Ph.D. from a university in a foreign country is not easy. The academic
demands are challenging enough without the added pressures of an unfamiliar language
and culture. While eager to return to my homeland, I am none-the-less proud of my
achievement at the University of Florida and thankful for all the support I have received
during my academic journey in America.
First and foremost, I wish to offer a tribute to Dr. John Wright who restored my
confidence and encouraged me to remain for the fourth year. He absolutely altered my
view of American teachers. Without him, I would have struggled continuously in
prospectus writing and sobbed in a dark comer of the past. He deserves all the warm
sunshine emanating from my heart.
To say that Dr. John Sutherland is my major advisor is an understatement, because
he not only advised me on my prospectus and dissertation writing, but guided me to a
new world without hatred and revenge. He shared his brilliant insights with me on mass
communication theories and philosophies of life. He also showed me how to behave like
a responsible teacher and researcher in the future. Dedicated to him for his strict but
friendly guidance in my final year is this Chinese motto: Be a teacher one day, be a
Father for ever!
Never is it enough to say "thank you" to Dr. Marilyn Roberts. She makes me feel
that she has more confidence in me than I do myself. All my knowledge, if any, about
n

political communication and American presidential campaigns is attributed to her caring
and warm teaching. She constantly draws a bright rainbow at the other side of the tunnel
whenever I feel downcast. The more I look back at the past, the greater debt I owe to her.
Tire original idea for this dissertation about crisis management came from Dr.
Baker, a public relations expert who provided useful tips on crisis communication and
literature. The theoretical framework of this dissertation was based upon her lectures and
ideas about crisis management.
Dr. Michael Martinez deserves special thanks for his insightful advice on
prospectus writing and instructions in understanding the dimensions of public opinion in
terms of political communication. Learning about political behavior became enjoyable
with his discipline.
My sincere kudos are paid to three doctoral colleagues: Walt McDowell, for
sparing no time in proofreading my prospectus and most of my dissertation text; Tom
Kelleher and Michelle OMalley, for their kindness in coding messages for the reliability
check. I also feel very lucky to have two friends in the Graduate Division, Jody Hedge
and Margi Hatch, who helped me with all the paper work with which I was unfamiliar
during my four years at the college.
I owe a lot to my lovely wife, who had to take care of my cute, but naughty,
children from the first day I arrived at this school. Without her, the pursuit of this
doctoral degree would have been out of the question. Finally, I also thank my parents,
who offer me all the support I need, financially and spiritually. They also exemplify why
they are the greatest parents of all in my mind. Hence, this doctoral degree is solely
dedicated to them.

TABLE OF CONTENTS
page
ACKNOWLEDGMENTS ii
LIST OF TABLES vi
LIST OF FIGURES viii
ABSTRACT ix
CHAPTERS
I. INTRODUCTION 1
Overview 1
Problem Defined 2
Research Purpose 3
Dissertation Outline 5
H. LITERATURE REVIEW 7
Review of Agenda-setting Theory 8
The Internet: A Brief History, Functions, and Current Situation 16
The Challenge of Monitoring the Internet 20
The Internet Research 23
m. CRISIS COMMUNICATION LITERATURE 28
Overview 28
Crisis Management Research and Public Opinion 31
Stages’ Public Opinion Model of Crisis Management 38
MacKuen’s Strategy Model of Public Dialogue 40
Research Questions 44
Research Hypotheses 46
IV. METHODOLOGY 50
Overview 50
Corporate Crisis Selection 51
iv

News Coverage Analysis 52
Media Selection for Analysis of News Stories 53
On-line Discussion Message Analysis 54
Definitions of Coding Categories 57
Conceptual Definitions 59
Operational Definitions 61
Independent and Dependent Variables 63
Introduction of Computer Software: DICTION 65
V. RESULTS 67
Descriptive Analysis 67
Mattel Cabbage Patch Doll Defect Case 70
Test of Hypotheses 72
Hudson Foods Meat Contamination Case 79
Test of Hypotheses 83
Post Hoc Analysis 89
VI. DISCUSSION 101
Summation of Findings 102
Research Questions 103
Proposed Crisis Communication Model 106
Limitations 108
Implications and Suggestions for Future Study 110
APPENDICES
A HOME PAGE OF DEJA NEWS SEARCH ENGINE 116
B POWER SEARCH PAGE OF DEJA NEWS SEARCH ENGINE 117
C CODING SCHEME FOR NEWS STORIES 118
D CODING SCHEME FOR ON-LINE MESSAGES 120
E SAMPLE OF REPORT FILE IN DICTION SOFTWARE 122
REFERENCES 125
BIOGRAPHICAL SKETCH 134
v

LIST OF TABLES
Table page
3-1. Payoff matrix for player encountering friend 42
3-2. Payoff matrix for player encountering friend 42
5-1. Frequencies of news stories on each news media in the Mattel case 71
5-2. Frequencies of messages appearing on each sampling day in the Mattel case 73
5-3. Results of Hypothesis 1 test in the Mattel case 74
5-4. Results of Hypothesis 2 test in the Mattel case 74
5-5. Results of first Hypothesis 4 test in the Mattel case 75
5-6. Results of second Hypothesis 4 test in the Mattel case 76
5-7. Results of Hypothesis 5 test in the Mattel case 76
5-8. Results of Hypothesis 6 test in the Mattel case 77
5-9. Results of Hypothesis 7 test in the Mattel case 77
5-10. Frequencies of news stories on each news media in Hudson Foods case 81
5-11. Frequencies of messages appearing on each sampling day in Hudson Foods case. .82
5-12. Results of Hypothesis 1 test in the Hudson Foods case 83
5-13. Results of Hypothesis 2 test in the Hudson Foods case 84
5-14. Results of comparison of three stages in Hypothesis 4 in the Hudson Foods Case .85
5-15. Results of comparison of two stages in Hypothesis 4 in the Hudson Foods Case. ..85
5-16. Results of comparison of two stages in Hypothesis 4 in the Hudson Foods Case ...86
vi

5-17. Results of Hypothesis 5 test in the Hudson Foods case 86
5-18. Results of comparison of two stages in Hypothesis 6 in the Hudson Foods Case ...87
5-19. Results of comparison of two stages in Hypothesis 6 in the Hudson Foods Case. ..88
5-20. Results of Hypothesis 7 test in the Hudson Foods case 88
5-21. Results of Chi-square test with combined data 99
Vll

LIST OF FIGURES
Figure page
3-1. Group opinion formation process 39
3-2. Sturges' model of public opinion with crisis management plan 40
3-3. Causal relationship between the media agenda and the public agenda 47
5-1. Bivariate correlation between the number of negative on-line messages and the
Optimism Score in the Mattel case 78
5-2. Bivariate correlation between the number of negative on-line messages and the
Optimism Score in the Hudson Foods case 89
5-3. Relationship between the media agenda and the public agenda in the Mattel case...90
5-4. Relationship between the media agenda and the public agenda in the Hudson
Foods case 90
5-5. Cubic relationship between the media agenda and the public agenda in the Hudson
Foods case 91
5-6. Distribution of numbers of on-line messages by types in the Mattel case 95
5-7. Distribution of numbers of on-line messages by types in the Hudson Foods case ....95
5-8. Frequencies of combined on-line messages and media stories 98
5-9. Frequencies of combined negative on-line messages and the Optimism Scores in
media stories 99
6-1. An optimal crisis communication model dealing with public agenda and media
agenda 108
viii

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
CORPORATE CRISES IN CYBERSPACE:
EXTENDING PUBLIC RELATIONS MEDIA MONITORING TO
TFEE PUBLIC DIALOGUES ON THE USENET
By
Tzong-Homg Dzwo
August 1998
Chairman: John C. Sutherland
Major Department: Mass Communication
Managing unexpected crises is a primary concern of the public relations
profession. With the remarkable growth of on-line communication using the Internet,
companies can no longer deal exclusively with conventional media such as newspapers,
radio, and television. Keeping a crisis situation under control has become an even greater
challenge in the age of cyberspace. This dissertation applies basic agenda-setting theory
to two crisis cases involving the interaction of conventional media coverage and the
USENET dialogues by on-line users. The on-line dialogues were sampled from a search
engine on the Internet. The conventional media stories were gathered from the
Lexis/Nexis data base. Two corporate crises selected for testing the hypotheses were the
Mattel Doll Defect case and the Hudson Foods Meat Contamination case.
The crisis communication model predicted that after a crisis happens, public
opinion about the company would be more negative than before its occurrence. As the
IX

company adopted interventions to ameliorate the detrimental impact caused by the crisis,
the number of negative on-line dialogues would decrease. Unlike the traditional method
of examining the public agenda in agenda-setting studies, content analysis of the public
dialogues on the USENET, a dependent variable, was employed in comparison to the
conventional media agenda, an independent variable. The number of on-line dialogues
was assumed to have a corresponding relationship with the media coverage of the crisis.
While the number of negative on-line messages about the company increased
after the crisis and were reduced due to intervention strategies adopted by each company,
the total number of on-line messages did not decrease after the intervention in the Ehidson
Foods case.
Results in the Mattel case revealed a significant correlation between the number
of media coverage and that of on-line messages. The significant correlation between the
tone of media stories and the number of negative on-line messages was also found in the
same case. However, this relationship that derived from agenda-setting theory was not
detected in the Hudson Foods case.
Despite the specific findings in each case, further examination of the combined
data supported the basic components of agenda-setting theory. There were correspondent
relationships between the media reports and the public agenda as well as between the
tone of media stories and the number of negatively toned on-line messages.
x

CHAPTER I
INTRODUCTION
Overview
The topic of media effects has been a primary research domain for over three
decades. Among the many theoretical approaches that have been proposed, one of the
most popular has been agenda-setting theory. Hundreds of studies have utilized aspects
of the agenda-setting framework to reveal positive correlations between media coverage
of issues and their distinctive importance in people's minds.
The development of agenda-setting theory can be understood from two levels of
research. The first level study can be traced back to 1972 when McCombs and Shaw
wanted to explain how and why people think about different social issues and rank their
importance. The findings showed a significant correlation between people's ranking of
issues' importance in their minds and the amount of media coverage, hi other words, the
salience of issues in people’s minds is greatly influenced by and transmitted from the
media coverage.
Besides the original focus on the issue salience transmitted from media to the
public’s mind, media effects in view of agenda-setting theory have also been found from
another perspective. The selection of objects and attributes of an issue in the news media
becomes a major concern. How news frames and agenda attributes impact the public
agenda plays a central role in the second level study of agenda-setting studies (McCombs,
1

2
1997). The research question in the second level has moved from “Does the amount of
media coverage transmit the issue salience to people’s minds?” to “To what extent is the
public's view of an issue shaped by the media, that frame objects or attributes of the issue
in different ways?” To further understand the media effects on people’s interpretation of
social issue(s), researchers look into the content of media coverage, instead of the number
of media stories on particular issues, and then examine how the news content has been
framed in ways that most affect people. The key difference between the first level and the
second level approaches to agenda-setting theory is in the way researchers study the
media coverage.
The central concept of agenda-setting theory is an explanation of the causal
influence of the media agenda on the public agenda. Regardless of the level of agenda¬
setting research that measures the media agenda from different aspects, the most common
methodology is the survey. For example, researchers will ask people “What is the most
important issue facing the society or yourself?” or “How do you think about a particular
issue or person?” These responses are compared with media content analyses to
determine if the media content influences people’s thinking about different issues.
Problem Defined
Whether the survey is conducted by telephone or mail, this method can be
somewhat obtrusive and may not capture the true opinions of subjects. What people
really think or say about the issue(s) in the interest of the researchers may not be
completely detected by asking conventional survey questions.

3
In contrast to surveying, content analysis has been available as a less obtrusive
method for analyzing messages for many years. Researchers have used content analysis
to evaluate the intentions or opinions of the professional communicators such as
newspapers, magazine, radio, television, and so on. However, no studies have used
content analysis to probe the expressed public opinions of the receivers of these mediated
messages about some particular issues within agenda-setting theory. With the
introduction of new interactive media technologies, particularly the Internet, ordinary
people can now express their views to their fellow citizens without the use of the
conventional media. Based upon the public relations crisis management concept
combined with new communications technology, this research will employ content
analysis to measure the public agenda for the purpose of extending the dimensions of
agenda-setting theory.
Research Purpose
Given the Internet’s dynamic characteristic of transforming information senders to
receivers, this new electronic delivery system provides both possibilities and problems for
the public relations professional. In particular, the Internet offers special challenges when
a company experiences some type of crisis.
The public relations field has long been criticized for its lack of a formal and
consistent theoretical framework to help its empirical practice; neither is there a mutual
understanding between public relations theorists and practitioners (Terry, 1989). This
tension is exacerbated when discussing crisis management communication. Because
public relations crisis communication demands an accurate depiction of public opinion

4
during the crisis, the Internet needs to become an integral part of contemporary agenda¬
setting theory.
By modifying the concepts of agenda-setting theory founded by McCombs and
Shaw in 1972, the current study will examine the relationship between the traditional
media coverage of a corporate crisis and public opinion expressed on the USENET, one
of the popular functions on the Internet. While previous agenda-setting studies examined
public opinion by either conducting cross-sectional telephone surveys or using secondary
public opinion surveys, this study will break new ground by abandoning survey
methodologies in favor of content analysis of the on-line public messages on the
USENET. This new type of analysis will permit public relations practitioners to “eaves
drop” on people’s conversations about the crisis event without disturbing their
communication. This method also will consolidate the theoretical framework of agenda¬
setting theory in the domain of the public agenda. In addition, the time frame effects of
media coverage of corporate crises on the on-line public discussions that occur daily on
the Internet are examined.
Another goal of this study is to understand the flow of on-line discussions when a
corporation is in the process of a crisis and how corporate crisis management strategy
impacts the flow of on-line discussions.
The examination of on-line messages as opposed to traditional media news
coverage will provide preliminary insights into understanding the formation of on-line
public opinion. Comparison of the media agenda and the public on-line agenda is needed
to have a better understanding of how these two factors interact.

5
After analysis of two case studies involving both the media agenda and the public
agenda concerning a corporate crisis, this research outlines a dynamic crisis
communication model that is intended to help companies prevent further damage as a
result of an organizational crisis. Based on a public opinion model of crisis management
proposed by Sturges (1994), this restructured model will provide greater insights into how
to effectively manage public opinion and control the crisis to the advantage of the
corporation.
Dissertation Outline
The following chapter describes the essential concepts and development of
agenda-setting theory that forms the basis of this study. Studies related to the Internet are
presented as well. In the third chapter, crisis management literature is reviewed to
understand the points of view of both scholars and public relations practitioners about
how to manage the corporate crises. To apply a general public opinion model to the flow
of on-line discussion messages, Sturges' (1994) model along with MacKuen's strategy
model of public dialogue is introduced. The research questions and hypotheses will
conclude the third chapter.
Chapter four discusses the methodology used in this study: content analyses of
both traditional media coverage and on-line discussion messages. It also explains the
intercoder reliability test, independent and dependent variables, and conceptual
definitions. The fifth chapter presents the findings of hypothesis testing for two corporate
crises and post-hoc analysis. In the final chapter, a summation of the results, the

6
discussion of findings, limitations of this study, and the practical implications and
suggestions for future research are discussed.

CHAPTER H
LITERATURE REVIEW
When a crisis occurs, the media can shape the general public’s perception about
the targeted organization and how it deals with a particular crisis. To better understand
how mass media affect people’s perceptions about the organization, a sound theoretical
background is necessary. A viable theory will help public relations professionals
understand not only how their company is portrayed in the media, but to what extent the
general public has been affected by the media coverage of the crisis.
Having been tested for nearly three decades in the mass communication field,
agenda-setting theory emphasizes a unidirectional impact from the media agenda to the
public agenda. Applying this theory to the crisis management communication, public
relations professionals are better able to detect the trend of public opinion and how
long-term media coverage influences people’s opinions as a crisis event evolves. After
understanding how the public thinks about an issue, the corporation can then adopt
appropriate strategies to manage and ultimately resolve the crisis.
The following chapter will first review the agenda-setting theory in detail,
followed by a discussion of pertinent studies of the Internet.
7

8
Review of Agenda-Setting Theory
First Level Research
Agenda-setting theory was originally proposed by McCombs and Shaw in 1972
when they conducted their pioneering study in Chapel Hill, NC. In that study, the amount
of media coverage of various issues within a period of time was compared with the
public’s ranking of their importance. The results revealed the powerful effects of media
on transmitting the salience of issues to people’s minds.
Three types of agenda are involved in the agenda-setting research: media agenda,
public agenda, and policy agenda. The hierarchical effects derived from this theory are
that the media agenda would influence the public agenda, which in turn may influence
policy agenda (Dearing & Rogers, 1996). Sometimes a fourth variable is studied in
agenda-setting studies: a real-world indicator, that Dearing & Rogers (1996) defines as "a
variable that measures more or less objectively the degree of severity or risk of social
problem" (p. 23). However, since the real-world indicator varies depending on the issue
topic, this variable to test agenda-setting theory is seldom used.
Measurement of Media and Public Agenda
Media agenda are usually measured by content analysis of the news media,
initially employed by McCombs and Shaw (1972) and Funkhouser (1973), to detennine
the number of news items about an issue or issues of study. The more news coverage is
devoted to an issue, the more salient that issue is considered to be for the media. In this
case, number counting is a very important way to measure media agenda. With growing
popularity of computer usage in recent years, computerized content analysis has become

9
an efficient alternative to counting news stories or identifying words in a text when they
appear together. It also facilitates qualitative analysis in some cases.
Public agenda were originally measured by asking a question in public opinion
surveys: "What is the most important problem facing this country today?" This type of
question has been used for decades in national public opinion surveys. It can indicate the
relative importance of an issue in the public's mind. In contrast to the measurement of
media agenda, the researchers will determine whether people take cues from the amount
of media coverage to answer the "MEP" (most important problem)-type question.
Intervening Variables in Affecting the Public Agenda
Although the original agenda-setting study has specified the direct relationship
between the media agenda and the public agenda, many other factors intervening in the
causal relationship have been found in various studies including intensity and variation in
news coverage. Brosius and Kepplinger (1990) found that agenda-setting effects were
most likely to occur when coverage was intense and when there was significant variation
in the coverage from month to month. In addition, the type of issue is another focus in
the agenda-setting study. Wanta and Hu (1993) found that press coverage, besides
increasing public concern with certain issues, can also decrease concerns. It mainly
depends upon recipients’ self-involvement and interest in the issues.
Media credibility is another determining factor. Rogers and Dearing (1988)
concluded that the Herald-Tribune, a local newspaper, transmitted more salience of a
news item to the subjects than the National Enquirer, a national tabloid magazine. Wanta
and Hu (1994) also indicated that individuals who perceive the media as more credible

10
would rely more on the media for information about issues and hence be more susceptible
to the media influencing their understanding of issues.
The extent of media exposure is another intervening variable. The more media
exposure, the more possible the public agenda will be influenced by the mass media
(Weaver et al., 1981; Mullins, 1977; Wanta & Hu, 1994). People’s need for orientation
also transfers the media agenda to the public agenda. Zucker (1978) pointed out that if
people have fewer direct experiences with an issue, the news media's influence on public
opinion about that issue will be greater than otherwise. By analyzing television and
newspaper issue coverage for four weeks, Wanta and Wu (1992) found that interpersonal
communication can reinforce the media agenda-setting effects on the public agenda when
the conversation deals with the same issue that media have emphasized.
Studies of Intermedia Influence
Not only do newspapers and television play a central role in setting media agenda,
but the wire services set the agenda for newspapers as well. The coefficient of
concordance between the contents of 24 Iowa daily newspapers across 13 categories and
those of the Associated Press amounted to 0.915 in Gold and Simmons’ (1965) study.
After analyzing 20 afternoon newspapers around the nation, Stempel (1964) found that
average use of the Associated Press news items by these papers was 22 percent, ranging
from a low of 11 percent by a New York paper to 34 percent by the Rochester Times-
Union.
Whitney and Becker (1982) examined the gatekeeping effects of wire service
news among 46 editorial managers from newspapers and commercial television stations.
They concluded that at the least local media are greatly influenced by “the decisions of a

11
relatively few editors operating at the regional, national and international bureaus of the
wire services” (p. 65). Reese and Danielian (1989) studied the media coverage of the
drug issue in 1986 and provided evidence of strong intermedia agenda-setting effects
from the New York Times to other television and newspaper organizations.
Time Frame Issues in Agenda-Setting Research
How long will an issue remain salient in people’s minds? One of the most
essential considerations in testing agenda-setting hypotheses is the time frame utilized by
the researchers in their studies (Eaton, 1989; Funkhouser, 1973; Mullins, 1977; Sohn,
1978; Stone & McCombs, 1981; Wanta & Hu, 1994). In general, time-lag selection is
important because it demonstrates the time-varying causal effects. The length of time
taking effects from the media agenda to the public agenda depends upon factors such as
nature of the issue, total amount of media coverage, personal relevance, and so on (Eyal,
1979; Mazur, 1987). Salwen (1988) urged that any time discrepancies in the
measurement of the public agenda may affect the public’s evaluations of issue salience.
Agenda-setting studies have to be concerned with the time frame over which
media coverage has the most impact on public opinion. Winter and Eyal (1981)
suggested the “optimal effect span” is between 4 and 6 weeks, whereas Stone and
McCombs (1981) thought that it takes two to six months for changes in the media agenda
to be fully “translated” to the public agenda. Shoemaker et al. (1989) used two months as
a basis to analyze the relationship between drug coverage and public opinion. Shoemaker
et al. (1989) proposed that coverage which recurs in emphasis on a three- or four-month
schedule may have the most influence on public opinion. Consequently, the causal
relationship between media and public opinion will be detected within a period of time

12
when media repeatedly cover one particular issue. As the above findings dealt with the
traditional media agenda-setting effects, no research exists in previous agenda-setting
studies regarding the time frame of on-line newsgroups and issue salience.
Direction of Causal Relationship between Media Agenda and Public Agenda:
With respect to the relationship between the media agenda and the public agenda,
Ito (1993) believed that “mass media effects will not take place unless media stand on the
majority side or the mainstream in the tripolar relationship among the mass media, the
government and the masses” (p. 123). According to Behr and Iyengar’s (1985) finding,
the direction of causality between media coverage and public awareness of the issue was
unidirectional, that is, media coverage on the issues had quite an impact on people’s
awareness of them. Beniger (1978) maintained that the public’s attitudes and opinions
may be more closely correlated with media coverage than with more objective social
conditions. Indeed, Zhu et al. (1993) found that the public’s issue priorities are much
more influenced by media coverage than by social interaction.
Willnat and Zhu (1996) also provided evidence of one-sided newspaper coverage
influencing public opinion about a governor’s overall performance when they compared
time-series data of public opinion polls with three leading newspapers in Hong Kong. To
investigate the impact of press coverage on the general public’s belief about HIV
transmission, Hertog and Fan (1995) discovered a significant causal relationship from
news contents to the public opinion, but not vice versa. From these findings, agenda¬
setting theory suggests to public relations professionals the importance of monitoring
media coverage of issues that have grabbed public attention.

13
On the first level of agenda-setting theory, the researchers put much emphasis on
what media say about different issues and then correlate it with how people rank the
importance of those issues. The content of media is not examined in particular. Most
studies focus on the amount of news coverage in the media versus public opinion about
the examined issues. The public agenda is measured by surveying people’s opinion,
instead of actually analyzing how and what people say about the issues. The core concept
of the first level agenda-setting theory is “where public opinion about an issue comes
from” (Whitney & Becker, 1982; McCombs & Shaw, 1972; Wanta & Hu, 1994; Zucker,
1978).
Second Level Research
In addition to the original focus on issue salience transmitted from the media to
the public’s mind, agenda-setting theory studies have moved to the second level. How
news stories are framed and issue attributes are presented play a central role in impacting
the public’s understanding of the issues (McCombs, 1997). On the second level, the
researchers attempt to answer the question: "To what extent is the public's view of an
issue shaped by the media that frame objects or attributes of the issue in different ways?"
To frame, as Entman (1993) explains, "is to select some aspects of a perceived
reality and make them more salient in a communicating text, in such a way as to promote
a particular problem definition, causal interpretation, moral evaluation and/or treatment
recommendation for the item described" (p. 52). Because all issues have different
dimensions, they can be interpreted from all kinds of perspectives. Through selection and
use of some attributes or objects of an issue, media can make some news prominent and
transmit that prominent aspect to the audience’s mind.

14
Given that the agenda-setting theory has progressed from counting numbers
(media agenda) and ranking issues (public agenda) to analyzing and explaining the textual
meanings of the media coverage, it can be assumed that the media are not only telling us
what to think about (Cohen, 1975), but also tell us how to think about some objects or
attributes of particular issues.
To understand the conceptualization of media frames and attributes, Ghanem
(1997) breaks down four dimensions of media frames (p. 10):
1. The topic of a news item (what is included in the frame);
2. Presentation (size and placement);
3. Cognitive attributes (details of what is included in the frame);
4. Affective attributes (tone of the picture).
Although the idea of news framing is quite novel in recent years, McCombs
(1997) provided some research studies confirming the effects of agenda-setting on the
second level. In regard to issue framing, Takeshita and Mikami (1995) found a highly
positive correlation between the salience of system-related aspects of reform on the public
agenda and people’s attentiveness to political news on both newspapers and television
news. Maher (1995) showed a high correspondence in the relative salience of the
attributes that defined the pictures of local environmental issues in the newspaper and
among the public.
In earlier studies, Weaver, Graber, McCombs, and Eyal (1981) found a striking
degree of correspondence between the agenda of attributes in the Chicago Tribune and
the agenda of attributes in Illinois voters' descriptions of Jimmy Carter and Jerry Ford.
The correspondence between the agenda of attributes in Newsweek and the agenda of

15
attributes in New York Democrats' descriptions of the contenders for the presidential
nomination has also been reported by Becker and McCombs (1978).
Concerning candidate attributes framed in the news stories, King (1997) compared
newspapers agenda on candidates for Taipei mayor with voters’ images of three
candidates and revealed six significant correlations ranged from 0.59 to 0.75. McCombs
et al. (1998) studied the second-level effects of agenda setting in the Spanish election and
found significant correspondence between various news and political advertising agendas
and the pictures of the parliamentary and mayoral candidates in voters' minds. In that
study, the strongest effects resulted from the affective correlations between news
coverage and voters' affective description of candidates.
We have already seen how the media agenda directly affect the public agenda, but
there is another aspect of the media that must also be considered: paid advertising. In
addition to news stories and commentary contained within programs, the general public is
often bombarded with advertising messages. For instance, to examine the effects of
political advertising, Roberts (1997) found that political advertising in the 1990's Texas
gubernatorial election campaign had great impact on media news items. Rank orders of
the media agenda were correlated with the political advertising agenda at two time
frames. The salience of candidates' image and political issues were also transmitted to
the voters' mind and further influenced their voting decision. This two-step process of
agenda-setting, as Roberts contended, reveals that candidates' paid political advertising
can shift the priority of the media agenda as well as shape the public agenda. The
agenda-setting process has gone beyond the influence on people's cognitive level to the
behavioral outcomes.

16
The idea of the second level agenda-setting research sheds much light on how and
what to examine in media effects on the direction of public opinion. The findings also set
a new path to delve into the relationship between the media and the public agenda.
Scholars focusing on the latter begin to look into the contexts of media in order to probe
why public opinion is formed. To understand people’s cognitive knowledge and affective
reactions to the issues has broadened the explanatory domain of agenda-setting theory.
Although the contexts of the media agenda have been thoroughly examined in
terms of the second level of agenda-setting theory, measurement of the public agenda still
relies on the survey to understand how people’s understanding of distinctive issues was
influenced by the media news framing. The growing development and adoption of new
communication technologies, which allow people to talk to each other interactively,
provide another useful tool for media researchers or professionals to analyze the public
agenda without directly surveying people.
The Internet: A Brief History, Functions, and Current Situation
The origin of the Internet can be traced back to 1969 when the first node was
installed at UCLA in California for research purposes. At that time, the Internet was used
only by scientists and government organizations. By December 1969, four nodes were
connected on the early computer network, namely ARPANET. With the establishment of
ARPANET, scientists and researchers could share one another's computer facilities at
different places. Ironically, the main traffic on ARPANET was not long-distance
computing. Instead, it was news and personal messages. Many researchers through
ARPANET were collaborating on projects, trading work experiences, and even

17
exchanging some gossip. Throughout the 1970s, ARPANET was expanding at a fast
pace.
Three decades later, the Internet has moved beyond its original functions for
military and research institutions, and entered the world of public education as well as the
commercial sectors. The Internet's pace of growth in the early 1990s is so rapid that
currently there are tens of thousands of nodes on the Internet that extend to over 40
countries in the world, with more coming on-line every day. The primary attraction that
the Internet has is that no official censorship has been implemented at least up until now.
Nobody controls this so-called "Information Highway."
Generally, people make use of the Internet in the following areas: World Wide
Web1 (WWW) surfing, electronic mail (e-mail), the USENET, Telnet, Gopher searching,
and file transfer.
To exchange information or ideas on the Internet, electronic bulletin board
systems (BBS), mailing lists (Listserv), and moderated newsgroups (USENET), have
been among the most popular electronic discussion groups for more than a decade.
Electronic bulletin boards function as a specialized medium and can be set up by
inter-connecting personal computers for the purpose of serving the debate, association,
and exchange function of that community (Thomsen, 1996). Thousands of people could
contribute to an open discussion or an informal dialogue at a specific time on this system.
A Listserv is a program that maintains one or more of mailing lists. A listserv
automatically distributes an e-mail message from one member of a list to all other
1 The World Wide Web is a protocol that allows users to easily make hypertext
information available to other users.

18
members on that list. Listservs maintain thousands of lists in the form of digests,
electronic journals, discussion groups and the like. In general, the topics discussed in the
Listservs are more serious than those in the newsgroups and tend to be more focused on
particular issues.
The USENET, also known as newsgroups, is a system where messages about any
subject can be posted, and other people on the Internet can reply to them. These on-line
discussion groups are a world of their own. Vast amounts of news information, debates
and arguments involving miscellaneous topics are transmitted and posted among more
than 5,000 discussion groups every day. The USENET also distributes various free
electronic journals and publications.
Another unique characteristic of the USENET is that senders and receivers do not
have to communicate with each other at the same time. Once posted on a newsgroup, the
messages are distributed to other readers immediately and can be responded to at any
time. Unlike the Listserv networks, the USENET has no central authority and each
contribution is passed throughout the system of interconnected hosts— systems that
receive and pass along each contribution they receive.
Although the above three electronic discussion groups provide a relatively free
environment for opinion or ideas exchange, informal but implicit rules of behavior called
“netiquette” have always been required when one person attempts to post his or her
message on the discussion groups. It is a user’s duty to post the messages relevant to the
topic issues of the appropriate discussion group without intentionally offending other
participants. Nevertheless, due to the user’s ability to conceal his or her identity in the
computer-mediated communication, a user might post the messages with no factual

19
information or provide excessively critical comments. This form of message is called
“flaming,”2 which often generates a series of public posts in which people flame one
another rather than contribute useful information. Hill and Hughes (1997) defined
flaming messages as personal attacks, usually accompanied by profanity. In his article,
Groper (1996) thought that flaming words or phrases are extreme criticism, sarcasm,
cursing, exclamations, multiple question marks, emoticons like :-(, which connote anger,
and block capital words. These socially undesirable acts mainly come from lack of
standard rules in the electronic communication world, namely, the Internet (McLaughlin
et al., 1995).
While radio took 30 years to reach an audience of 50 million, and television took
13, the Internet took just four years (Weise, 1998). In a recent report on its first major
study of the economic impact of the Internet, the Commerce Department of the United
States disclosed that "Net traffic is doubling every hundred days and electronic commerce
should reach $300 billion by 2002. More than 100 million people are now on line...."
(Weise, 1998).
How many users are surfing on the Internet today? According to a survey
conducted by the RelevantKnowledge company in late February of 1998, the Web users
in the United States have grown to 57,037,000, a growth of over 1.6 million users since
the company's last study in January 1998. Jeff Levy, CEO of the RelevantKnowledge
2 A public post or email message that expresses a strong opinion or criticism.
Flames can be fun when they allow people to vent their feelings, then return to the topic
at hand. Others are simply insulting and can lead to flame wars.

20
company, attributes the growth to an increase in awareness of the Internet among the
public in the United States.
The Challenge of Monitoring the Internet
As new technologies have changed life patterns in the past decades, the services
provided by these new technologies have grown in number. Providing the users more
control over the communication process, the interactive characteristics of the Internet
have changed the conventional communication paradigm that only sources have ability
and power to disseminate information. The communication capabilities of the Internet
are too enormous to be ignored. On the Internet, the devices such as electronic mail,
newsletters, and on-line discussion groups allow corporations around the world to
promote products, get feedback or comments from customers, and respond to questions
from a variety of publics. Similarly, the Internet has transfigured the arena in which the
public relations professionals operate.
For instance, when an unexpected corporate crisis event occurs, a story containing
inaccuracies about the crisis published on the electronic newspapers will stay on the Net
and be retrieved and used by other journalists as well as most on-line users. The
journalists can also obtain the first or longest story, correct or not, on any particular issue
and re-distribute it to other news sources or readers. Given the potential of instant and
interactive worldwide communication, the Internet can no longer be considered a
secondary communication medium.
In Kalish’s article appearing in the Reuters North American Wire (Feb. 13, 1997)
on public relations firms handling the Internet, he remarked that public relations veterans

21
agreed that about the worst thing a company could do is ignore persons who badmouth
them in cyberspace. Speaking of the advantages of the Internet for public relations,
Strenski (1995) pointed out that “professional public relations communicators can choose
from a blossoming array of cyberspace resources through which to channel their
messages” (p. 33).
Issue tracking is an integral part of public relations work. Whether it is called
issues management or crisis management, scanning the environment to seek public
opinion trends should be a top priority. There are many ways to scan and monitor the
evolution of issues. The most widely used method is the scanning of trade publications,
books, scholarly journals and, most importantly, the news media. In addition, themes
from popular entertainment such as movies, plays, novels and television shows should be
evaluated as well since they are consumed by millions of persons.
Besides traditional ways of monitoring the external environment, the Internet
offers another dynamic channel for public relations practitioners. One of the most
popular functions utilized by on-line users is called newsgroups. Thousands of electronic
forums have been established to exchange messages on countless issues or topics. The
subscribers to each newsgroup can remain anonymous, if desired, when they express their
own opinion. To explain the popularity of the newsgroups on the Internet, Berger (1995)
believes that the popularity stems from the loss of credibility of traditional media by the
general public. However, it is also recognized by most users that “much of the discussion
is trivial, and that, more often than not, debate rapidly disintegrates into verbal assaults
and idle banter” (Revah, 1995, p. 10). The dark side of the Internet is that “anyone can
say anything to anybody at anytime,” commented Don Middleberg, president of a public

22
relations firm (Kalish, 1997). As a result, some people are participating in a variety of
newsgroups that damage companies’ reputations.
To emphasize the importance of monitoring newsgroups discussion for references
to the business corporations, Kalish cited the suggestion of Adam Cooper, creative
manager at the Interactive Solutions Group of Edelman Public Relations Worldwide,
stating that, “We find that the newsgroups are much more interactive than the web, so the
newsgroups are really what we would consider the key place to listen to people” (Reuters
North American Wire, Feb. 13, 1997). Two examples involving corporate crises on the
Internet-related discussion bolster Cooper’s viewpoint.
One example was the crisis pertinent to the introduction of Intel’s new Pentium
computer processor. The Intel organization discovered the Pentium flaw in the summer
of 1994 but declined to acknowledge the problem publicly. By failing to recognize the
importance of the message posted on a CompuServe3 forum, Intel paid out over $450
million to replace the flawed chip. This is one example of ignoring the public’s concerns
about its own products, as well as its own image.
Another rumor-like incident occurred in November 1996 when people on the
Internet labeled fashion designer Tommy Hilfiger a racist and called for a boycott of his
products. In contrast to the Intel’s reaction, the fashion designer Tommy Hilfiger paid
attention to the offending Internet sites and responded by launching an e-mail campaign
that assured minorities on the Internet that they were valued customers and that the racist
rumors were totally groundless.
3 CompuServe has been one of major commercial on-line services in the United States for
over fifteen years.

23
Learning from the above examples, it is obvious that an Internet-caused crisis can
have profound influences on public opinion. On-line users can discuss any company in
different newsgroups and detractors can design a web page full of factitious information
about a specific company. These examples of business disruptions as a result of the
Internet discussions and the resulting media coverage can be influenced by an organized
group or a single individual operating in cyberspace. For whatever the reasons the
situation arises, the “on-line crisis” has the potential to damage a corporate image and
creates a corporate management problem for public relations practitioners.
The Internet Research
Is the Internet truly a mass medium? The answer is both yes and no. A properly
constructed website can address thousands of audiences while still dealing one-on-one as
an interactive medium. The more important issue is that these new websites must be
monitored in additional to the traditional media such as television, radio, and newspapers.
To what extent can the Internet be regarded as a mass medium? The Internet
encompasses interpersonal conversation and mass distribution of messages and each
message can reach a large audience. Morris and Ogan (1996) cited Valente’s idea about
the critical mass: “the critical mass is achieved when about 10 to 20 percent of the
population has adopted the innovation. When this level has been reached, the innovation
can be spread to the rest of the social system” (p. 45).
Before delving more into the impact of new communication technologies, it may
be worthwhile to reevaluate the definition of mass media. Does the Internet play the role
of an interactive medium? Rafaeli (1986) regarded this new type of communication

24
medium as similar to other media; however, it indeed integrates two interests in
communication with the technological innovation of computer networking: “interpersonal
conversation and mass propagation of messages” (p. 123). In Bucy’s (1995) article
discussing an objective measurement of interactivity, he suggested that the interactive
event should not only activate an interpersonal conversation, but it has to be “a form of
mass communication in which the sender broadcasts a message to a mass audience in the
process of responding to an individual” (p. 6). Hiebert (1992) considered message
transmission through a mass medium as an effective communication that elicits feedback
from the audience.
Rafaeli's (1988) definition of interactivity based on the issue of responsiveness
recognizes three pertinent levels: noninteractive communication sequences, quasi¬
interactive (reactive) communication, and fully interactive communication. The
distinction between quasi- and full interactivity depends on the nature of the
communication responses. In the fully interactive communication, sequential messages
“depend upon the reaction in earlier transactions, as well as on the content exchanged” (p.
118-9). Deriving from Rafaeli’s model, interactivity is “feedback that relates both to
previous messages and to the way previous messages related to those preceding them” (p.
120).
Rafaeli and LaRose’s (1993) study on electronic bulletin boards treated electronic
bulletin boards as a kind of collaborative mass medium, “a new type of mass
communications medium in which the audience acts both as the source and the receiver
of the message” (p. 277). An interactive medium is a “many-to-many” communication
channel unlike conventional mass media such as television, radio, and static print media,

25
which are a “one-to-mass” channel. Also, electronic bulletin boards represent an
interactive medium that demands users’ contribution as well as consumption.
Garramone, Harris, and Anderson (1986) explored the motivations for
participation on a political bulletin board. They observed no statistically significant
relationships between the use of BBS features and demographics or political partisan
predispositions. Also in their study, surveillance was ranked as the number one
motivation for BBS use, followed by curiosity, and knowledge of others’ opinions. In his
article defining electronic bulletin board users, Rafaeli (1986) concluded that the users of
boards could be characterized as both a faithful and active audience.
To further probe the success of electronic bulletin board systems, Rafaeli et al.
(1993) found that the diversity of board contents had positive impact on board users’
contribution levels and small group size contributed to adoption level of the BBSs. After
content analyzing public relations practitioners’ use of an electronic newsgroup called
PRForum4 Thomsen (1996) distinguished three main functions in using the PRForum:
exchange of information and advice, debate over issues affecting the profession, and
cultivation of a sense of self-validation and enhanced efficacy both at a personal and
professional level.
Among tools on the Internet that are influential in both public relations and
advertising fields, Bobbitt (1995) recommended the USENET newsgroup as a public
relations tool for both issues and audience research. Because study of issues and
4 As noted in Thomsen’s article, PRForum, founded in 1993, is an online discussion
group for public relations professionals. Updated in List search website
(http://www.lsoft.com/lists/list_q.html), there were 1,224 people subscribing to this group
as of May 13, 1998.

26
audience analysis is dependent on the credibility of the respondents, the researcher must
assume that expressed opinions or comments are indeed truthful. Treating the USENET
as a political discussion forum, Groper (1996) concluded that the political leadership did
not exist in the three sampled newsgroups. No single participant in the newsgroups
discussion was able to dominate the dialogues and provoke passion among his or her
followers. Hill and Hughes (1997) analyzed messages in 22 political newsgroups and
found that most political USENET groups exhibited the traditional characteristics of a
socially cohesive group. The authors further argued that although these groups became a
new form of political community for the purpose of political dialogue, it did not signalize
any paradigm shift. Instead, “people are merely moving their age-old patterns of
interaction into a new realm” (p. 25).
To what extent should the public relations people evaluate this new interactive
medium, the Internet? According to a comprehensive demographic survey released in
March of 1997 by CommerceNet and Nielsen Media Research, a large majority of Web
users—73 percent—spent some portion of their time online searching for a specific product
or service information. Given the growing popularity of the Internet, it is inevitable for
public relations professionals to add the Internet to the list of communication channels.
As two-way communication turns the monologue into an interactive dialogue
through the USENET discussion forums, the Internet offers a great opportunity to
examine the public agenda which has become more interactive than ever before. Since
the Internet discussion groups have the potential to galvanize public opinion in a matter
of hours due to asynchronous and rapid communication, the dynamic attitudes of on-line

27
users in the discussion groups toward the corporation during the crisis time need to be
investigated.
Even with many scholars and public relations practitioners emphasizing the
importance of monitoring media coverage of and public reactions to the corporate crises,
no formal study has yet been done to examine how people react to a corporate crisis on
the USENET in contrast to the conventional news media coverage. After analysis of both
media coverage and public discussion messages on the USENET, it would be of value to
track how the interpretation of the crisis varies among the news media and on-line users
over time.
From theoretical considerations, since the flow of public opinion is such an
essential component of crisis communication management, the USENET on the Internet
will function as an opinion forum to analyze the public agenda for not only public
relations practitioners, but also the agenda-setting researchers. The contexts of the public
conversations on the Internet can be further examined to test agenda-setting theory.
The next chapter will review the literature of crisis communication management
and introduce a crisis public opinion model proposed by Sturges (1994) and MacKuen's
(1990) model of public dialogue. At the end of the third chapter, research questions and
hypotheses are presented.

CHAPTER III
CRISIS COMMUNICATION LITERATURE
Overview
In recent decades in American business, there has often been a tension between
corporate public relations professionals and media journalists. When things are going
well, public relations managers desperately want media coverage, but “good news” is
seldom a high priority for people assigning reporters and writing headlines. However,
during a business crisis, the news media appear far more inspired to investigate the issue.
In these “damage control” situations, public relations managers must be aware totally of
what is being printed and said in all media. Often this information is inaccurate or biased
and the corporation must somehow correct the misconception and rehabilitate the
corporation’s public reputation (Birch, 1994; Traverso, 1992).
As with treating a spreading disease, early intervention is essential to successful
treatment of a corporate crisis. Prevention of something detrimental to the organization
remains a vital part of corporate public relations work. As no organization is exempt
from a crisis, the art and science of crisis management is no longer an unusual specialty
but an imperative job for public relations practitioners (Maggart, 1994). Rather than
coping with a crisis after it has reached overwhelming proportions in the media,
intelligent public relations professionals must leam to detect potential crises at the
embryonic stage. Dealing with an emerging crisis before it reaches the mass public is the
key to success.
28

29
Media play an essential role in helping people understand how the corporation
reacts to the crisis. Media also provide public relations professionals with information as
to how the general public feels about the crisis event, so the corporation can adopt
appropriate management tactics to ameliorate the crisis situation.
Effective communication with key publics in the society is the purpose of public
relations practice in a business corporation, especially when it encounters an unexpected
crisis. Among all key publics, the media stand out as the most significant in the public
relations field, because they disseminate information and serve as interpreters of social
phenomena. However, the media do not operate in isolation. Public relations
professionals can often affect the behavior of the mass media.
Not only should an organization pay attention to media coverage, which
sometimes generates negative effects on the public agenda if no prompt action has been
implemented by the corporations, but public perception of and opinion about the crisis
must be evaluated. Because most people perceive truth to be whatever public opinion
expresses at the crisis time, the essential purpose of crisis communications is to affect the
atmosphere of public opinion, hoping that anything good about the organization will
persist and bad connotations will soon disappear. Frequently asked questions in a crisis
by a professional public relations person are: “Does the general public feel the same way
as the media journalists?” and in terms of the perception of a crisis, “To what degree are
people affected by media coverage of the crisis and the organization?” The assumption
behind these questions posits that the general public is often influenced by the media
coverage.

30
In recent years a number of new media technologies have made the public
relations' goal of effectively reaching the general public more challenging. New
technologies affect and change the way people communicate with each other. The most
conspicuous distinction between old and new media is the control that new media venues
offer users over the communication process. But this new control also raises new
challenges and concerns for public relations practitioners, for they must broaden the
boundaries of surveying media contents. The new challenges go beyond conventional
media such as newspapers, radio and television and include new technologies where
information receivers can also become information senders.
In view of the explosive development of communication technology, new
message delivery systems such as cable and satellites offer dozens if not hundreds of
programming options, each of which has the potential of carrying a message about a
corporate crisis. While these message channels are almost always managed by
established companies, a new media technology that does not require a large company
infrastructure is the Internet. Given dollars and technical know-how, anyone can set up
an Internet web page for people all over the world to read. Furthermore, these web pages
can be interactive, meaning that individual users of a website can respond immediately to
the message senders. Breaking the prevalent concept of source-oriented media, the
Internet’s greatest advantage is its interactivity between information providers and
receivers. An audience member may become an active message producer, not just a
passive receiver.
From the crisis management perspective, the Internet also offers an excellent
opportunity for public relations professionals to gauge the impact of the crisis on the

31
Corporation covered by the media and to monitor the public’s reaction to it. The
exchange between the media, public relations professionals, and the public is imperative
for the modem corporation, especially in a crisis situation. By understanding these
interactions, the Internet can function as a more useful tool for public relations
practitioners.
When a corporate crisis occurs, the news media are tempted often to rely on
information not only from conventional journalistic sources such as wire services but on
information derived from the Internet websites (Birch, 1994). As a result, a company
faces a new source of pressure from both on-line users and the journalists who might seek
information related to the crisis through the Internet. As the Internet becomes another
important area in the public relations field, crisis managers must take a much broader
view of it as a potential problematic source.
Since monitoring of traditional media coverage about the corporate crisis has
always been a fundamental part of crisis management in the public relations field, the
advent of the Internet sets another path to understand the general public's opinion about
the crisis.
Crisis Management Research and Public Opinion
Nothing is more challenging and unnerving to public relations professionals than
the unpredictable crisis. A crises can cause a temporary disruption of activity with no
long-lasting consequences to the bottom line or it can permanently damage a
corporation’s reputation, resulting in reduced profitability. Understanding properly how
a crisis is contained and managed may in extreme cases determine the very survival of a
business organization.

32
Definition of Crisis
What is a crisis? No universally accepted definition has been adopted, but in
general, a crisis is a situation that threatens the normal activity of an organization. A
crisis can proceed from a mere disruption of activity to damaging the corporate reputation
and reducing profitability. Hayes (1985) believes that a crisis results from a major
incongruence between the expectations of a coiporation and what happens in the
environment. Weick (1988) thinks that “crises are characterized by low probability/high
consequence events that threaten the most fundamental goals of an organization” (p.
305). Dutton (1986) identifies three essential dimensions involved in a crisis in the study
of the decision-making processes between crisis and non-crisis issues in an organization:
importance, immediacy and uncertainty. Lerbinger (1997) treats a crisis as an event that
endangers an organization’s future profitability, growth, and even its survival. Feam-
Banks (1996) defines a crisis broadly as “... a major occurrence with a potentially
negative outcome affecting an organization, company, or industry, as well as its
publics...” (p. 1). The Institute for Crisis Management refers to crisis as a significant
business disruption, which results in extensive news media coverage and public scrutiny
(Irvine & Millar, 1996).
To sum up, a crisis can be something unexpected that occurs by surprise in any
type of coiporation (Woods, 1996); stems from the interaction of failures between the
corporation and the external environment (D'Aveni et ah, 1990); requires fast and
accurate reaction to neutralize high threat to corporate values (González-Herrero & Pratt,
1995); creates uncertainty (Mitchell, 1986); threatens the reputation and assets of the

33
organization (Barton, 1993); and causes publics, especially media, to scrutinize the
organization (Irvine & Millar, 1996).
Types of Crises
Although corporate crisis can occur suddenly without forewarning, it is wise to
identify the types of crises when they take place in order to adopt the proper strategic
plan. The Institute for Crisis Management puts the crisis into 16 categories and identifies
three origins of crises, management, employees, and other (Irvine & Millar, 1996). The
16 categories are business catastrophe, environmental damage, consumer action,
discrimination, financial damages, labor disputes, sexual harassment, white collar crime,
casualty accident, class action suits, defects/recalls, executive dismissal, hostile takeover,
mismanagement, whistle blowing, and workplace violence. Mitroff et al. (1996)
categorizes 11 types of crisis which include: criminal attacks; economic attacks; loss of
proprietary information; industrial disaster; natural disaster; equipment/plant
malfunction; legal problem; perceptual/reputational; human resources/occupational;
environmental/health; and regulatory. In her case study book, Kathleen Feam-Banks
(1996) lists the following five types of crisis: product tampering, environmental, natural
disasters, violence, and celebrities and crises. According to Lerbinger (1997), there are
seven crisis types: natural, technological, confrontation, malevolence, skewed
management values, deception, and management misconduct.
Environmental Monitoring in Public Relations
Business crises put corporate reputations and economic survival to the test.
Effective crisis communication planning not only results in good issue management but
in favorable public perception of the corporation. One of Hearit’s (1994) suggestions to

34
deal with crises was to diffuse hostility directed toward the corporation suffering from it.
But as a potential crisis is evolving, how can companies keep track of how they are being
portrayed and how can this information be intercepted before it reaches important
stakeholders? In the public relations field, environment monitoring involves a long-term
examination of media coverage before any issues related to the organization can
transform into a crisis that will detrimentally impact the business. Needless to say,
environment monitoring becomes an indispensable component of crisis management.
From a synthesis of the literature emphasizing the importance of environmental
monitoring in crisis management (Birch, 1994; Carney, 1993; Fombrun, 1996; Hayes,
1985; Unaansky, 1993; Wilcox et al., 1992), several reasons for monitoring media
coverage are identified:
1. The effectiveness of general issue management can be determined by analyzing the
tone of media coverage;
2. The information obtained from the media content also helps public relations people
examine public perception of an organization’s response to a crisis;
3. Monitoring and analyzing media coverage help public relations practitioners audit the
reputation of the organization;
4. Scrutinizing media coverage on issues provides public relations practitioners ideas
about how to position an organization in front of publics and how to modify the
strategies or tactics when handling a crisis.
5. Understanding media coverage is one part of formative research in the
implementation of public relations strategies.

35
Public Opinion in Public Relations
Public opinion research has always been a vital topic across a diversity of social
science fields such as political science, psychology, and mass communication. To
understand public attitudes, polls are often employed as an efficient way of measuring the
public agenda on major issues. Whether public opinion represents an aggregation of
individual views or a collective movement remains debatable (Price, 1992).
Public opinion, more often than not, can determine what governments and
corporations do when encountering the unexpected crisis. In a typical crisis situation, the
public perceives “the truth” to be whatever public opinion says that it is. Public relations
practitioners are responsible for surveying questions of public opinion. Therefore, it is
vital for an organization to prove to its stakeholders that the prevailing negative public
opinion is not accurate.
Public opinion can and must be shifted to the corporation’s advantage if an
organization desires to emerge from a crisis situation with a minimum of damage to its
reputation. It is difficult enough to deal with a situation in which the corporation is
indeed guilty, but even more frustrating is coping with bad information and false
accusations. Looking back to past crisis events under various circumstances, many
companies—such as Exxon, NASA, Denny's, and Intel—failed to communicate effectively
with the general public, including their loyal customers, when they faced a crisis or pre¬
crisis situation. Conventional press conferences or general press statements may not
subdue the anxieties of people who are directly influenced by a crisis. Corporations need
to go directly to the public to gauge how well public relations efforts are working; this is
where proactive tactics become essential.

36
Damage control is the key objective of crisis management. The more a public
relations practitioner knows about how a crisis is being interpreted by the media and the
general public, the more effective the damage control will be. The communication
challenge is to convince various audiences of the company's ability to alleviate the
damage, eliminate public doubts and position itself for future growth. The success of
crisis management is evaluated by how effective the company affects the public opinion
process.
Media and Crisis Management
The first lesson for a public relations practitioner to learn is that it is always better
to deal with a potential crisis before it blooms into a real crisis. Without tracing and
monitoring the shifts in public opinion, public relations programs become meaningless.
Public relations practitioners cannot afford to operate in a vacuum. Accurate information
concerning media messages and resultant public opinions is crucial for continued success.
Among the research devices used to accomplish this task are personal contacts, content
analysis of media coverage, field reports, and polling (Wilcox et ah, 1992). Media play a
crucial role in understanding public opinion trends on issues. Although the impact of
mass media on attitude formation and change has proven to be less influential than some
expected (Klapper, 1960), mass media still function as a primary resource for information
about many topics. Therefore, it becomes impossible to ignore what mass media cover or
say about an issue pertinent to an organization while simply probing public opinion.
Media impact on spreading a crisis is often time-dependent. It is better to control
an issue in the early stage than wait until it becomes a crisis management issue. As
recommended by Birch (1994), the use of tracking research, even daily in the early

37
stages, allows corporate management to know what the general public or key
stakeholders are really thinking and saying. Permanent monitoring of active publics,
getting continuous feedback through different channels from these publics, also facilitates
issue tracking (Umansky, 1993; Vendrell, 1993). In summary, when dealing with an
unexpected crisis, the two-way research, monitoring both the media and public agenda, is
pivotal.
During a corporate crisis, communicating with the general public is very
important. Crisis communication tends to evolve through three stages: mass media
publicity, public opinion arousal, and then public policy makers’ response (Mayer, 1991).
Public opinion stems from the media issue coverage and then impacts government policy
makers. At the initial phase of a crisis, an organization may have the power to influence
events. After the crisis has received large amounts of attention from both media and the
public, the ability of the company to control its public reputation diminishes. In her book
on crisis management, Feam-Banks (1996) terms “containment” as efforts exerted by the
organization to limit the duration of the crisis or keep it from spreading to other areas
affecting the organization.
An effective crisis communications plan must involve a sound media monitoring
program to ensure accuracy of media coverage. Traverso (1992) urged that a media
monitoring program is necessary when facing the crisis as media provide a fast channel to
deliver information to the public. Situated in such a dynamic information era, Strenski
(1995) pointed out that a key area in which public relations professionals must excel in
crisis control is to take responsibility for monitoring the on-line services such as BBSs
and all kinds of newsgroups that exist in cyberspace. He warned that ignoring these

38
groups is missing a major portion of the potential audience. Kotcher (1992) offered
reasoning why new communication media must not be neglected by public relations
practitioners. He states,
Communicators can no longer operate under the old assumption
that mass communication is passive in nature and that interactive,
personal communication can only really occur in small groups or
face-to-face. In addition, communicators should not assume their
present understanding of technology and its applications would be
sufficient at a time of crisis (p. 20).
Crises can start with rumors, which can be absolutely false or partly false without
substantial verification and credible sources. In her ten guidelines for reducing legal risks
in crisis management, Fitzpatrick (1995) stressed the urgent need to respond to rumors
spreading all over the variety of information channels that may turn out to cause a long¬
term, sustained crisis for a company.
Sturges’ Public Opinion Model of Crisis Management
Communication is an important tool in handling the impression of key
stakeholders at the time of crisis (D’Aveni et al., 1990). Sturges (1994) emphasizes the
value of communication during the crisis to the publics. The crisis does not happen in
isolation so the communication plan should consider how public opinion, interacting with
other social events and environment, is formed before, during and after the crisis. He
illustrates the evolution of group opinion through an eight-step process as shown in
Figure 3-1.

39
8. Social Norm
í
7. Social Action
í
6. Public’s
Opinion Forms
1. Latent Issue
2. Event Occurs
\
4. Debate
Occurs
FIGURE 3-1. Group opinion formation process
In the first stage, an issue must be salient but latent for the public members. As
public awareness is the necessary condition in the formulation of public opinion, an event
occurs and comes into people’s mind in the next step after they become aware of that
event. Then the pro and con factions among the members solidify in the third stage,
followed by public debate among the members (step 4). After a time lapse occurs (step
5), when enough has been said and done, public opinion thus emerges in step 6. In step
7, some action will be used to reinforce the public opinion formed in the previous step.
Finally, due to the performance of action, a social norm is going to be established.

40
Sturges applies this concept of opinion formation to the direction of public
opinion in the crisis situation as illustrated in Figure 3-2. The public opinion, either
negative or positive, expands as the crisis erupts and fades away while the crisis abates
and ends. Therefore, the objective of crisis communication is “to influence the public
opinion development to the point that opinions held in the post-crisis period are at the
same level or greater in positive opinions or lower in negative opinion” (p. 303) among
the members of the public.
— —Negative Opinion
"'** ^^^“Positive Opinion
Termination
FIGURE 3-2. Sturges' model of public opinion with crisis management plan
MacKuen’s Strategy Model of Public Dialogue
Regardless of the form (face-to-face, electronic bulletin boards, newsgroups, etc.)
one chooses to take part in the political conversation, MacKuen’s (1990) discussion about
personal strategies in the public dialogue to promote deliberative democracy is employed
to examine how on-line users get involved in the discussion groups. The basic idea
behind MacKuen’s strategy model for public dialogue is that deliberative democracy is a
function of political participants who interact with one another. These interactions set the
political and intellectual contexts for democratic decisions and may take two forms. One
✓ >.
/ \
^ \
Pre-crisis Build Up Break Out Abatement

41
occurs between those who hold similar views and another stems from individuals of
different views. However, only the latter allows genuine debate and an exchange of
ideas. As a result, the conversation which each individual creates becomes crucial in
shaping public opinion and then develops into deliberative democracy.
Regarding engagement in political discourse, MacKuen holds that one’s
willingness to engage in public discourse depends “not only on how much one enjoys
discussion or hates controversy, but also on the likelihood of running into sympathetic
and contrary viewpoints” (p. 61). Individuals can actively choose when and where to
participate without passively getting involved. Four strategies will affect the outcome of
interactive political conversations: TALK, CLAM, REACT, and SIGNAL. Either social
consensus or disagreement comes from different strategies that people employ when
encountering the political discussion. Similarly, the outward environment would play a
role in determining what strategy individuals may use while becoming involved in the
political dialogue.
MacKuen uses Expressivity (p. 64), the ratio of pleasure from rewarding
conversation to pain of disagreements, to depict how an individual feels about the
conversation and his or her own tolerance of opposing viewpoints. It also represents a
positive incentive to engage in the political conversation. Low levels of Expressivity
require friendly environments to make conversation worthwhile. By contrast, high levels
of Expressivity render political conversation attractive even when facing opposing
majorities.

42
Applications of Four Types of Strategies
The outcome of use of TALK/CLAM strategies in the political dialogue is shown
in Table 3-1 and 3-2.
TABLE 3-1.
Payoff matrix for player encountering friend
Friend’s Strategy
TALK
CLAM
Player’s Strategy
TALK
Reinforcement
Bluster
CLAM
Music
Pass
TABLE 3-2.
Payoff matrix for player encountering friend
Friend’s Strategy
TALK
REACT
CLAM
TALK
Reinforcement
Reinforcement
Bluster
Player’s
REACT
Reinforcement
Pass
Pass
Strategy
CLAM
Music
Pass
Pass

43
As seen in Table 3-1, the conversation that reinforces democratic dialogue
between two parties who reward self-expression, willingness to speak out, happens only
if both individuals choose TALK strategy. By the same token, if both individuals choose
CLAM strategy, a nonpolitical conversation (Pass) will follow. The theme in this matrix
of employment of both strategies is that all outcomes but Reinforcement and
Disagreement result in little or temporary value through the interactive conversation.
Other than the aforementioned two strategies, an individual might opt for the third
one— REACT.
As shown in Table 3-2, the unique advantage of this strategy is to avoid
disagreements when both player and friend adopt either TALK or REACT strategy in the
conversation. However, it also limits the attractiveness of the TALK option if an
opponent is involved. By using this strategy, a person can cheerfully respond to friendly
TALKERS and remaining silent before opposing viewpoints. The REACTOR never
initiates a political conversation, but simply follows the partners, with whom she or he
agrees. As the number of TALKing friends increases, the attractiveness of REACT will
also increase likewise.
The last strategy MacKuen mentions is called SIGNAL. The consequence of
signaling for individual choice produces a community of isolated groups and eliminates
meaningful public dialogue. So to speak, signaling not only prevents a consideration of
alternative views, but also generates a consistency of opinions around the symbols
embodied in the signal itself (p. 82).
Because the level of Expressivity in the public enviromnent determines an
individual’s willingness to engage in the political dialogue, MacKuen’s integrated

44
TALK/CLAM/REACT/SIGNAL models provide an insightful understanding about how
a person would contribute to the meaningful dialogues in any public discussion forum.
For instance, employment of the TALK/CLAM strategies is most likely to generate a
democratic dialogue; on the other hand, use of REACT/SIGNAL strategies will help to
frame a friendly interactive discussion environment without disagreement.
Research Questions
Studies on crisis or issue management have created stages or life cycles from the
emergence to the termination of a crisis (Feam-Banks, 1996; Gaunt et al., 1995;
González-Herrero et ah, 1995; Hainsworth, 1990; Meng, 1992). With the opinions
identified as positive or negative or combination of both, people tend to react positively
to good news and negatively to bad news (Sigelman et ah, 1993). Based upon Sturges’
(1994) public opinion formation model during crisis, negative public opinion about the
corporation at the breakout of the crisis reaches its highest point and then ebbs to the pre¬
crisis stage after the crisis.
As Sturges demonstrates earlier in Figure 3-2, it is anticipated that the highest
number of negative messages about an organization will be posted as the crisis erupts,
following the initial media coverage.
1. Flow will public opinion on an interactive communication channel, such as the
Internet, differ from Sturges’ model on public opinion during a crisis?
As commercial providers increase on the Internet and more political information
is provided, the problem of who sets the agenda for the new medium also becomes a
concern. Behr et ah (1985) maintained that public concerns about energy and inflation
were determined by television coverage, which in turn was dependent upon world

45
conditions. By comparing the contents of three local and regional daily newspapers with
the results surveying three waves of respondents drawn from the same study population
across eight months, Salwen (1988) detected a strong influence of media coverage on
environmental issues on the public agenda in the first and second wave. Despite time-
varying effects for agenda-setting, the direction of influence has traditionally been from
the media to the public.
2. Will this unidirectional influence from the media to the public hold when the
interactive public agenda is examined?
Famous public relations crisis cases such as the Intel Pentium flaw, the Pepsi
syringes tampering, and the Nike discrimination against Islam suggested the importance
of a quick and proper reaction to the potential crisis issue, whether it is only a rumor or
not. Communication with the target publics during the crisis becomes a top priority task
for public relations professionals; hence, preparation of complete crisis management
planning should be highly stressed (Barton, 1993; Mitroff, 1996).
3. To what degree is the frequency of on-line dialogues on the USENET affected by
an organization’s strategic crisis intervention?
MacKuen (1990) demonstrates that people who feel high levels of Expressivity
will more likely adopt a TALK strategy than those who do not. Although CLAM is one
of the strategies used by individuals to engage in the public dialogue, it is impossible to
detect who remains silent by adopting this strategy in the newsgroups discussion. CLAM
users are regarded as those who read the messages without posting their own opinions on
the newsgroups.

46
4. How will the frequency and tone of on-line dialogues on the USENET posted by
TALKERS and REACTORS vary in the light of MacKuen’s strategy model?
Research Hypotheses
Given that the Internet provides public relations professionals access to listening
to what people really think and say about the crisis or the targeted corporation on the
USENET, hypotheses in this study stress the change of number of on-line discussion
messages before, during and after the corporate crisis. Analysis of these messages will
also help advance the measurement of the public agenda in agenda-setting theory.
Sturges (1994) divides the time span during the crisis into several stages, that is,
pre-crisis, build-up, break out, abatement, and termination. He asserts that both negative
and positive opinions about the organization from the public will increase when the crisis
breaks out. To compare the difference in the number of on-line messages before and
after the crisis erupts, ten days will be used as a time frame to test the following
hypothesis:
Hypothesis 1: The average number of daily USENET dialogues regarding the
corporation will be greater during the ten days following the eruption of the crisis than
the average number of daily on-line dialogues within ten days before the crisis.
Also illustrated in Sturges' model is that public negative opinion will
comparatively increase more than public positive opinion since the crisis can have
negative impact on the society as well as the general public. Therefore, by using all the
sampling days after the crisis eruption as a time frame, a second hypothesis is formed.

47
Hypothesis 2: The average number of daily negative USENET dialogues will be
greater than the average number of daily positive USENET dialogues after the eruption
of the crisis.
Numerous agenda-setting studies lead to a common conclusion that news media
coverage of an issue has considerable impact on the public agenda. Concerning the first
level of agenda-setting theory, the more media coverage of an issue, the more salient that
issue will become in people's minds. The next hypothesis is hence made based upon the
agenda-setting concept as displayed in Figure 3-3.
FIGURE 3-3. Causal relationship between the media agenda and the public agenda
Hypothesis 3: The daily number of the USENET dialogues will be related to the
daily number of news stories about a crisis after its eruption.
In Sturges' model, to move to the abatement stage where the amount of public
negative opinion sharply decreases from the breakout stage, the corporation must employ
proper strategies to contain the crisis spreading. To test if change in the amount of public
negative on-line dialogues does occur after the company adopts the intervention
strategies, the following hypothesis is formed.
Hypothesis 4: The average number of daily negative USENET dialogues
following a crisis intervention strategy will be less than the negative USENET dialogues
prior to the intervention.
Public Opinion
on the USENET
Media \ ¡—y
Reports

48
Derived from the above hypothetical assumption, it is expected that the
intervention strategies will have positive influences on people's opinion about the
corporation. In Stages' model, the amount of public positive opinion about the company
remains similar at both pre-crisis and abatement stages, so the next hypothesis will be as
follows:
Hypothesis 5: Following a crisis intervention strategy, the average number of
daily positive on-line dialogues on the USENET concerning the company will be equal to
the average number of daily positive on-line dialogues before the crisis eruption.
After the intervention strategies, amount of both negative and positive opinion
from the public decreases in comparison with those when the crisis erupts according to
Stages' model. Thus, the hypothesis is developed as follows:
Hypothesis 6: The number of the USENET dialogues following a crisis
intervention will be less than the number of the USENET dialogues before the
intervention and after the crisis.
No study has applied MacKuen's strategy model of public dialogue to the on-line
users who express their opinion on the USENET. As TALKERS are inclined to express
their own opinion to initiate the public dialogue or disagree with other people, the next
hypothesis is suggested to test how both TALKERS and REACTORS voice their opinion
about the company after a crisis happens.
Hypothesis 7: The number of daily negative USENET dialogues posted by
TALKERS will be greater than that by REACTORS.
The concept and findings of the second-level agenda-setting research indicate that
content of media coverage of the corporate crisis can greatly affect the public opinion.

49
By using computer software to analyze the tone of media coverage, the hypothesis is
formed to test the relationship between the tone of the media agenda and the public on¬
line dialogues.
Hypothesis 8: The Optimism Score of media messages will be positively related
to the number of negatively toned USENET messages posted after the media stories were
released.
The next chapter will describe in detail the method used to test the hypotheses and
to answer the research questions.

CHAPTER IV
METHODOLOGY
Overview
To test the agenda-setting theory, most researchers have traditionally used content
analysis to examine the media agenda and public opinion surveys to examine the public
agenda. All studies investigating the media content have focused on the conventional
media such as television, magazines, newspapers, and radio. However, the Internet has
quickly evolved into another important communication medium not only as a corporate
communications tool but as a forum for public opinion. People have become more active
and willing to get on the Net to express their attitudes toward or opinion about particular
issues or persons. The Internet hence provides public relations professionals another
opportunity to understand the trend of public opinion about their own organization at
times of crisis. To better explore public opinion on the USENET, this study will employ
content analysis to analyze the on-line public dialogues as well as the media coverage of
the selected corporate crises.
This chapter will first briefly state the definition and functions of content analysis,
followed by detailed descriptions of how to select crisis cases, news media, and public
dialogues on the USENET. Following the listing of coding categories, both conceptual
and operational definitions of key terms stated in the hypotheses are provided. Then, all
independent and dependent variables are identified with an intercoder reliability check in
50

51
the next section. The final section introduces of a computer software called DICTION for
the purpose of testing the last hypothesis.
Description of Content Analysis
The most popular and frequently cited definition of content analysis is offered by
Berelson (1952), “Content analysis is a research technique for the objective, systematic,
and quantitative description of the manifest content of communication” (p. 18).
Krippendorff (1980) also defines content analysis as a research technique for making
replicable and valid inferences from data to their context. The inferences involve the
sender, receiver of the message, the message itself, as well as the communication
channels. This unobtrusive quantitative measure of texts serves many purposes such as
identification of intentions of the communicators, understanding of attitudinal and
behavioral responses to communications, and description of trends in communication
content (Webber, 1990).
Corporate Crisis Selection
This study is intended to predict the trend of public dialogues on the USENET
about corporate crises. The impact of news media coverage and the company’s
intervention strategy will be integrated into the model. Two examples of the same type
of crisis will be randomly selected. The first crisis will be analyzed and used to modify
the coding categories and to improve the ways of retrieving a sufficient number of media
stories and on-line discussion messages.
For the purposes of generalizability of findings and to understand the influences
of the crisis in the on-line USENET dialogues, product tampering was selected as the
target crisis. Feam-Banks (1996) considers product tampering crises as a result of claims

52
e
made against manufacturers’ products. After analyzing news stories about 16 types of
business crisis, Irvine and Millar (1996) found that crises of defects & recalls of products
has increased 41 percent between 1994 and 1995. Because product tampering has the
public safety and health at high risk, Lerbinger (1997) classifies it as a crisis of
malevolence, which often necessitates product recalls. Past examples include Johnson &
Johnson’s Tylenol murders in 1982, Gerber’s product tampering in 1986, Perrier’s
recalling of water bottles in 1990, and Pepsi’s syringe threat in 1993.
The time period from which the two crisis cases were selected was from March,
1995 to December, 1997. The crisis event was chosen for study because it met the
conditions that are pre-defined in the latter part of this chapter.
News Coverage Analysis
A corporate crisis has emerged when the news media report an issue in a negative
or adversary tone. Content analysis of crisis coverage by the media has usually been used
to guide an organization’s reaction or responses to handling the crisis and to measure how
the general public perceives the organization. Continuing coverage by the mass media
may serve to help comprehend the agenda-setting effects on the crisis and delineate a
crisis development model. Research on testing agenda-setting hypotheses has
traditionally focused on impact over time, which required at least two weeks to six
months to discover the power of media setting the public agenda (Eaton, 1989;
Funkhouser, 1973; Mullins, 1977; Salwen, 1988; Shoemaker et al., 1989; Sohn, 1978;
Stone & McCombs, 1981; Wanta & Hu, 1994; Winter et al., 1981).
With the considerable number of public dialogues on the USENET, the time
frame for analysis of media coverage is six weeks, 10 days before occurrence of a crisis

53
and 32 days following the crisis eruption. Although on-line discussion may or may not
end one month after occurrence of the crisis, the random error becomes larger as more
days of messages are retrieved when applying statistical analysis to test the effects of
news media coverage in relation to the number of on-line dialogues.
The on-line Lexis/Nexis data base served as a foundation of media news
coverage. The keywords for searching news articles in the Lexis/Nexis data base varied
depending on each crisis.
Media Selection for Analysis of News Stories
Television provides political information to most citizens in Western industrial
societies and has become the most credible source of information (Brosius & Kepplinger,
1990). Zucker (1978) maintained that at the national level the public may be more
influenced by the three networks’ newscasts than by newspapers because of television’s
accessibility. Shaw and McCombs (1977) argued that television news might have a
stronger short-term impact, however, newspaper content may have a more consistent
effect across longer periods of time.
The Lexis/Nexis data base was selected for capturing news media coverage. The
Lexis/Nexis contains both newspaper and television news transcripts that have national
circulation and audiences. The media chosen as a part of sampling frame included three
television news networks, and three elite nationally known newspapers. They were: the
CNN news, the ABC news, the NBC news, the New York Times, the Wall Street Journal,
the Washington Post. Because the CBS news transcripts are currently not in the
Lexis/Nexis data base, the CNN news, the ABC news, and the NBC news were selected
for the analysis of television network news analysis. The New York Times was chosen for

54
the study because, besides having a national circulation, it is thought to have an agenda¬
setting effect on smaller media that could be seen in survey data of a cross section of the
population. The New York Times was shown to set the agenda for the television networks
when the cocaine issue coverage was examined (Reese & Danielian, 1989). The Wall
Street Journal, a business-oriented national as well as international newspaper, devotes
considerable coverage to news about major corporations. Because news magazines were
not found to contribute as strong agenda-setting effects as newspapers and television
(Shoemaker, 1989), this type of media was not included in the analysis.
On-line Discussion Message Analysis
In addition to crisis coverage by the six news sources, on-line dialogues will also
be content analyzed. There are three advantages in using the unmodified USENET
discussion groups to study political behavior in cyberspace (Groper, 1996): these
newsgroups are simple to access, posts to them are free of charge, and messages in the
USENET groups can be downloaded instantly from the Internet without subscribing to
the groups. In contrast, both Listservs and BBSs make it difficult to download data
without subscribing to the group itself.
The time frame for downloading the messages from on-line newsgroups was the
same as sampling media news coverage, that is, 10 days before and 32 days after the
crisis eruption. A search engine called Deja News1 on the Internet (See its homepage in
1 The Deja News Inc., founded in May, 1995, is the first company to organize and
provide easy access to discussion groups. As of now, the company's
(http://www.dejanews.com) archives are extended back to March, 1995 and has about
138 million articles appearing in over 20,000 newsgroups, accounting for more than 180
gigabytes of disk space. Similar to the use of Lexis/Nexis data base, input of keywords

55
Appendix A.) was adopted to facilitate the message downloading. To have enough but
not too many redundant on-line discussion messages about the targeted corporation, the
keywords used the Deja News search engine were slightly different from those used in
the Lexis/Nexis data base.
Unit of Analysis
The unit of analysis is the single message by one reader regardless of its length
within each thread pooled from the Deja News search engine. Every sampled message
was coded for the manifest content with the assistance of pre-defined categories.
Although a thread has to include at least two response messages besides the original
subject message, based upon the concept of “responsive interactivity” proposed by
Rafaeli (1988), this study still retrieved and analyzed any single message under a thread
as long as it was related to the crisis event. In addition, the same message that was posted
to different newsgroups or on different days by the same on-line user was singled out and
counted once only.
Threads across newsgroups from the Deja News data base were retrieved for
analysis. Not all news stories were retrieved. For instance, a newsgroup called
"biz.clarinet.webnews" often posts the news stories directly from other media sources, so
this type of message was ignored. Nevertheless, if the news stories were cited or posted
by a self-identified on-line user who did not express any personal opinion, those
messages would be sampled. Some messages containing several segments were read and
treated as one single message.
will automatically display messages related to the topic interesting to the users. The
articles on the USENET can go back as far as 1979.

56
Steps to Download On-line Messages
Because on-line messages of 42-day duration would be downloaded for analysis,
the efficient way is to use the "power search" method in the Deja News search engine
(See Appendix B), which enables a user to specify a time frame without worrying about
having too many messages beyond the particular time span of 42 days. The following
describes how to search for the messages in the Deja News search engine step by step:
1. Type in the key words in the "Search for:" blank row;
2. Use “threaded” as the results format and sorted by “Date.”;
3. Leave "Group(s)", "Author(s)", and "Subject(s)" blank;
4. Specify the beginning and ending date in "Date from:" and "To:" blank rows;
5. Click on "Find" button.
After the search results showed on the screen, every message was downloaded
for further analysis. While downloading all pertinent messages based on above-
mentioned criteria, the researcher has also clicked on “view thread” within the listed
message to see exactly how many messages had been posted under that threaded topic.
Then, every single message under that threaded topic was examined to ensure that all
messages relevant to the crisis had been downloaded.
The reason for examining each single message under one thread is because the
content of on-line discussion is often diverted to other topics that can be totally irrelevant
to the crisis event or the involved company. Without careful inspection of each message,
the number of on-line messages can be erroneously expanded, thus threatening the
content validity of the study. Another rationale is that some on-line users may mention or

57
discuss the crisis as responding to other users' messages without mentioning any
keywords that are used to search for crisis-related messages.
Definitions of Coding Categories
The categories of discussion messages needed to be defined in order to compare
the progress of media coverage on the corporate crisis. Stempel III (1989) suggested that
it is efficient to use an established category system that had been found functional and
manageable by other researchers. Although a large number of articles emphasize how
crucial evaluation of public reaction to the crisis is, specific definitions of positive or
negative public reaction are unavailable. Since no preset categories related to this study
were available for the on-line message analysis, the researcher synthesized various
articles dealing with either the public reaction to a crisis or corporate planning for crisis
cases (Birch, 1994; Carney et al., 1993; Feam-Banks, 1996; Fitzpatrick, 1995; González-
Herrero etal., 1995; Hearit, 1994; Lerbinger, 1997; Mitroff, 1996; Stanton, 1995;
Thomsen, 1996; Traverso, 1992; Wilcox et al., 1992; Woods, 1996).
As noted in the second chapter, it is not unusual to receive so-called “flaming”
messages from the newsgroups discussions (Buey, 1995); therefore, the flaming
messages were also coded. Coding categories were tested and re-tested for reliability.
When training the coders for reliability, the researcher modified the coding items, if
necessary, to make them mutually exclusive and exhaustive and create a common frame
of reference between coders and the researcher. When a negative or positive message
involved more than one item listed below, it was coded as one single number for that
category. For example, if one negative message mentioned the adoption of legal action

58
and boycott of the product or service, this message would be coded as “1”, which
indicates a negative message. This rule was also applied to positive messages.
The coding schemes for news stories and on-line messages are shown in
Appendix C and Appendix D.
1. Negative posted messages will be coded as those that
a) request legal action against the company,
b) request compensation from the company,
c) mention of personal or others’ experiences suffering from the crisis
event,
d) urge the boycott of the company’s product or service,
e) give sarcastic jokes or harsh comments about the company, or
f) blame or accuse the company for its wrongdoing or its defective
product(s).
2. Positive posted messages will be coded as those that
a) urge end of a boycott,
b) compliment the company,
c) mention the laws as unfair to the organization,
d) praise company’s intervening actions, or
e) deny that the crisis event is solely due to the company.
3. Neutral posted messages will be coded as those including
a) forwarded media reports without expressing opinion,
b) forwarded website information without expressing opinion,

59
c) personal questions or responses to others’ messages without expressing any opinions,
or
d) statements that take the crisis issue for granted without blaming or praising the
company.
Conceptual Definitions
In the following two sections, some special terms stated in the hypotheses are
defined both conceptually and operationally along with concise reviews of previous
studies germane to the terms.
On-line Dialogues on the USENET
The idea of dialogue has been extended to apply to the messages on the
computer-mediated communication channels, such as the USENET, BBS, and e-mail
system. As far as messages on the USENET are concerned, MacKinnon (1992) points
out that these newsgroup postings have the quality of spontaneity and are uncensored,
functioning more like true dialogues. Unlike traditional conversation in the real world,
the feelings expressed in the postings are not mirrored either physically or symbolically,
and the dialogue is “absent from physical proximity, face-to-face interaction, and
nonverbal cues” (p. 114). Morris and Ogan (1996) regard the USENET as a many-to-
many asynchronous communication. So to speak, two or more on-line users
communicating with each other do not have to be on-line at the same time. Each could
leave messages to others and retrieve replies at his or her convenience (Groper, 1996).
To contrast with dialogic communication mainly applied to interpersonal communication,
Thomsen (1996) cited Ball-Rokeach and Reardon’s concept of “telelogic
communication” for the analysis of bulletin board systems. According to Ball-Rokeach

60
and Reardon (1988), telelogic communication “involves alternating dialogue between
people at a distance who use both conventional and unconventional language and
electronic or optical channels” (p. 135). This type of communication is often
geographically and temporally unbounded. Buey (1995) considers mediated exchanges
of messages as a form of mass communication in which the sender broadcasts a message
to a mass audience in the process of responding to an individual.
All messages on the USENET have a subject heading and those which carry the
same subject line are called a “thread.” Hill and Hughes (1997) define thread as
responses to the original message, or to something said in someone else’s reply. Once a
thread has been started though, it is no longer under the sender’s control.
For this study, on-line dialogues on the USENET are defined as—
Multiple messages about the same subject matter, with a single sender to a
group not in the same physical space which was formed because of its
interest in the subject matter, who can respond immediately to the message
and who can direct that response to the entire group.
A message is considered to be the total written materials contained in the text.
This includes all words, sentences, phrases and paragraphs. The subject matter in the
multiple messages refers to the discussion topic of one single thread. A thread is defined
as messages responding to the original message or to other messages under the same
discussion topic.
Crisis
The media coverage of a corporate crisis provides people most information about
the crisis event itself and, thus, shapes the public's perception of the organization.
Aiming to examine the relationship between the media coverage of the crisis and on-line

61
public opinion about it, a crisis in this study is defined as similar to that by Irvine and
Millar (1996)—
A significant business disruption, which negatively impacts its
stakeholders, such as employees, stockholders, customers, or community
members and results in extensive news media coverage and public
scrutiny.
Crisis intervention
Crisis communication results from a need to offset potential negative
consequences of not communicating (Sturges, 1994). Hearit (1994) suggests that to
intervene in the crisis is to diffuse hostility directed toward the corporation suffering from
it. Resembling the idea of crisis intervention, Feam-Banks (1996) uses “containment” to
refer to “the effort to limit the duration of the crisis or keep it from spreading to other
areas affecting the corporation” (p. 7). Lerbinger (1997) regards crisis intervention as an
action that helps remove ambiguity during a crisis and improves the understanding of the
public and the media.
For this study, crisis intervention is defined as—
When the organization acts strategically to reduce the negative impact of
the crisis on stakeholders after the occurrence of the crisis.
Operational Definitions
The operational definitions of key terms are listed as follows:
• On-line Dialogues on the USENET: “number of messages from different individuals
within one subject thread.” Here the thread is defined as messages responding to the
original message or to other messages under the same discussion topic or subject.

62
• Crisis: "the events that
• negatively impact the target subjects, e.g.: the general public, customers, and
employees,
• are covered by both a national elite newspaper and a national television
network,
• continue to be reported by major national news media for at least a one-week
duration and covered by some news sources for a total of seven days during
the month, and
• are discussed by an Internet newsgroup periodically for at least two weeks
since the initial news media coverage."
The crisis is considered to have erupted on the first day that any of the selected
news media reported the crisis event.
• Corporate Crisis Interventions: "the actions that the corporation adopts after the
media coverage of the negative event and which are intended to reduce the damage
caused by the media coverage."
The following would be coded as crisis interventions:
1. evacuation of employees or particular publics involved in the crisis for the sake of
safety,
2. hosting a news conference to admit or disclose happening of the crisis,
3. setting up hot lines to answer requests from media and publics,
4. provision of medical services to injured or dead people,
5. preparation for negotiation with interest groups or specific publics concerning the
crisis,

63
6. product recalls or destruction,
7. decision to enforce laws to combat the crisis,
8. offer of monetary or product compensation,
9. making a public apology,
10. acknowledgement of wrongdoing,
11. replacement of persons who caused the crisis, or
12. promise to prevent a similar crisis in the future.
• TALKER: "the on-line users who either respond to without expressing agreement, or
express different opinions, or initiate the subject message within one thread."
° REACTOR: "the on-line users who express their agreement to the initial message or
merely reply to other users without voicing their opinions about others' statements."
Independent and Dependent Variables
Four independent variables and two dependent variables are included in this
study. The number of news stories covered by the media about the crisis event is the first
independent variable. The second independent variable is considered the eruption of a
crisis. The contents of selected media will be analyzed to determine when the crisis
occurs. The first day that any one of the media reports the crisis event is going to be the
day of the crisis eruption. As the primary purpose of the crisis intervention is to alleviate
the damage done by that crisis to the corporation, how a corporation manages the crisis
intervention becomes indispensable in changing or reducing negative public opinion
about the company. The third independent variable is the direct action(s) taken by the
corporation during the crisis period. The last independent variable is the number of on¬
line messages posted by both TALKERS and REACTORS.

64
Since the number of negative as well as positive on-line dialogues on the
USENET is the main item examined in the hypotheses, which presumably varies with the
other variables, the number of negative and positive on-line dialogues on the USENET
become the two dependent variables.
Reliability Test
In regard to content analysis, reliability is the fundamental requirement for
yielding similar results with the same instrument on a given data sample. Intercoder
reliability represents the consistency of shared understandings or meanings held by
coders. Weber (1990) urged that this type of reliability be a minimum standard for
content analysis.
To test the intercoder reliability, 20 messages randomly selected from the first
case were given to two coders, who are native English speakers. They were instructed in
detail on how to code a nominal variable—tone of message. New items suggested by the
coders were added to the coding scheme.
Although there is no agreement on a minimum intercoder reliability coefficient,
most content analysts suggest that the researcher should look at whether or not the extent
of agreement exceeds chance (Stempel III, 1989). Krippendorff (1980) found that an
agreement correlation of less than 0.7 tended to be statistically insignificant and that 90
percent of agreement is not plausible in most cases. For this study, the pre-test intercoder
reliability coefficient was expected to reach at least 80 percent before the researcher
started to code messages in the second case.
According to Scott's (1955) formula, reliability reached 0.76; that is, both coders
had consistent agreements on 16 out of 20 messages in the same category. Because they

65
added some reasons why they coded the negative tone of messages, two more items were
added to the negative tone category. As a result, both coders were given the modified
coding scheme to re-code another 20 messages from the second case. The reliability this
time improved to 0.82.
Introduction of Computer Software: DICTION
In order to test the eighth hypothesis that the Optimism Score of media messages
will be positively related to the number of negatively toned USENET messages posted
after the media stories were released, a computer software, DICTION, designed by Hart
(1997) for textual analysis was employed to calculate the number of negative words
appearing in each news story.
After selected articles are analyzed, this textual analysis program will calculate
scores of five “Master Variables”: Certainty, Optimism, Activity, Realism, and
Commonality. Every "Master Variable" score has its own formula. Take "Optimism" as
an example. The definition of the Optimism Score is "language endorsing or highlighting
the positive entailments of some person, group, concept, or event" (Hart, 1997, p. 46).
The formula for this variable is (Praise + Satisfaction + Inspiration) - (Blame + Hardship
+ Denial). Terms of Praise, Satisfaction, Inspiration, Blame, Hardship, and Denial are
consisted of different words stored in the dictionaries in the software.
To see to what degree a news story covered the crisis event in a negative tone, the
Optimism Score was used. However, to simplify the formula, only numbers of words
representing above six terms were put into the formula, instead of the standardized score
displayed in the report file (See Appendix E). Once the Optimism Score for each news

66
story was obtained, it was compared with the number of daily negative USENET
messages to test hypothesis 8.
The next chapter will first present descriptive analysis. Then, description of each
crisis case along with results of downloading media stories and USENET messages is
provided, followed by the discussion of outcomes of hypotheses testing. Post-hoc
analysis will end the chapter.

CHAPTER V
RESULTS
Two corporate crises were selected for testing the hypotheses. The first case is
"Mattel Cabbage Patch Doll Defect," which occurred at the end of 1996. The second
crisis case, "Hudson Foods Meat Contamination," happening in the mid-1997 was
analyzed on the basis of modified coding categories derived from the Mattel case.
This chapter begins with descriptive analysis. Then, the results of downloading
media stories and on-line discussion messages and test of hypotheses are presented for
each case. Post-hoc analysis concludes this chapter.
Descriptive Analysis
Media Stories
The media stories covering both crises were retrieved solely from the Lexis/Nexis
data base. Except for a few news stories with the same content, all the news articles
appearing in the search results had been gathered and served as a sample of the media
agenda. Consequently, 46 media stories covering the Mattel case and 172 stories about
the Hudson Foods company were analyzed.
On-line Messages on the USENET
The Deja News search engine functions as the only source for downloading the
on-line messages concerning each crisis event. With different keywords used in each
case, 488 messages of the Mattel case and 250 of the Hudson Foods case appeared on the
67

68
results pages. However, because not all the messages displaying on the results pages
were related to the crisis, many messages were eliminated. Finally, 139 and 154 on-line
messages from the Mattel and the Hudson Foods cases were sampled through detailed
examination. They were used as the basis for the public agenda to test the hypotheses.
Tone of On-line Messages
Tone of on-line messages was coded as either positive, negative, neutral or
unidentified by coders. In the Mattel case, 59 negative (42%) and 5 positive messages
(4%) were posted discussing the crisis. In the Hudson Foods case, 66 messages (43%)
were considered negative and 11 messages positive (7%). The remaining messages in
both cases were either neutral or unidentified. Interestingly, percentages of negative and
positive messages in both cases were very close.
Frequency Appearance of User Names in the Discussion Messages
According to Groper’s (1996) findings, no single participant in the newsgroups
discussion was able to dominate the dialogues and provoke passion among his or her
followers. After counting the user name in the e-mail address, it was found that only two
persons posted more than three times in all 139 messages in the Mattel case. In the
Hudson Foods case, six on-line users posted more than three messages and some even
only forwarded media reports without expressing opinions. The majority of on-line users
involved in the crisis discussion posted one or two messages. Even though some people
might post one single message in different newsgroups, this phenomenon may imply that
the corporate crisis did not arouse much self-interest from on-line users who often post
messages to the newsgroups.

69
Users' Strategies
Two types of users' strategy were categorized. One is TALKER strategy applied
to users who initiate the discussion thread or express disagreement with other users.
Another is REACTOR strategy adopted by users who tend to agree with other people's
opinion or do not express their own opinions.
In the Mattel case, 38 users adopted TALKER strategy, while REACTOR
strategy was employed by 78 users, hi the Hudson Foods case, 46 users were identified
to have a TALKER strategy, and 83 employed REACTOR strategy. Because some on¬
line users had posted more than one message, they could adopt both TALKER and
REACTOR strategies in various postings. In total, 83 percent of users, 154 for Mattel
and 139 for Hudson Foods, were classified as either TALKERS or REACTORS.
Frequencies of Flaming Messages
As discussed earlier in Chapter two, flaming messages are often posted and seen
in the newsgroups, a public discussion forum without a central authority to regulate
users’ behavior. Contrary to expectation, the analysis of the on-line messages in the two
cases revealed that very few messages could be coded as flaming.
In the Mattel case, 14 flaming messages (10%) were analyzed: 9 targeted on the
Mattel company, 4 on other users, and 1 on both the company and users. As to the
Hudson Foods case, flaming messages were even fewer. Only 7 were retrieved: the
government was the flaming target in 3, 3 messages targeted on other users, and 1
targeted on both the government and users.

70
Mattel Cabbage Patch Doll Defect Case
Case Description
The Mattel toy company encountered a product defect crisis in the United States
at the end of 1996. The following story is based on the media stories retrieved from the
Lexis/Nexis data base.
On December 28, 1996, the New York Times reported that a battery-operated
cabbage patch doll manufactured by Mattel, an international toy company headquartered
in HongKong, munched a seven-year-old girl's hair and would not let go until her family
worked to dissemble the components of the doll. After the news was disclosed in the
conventional media, a company spokesman defended the toy, saying that the toy was safe
and that no other complaints had been presented. However, in the next few days, more
than nine accidents had been reported to the company and the news media as well. The
Consumer Product Safety Commission then started to investigate the doll accidents to
determine if legal action would be needed.
After January 1, 1997, Mattel decided to take action to resolve the crisis. First,
they stuck warning labels on the toy and then recalled all the defective cabbage patch
dolls from the retailers, offering 40 dollars as a refund to the customers who had bought
this type of doll.
After these interventions, no further serious damage case was discovered or
reported by mass media after January 10, 1997. News stories concerning the Mattel
company after January 10 had nothing to do with this crisis.
The eruption day of the crisis was December 28 and adoption of the intervention
strategies was first reported on January 1, 1997.

71
Retrieval of Media Stories from the Lexis/Nexis Data Base
As this crisis was solely related to the Mattel company, the keyword for searching
the news stories about the crisis was "mattel." After identifying the date when any of six
major media first reported the crisis event, the time fame was set up to retrieve news
stories. In this case, the date was December 28, 1996 when the New York Times
disclosed this crisis event on Section 1. According to the preset criteria, news stories that
appeared on the selected media 10 days before and 32 days after the crisis had to be
retrieved from the Lexis/Nexis data base. The time ñame hence was from December 18,
1996 to January 28, 1997. The search string was "mattel and date (aft 12/18/1996 & bef
01/28/1997)."
Forty-six news stories were downloaded and served as the basis of the media
agenda. Frequencies of news stories appearing in these media sources are shown in Table
5-1.
TABLE 5-1.
Frequencies of news stories on each news media in the Mattel case
Frequency
Percent
Valid
Percent
Cumulative
Percent
Valid
ABC News
4
8.7
8.7
8.7
CNN
8
17.4
17.4
26.1
NBC News
4
8.7
8.7
34.8
New York
Times
11
23.9
23.9
58.7
Wall Street
Journal
14
30.4
30.4
89.1
Washington
Post
5
10.9
10.9
100.0
Total
46
100.0
100.0
Total
46
100.0

72
In this case, most media stories came from two major news sources, the New York
Times and the Wall Street Journal; 54 % of total 46 news articles were gathered from
these two media.
Downloading of On-line Discussion Messages from Deja News Search Engine
The Deja News search engine provides different ways of retrieving on-line
messages that interest the readers. To simplify the search process, the method called
“Power Search” was employed, which allows on-line users to specify a time frame and
type in the keywords in order to get limited as well as desirable messages related to users'
topics of interest.
The time frame for the on-line search was identical to the one established for the
conventional media: from December 18, 1996 to January 28, 1997. By using the
keywords, "mattel AND (cabbage OR doll)," 488 messages displayed on the results
pages. To further examine each message, 139 were actually related to the crisis
discussion and therefore were used for testing the hypotheses. Frequencies of messages
appearing on each sampling day are shown in Table 5-2.
Test of Hypotheses
Hypothesis 1: The average number of daily USENET dialogues regarding the
corporation will be greater during the ten days following the eruption of the crisis
than the average number of daily on-line dialogues within ten days before the
crisis.
As people did not discuss the crisis event until it erupted, all 488 messages
downloaded by using the keywords were used to test this hypothesis. Consequently,
there were 66 and 136 discussion messages 10 days before and after the crisis eruption,

73
respectively. Table 5-3 shows a significant difference between the two groups.
Therefore, this hypothesis was supported.
TABLE 5-2.
Frequencies of messages appearing on each sampling day in the Mattel case
Valid Cumulative
Frequency
Percent
Percent
Percent
12/28/96
4
2.9
2.9
2.9
12/29/96
8
5.8
5.8
8.6
12/30/96
17
12.2
12.2
20.9
12/31/96
17
12.2
12.2
33.1
01/01/97
5
3.6
3.6
36.7
01/02/97
5
3.6
3.6
40.3
01/03/97
11
7.9
7.9
48.2
01/04/97
10
7.2
7.2
55.4
01/05/97
9
6.5
6.5
61.9
01/06/97
1
.7
.7
62.6
01/07/97
10
7.2
7.2
69.8
01/08/97
3
2.2
2.2
71.9
01/09/97
6
4.3
4.3
76.3
01/10/97
1
.7
.7
77.0
01/11/97
11
7.9
7.9
84.9
01/12/97
2
1.4
1.4
86.3
01/13/97
5
3.6
3.6
89.9
01/14/97
8
5.8
5.8
95.7
01/16/97
5
3.6
3.6
99.3
01/17/97
1
.7
.7
100.0
Total
139
100.0
100.0
139
100.0
Total

74
TABLE 5-3.
Results of Hypothesis 1 test in the Mattel case
N
Mean
Std.
Deviation
Std. Error
t_- value
10 days
before
crisis
66
6.6
5.1
1.61
-2.756*
10 days
after
crisis
136
13.6
6.2
1.96
*p < 0.05 level
Hypothesis 2: The average number of daily negative USENET dialogues will be
greater than the average number of daily positive USENET dialogues after the
eruption of the crisis.
The numbers of positive and negative messages during 32 days after the crisis
eruption were compared. Table 5-4 shows a significant difference between the average
numbers of daily negative and positive USENET dialogues. It thus supported the
hypothesis.
TABLE 5-4.
Results of Hypothesis 2 test in the Mattel case
N
Mean
Std.
Deviation
Std. Error
t_- value
Positive
Messages
5
0.16
0.51
0
3.59
Negative
Messages
58
1.81
2.76
0.49
*p < 0.05 level
Hypothesis 3: The daily number of the USENET dialogues will be related to the
daily number of news stories about a crisis after its eruption.
Consistent with the concept of agenda-setting theory, the more news stories are
about a crisis, the more on-line discussion messages should be posted on the USENET.

75
Simple regression analysis was utilized to test this hypothesis. The number of news
stories on each day after the crisis eruption was the independent variable, whereas that of
the on-line dialogues was the dependent variable. As r was 0.635 (r2 = 0.403) and
significant (F (1, 31) = 20.24, p < 0.000), this supported the hypothesis. While the
correlation was moderately strong, the relationship was significant and in the expected
direction.
Hypothesis 4: The average number of daily negative USENET dialogues
following a crisis intervention strategy will be less than the negative USENET
dialogues prior to the intervention.
Twenty-seven negative on-line dialogues in four days had been posted before the
adoption of a crisis intervention and 31 messages had been posted in 28 days after its
intervention. Table 5-5 reveals a significant difference that supports the hypothesis.
Nevertheless, to compare the average number of negative on-line dialogues four days
before and after the intervention, the results shown in Table 5-6 differed and the
hypothesis was not supported.
TABLE 5-5.
Results of first Hypothesis 4 test in the Mattel case
N
Mean
Std.
Deviation
Std. Error
t_- value
4 days before
intervention
27
6.75
2.75
1.38
%
3.96
28 days after
intervention
31
1.11
1.95
0.37
*p < 0.05 level

76
TABLE 5-6.
Results of second Hypothesis 4 test in the Mattel case
N
Mean
Std.
Deviation
Std. Error
t_- value
4 days before
intervention
27
6.75
2.75
1.38
1.075NS
4 days after
intervention
19
4.75
2.5
1.25
Hypothesis 5: Following a crisis intervention strategy, the average number of
daily positive on-line dialogues on the USENET concerning the company will be
equal to the average number of daily positive on-line dialogues before the crisis
eruption.
As no positive message was posted before the crisis eruption and three were
posted after the adoption of intervention, no statistical significance was found in Table
5-7. The hypothesis was supported, although the positive messages were very few.
TABLE 5-7.
Results of Hypothesis 5 test in the Mattel case
N
Mean
Std.
Deviation
Std. Error
t_- value
10 days
before crisis
0
0
0
0
-1.362NS
28 days after
intervention
3
0.11
0.42
0
Hypothesis 6: The number of the USENET dialogues following a crisis
intervention will be less than the number of the USENET dialogues before the
intervention and after the crisis.
Comparing the average number of on-line dialogues on each day, four days before
(nj = 46) versus 28 days after the intervention adoption (n2 = 93), the hypothesis was
supported as demonstrated in Table 5-8.

77
TABLE 5-8.
Results of Hypothesis 6 test in the Mattel case
N
Mean
Std.
Deviation
Std. Error
t_- value
4 days after crisis
and before
intervention
46
11.5
6.56
3.28
3.544
28 days after
intervention
93
3.32
3.99
0.75
*p < 0.05 level
Hypothesis 7: The number of daily negative USENET dialogues posted by
TALKERS will be greater than that by REACTORS.
TALKERS posted 22 negative on-line messages, whereas 34 negative on-line
messages were posted by REACTORS. By using a Chi-square test, this hypothesis was
not supported as shown in Table 5-9, %2 (1, N = 61) = 0.727, p < 0.394.
TABLE 5-9.
Results of Hypothesis 7 test in the Mattel case
Negative
Positive
Total
df
Chi-square
value
TALKER
22
1
23
1
0.727NS
REACTOR
34
4
38
Total
56
5
61
Hypothesis 8: The Optimism Score of media messages will be positively related
to the number of negatively toned USENET messages posted after the media
stories were released.

78
To test this hypothesis, the correlation was examined between the average number
of negative on-line messages each day and the Optimism Score, which was calculated by
subtracting the number of negative words from the number of positive words found in
each media story. The Optimism Score is a summed score for all media messages on a
particular day.
Simple regression analysis revealed that the Optimism Scores in news stories
were positively related to the number of negative on-line messages posted each day as the
correlation coefficient was 0.467, ¿ = 0.22, p < 0.02. Although the variance explained
was small, this hypothesis was supported.
o
as
UJ
O)
05
U)
(/)
(D
10
â–¡
â–¡
s
03
a>
03
0)
X!
E
13
6
4
2
0
-2
-40
â–¡
â–¡
â–¡
â–¡ â–¡
â–¡ â–¡ â–¡
â–¡ â–¡
â–¡â–¡ â–¡â–¡â–¡â–¡â–¡
â–¡ â–¡
â–¡
-30 -20 -10 0 10 20
Optimism Score for All News Stories Each Day
FIGURE 5-1. Bivariate correlation between the number of negative on-line messages
and the Optimism Score in the Mattel case

79
Hudson Foods Meat Contamination Case
Case Description
The Hudson Foods company encountered a serious meat contamination crisis
across the United States after July of 1997. The following story is based upon the media
stories retrieved from the Lexis/Nexis data base.
On August 12, 1997, the ABC World News Tonight had the following coverage:
"..., an Arkansas meat processor is recalling 20,000 pounds of frozen hamburger patties
which may be contaminated with the E. coli bacteria. Hudson Foods distributed the
hamburger nationwide. The problem was first spotted in Colorado when several people
who had eaten the hamburger became ill." This news article is the earliest one that
showed "Hudson Foods" in all news documents of the Lexis/Nexis data base in both July
and August of 1997. Therefore, the beginning day of this crisis was considered to be
August 12, 1997.
After the government urged more recall of the contaminated meat, Hudson Foods
recalled about 1.2 million beef products from the market on August 16, 1997, closed its
plant in Nebraska and recalled 25 million beef patties on August 21, 1997, according to
the ABC news reports.
In the beginning of September, 1997, Hudson Foods agreed to merge with its rival
company, Tyson Foods.
As to this crisis case, the eruption day was August 12, 1997 and the adoption of
the intervention strategies occurred twice. The first was on August 16, 1997, when the
company recalled one million pounds of beef patties and the second was on August 21,

80
1997, when about 25 million beef patties were recalled and a food plant in Nebraska was
shut down.
The intensity of this crisis extended beyond the 42 days framed in this study. The
U.S. News & World Report and the Newsweek in their September editions had stories
featuring the food contamination cases in the past decade to demonstrate the importance
of food safety. The New York Times had the following news story in Section A of its late
edition dated on October 2, 1997:
A Federal grand jury is investigating whether a
meatpacking company tried to hide the extent of potential E. coli
bacteria contamination in what became the largest meat recall.
Tom Monaghan, the United States Attorney for Nebraska,
issued a statement saying the investigation had arisen from
information received last month from the Agriculture Department's
inspector general regarding the August recall of 25 million pounds
of hamburger produced at the Hudson Foods plant in Columbus,
Neb.
Mr. Monaghan would provide no other details, but
company officials confirmed today that they had received a
subpoena for documents related to the recall and expected some
employees to be called.
Hudson Foods, based in Rogers, Ark., said in a statement
today, "As it has in the past, Hudson Foods will continue to
cooperate frilly."
Hudson shut down the Columbus plant at Agriculture
Department insistence after the amount of meat recalled due to
possible E. coli contamination rose from 20,000 pounds initially to
25 million pounds.
Retrieval of Media Stories from the Lexis/Nexis Data Base
To search for news articles about this crisis event in the Lexis/Nexis data base, the
company's name "Hudson foods" did not appear until August 12, 1997 by using the
keywords “Hudson foods and date (bef08/31/1997 and aft 07/01/1997)/’’ Following the
preset criteria, news stories that appear on the selected media 10 days before and 32 days
after the crisis need to be retrieved from the Lexis/Nexis data base. The time frame

81
hence was from August 2, 1997 to September 12, 1997 and 172 news stories were
downloaded and served as the basis of the media agenda.
TABLE 5-10.
Frequencies of news stories on each news media in Hudson Foods case
Frequency
Percent
Valid
Percent
Cumulative
Percent
Valid
ABC News
20
11.6
11.6
11.6
CNN
51
29.7
29.7
41.3
NBC News
29
16.9
16.9
58.1
New York
Times
32
18.6
18.6
76.7
Wall Street
Journal
20
11.6
11.6
88.4
Washington
Post
20
11.6
11.6
100.0
Total
172
100.0
100.0
Total
172
100.0
As Table 5-10 shown above indicates, CAW had 51 stories related to the Hudson
Foods company and the New York Times had the second most, 32 news stories covering
this company. Interestingly, the ABC did not have any news stories in September
containing the keywords oí" Hudson foods," so only 20 news stories were analyzed from
the ABC news media.
Downloading of On-line Discussion Messages from Deja News Search Engine
As similar to the Mattel case, the search function of “Power Search” in the Deja
News search engine was used to obtain on-line discussion messages pertaining to the
crisis. Because the last news article was retrieved on September 12, 1997, discussion
messages of one more day were downloaded for time-lag analysis. The time frame then
started from August 2, 1997 to September 13, 1997. Keywords were "hudson AND foods

82
OR beef OR meat." That is, any messages containing Hudson foods or Hudson and beef
or Hudson and meat would have been retrieved and examined to determine if they were
relevant to the crisis discussion. As a result, 154 messages were regarded as related to
the crisis and became the basis of the public agenda. Frequencies of messages appearing
on each sampling day are displayed in Table 5-11.
TABLE 5-11.
Frequencies of messages appearing on each sampling day in Fludson Foods case
Frequency
Percent
Valid
Percent
Cumulative
Percent
Valid
08/08/97
1
.6
.6
.6
08/13/97
1
.6
.6
1.3
08/15/97
2
1.3
1.3
2.6
08/16/97
3
1.9
1.9
4.5
08/17/97
3
1.9
1.9
6.5
08/18/97
5
3.2
3.2
9.7
08/19/97
7
4.5
4.5
14.3
08/20/97
2
1.3
1.3
15.6
08/21/97
2
1.3
1.3
16.9
08/22/97
11
7.1
7.1
24.0
08/23/97
11
7.1
7.1
31.2
08/24/97
8
5.2
5.2
36.4
08/25/97
1
.6
.6
37.0
08/26/97
1
.6
.6
37.7
08/27/97
3
1.9
1.9
39.6
08/28/97
8
5.2
5.2
44.8
08/29/97
6
3.9
3.9
48.7
08/30/97
4
2.6
2.6
51.3
08/31/97
3
1.9
1.9
53.2
09/01/97
5
3.2
3.2
56.5
09/02/97
3
1.9
1.9
58.4
09/03/97
1
.6
.6
59.1
09/04/97
5
3.2
3.2
62.3
09/05/97
3
1.9
1.9
64.3
09/06/97
3
1.9
1.9
66.2
09/07/97
3
1.9
1.9
68.2
09/08/97
9
5.8
5.8
74.0
09/09/97
6
3.9
3.9
77.9
09/10/97
6
3.9
3.9
81.8
09/11/97
11
7.1
7.1
89.0
09/12/97
14
9.1
9.1
98.1
09/13/97
3
1.9
1.9
100.0
Total
154
100.0
100.0
Total
154
100.0

83
Test of Hypotheses
Hypothesis 1: The average number of daily USENET dialogues regarding the
coiporation will be greater during the ten days following the eruption of the crisis
than the average number of daily on-line dialogues within ten days before the
crisis.
One message regarding the company was posted ten days before the crisis
eruption and 25 were posted ten days after the crisis. The t-test shows a significant
difference in Table 5-12. Therefore, this hypothesis was supported.
TABLE 5-12.
Results of Hypothesis 1 test in the Hudson Foods case
N
Mean
Std.
Deviation
Std. Error
t_- value
10 days
before
crisis
1
0.1
0.32
0
-3.456*
10 days
after
crisis
25
2.5
2.17
0.69
*p < 0.05 level
Hypothesis 2: The average number of daily negative USENET dialogues will be
greater than the average number of daily positive USENET dialogues after the
eruption of the crisis.
The numbers of positive and negative messages during 32 days after the crisis
eruption were compared. The t-test reveals a significant difference between the average
numbers of daily negative and positive on-line dialogues in Table 5-13. Consequently,
this hypothesis was supported.

84
TABLE 5-13.
Results of Hypothesis 2 test in the Hudson Foods case
N
Mean
Std.
Deviation
Std. Error
t_- value
Positive
Messages
11
0.33
0.54
0
* —
5.71
Negative
Messages
68
2.06
1.85
0.32
*p < 0.05 level
Hypothesis 3: The daily number of the USENET dialogues will be related to the
daily number of news stories about a crisis after its eruption.
Different from the finding in the Mattel case, this result, r = 0.231, p < 0.197,
F (1, 32) = 1.742, r = 0.053, does not support this hypothesis. The relationship between
number of news stories and that of on-line messages was not significant.
Hypothesis 4: The average number of daily negative USENET dialogues
following a crisis intervention strategy will be less than the negative USENET
dialogues prior to the intervention.
To test this hypothesis, a one-way ANOVA was used. Unlike Mattel, which had
only one intervention, Hudson Foods adopted intervention strategies at two different
times. In the Hudson Foods case, hypothesis 4 was tested in three ways.
The first comparison was the average number of daily negative USENET
dialogues among three stages, before the first intervention, after the first intervention but
before the second intervention, and after the second intervention. Table 5-14 shows no
significant relationship among three stages, F (2, 32) = 2.966, p < 0.067.

85
TABLE 5-14.
Results of comparison of three stages in Hypothesis 4 in the Hudson Foods Case
Sum of
Squares
df
Mean
Square
F
Sig.
Number of Between
negative Groups
18.140
2
9.070
2.966
.067
message Within
Groups
91.739
30
3.058
Total
109.879
32
a. Non-Significant
The second test was to compare the number of negative USENET messages
before and after the first intervention stages. Table 5-15 reveals a significant
relationship, F (1, 32) = 5.38, p < 0.027.
TABLE 5-15.
Results of comparison of two stages in Hypothesis 4 in the Hudson Foods Case
Sum of
Mean
Squares
df
Square
F
Sig.
Number of
negative
Between
Groups
16.250
1
16.250
5.380
,027a
message
Within
Groups
93.629
31
3.020
Total
109.879
32
a- p < 0.05 level
The final test was to compare the number of negative USENET messages before
and after the second intervention stages. A significant relationship was found in Table 5-
16, F (1,32) = 4.247, p< 0.048.

86
TABLE 5-16.
Results of comparison of two stages in Hypothesis 4 in the Hudson Foods Case
Sum of
Mean
Squares
df
Square
F
Sig.
Number of
negative
Between
Groups
13.240
1
13.240
4.247
,048a
message
Within
Groups
96.639
31
3.117
Total
109.879
32
a p < 0.05 level
Hypothesis 5: Following a crisis intervention strategy, the average number of
daily positive on-line dialogues on the USENET concerning the company will be
equal to the average number of daily positive on-line dialogues before the crisis
eruption.
The ANOVA results in Table 5-17 supported this hypothesis, m = 1, x\2 - 11,
F (1, 37) = 2.375, p < 0.132. That is, whether the company adopted the intervention or
not, no significant difference was found in the average number of daily positive on-line
dialogues after the intervention and before the crisis regardless of the few positive on-line
messages.
TABLE 5-17.
Results of Hypothesis 5 test in the Hudson Foods case
Sum of
Squares
df
Mean
Square
F
Sig.
Average # of Positive Between
On-line Messages Groups
Each Day within
Groups
Total
.632
9.579
10.211
1
36
37
.632
.266
2.375
. _ _ 3
.132
a. Non-Significant

87
Hypothesis 6: The number of the USENET dialogues following a crisis
intervention will be less than the number of the USENET dialogues before the
intervention and after the crisis.
As Hudson Foods adopted interventions at two different times, this hypothesis
was tested in two ways. The first test was to see if the total number of on-line dialogues
decreases after the first intervention. By the t-test, Table 5-18 did not support the
hypothesis. In other words, the total number of on-line dialogues did not decrease but
actually increased with adoption of the first crisis intervention.
TABLE 5-18.
Results of comparison of two stages in Hypothesis 6 in the Hudson Foods Case
N
Mean
Std.
Deviation
Std. Error
t_- value
5 days after crisis
and before 1 st
intervention
6
1.2
1.3
0.58
-4.607*
28 days after 1st
intervention
147
5.25
3.48
0.66
*p < 0.05 level
The other time frame to test this hypothesis was before and after the second
intervention. Similarly, the t-test results in Table 5-19 showed no evidence of a decrease
in the total number of USENET dialogues after the second crisis intervention. The
number of USENET discussion dialogues increased after the second intervention.
Hypothesis 7: The number of daily negative USENET dialogues posted by
TALKERS will be greater than that by REACTORS.
TALKERS posted 23 negative USENET dialogues and REACTORS posted 42.
By using a Chi-square test, this hypothesis was not supported in Table 5-20, x2 (1, N =
76) = 0.004, p< 0.95.

88
TABLE 5-19.
Results of comparison of two stages in Hypothesis 6 in the Hudson Foods Case
N
Mean
Std.
Deviation
Std. Error
t_- value
10 days after
crisis and before
2nd intervention
25
2.5
2.17
0.69
-2.982*
23 days after 2nd
intervention
128
5.57
3.67
0.76
*p < 0.05 level
TABLE 5-20.
Results of Hypothesis 7 test in the Hudson Foods case
Negative
Positive
Total
df
Chi-square
value
TALKER
23
4
27
1
0.004NS
REACTOR
42
7
49
Total
65
11
76
Hypothesis 8: The Optimism Score of media messages will be positively related
to the number of negatively toned USENET messages posted after the media
stories were released.
To test this hypothesis, bivariate correlation between the average number of
negative on-line messages each day and the Optimism Score was examined as shown in
Figure 5-2.

89
a
>*
to
D
.c
o
ffl
LU
C/>
CD
3
(/)
I
to
B>
£
z
0
-2
â–¡ o â–¡ on â–¡
â–¡ â–¡ CD â–¡
â–¡ â–¡ â–¡ â–¡
-120 -100 -80 -60 -40 -20 0 20
Optimism Score for All News Stories Each Day
FIGURE 5-2. Bivariate correlation between the number of negative on-line messages
and the Optimism Score in the Hudson Foods case
The above figure suggested that the Optimism Scores in news stories were not
related to the number of negative messages posted each day as r = 0.112 (r2 = 0.012) with
P < 0.536. As a result, this hypothesis was not supported.
Post Hoc Analysis
According to the agenda-setting theory, there should be a positive correspondent
relationship between the media agenda and the public agenda. In the Mattel case, this
relationship has been found. The relationship was not found in the Hudson Foods case.
To examine why the results of Hypothesis 3 differed between the two cases, additional
analyses were conducted. Figures 5-3 and 5-4 show the patterns of USENET messages
and the number of news stories.

90
1 8
FIGURE 5-3. Relationship between the media agenda and the public agenda in the
Mattel case
FIGURE 5-4. Relationship between the media agenda and the public agenda in the
Hudson Foods case
The lack of a significant relationship between the media agenda and the public
agenda in the Hudson Foods case is evident (r = 0.012). Using SPSS curve-fitting, a
cubic relationship was found to improve the explanatory power of the media agenda

91
impacting the public agenda from 0.012 to 0.14 as demonstrated in Figure 5-5. The
Mattel data illustrate the expected linear pattern of news stories and USENET messages.
Rsq = 0.1415
Number of News Stories on Each Day
FIGURE 5-5. Cubic relationship between the media agenda and the public agenda in
the Hudson Foods case
Analysis of Message Type
In order to better understand the curvilinear relationship between the public
agenda and the media agenda, contents of all on-line messages for both cases were re¬
examined and categorized.
The coding categories were created based on Thomsen’s (1996) article analyzing
the discussion messages in the PRForum. Five categories were formed:
1. Question/Answer: exchange of information about the crisis or the target company
Example:

92
Subject: Re: How much Tuna is safe?
From: atcshane@aol.com (ATC Shane)
Date: 1997/08/18
Newsgroups: alt.sport.weightlifting
>1 was recently told that eating tuna more than 3 times a week could cause
>mercury poisoning; is there any truth to this?
Hmmm. Sounds like someone is hitting you with some paranoia. I think if it
was that easy to get poisoned from tuna, the FDA wouldn't allow it on the
market. Heck, look at Hudson Beef. It's like one girl got EColi poisoning
and the gov pulled over a million pounds of beef off the market!
2. Opinion: discussion, debates or arguments concerning the crisis, expression of
personal experiences
Example:
Subject: Re: [OT] US News Finally Finds a 'Clinton Watch' Scandal
From: gcruse@worldnetnospam.att.net (gary cruse)
Date: 1997/08/23
Newsgroups: alt.current-events.clinton.whitewater
On Sat, 23 Aug 1997 13:07:13 -0500, Bill Nalty
wrote:
>http://www.usnews.com/usnews/issue/970901/lfeed.htm
>
>U.S NEWS & WORLD REPORT
>September 1, 1997
>
>
> The next bad beef scandal?
>
> Cattle feed now contains things like
> manure and dead cats
>
>
I'm not sure I wanted to know about this.
Well, maybe. But, it sure puts a new light
on how "removing agricultural subsidies
will put the small farmers out of business."
Since, according to the article, this
manure recycling is practiced by small

93
farmers and not agribusiness, I am even
less convinced that individual farming
should be subsidized.
3. Media Report: forwarding media story related to the crisis, updates of media reports
Example:
Subject: From AM News Abuse for August 30/31/Sep 1
From: Daisy
Date: 1997/08/31 ’
Newsgroups: alt.culture.fabulous
Dear Is The Irradiator Ready:
- One result of the Hudson Foods hamburger contamination fiasco is
renewed http://www.pmac.net/foodirr.htm discussion about why the
beef industry doesn't have FDA approval to use
http://www.eatright.org/airradi.html irradiation to kill bacteria like
the pork and chicken industries do. Forty countries use it regularly.
4. Crisis Cited as Example: mention of the crisis without expressing any comments, use
of the crisis as a sarcastic or Finny joke
Example:
Subject: Re: How Damned Convenient
From:
Date: 1997/09/07
Newsgroups: alt.conspiracy
No money to the DNC works or maybe nothing more complicated than a gross
power grab by the federal government. AFer all Hudson Foods immediately shut
aFer the FDA said there was a problem. The very same day the Director of the
FDA held a news conference stating he wants a law to force closure. Doesn't
make sense to me, unless the government was just taking more power.
Everyday American's rights are being removed by an oppressive government and
America sleeps. Everyday we talk about how bad it is. Talk is cheap and our
rights slip away daily.
5. Combination of category 1, 2 or 3.
Example:

94
Subject: food irradiation
From: John McCarthy
Date: 1997/08/31
Newsgroups: sci.environment
Here's a Reuters story.
It followed the NYTimes Op-Ed by Richard Rhodes by only a few days.
Being a news story and not an Op-Ed it has direct quotes from various
people. The Nader minion, Caroline Smith, was negative but was rather
mild for a Naderite. She was sufficiently mild it may relieve fears
among politicians of an environmentalist attack. There was also the
usual expert Gary Smith who said that if only the consumers would cook
the food enough, it would be safe. [I won't do that, because I like
meat rare, and I think the number of food poisoning incidents is low
enough so that eating rare meat or even steak tartare is one of the
lesser risks I face.]
The Nader minion says irradiation changes the taste of meat, whereas
Rhodes says it doesn't. It would be interesting to know if anyone has
ever done an actual test.
WASHINGTON (Reuter) - Irradiation. It may sound sinister but
food safety experts say it could become a key tool to keep
sickness-causing bacteria out of the U.S. meat supply....
Mattel Case
As shown in Figure 5-6 below, 119 out of the total 139 messages (86%) were
coded in the category of either "question/answer" or "opinion." While numbers of the
question/answer type messages were equally distributed until January 15, 1997, the
number of opinion type messages reached its peak before January 7, 1997. This suggests
that the on-line users' concerns about the crisis diminished after it has progressed to the
"abate" stage. Overall, the distribution of frequency of the messages corresponded with
media coverage of the crisis as previously seen in Figure 5-3.

95
-- Q & A
O pinion
Sampling Date
FIGURE 5-6. Distribution of numbers of on-line messages by types in the Mattel
case
Hudson Foods Case
An interesting finding arose from the Hudson Foods case when contrasting the
public agenda with the media agenda: there are more media stories than USENET
messages. The distribution of numbers of messages by types during the sampling period
is demonstrated in Figure 5-7.
Q & A
O pinions
I \
FIGURE 5-7. Distribution of numbers of on-line messages by types in the Hudson Foods
case
Analysis of Figure 5-7 suggested that the occurrence of the public on-line
messages can be divided into three stages. The first stage runs from the first sampling

96
day to August 26, 1997, the second stage then begins after August 27 and runs to
September 7, 1997. The last stage is after September 8, 1997. An examination of the
message content shows that USENET dialogues were primarily question/answer after the
crisis erupted and before Hudson Foods implemented intervention strategies. The
question/answer content decreased until the final stage, where on-line users again posted
more questions and answers. A possible explanation can be found in the complexity of
this crisis. The Hudson Foods case involved special terms of food bacteria such as E-coli
(main topic of questions in the first stage) and handling of the contaminated meat (main
topic of questions in the final stage). This suggests that the public agenda may be
questions and answers rather than expressing opinions when the media content is a
complicated issue about which the public lacks the knowledge to fonn an opinion.
Analysis of Combined Data from Two Cases
The central research question in the study is "Does the media agenda have
powerful impacts on the public agenda in terms of agenda-setting theory applied to the
crisis communication management?" The answer to the above question is yes in the
Mattel case, but no in the Hudson Foods case given the test of Hypothesis 3 yielding
distinctive findings. Because both cases belonged to a type of product tampering crisis,
data of several variables in both cases were aggregated to provide an alternative answer
to this question. The variables being aggregated included number of on-line messages,
number of media stories, number of negative on-line messages, and the Optimism Score
of media stories on each day.
Because the Mattel case did not have media stories and USENET messages after
the 40th day, the data were aggregated across 43 days in this case. The method of

97
aggregating data in one variable was to add them together from two cases and then to
divide the sum by two.
Along with those four variables, tone of messages and users' strategy were also
selected to answer another question. For the two categorical variables, 293 cases were
generated.
With six variables combined, three questions are formed:
Q1: Is the number of on-line messages on each day positively related to the number of
media stories on each day?
Q2: Is the number of negative on-line messages on each day positively related the
negative tone that media used in their news stories on each day?
Q3: Is there any difference between TALKERS and REACTORS in posting the negative
as well as positive on-line messages on the USENET?
Simple regression analysis was used to answer the first and second questions and
the third question was tested with a Chi-square test.
For the first question, r = 0.599, p < 0.000, F (1, 42) = 22.99, r2 = 0.359. There
was a strong significant correlation between the number of on-line messages and the
number of media stories. Figure 5-8 shows that most on-line messages were posted
between day 10 to day 30, that is, on-line users paid attention to the corporate crisis in the
first 20 days and then their interests waned.

98
# of Combined On-line Messages
'# of Combined Media Stories
Sequence of Days
FIGURE 5-8. Frequencies of combined on-line messages and media stories
Frequencies of combined negative on-line messages and the Optimism Scores in
media stories are illustrated in Figure 5-9. The correlation coefficient indicated a
moderately significant relationship between the two variables, r = 0.487, p < 0.001,
F (1,42)= 12.757, r2 = 0.237.
Concerning the third question, a Chi-square test in Table 5-21 displayed a non¬
significant difference between TALKERS and REACTORS in posting the negative as
well as positive on-line messages on the USENET, x2 (1, N = 137) = 0.215, p < 0.643.
This result was consistent with what has been found in each crisis case as TALKERS did
not particularly post more negative messages than REACTORS.

99
20
10
0
—•—# of Combined Negative On¬
line Messages
—a®—Combined Optimism Scores
-10
%
-20
-D
a
s
Z
-30
a -40
-50
-60
-70
Sequence of Days
FIGURE 5-9. Frequencies of combined negative on-line messages and the Optimism
Scores in media stories
TABLE 5-21.
Results of Chi-square test with combined data
Negative
Positive
Total
df
Chi-square
value
TALKER
45
5
50
1
0.215NS
REACTOR
76
11
87
Total
121
16
137

100
The summation of findings and answers to the research questions will be
presented in the next chapter. The crisis communication model based upon the public
agenda examined in both cases will also be discussed, followed by limitations of the
study. Implications and suggestions for public relations professionals as well as
academic researchers conclude the final chapter.

CHAPTER VI
DISCUSSION
Contrary to the traditional way of surveying people about their opinions, this
research employed content analysis to investigate what people really think and say about
a corporation involved in an unexpected crisis. Agenda-setting theory functioned as a
theoretical framework for developing research questions. In addition, Sturges’ public
opinion model and MacKuen’s strategy model of public dialogue were used to
understand the trend of on-line public opinion concerning two selected corporate crises.
Four research questions and eight hypotheses were presented in Chapter three to
examine to what degree the media coverage of a crisis and the corporate intervention
strategies affect the public discussion on the USENET. The tone of the USENET
dialogues was the major focus. The numbers of negatively and positively toned messages
were the dependent variables, forming the wave line of a crisis communication model
proposed later in this chapter.
The results of hypothesis testing are summarized first, followed by answers to the
research questions. The crisis communication model generated from the findings is
outlined next, followed by limitations of the study. This chapter ends with implications
and suggestions for future research.
101

102
Summation of Findings
Mattel Case
Seven hypotheses out of 8 in the Mattel crisis case were supported. The number
of negative USENET messages increased after the crisis erupted. There were 46
messages discussing the crisis 4 days before the company decided to take intervention
actions. Among them, 27 (59%) were negative. However, after the company adopted the
intervention strategies, such as recall of defective products, refunding, and a warning
label attached to the product, the total number of discussion messages was reduced.
With regard to application of agenda-setting theory for testing the hypotheses, the
number of media stories covering this crisis supports the notion of causal influences on
the number of on-line messages. Even the negative tone of news articles was
significantly correlated with the number of negative on-line messages, r = 0.467, ¿ =
0.22, p< 0.02.
The only hypothesis not supported was the test of bivariate correlation between
users' strategies and the tone of on-line messages. There was no significant correlation
between employment of users' strategies and the number of negative on-line messages,
X2(1,N = 61) = 0.727,p <0.394.
Hudson Foods Case
Only three hypotheses were supported in this meat contamination case. The total
number of on-line messages about the Hudson Foods company and negative on-line
messages increased after the crisis erupted. With a few positive on-line messages
analyzed, their numbers were similar after the intervention actions were taken, F (1, 37) =
2.375, g <0.132.

103
Unlike the findings in the Mattel case, the number of media stories did not
causally impact the number of on-line messages, r = 0.231 , p < 0.197, F (1, 32) = 1.742,
2 9 t
r = 0.053, with 172 media stories and 154 on-line messages analyzed. Neither did the
tone of media stories have a significant correlation with the number of negative on-line
messages, r = 0.112, r2 = 0.012, p < 0.536.
As in the Mattel case, no significant correlation between users' strategy and the
number of negative on-line messages was found, x,2 (T N = 76) = 0.004, p < 0.95.
Data Aggregation in Both Cases
Results from analysis of the aggregated data corresponded with concepts of
agenda-setting theory. The finding, r = 0.599, p < 0.000, F (1, 42) = 22.99, r2 = 0.359,
supported the assumption that the conventional media coverage of a corporate crisis one-
sidedly affects the public discussion about the crisis or the target company. The tone that
the media used to cover the crisis events also had a moderately significant correlation
with the number of negative on-line messages, r = 0.487, p < 0.001, F (1, 42) = 12.757, r2
= 0.237. Again, no significant correlation between users’ strategy and the number of
negative on-line messages was found, % (1, N = 137) = 0.215, p < 0.643.
Research Questions
Research questions in this study were based on the Sturges’ public opinion model,
past agenda-setting studies, and MacKuen’s model of public dialogue. They are
summarized from Chapter III as follows:
1. Does the trend of on-line public opinion during the corporate crisis follow
Sturges’ model?

104
2. Does the media agenda affect the public agenda in terms of the first and
second level of agenda-setting theory?
3. Do on-line users applied to MacKuen’s concept of users’ strategies post
different numbers of negative messages?
The trend of public opinion during the time of the Mattel crisis matched Sturges’
public opinion model of crisis communication. The number of on-line discussion
messages moved up and down with the advent of the crisis and adoption of the
interventions. The development of public opinion was similar to what was predicted by
Sturges (1994) from pre-crisis to the termination stage (See Figure 3-2 on p. 40). To
compare the numbers of negative and total on-line messages with the media reports, the
media impact on the public agenda was consistent with the agenda-setting studies.
Therefore, answers to the first two questions are positive.
In the Hudson Foods case, the trend of public opinion did not look like that in the
Mattel case, although the number of negative or total on-line messages indeed grew in the
beginning of the crisis. The number of on-line messages did not decline even after
Hudson Foods took intervention actions. The public still felt very concerned about this
crisis. The reason for this could result from the prolonged media reports about food
safety. For instance, 34 news articles between September 14 and October 15, 1997 about
the Hudson Foods company or food safety in major newspapers were found through a
search in the Lexis/Nexis data base. The media agenda has gone beyond the 43-day time
frame in this case study. As a matter of fact, the meat contamination accident had been
reported dating back to June, 1997.

105
Another reason could be due to the intense and broad impact of this crisis on the
consumers as well as the food industry, motivating people to feel more uncertain and
concerned about the product safety of daily food than about toy dolls in the Mattel crisis.
To answer the second research question, the number of media stories did not
increase the number of on-line messages. The explanation for this finding would be due
to the uneven numbers of retrieved messages, equally stretched out to all sampling days,
while the majority (48%) of 154 media stories appeared from August 21 to August 26. In
fact, if comparing both variables within another time frame, the finding differs. For
example, if the average number of news stories is compared with that of on-line messages
each day before the first intervention, the correlation is significant, r = 0.943, p < 0.016, F
(1, 4) = 24.14, r2= 0.889.
As discussed earlier, answers to the third research question in both cases is "no."
No significant correlation between users’ strategies, TALKER and REACTOR, and the
number of negative on-line messages was found, even with data combined. As a result,
there was no difference among on-line users being identified as TALKERS or
REACTORS in posting negative messages about the company.
Failure to detect the significant correlation between users’ strategies and the
number of negatively toned on-line messages could rest on the difference of the settings
where the public dialogues take place on a daily basis. MacKuen urged that democracy
will progress when people use TALK/CLAM strategies in a public discussion forum to
become involved in meaningful dialogues. However, in order to frame a friendly
situation, people will likely adopt REACT/SIGNAL strategies without expressing
disagreement.

106
Situated in the so-called “virtual reality” of the Internet, anonymous users by
using TALKER or REACTOR strategies are allowed to freely post their opinions either
agreeing or disagreeing with other people without worrying about getting into an
unfriendly environment that impedes public discussion or mutual communication.
Proposed Crisis Communication Model
In times of crisis, monitoring the media agenda as well as the public agenda is the
key to successful crisis management. Considering the urgency of the crisis and pressure
on the organization, a communication model that facilitates control over trends of public
opinion is indispensable.
Any crisis communication model based on this examination of public discussion
messages on the USENET must integrate several factors. The fourth hypothesis test
revealed the importance of adopting intervention strategies that can reduce the number of
negative on-line messages about the corporation. Consequently, the earlier the
interventions are employed, the quicker the negative public opinion about the company is
neutralized.
In both cases examined, the content of on-line messages was re-examined for the
qualitative analysis. The results showed that not all messages posted on the newsgroups
expressed opinion. Some of them were information exchange or questions/answers
posting. The corporation should not only feel concerned about what people say about the
crisis, but about what people want to know about the crisis situation. It then can respond
to the questions posted on the Internet.
Take Hudson Foods as an example. Some on-line users asked about how much
harm the E-coli bacteria could cause to human beings. Some felt anxious about the

107
handling methods of recalled meats. Some even desired to know how to properly cook
the beef to avoid being infected by the E-coli bacteria. Suppose that Hudson Foods had
been able to respond to these questions through the traditional media and the Internet, the
duration of the crisis might have been much shorter. Taking advantage of the
spontaneous and interactive characteristics of the Internet, the Tommy Hilfiger company
set a good example for fighting against a racism rumor crisis in early 1997.
Analysis of combined data from the two cases demonstrated the causal impact
from the traditional media coverage to the number of on-line messages. In view of
second level agenda-setting theory, the tone of media stories had a slightly significant
impact on the number of negative on-line messages, too. It is assumed that the public
agenda on the USENET will correspond to the media agenda in most, if not all, corporate
crises.
If a corporation reacts to the on-line public agenda, which is influenced by the
media coverage, with proper intervention strategies, it should expect the total number of
public on-line messages to decrease along with the number of media stories about the
crisis. Figure 6-1 is an optimal model of crisis management communication for two
weeks after a crisis occurs. On day 4, media reports increase because the corporation has
taken appropriate intervention actions fast enough to contain the negative impact of the
crisis on the company. Then, hopefully, both the media as well as the public agenda will
gradually move to the "dormant" stage after day 11.

108
A
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
U nit of D ay
FIGURE 6-1. An optimal crisis communication model dealing with public agenda and
media agenda
Limitations
Despite growing and innovative uses of the new media, the Internet, the sampling
method used to download the on-line discussion messages through Deja News search
engine ran a risk of violating the content validity. Although the Deja News company
claims that it has stored hundreds of millions of articles posted on miscellaneous
newsgroups in its archives, it also implies that most spamming or flaming messages were
filtered out. If so, it is likely that some messages about the crises that were regarded as
flaming or profane by the Deja News company had been excluded from the data base.
The search keywords used in the "Power Search" function would also contribute to the
possible omission of some pertinent messages. Because numerous on-line users post
their messages to the USENET every day, the population flame becomes unknown when
it comes to sampling error.

109
In order to explore new theoretical ground work and obtain good control over the
analytical data, only two crises were selected to test the hypotheses. Both cases were of
the same type, product tampering. Small sample size may limit the reliability and
validity of the findings; caution should be taken before generalizing these results to other
types of corporate crises.
The crises selected in the study were originated from the media coverage that led
the public discussion on the USENET. Nevertheless, because people can anonymously
post messages on the Internet, a few crises could first start on this new media and then be
covered by the conventional media. Findings need to be strictly applied to those crises
first reported by the traditional media.
According to the preset criteria, the crisis was deemed as being covered by major
media and discussed by on-line users in the newsgroups for two weeks. Other corporate
crises that did not last two weeks of media coverage were missed. Those could be useful
to establish a sound crisis communication model.
The number of media sources is another limitation to analysis of the media
agenda. To add other types of media, such as electronic newspapers and magazines on
the Internet, is necessary to expand the scope of the media agenda as more and more
people surf the Internet for information from these on-line media.
The television news programs were examined by coding the textual transcripts
obtained from the Lexis/Nexis data base. To improve the reliability of media content
analysis and broaden the dimensions of the second level agenda-setting theory, it is better
to watch the live television broadcasts and analyze the crisis news items or attributes
framed in the program content.

110
Because coding categories on the tone of public opinion are self-defined, these
coding items would vary contingent upon the nature of the crisis. Development of
consistent and unified coding categories applied to different types of corporate crises will
facilitate future qualitative analysis of on-line discussion messages for both public
relations professionals and researchers.
Implications and Suggestions for Future Study
Implications
Agenda-setting theory has been tested in various fields in the social sciences. The
central idea is the one-sided influence of the media coverage of issues on public opinion
about or perception of issues. Since 1972, the media agenda has been examined through
content analysis of media reports, whereas analysis of the public agenda has relied on
opinion surveys or secondary survey data. This study adds one small piece to the
literature of public opinion by employing content analysis of public discussion messages
on the USENET.
This change in method from the traditional survey of public opinion encourages
public relations professionals to explore more about what people say and think about the
unexpected crisis they encounter. For both media professionals and academic
researchers, a variety of public discussion forums on the Internet offer an excellent
opportunity to measure the general public’s opinion about all kinds of issues. If time and
cost permit, it is suggested that both personal surveys and content analysis of the public
messages posted on the on-line discussion forums be utilized to grasp trends of public
opinion more objectively and accurately.

Ill
The crisis communication model proposed in this study was outlined from two
corporate crises focusing on the on-line public agenda integrated with the conventional
media agenda. To take another look at the content of on-line messages, many factors
might affect the trend of on-line public opinion. For instance, both cases were involved
in the governmental inspection of products or manufacturing plants. A crisis such as the
Hudson Foods case can be complex enough for people to comprehend that it will extend
the duration of the media reports as well as the public’s concerns. The extent to which a
crisis will impact the general public and society will also prolong the time line for the
public on-line discussion.
Although only public on-line dialogues were analyzed as a surrogate of the public
agenda, it was true that adoption of corporate intervention strategies is the key factor to
succeed in neutralizing or reversing people’s negative perception of the company. This
finding also corresponds with recommendations provided in previous crisis
communication literature.
The capability of rapidly responding to public concerns about a crisis expressed
on the Internet is a widely recognized strength that the Internet brings to the public
relations field. Only through sound planning for media monitoring is a public relations
program able to reduce the damages caused by the crisis. Ironically, not one message
posted in the newsgroups was found from anyone representing Mattel or Hudson Foods.
However, several messages analyzed from the Hudson Foods crisis were posted by an
employee from the Tyson Foods company, which was merged with Hudson Foods and
mistakenly treated as guilty as the Hudson Foods company. This example illustrates that
not only should public relations people monitor the public discussion forums on the

112
Internet, but more proactive steps need to be taken to clarify any doubts or
misunderstandings appearing on the Internet among on-line users.
Although on-line users discussing both crises represent only a tiny segment of the
general public, their voices should not be ignored by public relations practitioners. One
reason is because it is very likely that one person’s badmouthing on the Internet may
cause irreparable damages to the company beyond what would be expected. Another
reason is that other on-line users who did not participate in the discussion or media
reporters who looked for news information about the crisis might actually substantiate the
harmful messages and further spread them all over the Internet and other media. Thus,
public relations professionals need to remain alert for unexpected on-line crises when
dealing with potential target audiences on the Internet.
Suggestions for Future Studies
The popularity of new communication technologies provides an opportunity for
both academia and professionals to examine the relationship between the media agenda
and the dynamic public opinion expressed in the new media. Content analysis of public
discussion messages on the Internet paves the way for improving aspects of agenda¬
setting theory and sheds light on understanding the impact of this dynamic new media,
the Internet.
This research demonstrates that the traditional concept of agenda-setting theory
can be applied to the new media environment with statistically significant results as in the
Mattel case and in the analysis of combined data from both cases. Prior to this, it was
only assumed that its applicability was transferable among the mainstream media. As far
as communication theories are concerned, agenda-setting theory will serve as a

113
supplemental theoretical framework to help public relations professionals implement
effective crisis management planning for dealing with on-line public discussion.
Concerning the second level study of agenda-setting theory, a textual analysis
software, DICTION, was used to analyze the tone of media stories. Although only
numbers of negative and positive words in media stories were used in calculation of the
Optimism Score stated in Hypothesis 8, this software is still of considerable value to
examine the contents of communication texts. Given huge amounts of communication
messages or information to be examined from the Internet, it is foreseeable that the
computerized software will be broadly used in order to gain more reliability and validity
in content analysis.
For future testing of the proposed public opinion model, three suggestions are
made. First, as this research exclusively focused on examination of the USENET
dialogues as a basis of the public agenda, to expand the explanatory and predictive power
of this crisis communication model, other interactive communication channels emerging
from the new technologies for expressing public opinion should be considered along with
on-line newsgroups. The communication channels such as on-line chat rooms, local
BBSs, corporate web sites, govennnent sites to which people direct their complaints
about the crisis, and even commercial services like America Online and Microsoft
Network will help reveal to all how the public perceives the crisis event and the target
company. Public relations professionals should expand their monitoring outlets so that
they can make more precise judgements in planning the management of crisis
communication that will benefit both the corporation and the general public.

114
Second, because the impact of each corporate crisis on the society and the public
varies, duration of the public’s interests and attention differs as well. Extending this
model to other types of crises other than a product tampering crisis may shed light on
how to manage trends of public opinion while experiencing distinctive corporate crises.
Third, for theoretical researchers to compare and contrast content analysis of
public opinion with general personal surveys will verify the convergent validity of public
opinion measurement, that in turn also solidifies establishment of a public opinion model.
Although the crises in the study did not originate from the Internet like the Intel
Pentium Flaw crisis, it is advisable that issue management becomes even more important
before that issue evolves into a harmful crisis spreading over the Internet. Investigating
corporate crises originating from the interactive media instead of from the traditional
media in the future can improve understanding of the mutual influences of the traditional
and new media.
Various types of crises need to be selected to re-test the public opinion model
outlined earlier. It would also be important in selecting a local-based crisis or an
international crisis to test whether trends of public opinion about the crises are formed in
the same way as examined in this study.
The main purpose of this study was to examine the relationship between one
independent variable, the media agenda, and one dependent variable, the public agenda.
As corporate agendas are covered by the wire services, which in turn might set an agenda
for other types of media, it becomes necessary for the academics to further understand the
effects of the corporate agendas on the media agenda and the public agenda expressed on
the new media. To enrich public relations crisis management literature and to solidify

115
agenda-setting theory, researchers can incorporate information about the crisis from the
news wire services, corporate reactions publicized in the news releases, and the
governmental reactions to the public agenda to better understand the inter-relationships
between these independent variables and trends of public opinion. After all, public
opinion does not change only because of the single factor of media coverage.
Agenda-setting theory proved useful in predicting development of public opinion
about the corporate crisis expressed on the USENET through examination of the news
media coverage. Content analysis of the public agenda provides another outlook of
public opinion on the Internet. For public relations professionals, most crises happen
unexpectedly, but hopefully, this content analysis of on-line messages will offer insight
into how they should handle the crisis in terms of monitoring the interactive new media.
For the academic researchers, content analysis of the public agenda can improve the
agenda-setting theoretical framework in view of the first and second level research.
Because most findings corresponded with the ideas of agenda-setting theory, this study
should contribute to enriching the empirical as well as theoretical applicability of this
theory, at both the first and second levels, to the new media research.

APPENDIX A
HOME PAGE OF DEJA NEWS SEARCH ENGINE
In JritemfltDiKUislón
Ji> *swti IMih&Kt; Kite
rncJJT %
icik v»w"Ríasiiw
5'J;'
;Sksjsí&us*
• i&ary
'ESE
■ Jss¿ :
>?Ub usitiwni.^^ • ,.
Windows 98: Is It
Upgrade-worthy?
IC* >■ ~.V...<» *1 .VfidwC<í lo
U b'nvjvrfef.nra. caj ba.>iy won
VAndawiK’
filama ' .MStnumivnij »r-i*< ifto oissj**
Olytiiortgn ¡ f-jgilj!N.vi.U V.,.yl j ÍUxÚLiAs ¡ | «Wptl
NSW Uhi* - • AÜ I II» .ailwjnqa
.. tfK wWJj&ál • J±b.f=!i°?y?^^i - &gjS5aMtjQ«iükBua
'l|i=;r - &Mk Jfcxm ' i!ii£ UKS££y«S
d«n.a.?
Ctof'rftfri: C lí'Pí-H Pi-j i H>ir,y..(as.. All sitter
116

APPENDIX B
POWER SEARCH PAGE OF DEJA NEWS SEARCH ENGINE
IllllüÜ?
Thú In Intsmírt: fsi« nia *> «1 fai ra
ray V
Sittil, irá* «nldniB MrSeEHaa. IKi
Power Search
Slülldt fur
foasplt; uii:AN l! (3ltMTO?SM,a
s Ml!
| Fnd íiúdíUÜííysb
IMwEeifer
fteKifeikaea
AnshiVi* co-iiptfa |7|
Kcy“..¡r4í ,..¡ |71
**a-l£ií£ll3 ; *-
Exampk; éfc.iv.vhls* nr "k fees*
AulliOri,íj:_
NuaifccF'M ?t; FT]
miilíHei - “ 1 1
ftísuMí
Earmátl -
leil
Cerne
gg. li
ñubjccb!j);_
F txm¡l!r:: :l:;:;ia^l{«ijaiÉV«S.£10ni
Kv:
Sea* If
Hato
Imin!-
ixiiltlplis FAQ or (Fwipjimiv AsfeJQuwííuir)
I Tw i
Lureple ferwli Ay I I ’JJ •
OiReeonc! i—-0 People Tjaffii 1 Cl*wj|ki!s I Vdlim Piwra, | Bieislc
YUlir. Uuimin ~V*mr
Mnv i ':<■:■ Afeué üfju Nc»í • Aó mfe • O-¡ AOvngtaers
PrcoWcbEmiií Lilli: lo lisia feí' C'fW Yarr OwnjfeEi
Heme • Sí?.*** • • Fu-i • tdxiikÚJtel Itólfi Mowráws
sÉÍg£:
Copyright 9 iV9i-9i reals Mciw. la:. a:I lipjii» wuurf Óaidjijiaifr lifwv
117

APPENDIX C
CODING SCHEME FOR NEWS STORIES
News story identification #: #
Media source:
â–¡ ABC News â–¡ CNN â–¡ NBC News â–¡ New York Times
â–¡ Wall Street Journal â–¡ Washington Post
Crisis target audience:
â–¡ employees â–¡ customers â–¡ stockholders â–¡ all publics
â–¡ community members
News story dated:
Month
Day
Year
Intervention dated:
Month
Day
Year
Intervention type(s)
1. evacuation of employees, particular publics involved in the crisis for the sake of
safety,
2. hosting news conference to admit or disclose happening of the crisis,
3. setting up hot lines to answer requests from media and publics,
4. provision of medical services to the injured or dead people,
5. preparation or ready for negotiation with interest groups or specific publics
concerning the crisis,
6. product recalls or destroy,
7. decision to enforce laws to combat the crisis,
8. offer of monetary or product compensation,
118

119
9. making public apology,
10. acknowledgement of wrongdoing,
11. replacement of persons who causes the crisis, or
12. promise to preventing from occurrence of the similar crisis.

APPENDIX D
CODING SCHEME FOR ON-LINE MESSAGES
On-line message identification #:
On-line message dated:
Target of flaming message:
User’s strategy: □ TALKER
User’s name in the e-mail:
Tone of on-line message:
□ Negative— Why?
a) request legal action against the company,
b) request compensation from the company,
c) mention of personal or others experiences suffering from the crisis event,
d) urge the boycott of the company’s product or service,
e) give sarcastic jokes or harsh comments about the company, or
f) blame or accuse the company for its wrong doing or its defected product(s).
#
Month Day Year
â–¡ Government â–¡ Company
â–¡ Other Users â–¡ Some or all of above
â–¡ Others
â–¡ REACTOR â–¡ Unidentified
120

121
□ Positive— Why?
a) urge end of a boycott,
b) compliment the company,
c) mention the laws as unfair to the organization,
d) praise company’s intervening actions, or
e) deny that the crisis event is solely due to the company.
□ Neutral— Why?
a) forwarded media reports without expressing opinion,
b) forwarded website information without expressing opinion,
c) personal questions or responses to others’ messages without expressing any opinions,
or
d) statements that take the crisis issue for granted without blaming or praising the
company.
â–¡ Non-applicable
User’s identification: □ Company’s employee
□ Company’s management
â–¡ Other occupation
â–¡ Unable to identify

APPENDIX E
SAMPLE OF REPORT FILE IN DICTION SOFTWARE
TOTAL NUMBER OF WORDS ANALYZED = 26
NUMBER OF CHARACTERS IN PASSAGE =158
CHARACTERS PER WORD = 4.62
NUMBER OF DIFFERENT WORDS = 24
ALPHA-NUMERIC IDENTIFIERS:
00000000
DESCRIPTIVE IDENTIFIER:

INPUT TEXT
Geez, Chuckie LIVES!! Actually, can you imagine how frightening
something like that would be to a little kid. Attacked by a doll
that eats it's owners?!!
* This passage contains fewer than 500 words. *
* Because DICTION'S norms are based on *
* 500-word samples, the raw scores reported *
* here cannot be compared to the nonnal range *
* of scores provided. *
DICTIONARY TOTALS
VARIABLE
FREQ.
(FREQ/26)
PER CENT
NORMAL RANGE
OF SCORES (#)
STANDARD¬
IZED SCORES
Numerical Terms
0.00
0.00
0.30- 15.04
-1.04 *
Ambivalence
2.00
7.69
6.49- 19.21
-1.71 *
Self-reference
0.00
0.00
0.00- 15.10
-0.86
122

123
Tenacity
2.00
7.69
23.32- 39.76
-3.59
He
Leveling
1.00
3.85
5.02- 12.76
-2.04
*
Collectives
0.00
0.00
4.04- 14.46
-1.78
=1=
Praise
0.00
0.00
2.77- 9.59
-1.81
*
Satisfaction
1.00
3.85
0.47- 6.09
-0.81
Inspiration
0.00
0.00
1.56- 11.10
-1.33
*
Blame
1.00
3.85
0.06- 4.16
-0.54
Hardship
0.00
0.00
1.26- 10.48
-1.27
*
Aggression
1.00
3.85
1.07- 9.79
-1.02
*
Accomplishment
0.00
0.00
4.96- 23.78
-1.53
*
Communication
0.00
0.00
2.21- 11.79
-1.46
*
Cognitive Terms
1.00
3.85
4.43- 14.27
-1.70
*
Passivity
0.00
0.00
2.10- 8.08
-1.70
*
Spatial Awareness 0.00
0.00
4.17- 19.85
-1.53
*
Familiarity
6.00
23.08
117.87-147.19
-8.63
*
Temporal Awareness 0.00
0.00
8.36-21.82
-2.24
*
Present Concern
1.83
7.05
7.02- 16.60
-2.08
*
Human Interest
1.65
6.35
18.13- 45.49
-2.20
*
Concreteness
1.00
3.85
10.70-28.50
-2.09
*
Past Concern
1.00
3.85
0.97- 6.19
-0.99
Centrality
0.00
0.00
1.19- 7.54
-1.37
=1=
Rapport
0.00
0.00
0.42- 4.26
-1.22
*

124
Cooperation
0.00
0.00
0.36- 8.44
-1.09
*
Diversity
0.00
0.00
0.07- 3.81
-1.04
*
Exclusion
0.00
0.00
0.00- 4.31
-0.99
Liberation
0.00
0.00
0.00- 4.72
-0.82
Denial
0.00
0.00
2.57- 10.35
-1.66
*
Motion
0.00
0.00
0.17- 4.35
-1.08
*
USER DICTIONARY TOTALS
DICTIONARY NAME FREQUENCY

WORDS DESIGNATED TO COMPOSE INSISTENCE SCORE: None.
CALCULATED
VARIABLES
SCORE
NORMAL RANGE
OF SCORES (#)
STANDARD¬
IZED SCORES
Insistence
0.00
6.71- 79.67
-1.18 *
Embellishment
0.52
0.18- 1.10
-0.26
Variety
0.92
0.45- 0.53
10.83 *
Complexity
4.62
4.31- 5.01
0.02
COMPOSITE
VARIABLES
SCORE
Activity
48.58
47.18- 52.46
Optimism
49.53
46.90- 54.00
Certainty
34.19
47.48- 53.50 *
Realism
32.19
48.59- 54.37 *
Commonality
49.17
47.80- 53.42

REFERENCES
(1997). Hilfiger denies Internet accusations of racism : Associated Press.
Ball-Rokeach, S. J., & Reardon, K. (1988). Monologue, dialogue, and telelog. In R. P.
Hawkins, J. M. Wiemann, & S. Pingree (Eds.), Advancing Communication
Science: Merging Mass and Interpersonal Processes (pp. 135-161). Newbury Park,
CA.: Sage.
Barton, L. (1993). Crisis in Organizations. Cincinnati, OH.: South-Western Publishing
Co.
Becker, L., & McCombs, M. (1978). The role of the press in determining voter reactions
to presidential primaries. Human Communication Research. 4. 301-307.
Behr, R. L., & Iyengar, S. (1985). Television news, real-world cues, and changes in the
public agenda. Public Opinion Quarterly. 49(1). 38-57.
Beniger, J. R. (1978). Media content as social indicators: The Greenfield Index of
agenda-setting. Communication Research. 5(4), 437-451.
Berelson, B. (1952). Content Analysis in Communications Research. New York, NY.:
Free Press.
Berger, J. (1995, December 30). Critics of traditional media are flocking to the Web.
Editor & Publisher. 128. 40.
Birch, J. (1994). New factors in crisis planning and response. Public Relations Quarterly,
39(1), 31-34.
Bobbitt, R. (1995). An Internet primer for public relations. Public Relations Quarterly,
40(3), 27-32.
Brosius, H.-B., & Kepplinger, H. M. (1990). The agenda-setting function of television
news: Static and dynamic views. Communication Research. 17(2). 183-211.
Buey, E. P. (1995, May 29). Where interactivity resides: An agenda for 'new media'
research in political campaign. Paper presented at the International
Communication Association, Albuquerque, NM.
125

126
Carney, A., & Jorden, A. (1993). Prepare for business-related crises. Public Relations
Journal, 49(8). 34-35.
Cohen, D. (1975, August). A report on a non-election agenda-setting study. Paper
presented at the Association for Education in Journalism, Ottawa, Canada.
D'Aveni, R. A., & MacMillan, 1. C. (1990). Crisis and the content of managerial
communications: A study of the focus of attention of top managers in surviving
and failing firms. Administrative Science Quarterly, 35141, 634-657.
Dearing, J. W., & Rogers, E. M. (1996). Agenda-Setting. Thousand Oaks, CA.: Sage
Publications.
Dutton, J. E. (1986). The processing of crisis and non-crisis strategic issues. Journal of
Management Studies. 2315'). 501-517.
Dyer, S. C., Jr., Miller, M. M., & Boone, J. (1991). Wire service coverage of the Exxon
Valdez crisis. Public Relations Review, 17(1), 27-36.
Eaton, H. (1989). Agenda-setting with bi-weekly data on content of three national media.
Journalism Quarterly, 66, 942-8, 959.
Egelhoff, W. G., & Sen, F. (1992). An information-processing model of crisis
management. Management Communication Quarterly, 5(4), 443-484.
Entman, R. (1993). Framing: Toward clarification of a fractured paradigm. Journal of
Communication, 43(4), 51-58.
Eyal, C. H. (1979). Time-frame in agenda-setting research: A study of the conceptual and
methodological factors affecting the time frame context of the agenda-setting
process. Doctoral dissertation, Syracuse University, Syracuse, NY.
Feam-Banks, K. (1996). Crisis Communications: A Casebook Approach. Mahwah, NJ:
Lawrence Erlbaum Associates.
Fitzpatrick, K. R. (1995). Ten guidelines for reducing legal risks in crisis management.
Public Relations Quarterly, 40(21, 33-38.
Fombrun, C. J. (1996). The reputational audit, Reputation: Realizing Value from the
Corporate Image (pp. 192-209). Boston, MA: Harvard Business School Press.
Funkhouser, G. R. (1973). The issues of the sixties: An exploratory study in the dynamics
of public opinion. Public Opinion Quarterly, 37, 62-75.

127
Garramone, G. M., Harris, A. C., bulletin boards. Journal of Broadcasting & Electronic Media, 30(3). 325-339.
Gaunt, P., & Ollenburger, J. (1995). Issues management revisited: A tool that deserves
another look. Public Relations Review. 21(3). 199-210.
Ghanem, S (1997). Filling in the tapestry: The second level of agenda setting. In M.
McCombs, D. L. Shaw, & D. Weaver (Eds.), Communication and Democracy:
Exploring the Intellectual Frontiers in Agenda-Setting Theory (pp. 3-14).
Mahwah, NJ: Lawrence Erlbaum Associates, Inc.
Gold, D., & Simmons, J. L. (1965). News selection patterns among Iowa dailies. Public
Opinion Quarterly, 29(3), 425-430.
González-Herrero, A., & Pratt, C. B. (1995). How to manage a crisis before- or
whenever- it hits. Public Relations Quarterly, 40(1), 25-29.
Groper, R. (1996, August/September). Political behavior of actors on electronic
discussion groups. Paper presented at the 92nd Annual Meeting of American
Political Science Association, San Francisco, CA.
Grunig, J. E. (1989). Symmetrical Presuppositions as a Framework for Public Relations
Theory. In C. H. Botan «fe V. Hazleton Jr. (Eds.), Public Relations Theory (pp. 17-
44). Hillsdale, NJ.: Lawrence Erlbaum Associates Inc.
Hainsworth, B. (1990). The distribution of advantages and disadvantages. Public
Relations Review, 16(1), 33-39.
Hart, R. P. (1997). DICTION 4.0: The Text-Analysis Program. Thousand Oaks, CA.:
Sage Publications Software.
Hayes, R. E. (1985). Corporate crisis management as adaptive control. In S. J. Andriole
(Ed.), Corporate Crisis Management (pp. 316). Princeton, NJ: Petrocelli.
Hearit, K. M. (1994). Apologies and public relations crises at Chrysler, Toshiba, and
Volvo. Public Relations Review, 20(2), 113-125.
Hertog, J. K., case of HIV transmission. Communication Research, 22(5), 545-574.
Hiebert, R. E. (1992). Global public relations in a post-communist world: A new model.
Public Relations Review. 18(2), 117-126.
Hill, K. A., USENET and political communication. Political Communication, 14(1), 3-27.

128
Hoffman, P. E. (1994). Internet Instant Reference. Alameda, CA: SYBEX Inc.
Irvine, R. B., & Millar, D. P. (1996, August 8). Debunking the stereotypes of crisis
management: The nature of business crises in the 1990's. Paper presented at the
5th Annual Conference on Crisis Management, Las Vegas, NV.
Ito, Y. (1993). New directions in communication research from a Japanese perspective. In
P. Gaunt (Ed.), Beyond Agendas: New Directions in Communication Research
(pp. 119-135). Westport, CT: Greenwood Press.
Kalish, J. (1997). P.R. firms surf the Net. Reuters North American Wire.
King, P.-t. (1997). The press, candidate images, and voter perceptions. In M. McCombs,
D. L. Shaw, & D. Weaver (Eds.), Communication and Democracy: Exploring the
Intellectual Frontiers in Agenda-Setting Theory (pp. 29-40). Mahwah, NJ:
Lawrence Erlbaum Associates, Inc.
Klapper, J. T. (1960). The Effects of Mass Communication. New York: NY: Free Press.
Kotcher, R. L. (1992). The technological revolution has transformed crisis
communication. Public Relations Quarterly, 37(3). 19-21.
Krippendorff, K. (1980). Content Analysis: An Introduction to its Methodology. Beverly
Hills, London: Sage Publications.
Lerbinger, O. (1997). Part I: Communicating in an era of crises, The Crisis Manager:
Facing Risk and Responsibility (pp. 1-54). Mahwah, NJ: Lawrence Erlbaum
Associates.
MacKinnon, R. C. (1995). Searching for the Leviathan in Usenet. In S. G. Jones (Ed.),
CyberSociety: Computer-Mediated Communication and Community (pp. 112-
137). Thousand Oaks, CA: Sage Publications.
MacKuen, M. (1990). Speaking of politics: Individual conversational choice, public
opinion, and the prospects for deliberative democracy. In J. A. Ferejohn & J. H.
Kuklinski (Eds.), Information and Democratic Processes (pp. 59-99). Urbana, IL.:
University of Illinois Press.
Maggart, L. (1994). Bowater incorporated- a lesson in crisis communication. Public
Relations Quarterly, 39(3), 29-31.
Malier, M. (1995). Media framing and salience of the population issue: A multi-method
approach. Doctoral dissertation, University of Texas, Austin.

129
Mayer, R. N. (1991). Gone yesterday, here today: Consumer issues in the agenda-setting
process. Journal of Social Issues, 47(1), 21-39.
Mazur, A. (1987). Putting radon on the public's risk agenda. Science. Technology, and
Human Values. 12(3-4). 86-93.
McCombs, M. (1997,). New frontiers in agenda setting: Agendas of attributes and
frames. Paper presented at the Association for Education in Journalism and Mass
Communication, Chicago, IL.
McCombs, M., & Shaw, D. (1972). The agenda-setting function of mass media. Public
Opinion Quarterly. 36. 176-187.
McCombs, M., Danielian, L., & Wanta, W. (1995). Issues in the news and public agenda:
The agenda-setting tradition. In T. L. Glasser & C. T. Salmon (Eds.), Public
Opinion and the Communication of Consent (pp. 301-322). New York City, NY:
The Guilford Press.
McCombs, M., Llamas, J. P., Lopez-Escobar, E., & Rey, F. (1998). Candidate images in
Spanish elections: Second-level agenda-setting effects. Journalism & Mass
Communication Quarterly. 74(4), 703-717.
McLaughlin, M. L., Osborne, K. K., & Smith, C. B. (1995). Standards of conduct on
Usenet. In S. G. Jones (Ed.), CyberSociety: Computer-Mediated Communication
and Community (pp. 90-111). Thousand Oaks, CA: Sage Publications.
Meng, M. (1992). Issue life cycle has five stages. Public Relations Journal, 48(3), 23.
Mitchell, T. H. (1986). Coping with a corporate crisis. Canadian Business Review, 13,
17-20.
Mitroff, 1.1., Pearson, C. M., & Harrington, K. L. (1996). The Essential Guide to
Managing Corporate Crises: A Step-By-Step Handbook for Surviving Major
Catastrophes. Oxford, NY: Oxford University Press.
Morris, M., & Ogan, C. (1996). The Internet as mass medium. Journal of
Communication, 46(1), 39-50.
Mullins, E. L. (1977). Agenda-setting and the young voter. In D. L. Shaw & M. E.
McCombs (Eds.), The Emergence of American Political Issues: The Agenda-
Function of the Press (pp. 133-148). St. Paul, MN: West Publishing Co.
Ogan, C. (1993). Listserver communication during the Gulf War: what kind of medium is
the electronic bulletin board? Journal of Broadcasting & Electronic Media, 37121,
177-196.

130
Price, V. (1992). Communication Concepts 4: Public Opinion. Newbury Park, CA: Sage.
Priest, S. H. (1998). Public opinion, expert opinion, and the illusion of consensus:
Gleaning points of view electronically. In D. L. Borden & K. Harvey (Eds.), The
Electronic Grapevine: Rumor. Reputation, and Reporting in the New On-Line
Environment (pp. 23-29). Mahwah, NJ: Lawrence Erlbaum Associates.
Rafaeli, S. (1986). The electronic bulletin board: A computer-driven mass medium.
Computers and the Social Sciences. 2(3). 123-136.
Rafaeli, S. (1988). Interactivity: From new media to communication. In R. Hawkins, J.
Wiemann, & S. Pingree (Eds.), Advancing Communication Science: Merging
Mass and Interpersonal Processes (pp. 110-134). Newbury Park, CA: Sage.
Rafaeli, S., & LaRose, R. J. (1993). Electronic bulletin boards and "public goods"
explanations of collaborative mass media. Communication Research. 20(2). 277-
297.
Reese, S. D., & Danielian, L. H. (1989). A closer look at intermedia influences on agenda
setting: The cocaine issue of 1986. In P. J. Shoemaker (Ed.), Communication
Campaigns about Drugs: Government. Media, and the Public (pp. 47-66).
Hillsdale, NJ: Lawrence Erlbaum Associates.
Reese, S. D., & Danielian, L. H. (1989). Intermedia influence and the drug issue:
Converging on cocaine. In P. J. Shoemaker (Ed.), Communication Campaigns
about Drugs: Government, Media, and the Public (pp. 29-45). Hillsdale, NJ:
Lawrence Erlbaum Associates.
Revah, S. (1995). It's a jungle out there in cyberspace. American Journalism Review,
17(2), 10-11.
Roberts, M. (1997). Political advertising's influence on news, the public and their
behavior. In M. McCombs, D. L. Shaw, & D. Weaver (Eds.), Communication and
Democracy: Exploring the Intellectual Frontiers in Agenda-Setting Theory (pp.
85-96). Mahwah, NJ: Lawrence Erlbaum Associates, Inc.
Rogers, E. M., & Dearing, J. W. (1988). Agenda-setting research: Where has it been?
Where is it going? In J. A. Anderson (Ed.), Communication Yearbook 11 (pp.
555-594). Newbury Park: Sage.
Salwen, M. B. (1988). Effects of accumulation of coverage on issue salience in agenda¬
setting. Journalism Quarterly, 65. 100-106.
Scott, W. A. (1955). Reliability of content analysis: The case of nominal scale coding.
Public Opinion Quarterly. 19(3L 321-325.

131
Shaw, C. (1988, April 3). The AP: It's everywhere and powerful. Los Angeles Times.
pp. 1.
Shaw, D. L., & McCombs, M. E. (1977). The Emergence of American Political Issues:
The Agenda-Setting Function of the Press. St. Paul, MN: West Publishing.
Shoemaker, P. J., Wanta, W., & Leggett, D. (1989). Drug coverage and public opinion. In
P. J. Shoemaker (Ed.), Communication Campaigns about Drugs: Government.
Media, and the Public (pp. 67-80). Hillsdale, NJ: Lawrence Erlbaum Associates.
Sigelman, L., Lebovic, J., Wilcox, C., & Allsop, D. (1993). As time goes by: Daily
opinion change during the Persian Gulf Crisis. Political Communication, 10. 353-
367.
Sohn, A. B. (1978). A longitudinal analysis of local non-political agenda-setting effects.
Journalism Quarterly, 55, 325-333.
Stanton, A. (1995, Feb 20). Pentium brouhaha a marketing lesson: Better
communications would have helped. Advertising Age, 66, 18.
Starbuck, W. H., Greve, A., & Hedberg, B. L. T. (1978). Responding to crisis. Journal of
Business Administration, 9(2), 112-137.
Stempel m, G. H. (1964). How newspapers use the Associated Press afternoon A-Wire.
Journalism Quarterly, 41, 380-384.
Stempel III, G. H., & Westley, B. H. (1989). Research Methods in Mass Communication.
(2nd ed.). Englewood Cliffs, NJ.: Prentice Hall.
Stone, G. C., & McCombs, M. E. (1981). Tracing the time lag in agenda-setting.
Journalism Quarterly, 58. 51-55.
Strenski, J. B. (1995). The ethics of manipulated communication. Public Relations
Quarterly, 40(3), 33-35.
Sturges, D. L. (1994). Communicating through crisis: A strategy for organizational
survival. Management Communication Quarterly, 7(3). 297-316.
Takeshita, T., & Mikami, S. (1995). How did the mass media influence the voter's choice
in the 1993 general election in Japan? A study of agenda-setting. Keio
Communication Review, 17, 27-41.
Terry, K. E. (1989). Educator and practitioner differences on the role of theory in public
relations. In C. H. Botan & V. Hazleton, Jr. (Eds.), Public Relations Theory (pp.
281-298). Hillsdale, NJ.: Lawrence Erlbaum Associates.

132
Thomsen, S. R. (1996). "@Work in cyberspace": Exploring practitioner use of the
PRForum. Public Relations Review, 22(2), 115-131.
Traverso, D. K. (1992). Opening a credible dialogue with your community. Public
Relations Journal., 48(8), 30-32.
Umansky, D. (1993). How to survive and prosper when it hits the fan. Public Relations
Quarterly, 38(4), 32-34.
Vendrell, I. B. (1993). Oil spills show lessons still not learned. Public Relations Journal,
49(12), 40-42.
Wanta, W., & Wu, Y.-C. (1992). Interpersonal communication and the agenda-setting
process. Journalism Quarterly, 69(4), 847-849.
Wanta, W., & Hu, Y.-W. (1993). The agenda-setting effects of international news
coverage: An examination of differing news frames. International Journal of
Public Opinion Research, 5(3), 250-263.
Wanta, W., & Hu, Y.-W. (1994). Time-lag differences in the agenda-setting process: An
examination of five news media. International Journal of Public Opinion
Research, 6(3), 225-240.
Weaver, D., Graber, D., McCombs, M., & Eyal, C. (1981). Media Agenda Setting in a
Presidential Election: Issues, Images and Interest. New York, NY.: Praeger.
Webber, R. P. (1990). Basic Content Analysis. (2nd ed.). Newbury Park, CA: Sage
Publications.
Weick, K. E. (1988). Enacted sensemaking in crisis situations. Journal of Management
Studies, 25(4), 305-317.
Weise, E. (1998, April 16). Net use doubling every 100 days. USA Today.
White, D. M. (1949). The "Gate Keeper": A case study in the selection of news.
Journalism Quarterly, 27, 383-390.
Whitney, D. C., & Becker, L. B. (1982). 'Keeping the Gates' for gatekeepers: The effects
of wire news. Journalism Quarterly, 59, 60-65.
Wilcox, D. L., Ault, P. H., & Agee, W. K. (1992). Public Relations: Strategies and
Tactics. (3rd ed.). New York, NY: HarperCollins.

133
Willnat, L., & Zhu, J.-H. (1996). Newspaper coverage and public opinion in Hong Kong:
A time-series analysis of media priming. Political Communication. 13(2'). 231-
246.
Winter, J. P., & Eyal, C. H. (1981). Agenda setting for the civil rights issue. Public
Opinion Quarterly, 4563^), 376-383.
Woods, G. B. (1996, June 10). It could happen to you: A framework for understanding
and managing crises involving race. Paper presented at the Public Relations
Society of America Educators Section.
Zhu, J.-H., Watt, J. H., Snyder, L. B., Yan, J., & Jiang, Y. (1993). Public issue priority
formation: Media agenda-setting and social interaction. Journal of
Communication. 43(1), 8-30.
Zucker, H. G. (1978). The variable nature of news media influence. In B. D. Ruben (Ed.),
Communication Yearbook (Vol. 2, pp. 225-240). New Brunswick: Sage
Publications.

BIOGRAPHICAL SKETCH
Tzong-Homg Dzwo was bom in a southern city of Taiwan, on October 25, 1964,
and has lived in Taiwan nearly 30 years before coming to the United States for graduate
study. He received a bachelor’s degree in British and American Literature from Tung-hai
University in Tai-Chung City of Taiwan. After two-year military service from 1986 to
1988, he worked as a sales person and account executive for a domestic shipping
company for two and a half years.
Gaining three years of working experiences in Taiwan, in 1992 Dzwo came to the
United States to pursue his master’s degree in mass communication with a concentration
in public relations at the University of Georgia. In the fall of 1994, Dzwo continued the
doctoral studies at the University of Florida, where he began to focus on applied research
of new media in the public relations field. For three years since 1994, he worked as a
research assistant as well as teaching assistant to help students leam more about computer
programs and statistics related to mass communication research. Awarded the academic
achievement for the international students twice while at the University of Florida, Dzwo
will return to his mother land, Taiwan, to assume a teaching position in August 1998.
Dzwo, his wife and their three kids will live in Taiwan after August 1998 and can
be reached at (886) 8-766-2674, #1-1, Lane 293, Chung-wu Rd., Ping-Tung City,
Taiwan.
134

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
d
)hn C. Sutherland, Chair
3rofessor of Journalism and
Communications
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Marilyn S. ‘Roberts
Associate Professor of Journalism and
Communications
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Gail Baker
Associate Professor of Journalism and
Communications
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Michael D. Martinez |
Associate Professor of
Political Science

This dissertation was submitted to the Graduate Faculty of the College of
Journalism and Communications and to the Graduate School and was accepted as partial
fulfillment of the requirements for the degree of Doctor of Philosophy.
August 1998
Dean, College of Jourhalism and
Communications
Dean, Graduate School



xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E80YDUR22_YAMUNX INGEST_TIME 2014-08-27T20:15:15Z PACKAGE AA00024975_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES



PAGE 1

&25325$7( &5,6(6 ,1 &<%(563$&( (;7(1',1* 38%/,& 5(/$7,216 0(',$ 021,725,1* 72 7+( 38%/,& ',$/2*8(6 21 7+( 86(1(7 %\ 7=21*+251* '=:2 $ ',66(57$7,21 35(6(17(' 72 7+( *5$'8$7( 6&+22/ 2) 7+( 81,9(56,7< 2) )/25,'$ ,1 3$57,$/ )8/),//0(17 2) 7+( 5(48,5(0(176 )25 7+( '(*5(( 2) '2&725 2) 3+,/2623+< 81,9(56,7< 2) )/25,'$

PAGE 2

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

PAGE 3

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

PAGE 4

7$%/( 2) &217(176 SDJH $&.12:/('*0(176 LL /,67 2) 7$%/(6 YL /,67 2) ),*85(6 YLLL $%675$&7 L[ &+$37(56 ,1752'8&7,21 2YHUYLHZ 3UREOHP 'HILQHG 5HVHDUFK 3XUSRVH 'LVVHUWDWLRQ 2XWOLQH + /,7(5$785( 5(9,(: 5HYLHZ RI $JHQGDVHWWLQJ 7KHRU\ 7KH ,QWHUQHW $ %ULHI +LVWRU\ )XQFWLRQV DQG &XUUHQW 6LWXDWLRQ 7KH &KDOOHQJH RI 0RQLWRULQJ WKH ,QWHUQHW 7KH ,QWHUQHW 5HVHDUFK P &5,6,6 &20081,&$7,21 /,7(5$785( 2YHUYLHZ &ULVLV 0DQDJHPHQW 5HVHDUFK DQG 3XEOLF 2SLQLRQ 6WDJHVf 3XEOLF 2SLQLRQ 0RGHO RI &ULVLV 0DQDJHPHQW 0DF.XHQfV 6WUDWHJ\ 0RGHO RI 3XEOLF 'LDORJXH 5HVHDUFK 4XHVWLRQV 5HVHDUFK +\SRWKHVHV ,9 0(7+2'2/2*< 2YHUYLHZ &RUSRUDWH &ULVLV 6HOHFWLRQ LY

PAGE 5

1HZV &RYHUDJH $QDO\VLV 0HGLD 6HOHFWLRQ IRU $QDO\VLV RI 1HZV 6WRULHV 2QOLQH 'LVFXVVLRQ 0HVVDJH $QDO\VLV 'HILQLWLRQV RI &RGLQJ &DWHJRULHV &RQFHSWXDO 'HILQLWLRQV 2SHUDWLRQDO 'HILQLWLRQV ,QGHSHQGHQW DQG 'HSHQGHQW 9DULDEOHV ,QWURGXFWLRQ RI &RPSXWHU 6RIWZDUH ',&7,21 9 5(68/76 'HVFULSWLYH $QDO\VLV 0DWWHO &DEEDJH 3DWFK 'ROO 'HIHFW &DVH 7HVW RI +\SRWKHVHV +XGVRQ )RRGV 0HDW &RQWDPLQDWLRQ &DVH 7HVW RI +\SRWKHVHV 3RVW +RF $QDO\VLV 9, ',6&866,21 6XPPDWLRQ RI )LQGLQJV 5HVHDUFK 4XHVWLRQV 3URSRVHG &ULVLV &RPPXQLFDWLRQ 0RGHO /LPLWDWLRQV ,PSOLFDWLRQV DQG 6XJJHVWLRQV IRU )XWXUH 6WXG\ $33(1',&(6 $ +20( 3$*( 2) '(-$ 1(:6 6($5&+ (1*,1( % 32:(5 6($5&+ 3$*( 2) '(-$ 1(:6 6($5&+ (1*,1( & &2',1* 6&+(0( )25 1(:6 6725,(6 &2',1* 6&+(0( )25 21/,1( 0(66$*(6 ( 6$03/( 2) 5(3257 ),/( ,1 ',&7,21 62)7:$5( 5()(5(1&(6 %,2*5$3+,&$/ 6.(7&+ Y

PAGE 6

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

PAGE 7

5HVXOWV RI +\SRWKHVLV WHVW LQ WKH +XGVRQ )RRGV FDVH 5HVXOWV RI FRPSDULVRQ RI WZR VWDJHV LQ +\SRWKHVLV LQ WKH +XGVRQ )RRGV &DVH 5HVXOWV RI FRPSDULVRQ RI WZR VWDJHV LQ +\SRWKHVLV LQ WKH +XGVRQ )RRGV &DVH 5HVXOWV RI +\SRWKHVLV WHVW LQ WKH +XGVRQ )RRGV FDVH 5HVXOWV RI &KLVTXDUH WHVW ZLWK FRPELQHG GDWD 9OO

PAGE 8

/,67 2) ),*85(6 )LJXUH SDJH *URXS RSLQLRQ IRUPDWLRQ SURFHVV 6WXUJHVn PRGHO RI SXEOLF RSLQLRQ ZLWK FULVLV PDQDJHPHQW SODQ &DXVDO UHODWLRQVKLS EHWZHHQ WKH PHGLD DJHQGD DQG WKH SXEOLF DJHQGD %LYDULDWH FRUUHODWLRQ EHWZHHQ WKH QXPEHU RI QHJDWLYH RQOLQH PHVVDJHV DQG WKH 2SWLPLVP 6FRUH LQ WKH 0DWWHO FDVH %LYDULDWH FRUUHODWLRQ EHWZHHQ WKH QXPEHU RI QHJDWLYH RQOLQH PHVVDJHV DQG WKH 2SWLPLVP 6FRUH LQ WKH +XGVRQ )RRGV FDVH 5HODWLRQVKLS EHWZHHQ WKH PHGLD DJHQGD DQG WKH SXEOLF DJHQGD LQ WKH 0DWWHO FDVH 5HODWLRQVKLS EHWZHHQ WKH PHGLD DJHQGD DQG WKH SXEOLF DJHQGD LQ WKH +XGVRQ )RRGV FDVH &XELF UHODWLRQVKLS EHWZHHQ WKH PHGLD DJHQGD DQG WKH SXEOLF DJHQGD LQ WKH +XGVRQ )RRGV FDVH 'LVWULEXWLRQ RI QXPEHUV RI RQOLQH PHVVDJHV E\ W\SHV LQ WKH 0DWWHO FDVH 'LVWULEXWLRQ RI QXPEHUV RI RQOLQH PHVVDJHV E\ W\SHV LQ WKH +XGVRQ )RRGV FDVH )UHTXHQFLHV RI FRPELQHG RQOLQH PHVVDJHV DQG PHGLD VWRULHV )UHTXHQFLHV RI FRPELQHG QHJDWLYH RQOLQH PHVVDJHV DQG WKH 2SWLPLVP 6FRUHV LQ PHGLD VWRULHV $Q RSWLPDO FULVLV FRPPXQLFDWLRQ PRGHO GHDOLQJ ZLWK SXEOLF DJHQGD DQG PHGLD DJHQGD YLLL

PAGE 9

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

PAGE 10

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

PAGE 11

&+$37(5 ,1752'8&7,21 2YHUYLHZ 7KH WRSLF RI PHGLD HIIHFWV KDV EHHQ D SULPDU\ UHVHDUFK GRPDLQ IRU RYHU WKUHH GHFDGHV $PRQJ WKH PDQ\ WKHRUHWLFDO DSSURDFKHV WKDW KDYH EHHQ SURSRVHG RQH RI WKH PRVW SRSXODU KDV EHHQ DJHQGDVHWWLQJ WKHRU\ +XQGUHGV RI VWXGLHV KDYH XWLOL]HG DVSHFWV RI WKH DJHQGDVHWWLQJ IUDPHZRUN WR UHYHDO SRVLWLYH FRUUHODWLRQV EHWZHHQ PHGLD FRYHUDJH RI LVVXHV DQG WKHLU GLVWLQFWLYH LPSRUWDQFH LQ SHRSOHnV PLQGV 7KH GHYHORSPHQW RI DJHQGDVHWWLQJ WKHRU\ FDQ EH XQGHUVWRRG IURP WZR OHYHOV RI UHVHDUFK 7KH ILUVW OHYHO VWXG\ FDQ EH WUDFHG EDFN WR ZKHQ 0F&RPEV DQG 6KDZ ZDQWHG WR H[SODLQ KRZ DQG ZK\ SHRSOH WKLQN DERXW GLIIHUHQW VRFLDO LVVXHV DQG UDQN WKHLU LPSRUWDQFH 7KH ILQGLQJV VKRZHG D VLJQLILFDQW FRUUHODWLRQ EHWZHHQ SHRSOHnV UDQNLQJ RI LVVXHVn LPSRUWDQFH LQ WKHLU PLQGV DQG WKH DPRXQW RI PHGLD FRYHUDJH KL RWKHU ZRUGV WKH VDOLHQFH RI LVVXHV LQ SHRSOHfV PLQGV LV JUHDWO\ LQIOXHQFHG E\ DQG WUDQVPLWWHG IURP WKH PHGLD FRYHUDJH %HVLGHV WKH RULJLQDO IRFXV RQ WKH LVVXH VDOLHQFH WUDQVPLWWHG IURP PHGLD WR WKH SXEOLFfV PLQG PHGLD HIIHFWV LQ YLHZ RI DJHQGDVHWWLQJ WKHRU\ KDYH DOVR EHHQ IRXQG IURP DQRWKHU SHUVSHFWLYH 7KH VHOHFWLRQ RI REMHFWV DQG DWWULEXWHV RI DQ LVVXH LQ WKH QHZV PHGLD EHFRPHV D PDMRU FRQFHUQ +RZ QHZV IUDPHV DQG DJHQGD DWWULEXWHV LPSDFW WKH SXEOLF DJHQGD SOD\V D FHQWUDO UROH LQ WKH VHFRQG OHYHO VWXG\ RI DJHQGDVHWWLQJ VWXGLHV 0F&RPEV

PAGE 12

f 7KH UHVHDUFK TXHVWLRQ LQ WKH VHFRQG OHYHO KDV PRYHG IURP f'RHV WKH DPRXQW RI PHGLD FRYHUDJH WUDQVPLW WKH LVVXH VDOLHQFH WR SHRSOHfV PLQGV"f WR f7R ZKDW H[WHQW LV WKH SXEOLFnV YLHZ RI DQ LVVXH VKDSHG E\ WKH PHGLD WKDW IUDPH REMHFWV RU DWWULEXWHV RI WKH LVVXH LQ GLIIHUHQW ZD\V"f 7R IXUWKHU XQGHUVWDQG WKH PHGLD HIIHFWV RQ SHRSOHfV LQWHUSUHWDWLRQ RI VRFLDO LVVXHVf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n VHWWLQJ UHVHDUFK WKDW PHDVXUHV WKH PHGLD DJHQGD IURP GLIIHUHQW DVSHFWV WKH PRVW FRPPRQ PHWKRGRORJ\ LV WKH VXUYH\ )RU H[DPSOH UHVHDUFKHUV ZLOO DVN SHRSOH f:KDW LV WKH PRVW LPSRUWDQW LVVXH IDFLQJ WKH VRFLHW\ RU \RXUVHOI"f RU f+RZ GR \RX WKLQN DERXW D SDUWLFXODU LVVXH RU SHUVRQ"f 7KHVH UHVSRQVHV DUH FRPSDUHG ZLWK PHGLD FRQWHQW DQDO\VHV WR GHWHUPLQH LI WKH PHGLD FRQWHQW LQIOXHQFHV SHRSOHfV WKLQNLQJ DERXW GLIIHUHQW LVVXHV 3UREOHP 'HILQHG :KHWKHU WKH VXUYH\ LV FRQGXFWHG E\ WHOHSKRQH RU PDLO WKLV PHWKRG FDQ EH VRPHZKDW REWUXVLYH DQG PD\ QRW FDSWXUH WKH WUXH RSLQLRQV RI VXEMHFWV :KDW SHRSOH UHDOO\ WKLQN RU VD\ DERXW WKH LVVXHVf LQ WKH LQWHUHVW RI WKH UHVHDUFKHUV PD\ QRW EH FRPSOHWHO\ GHWHFWHG E\ DVNLQJ FRQYHQWLRQDO VXUYH\ TXHVWLRQV

PAGE 13

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f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f 7KLV WHQVLRQ LV H[DFHUEDWHG ZKHQ GLVFXVVLQJ FULVLV PDQDJHPHQW FRPPXQLFDWLRQ %HFDXVH SXEOLF UHODWLRQV FULVLV FRPPXQLFDWLRQ GHPDQGV DQ DFFXUDWH GHSLFWLRQ RI SXEOLF RSLQLRQ

PAGE 14

GXULQJ WKH FULVLV WKH ,QWHUQHW QHHGV WR EHFRPH DQ LQWHJUDO SDUW RI FRQWHPSRUDU\ DJHQGDn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fHDYHV GURSf RQ SHRSOHfV FRQYHUVDWLRQV DERXW WKH FULVLV HYHQW ZLWKRXW GLVWXUELQJ WKHLU FRPPXQLFDWLRQ 7KLV PHWKRG DOVR ZLOO FRQVROLGDWH WKH WKHRUHWLFDO IUDPHZRUN RI DJHQGDn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

PAGE 15

$IWHU DQDO\VLV RI WZR FDVH VWXGLHV LQYROYLQJ ERWK WKH PHGLD DJHQGD DQG WKH SXEOLF DJHQGD FRQFHUQLQJ D FRUSRUDWH FULVLV WKLV UHVHDUFK RXWOLQHV D G\QDPLF FULVLV FRPPXQLFDWLRQ PRGHO WKDW LV LQWHQGHG WR KHOS FRPSDQLHV SUHYHQW IXUWKHU GDPDJH DV D UHVXOW RI DQ RUJDQL]DWLRQDO FULVLV %DVHG RQ D SXEOLF RSLQLRQ PRGHO RI FULVLV PDQDJHPHQW SURSRVHG E\ 6WXUJHV f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n f PRGHO DORQJ ZLWK 0DF.XHQn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

PAGE 16

GLVFXVVLRQ RI ILQGLQJV OLPLWDWLRQV RI WKLV VWXG\ DQG WKH SUDFWLFDO LPSOLFDWLRQV DQG VXJJHVWLRQV IRU IXWXUH UHVHDUFK DUH GLVFXVVHG

PAGE 17

&+$37(5 + /,7(5$785( 5(9,(: :KHQ D FULVLV RFFXUV WKH PHGLD FDQ VKDSH WKH JHQHUDO SXEOLFfV SHUFHSWLRQ DERXW WKH WDUJHWHG RUJDQL]DWLRQ DQG KRZ LW GHDOV ZLWK D SDUWLFXODU FULVLV 7R EHWWHU XQGHUVWDQG KRZ PDVV PHGLD DIIHFW SHRSOHf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fV RSLQLRQV DV D FULVLV HYHQW HYROYHV $IWHU XQGHUVWDQGLQJ KRZ WKH SXEOLF WKLQNV DERXW DQ LVVXH WKH FRUSRUDWLRQ FDQ WKHQ DGRSW DSSURSULDWH VWUDWHJLHV WR PDQDJH DQG XOWLPDWHO\ UHVROYH WKH FULVLV 7KH IROORZLQJ FKDSWHU ZLOO ILUVW UHYLHZ WKH DJHQGDVHWWLQJ WKHRU\ LQ GHWDLO IROORZHG E\ D GLVFXVVLRQ RI SHUWLQHQW VWXGLHV RI WKH ,QWHUQHW

PAGE 18

5HYLHZ RI $JHQGD6HWWLQJ 7KHRU\ )LUVW /HYHO 5HVHDUFK $JHQGDVHWWLQJ WKHRU\ ZDV RULJLQDOO\ SURSRVHG E\ 0F&RPEV DQG 6KDZ LQ ZKHQ WKH\ FRQGXFWHG WKHLU SLRQHHULQJ VWXG\ LQ &KDSHO +LOO 1& ,Q WKDW VWXG\ WKH DPRXQW RI PHGLD FRYHUDJH RI YDULRXV LVVXHV ZLWKLQ D SHULRG RI WLPH ZDV FRPSDUHG ZLWK WKH SXEOLFfV UDQNLQJ RI WKHLU LPSRUWDQFH 7KH UHVXOWV UHYHDOHG WKH SRZHUIXO HIIHFWV RI PHGLD RQ WUDQVPLWWLQJ WKH VDOLHQFH RI LVVXHV WR SHRSOHfV PLQGV 7KUHH W\SHV RI DJHQGD DUH LQYROYHG LQ WKH DJHQGDVHWWLQJ UHVHDUFK PHGLD DJHQGD SXEOLF DJHQGD DQG SROLF\ DJHQGD 7KH KLHUDUFKLFDO HIIHFWV GHULYHG IURP WKLV WKHRU\ DUH WKDW WKH PHGLD DJHQGD ZRXOG LQIOXHQFH WKH SXEOLF DJHQGD ZKLFK LQ WXUQ PD\ LQIOXHQFH SROLF\ DJHQGD 'HDULQJ t 5RJHUV f 6RPHWLPHV D IRXUWK YDULDEOH LV VWXGLHG LQ DJHQGDVHWWLQJ VWXGLHV D UHDOZRUOG LQGLFDWRU WKDW 'HDULQJ t 5RJHUV f GHILQHV DV D YDULDEOH WKDW PHDVXUHV PRUH RU OHVV REMHFWLYHO\ WKH GHJUHH RI VHYHULW\ RU ULVN RI VRFLDO SUREOHP S f +RZHYHU VLQFH WKH UHDOZRUOG LQGLFDWRU YDULHV GHSHQGLQJ RQ WKH LVVXH WRSLF WKLV YDULDEOH WR WHVW DJHQGDVHWWLQJ WKHRU\ LV VHOGRP XVHG 0HDVXUHPHQW RI 0HGLD DQG 3XEOLF $JHQGD 0HGLD DJHQGD DUH XVXDOO\ PHDVXUHG E\ FRQWHQW DQDO\VLV RI WKH QHZV PHGLD LQLWLDOO\ HPSOR\HG E\ 0F&RPEV DQG 6KDZ f DQG )XQNKRXVHU f WR GHWHQQLQH WKH QXPEHU RI QHZV LWHPV DERXW DQ LVVXH RU LVVXHV RI VWXG\ 7KH PRUH QHZV FRYHUDJH LV GHYRWHG WR DQ LVVXH WKH PRUH VDOLHQW WKDW LVVXH LV FRQVLGHUHG WR EH IRU WKH PHGLD ,Q WKLV FDVH QXPEHU FRXQWLQJ LV D YHU\ LPSRUWDQW ZD\ WR PHDVXUH PHGLD DJHQGD :LWK JURZLQJ SRSXODULW\ RI FRPSXWHU XVDJH LQ UHFHQW \HDUV FRPSXWHUL]HG FRQWHQW DQDO\VLV KDV EHFRPH

PAGE 19

DQ HIILFLHQW DOWHUQDWLYH WR FRXQWLQJ QHZV VWRULHV RU LGHQWLI\LQJ ZRUGV LQ D WH[W ZKHQ WKH\ DSSHDU WRJHWKHU ,W DOVR IDFLOLWDWHV TXDOLWDWLYH DQDO\VLV LQ VRPH FDVHV 3XEOLF DJHQGD ZHUH RULJLQDOO\ PHDVXUHG E\ DVNLQJ D TXHVWLRQ LQ SXEOLF RSLQLRQ VXUYH\V :KDW LV WKH PRVW LPSRUWDQW SUREOHP IDFLQJ WKLV FRXQWU\ WRGD\" 7KLV W\SH RI TXHVWLRQ KDV EHHQ XVHG IRU GHFDGHV LQ QDWLRQDO SXEOLF RSLQLRQ VXUYH\V ,W FDQ LQGLFDWH WKH UHODWLYH LPSRUWDQFH RI DQ LVVXH LQ WKH SXEOLFnV PLQG ,Q FRQWUDVW WR WKH PHDVXUHPHQW RI PHGLD DJHQGD WKH UHVHDUFKHUV ZLOO GHWHUPLQH ZKHWKHU SHRSOH WDNH FXHV IURP WKH DPRXQW RI PHGLD FRYHUDJH WR DQVZHU WKH 0(3 PRVW LPSRUWDQW SUREOHPfW\SH TXHVWLRQ ,QWHUYHQLQJ 9DULDEOHV LQ $IIHFWLQJ WKH 3XEOLF $JHQGD $OWKRXJK WKH RULJLQDO DJHQGDVHWWLQJ VWXG\ KDV VSHFLILHG WKH GLUHFW UHODWLRQVKLS EHWZHHQ WKH PHGLD DJHQGD DQG WKH SXEOLF DJHQGD PDQ\ RWKHU IDFWRUV LQWHUYHQLQJ LQ WKH FDXVDO UHODWLRQVKLS KDYH EHHQ IRXQG LQ YDULRXV VWXGLHV LQFOXGLQJ LQWHQVLW\ DQG YDULDWLRQ LQ QHZV FRYHUDJH %URVLXV DQG .HSSOLQJHU f IRXQG WKDW DJHQGDVHWWLQJ HIIHFWV ZHUH PRVW OLNHO\ WR RFFXU ZKHQ FRYHUDJH ZDV LQWHQVH DQG ZKHQ WKHUH ZDV VLJQLILFDQW YDULDWLRQ LQ WKH FRYHUDJH IURP PRQWK WR PRQWK ,Q DGGLWLRQ WKH W\SH RI LVVXH LV DQRWKHU IRFXV LQ WKH DJHQGDVHWWLQJ VWXG\ :DQWD DQG +X f IRXQG WKDW SUHVV FRYHUDJH EHVLGHV LQFUHDVLQJ SXEOLF FRQFHUQ ZLWK FHUWDLQ LVVXHV FDQ DOVR GHFUHDVH FRQFHUQV ,W PDLQO\ GHSHQGV XSRQ UHFLSLHQWVf VHOILQYROYHPHQW DQG LQWHUHVW LQ WKH LVVXHV 0HGLD FUHGLELOLW\ LV DQRWKHU GHWHUPLQLQJ IDFWRU 5RJHUV DQG 'HDULQJ f FRQFOXGHG WKDW WKH +HUDOG7ULEXQH D ORFDO QHZVSDSHU WUDQVPLWWHG PRUH VDOLHQFH RI D QHZV LWHP WR WKH VXEMHFWV WKDQ WKH 1DWLRQDO (QTXLUHU D QDWLRQDO WDEORLG PDJD]LQH :DQWD DQG +X f DOVR LQGLFDWHG WKDW LQGLYLGXDOV ZKR SHUFHLYH WKH PHGLD DV PRUH FUHGLEOH

PAGE 20

ZRXOG UHO\ PRUH RQ WKH PHGLD IRU LQIRUPDWLRQ DERXW LVVXHV DQG KHQFH EH PRUH VXVFHSWLEOH WR WKH PHGLD LQIOXHQFLQJ WKHLU XQGHUVWDQGLQJ RI LVVXHV 7KH H[WHQW RI PHGLD H[SRVXUH LV DQRWKHU LQWHUYHQLQJ YDULDEOH 7KH PRUH PHGLD H[SRVXUH WKH PRUH SRVVLEOH WKH SXEOLF DJHQGD ZLOO EH LQIOXHQFHG E\ WKH PDVV PHGLD :HDYHU HW DO 0XOOLQV :DQWD t +X f 3HRSOHfV QHHG IRU RULHQWDWLRQ DOVR WUDQVIHUV WKH PHGLD DJHQGD WR WKH SXEOLF DJHQGD =XFNHU f SRLQWHG RXW WKDW LI SHRSOH KDYH IHZHU GLUHFW H[SHULHQFHV ZLWK DQ LVVXH WKH QHZV PHGLDnV LQIOXHQFH RQ SXEOLF RSLQLRQ DERXW WKDW LVVXH ZLOO EH JUHDWHU WKDQ RWKHUZLVH %\ DQDO\]LQJ WHOHYLVLRQ DQG QHZVSDSHU LVVXH FRYHUDJH IRU IRXU ZHHNV :DQWD DQG :X f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f f VWXG\ $IWHU DQDO\]LQJ DIWHUQRRQ QHZVSDSHUV DURXQG WKH QDWLRQ 6WHPSHO f IRXQG WKDW DYHUDJH XVH RI WKH $VVRFLDWHG 3UHVV QHZV LWHPV E\ WKHVH SDSHUV ZDV SHUFHQW UDQJLQJ IURP D ORZ RI SHUFHQW E\ D 1HZ
PAGE 21

UHODWLYHO\ IHZ HGLWRUV RSHUDWLQJ DW WKH UHJLRQDO QDWLRQDO DQG LQWHUQDWLRQDO EXUHDXV RI WKH ZLUH VHUYLFHVf S f 5HHVH DQG 'DQLHOLDQ f VWXGLHG WKH PHGLD FRYHUDJH RI WKH GUXJ LVVXH LQ DQG SURYLGHG HYLGHQFH RI VWURQJ LQWHUPHGLD DJHQGDVHWWLQJ HIIHFWV IURP WKH 1HZ
PAGE 22

ZKHQ PHGLD UHSHDWHGO\ FRYHU RQH SDUWLFXODU LVVXH $V WKH DERYH ILQGLQJV GHDOW ZLWK WKH WUDGLWLRQDO PHGLD DJHQGDVHWWLQJ HIIHFWV QR UHVHDUFK H[LVWV LQ SUHYLRXV DJHQGDVHWWLQJ VWXGLHV UHJDUGLQJ WKH WLPH IUDPH RI RQOLQH QHZVJURXSV DQG LVVXH VDOLHQFH 'LUHFWLRQ RI &DXVDO 5HODWLRQVKLS EHWZHHQ 0HGLD $JHQGD DQG 3XEOLF $JHQGD :LWK UHVSHFW WR WKH UHODWLRQVKLS EHWZHHQ WKH PHGLD DJHQGD DQG WKH SXEOLF DJHQGD ,WR f EHOLHYHG WKDW fPDVV PHGLD HIIHFWV ZLOO QRW WDNH SODFH XQOHVV PHGLD VWDQG RQ WKH PDMRULW\ VLGH RU WKH PDLQVWUHDP LQ WKH WULSRODU UHODWLRQVKLS DPRQJ WKH PDVV PHGLD WKH JRYHUQPHQW DQG WKH PDVVHVf S f $FFRUGLQJ WR %HKU DQG ,\HQJDUfV f ILQGLQJ WKH GLUHFWLRQ RI FDXVDOLW\ EHWZHHQ PHGLD FRYHUDJH DQG SXEOLF DZDUHQHVV RI WKH LVVXH ZDV XQLGLUHFWLRQDO WKDW LV PHGLD FRYHUDJH RQ WKH LVVXHV KDG TXLWH DQ LPSDFW RQ SHRSOHfV DZDUHQHVV RI WKHP %HQLJHU f PDLQWDLQHG WKDW WKH SXEOLFfV DWWLWXGHV DQG RSLQLRQV PD\ EH PRUH FORVHO\ FRUUHODWHG ZLWK PHGLD FRYHUDJH WKDQ ZLWK PRUH REMHFWLYH VRFLDO FRQGLWLRQV ,QGHHG =KX HW DO f IRXQG WKDW WKH SXEOLFfV LVVXH SULRULWLHV DUH PXFK PRUH LQIOXHQFHG E\ PHGLD FRYHUDJH WKDQ E\ VRFLDO LQWHUDFWLRQ :LOOQDW DQG =KX f DOVR SURYLGHG HYLGHQFH RI RQHVLGHG QHZVSDSHU FRYHUDJH LQIOXHQFLQJ SXEOLF RSLQLRQ DERXW D JRYHUQRUfV RYHUDOO SHUIRUPDQFH ZKHQ WKH\ FRPSDUHG WLPHVHULHV GDWD RI SXEOLF RSLQLRQ SROOV ZLWK WKUHH OHDGLQJ QHZVSDSHUV LQ +RQJ .RQJ 7R LQYHVWLJDWH WKH LPSDFW RI SUHVV FRYHUDJH RQ WKH JHQHUDO SXEOLFfV EHOLHI DERXW +,9 WUDQVPLVVLRQ +HUWRJ DQG )DQ f GLVFRYHUHG D VLJQLILFDQW FDXVDO UHODWLRQVKLS IURP QHZV FRQWHQWV WR WKH SXEOLF RSLQLRQ EXW QRW YLFH YHUVD )URP WKHVH ILQGLQJV DJHQGDn VHWWLQJ WKHRU\ VXJJHVWV WR SXEOLF UHODWLRQV SURIHVVLRQDOV WKH LPSRUWDQFH RI PRQLWRULQJ PHGLD FRYHUDJH RI LVVXHV WKDW KDYH JUDEEHG SXEOLF DWWHQWLRQ

PAGE 23

2Q WKH ILUVW OHYHO RI DJHQGDVHWWLQJ WKHRU\ WKH UHVHDUFKHUV SXW PXFK HPSKDVLV RQ ZKDW PHGLD VD\ DERXW GLIIHUHQW LVVXHV DQG WKHQ FRUUHODWH LW ZLWK KRZ SHRSOH UDQN WKH LPSRUWDQFH RI WKRVH LVVXHV 7KH FRQWHQW RI PHGLD LV QRW H[DPLQHG LQ SDUWLFXODU 0RVW VWXGLHV IRFXV RQ WKH DPRXQW RI QHZV FRYHUDJH LQ WKH PHGLD YHUVXV SXEOLF RSLQLRQ DERXW WKH H[DPLQHG LVVXHV 7KH SXEOLF DJHQGD LV PHDVXUHG E\ VXUYH\LQJ SHRSOHfV RSLQLRQ LQVWHDG RI DFWXDOO\ DQDO\]LQJ KRZ DQG ZKDW SHRSOH VD\ DERXW WKH LVVXHV 7KH FRUH FRQFHSW RI WKH ILUVW OHYHO DJHQGDVHWWLQJ WKHRU\ LV fZKHUH SXEOLF RSLQLRQ DERXW DQ LVVXH FRPHV IURPf :KLWQH\ t %HFNHU 0F&RPEV t 6KDZ :DQWD t +X =XFNHU f 6HFRQG /HYHO 5HVHDUFK ,Q DGGLWLRQ WR WKH RULJLQDO IRFXV RQ LVVXH VDOLHQFH WUDQVPLWWHG IURP WKH PHGLD WR WKH SXEOLFfV PLQG DJHQGDVHWWLQJ WKHRU\ VWXGLHV KDYH PRYHG WR WKH VHFRQG OHYHO +RZ QHZV VWRULHV DUH IUDPHG DQG LVVXH DWWULEXWHV DUH SUHVHQWHG SOD\ D FHQWUDO UROH LQ LPSDFWLQJ WKH SXEOLFfV XQGHUVWDQGLQJ RI WKH LVVXHV 0F&RPEV f 2Q WKH VHFRQG OHYHO WKH UHVHDUFKHUV DWWHPSW WR DQVZHU WKH TXHVWLRQ 7R ZKDW H[WHQW LV WKH SXEOLFnV YLHZ RI DQ LVVXH VKDSHG E\ WKH PHGLD WKDW IUDPH REMHFWV RU DWWULEXWHV RI WKH LVVXH LQ GLIIHUHQW ZD\V" 7R IUDPH DV (QWPDQ f H[SODLQV LV WR VHOHFW VRPH DVSHFWV RI D SHUFHLYHG UHDOLW\ DQG PDNH WKHP PRUH VDOLHQW LQ D FRPPXQLFDWLQJ WH[W LQ VXFK D ZD\ DV WR SURPRWH D SDUWLFXODU SUREOHP GHILQLWLRQ FDXVDO LQWHUSUHWDWLRQ PRUDO HYDOXDWLRQ DQGRU WUHDWPHQW UHFRPPHQGDWLRQ IRU WKH LWHP GHVFULEHG S f %HFDXVH DOO LVVXHV KDYH GLIIHUHQW GLPHQVLRQV WKH\ FDQ EH LQWHUSUHWHG IURP DOO NLQGV RI SHUVSHFWLYHV 7KURXJK VHOHFWLRQ DQG XVH RI VRPH DWWULEXWHV RU REMHFWV RI DQ LVVXH PHGLD FDQ PDNH VRPH QHZV SURPLQHQW DQG WUDQVPLW WKDW SURPLQHQW DVSHFW WR WKH DXGLHQFHfV PLQG

PAGE 24

*LYHQ WKDW WKH DJHQGDVHWWLQJ WKHRU\ KDV SURJUHVVHG IURP FRXQWLQJ QXPEHUV PHGLD DJHQGDf DQG UDQNLQJ LVVXHV SXEOLF DJHQGDf WR DQDO\]LQJ DQG H[SODLQLQJ WKH WH[WXDO PHDQLQJV RI WKH PHGLD FRYHUDJH LW FDQ EH DVVXPHG WKDW WKH PHGLD DUH QRW RQO\ WHOOLQJ XV ZKDW WR WKLQN DERXW &RKHQ f EXW DOVR WHOO XV KRZ WR WKLQN DERXW VRPH REMHFWV RU DWWULEXWHV RI SDUWLFXODU LVVXHV 7R XQGHUVWDQG WKH FRQFHSWXDOL]DWLRQ RI PHGLD IUDPHV DQG DWWULEXWHV *KDQHP f EUHDNV GRZQ IRXU GLPHQVLRQV RI PHGLD IUDPHV S f 7KH WRSLF RI D QHZV LWHP ZKDW LV LQFOXGHG LQ WKH IUDPHf 3UHVHQWDWLRQ VL]H DQG SODFHPHQWf &RJQLWLYH DWWULEXWHV GHWDLOV RI ZKDW LV LQFOXGHG LQ WKH IUDPHf $IIHFWLYH DWWULEXWHV WRQH RI WKH SLFWXUHf $OWKRXJK WKH LGHD RI QHZV IUDPLQJ LV TXLWH QRYHO LQ UHFHQW \HDUV 0F&RPEV f SURYLGHG VRPH UHVHDUFK VWXGLHV FRQILUPLQJ WKH HIIHFWV RI DJHQGDVHWWLQJ RQ WKH VHFRQG OHYHO ,Q UHJDUG WR LVVXH IUDPLQJ 7DNHVKLWD DQG 0LNDPL f IRXQG D KLJKO\ SRVLWLYH FRUUHODWLRQ EHWZHHQ WKH VDOLHQFH RI V\VWHPUHODWHG DVSHFWV RI UHIRUP RQ WKH SXEOLF DJHQGD DQG SHRSOHfV DWWHQWLYHQHVV WR SROLWLFDO QHZV RQ ERWK QHZVSDSHUV DQG WHOHYLVLRQ QHZV 0DKHU f VKRZHG D KLJK FRUUHVSRQGHQFH LQ WKH UHODWLYH VDOLHQFH RI WKH DWWULEXWHV WKDW GHILQHG WKH SLFWXUHV RI ORFDO HQYLURQPHQWDO LVVXHV LQ WKH QHZVSDSHU DQG DPRQJ WKH SXEOLF ,Q HDUOLHU VWXGLHV :HDYHU *UDEHU 0F&RPEV DQG (\DO f IRXQG D VWULNLQJ GHJUHH RI FRUUHVSRQGHQFH EHWZHHQ WKH DJHQGD RI DWWULEXWHV LQ WKH &KLFDJR 7ULEXQH DQG WKH DJHQGD RI DWWULEXWHV LQ ,OOLQRLV YRWHUVn GHVFULSWLRQV RI -LPP\ &DUWHU DQG -HUU\ )RUG 7KH FRUUHVSRQGHQFH EHWZHHQ WKH DJHQGD RI DWWULEXWHV LQ 1HZVZHHN DQG WKH DJHQGD RI

PAGE 25

DWWULEXWHV LQ 1HZ
PAGE 26

7KH LGHD RI WKH VHFRQG OHYHO DJHQGDVHWWLQJ UHVHDUFK VKHGV PXFK OLJKW RQ KRZ DQG ZKDW WR H[DPLQH LQ PHGLD HIIHFWV RQ WKH GLUHFWLRQ RI SXEOLF RSLQLRQ 7KH ILQGLQJV DOVR VHW D QHZ SDWK WR GHOYH LQWR WKH UHODWLRQVKLS EHWZHHQ WKH PHGLD DQG WKH SXEOLF DJHQGD 6FKRODUV IRFXVLQJ RQ WKH ODWWHU EHJLQ WR ORRN LQWR WKH FRQWH[WV RI PHGLD LQ RUGHU WR SUREH ZK\ SXEOLF RSLQLRQ LV IRUPHG 7R XQGHUVWDQG SHRSOHfV FRJQLWLYH NQRZOHGJH DQG DIIHFWLYH UHDFWLRQV WR WKH LVVXHV KDV EURDGHQHG WKH H[SODQDWRU\ GRPDLQ RI DJHQGDVHWWLQJ WKHRU\ $OWKRXJK WKH FRQWH[WV RI WKH PHGLD DJHQGD KDYH EHHQ WKRURXJKO\ H[DPLQHG LQ WHUPV RI WKH VHFRQG OHYHO RI DJHQGDVHWWLQJ WKHRU\ PHDVXUHPHQW RI WKH SXEOLF DJHQGD VWLOO UHOLHV RQ WKH VXUYH\ WR XQGHUVWDQG KRZ SHRSOHf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nV FRPSXWHU IDFLOLWLHV DW GLIIHUHQW SODFHV ,URQLFDOO\ WKH PDLQ WUDIILF RQ $53$1(7 ZDV QRW ORQJGLVWDQFH FRPSXWLQJ ,QVWHDG LW ZDV QHZV DQG SHUVRQDO PHVVDJHV 0DQ\ UHVHDUFKHUV WKURXJK $53$1(7 ZHUH FROODERUDWLQJ RQ SURMHFWV WUDGLQJ ZRUN H[SHULHQFHV DQG HYHQ

PAGE 27

H[FKDQJLQJ VRPH JRVVLS 7KURXJKRXW WKH V $53$1(7 ZDV H[SDQGLQJ DW D IDVW SDFH 7KUHH GHFDGHV ODWHU WKH ,QWHUQHW KDV PRYHG EH\RQG LWV RULJLQDO IXQFWLRQV IRU PLOLWDU\ DQG UHVHDUFK LQVWLWXWLRQV DQG HQWHUHG WKH ZRUOG RI SXEOLF HGXFDWLRQ DV ZHOO DV WKH FRPPHUFLDO VHFWRUV 7KH ,QWHUQHWnV SDFH RI JURZWK LQ WKH HDUO\ V LV VR UDSLG WKDW FXUUHQWO\ WKHUH DUH WHQV RI WKRXVDQGV RI QRGHV RQ WKH ,QWHUQHW WKDW H[WHQG WR RYHU FRXQWULHV LQ WKH ZRUOG ZLWK PRUH FRPLQJ RQOLQH HYHU\ GD\ 7KH SULPDU\ DWWUDFWLRQ WKDW WKH ,QWHUQHW KDV LV WKDW QR RIILFLDO FHQVRUVKLS KDV EHHQ LPSOHPHQWHG DW OHDVW XS XQWLO QRZ 1RERG\ FRQWUROV WKLV VRFDOOHG ,QIRUPDWLRQ +LJKZD\ *HQHUDOO\ SHRSOH PDNH XVH RI WKH ,QWHUQHW LQ WKH IROORZLQJ DUHDV :RUOG :LGH :HE :::f VXUILQJ HOHFWURQLF PDLO HPDLOf WKH 86(1(7 7HOQHW *RSKHU VHDUFKLQJ DQG ILOH WUDQVIHU 7R H[FKDQJH LQIRUPDWLRQ RU LGHDV RQ WKH ,QWHUQHW HOHFWURQLF EXOOHWLQ ERDUG V\VWHPV %%6f PDLOLQJ OLVWV /LVWVHUYf DQG PRGHUDWHG QHZVJURXSV 86(1(7f KDYH EHHQ DPRQJ WKH PRVW SRSXODU HOHFWURQLF GLVFXVVLRQ JURXSV IRU PRUH WKDQ D GHFDGH (OHFWURQLF EXOOHWLQ ERDUGV IXQFWLRQ DV D VSHFLDOL]HG PHGLXP DQG FDQ EH VHW XS E\ LQWHUFRQQHFWLQJ SHUVRQDO FRPSXWHUV IRU WKH SXUSRVH RI VHUYLQJ WKH GHEDWH DVVRFLDWLRQ DQG H[FKDQJH IXQFWLRQ RI WKDW FRPPXQLW\ 7KRPVHQ f 7KRXVDQGV RI SHRSOH FRXOG FRQWULEXWH WR DQ RSHQ GLVFXVVLRQ RU DQ LQIRUPDO GLDORJXH DW D VSHFLILF WLPH RQ WKLV V\VWHP $ /LVWVHUY LV D SURJUDP WKDW PDLQWDLQV RQH RU PRUH RI PDLOLQJ OLVWV $ OLVWVHUY DXWRPDWLFDOO\ GLVWULEXWHV DQ HPDLO PHVVDJH IURP RQH PHPEHU RI D OLVW WR DOO RWKHU 7KH :RUOG :LGH :HE LV D SURWRFRO WKDW DOORZV XVHUV WR HDVLO\ PDNH K\SHUWH[W LQIRUPDWLRQ DYDLODEOH WR RWKHU XVHUV

PAGE 28

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f§ V\VWHPV WKDW UHFHLYH DQG SDVV DORQJ HDFK FRQWULEXWLRQ WKH\ UHFHLYH $OWKRXJK WKH DERYH WKUHH HOHFWURQLF GLVFXVVLRQ JURXSV SURYLGH D UHODWLYHO\ IUHH HQYLURQPHQW IRU RSLQLRQ RU LGHDV H[FKDQJH LQIRUPDO EXW LPSOLFLW UXOHV RI EHKDYLRU FDOOHG fQHWLTXHWWHf KDYH DOZD\V EHHQ UHTXLUHG ZKHQ RQH SHUVRQ DWWHPSWV WR SRVW KLV RU KHU PHVVDJH RQ WKH GLVFXVVLRQ JURXSV ,W LV D XVHUfV GXW\ WR SRVW WKH PHVVDJHV UHOHYDQW WR WKH WRSLF LVVXHV RI WKH DSSURSULDWH GLVFXVVLRQ JURXS ZLWKRXW LQWHQWLRQDOO\ RIIHQGLQJ RWKHU SDUWLFLSDQWV 1HYHUWKHOHVV GXH WR WKH XVHUfV DELOLW\ WR FRQFHDO KLV RU KHU LGHQWLW\ LQ WKH FRPSXWHUPHGLDWHG FRPPXQLFDWLRQ D XVHU PLJKW SRVW WKH PHVVDJHV ZLWK QR IDFWXDO

PAGE 29

LQIRUPDWLRQ RU SURYLGH H[FHVVLYHO\ FULWLFDO FRPPHQWV 7KLV IRUP RI PHVVDJH LV FDOOHG f\ODPLQJf ZKLFK RIWHQ JHQHUDWHV D VHULHV RI SXEOLF SRVWV LQ ZKLFK SHRSOH IODPH RQH DQRWKHU UDWKHU WKDQ FRQWULEXWH XVHIXO LQIRUPDWLRQ +LOO DQG +XJKHV f GHILQHG IODPLQJ PHVVDJHV DV SHUVRQDO DWWDFNV XVXDOO\ DFFRPSDQLHG E\ SURIDQLW\ ,Q KLV DUWLFOH *URSHU f WKRXJKW WKDW IODPLQJ ZRUGV RU SKUDVHV DUH H[WUHPH FULWLFLVP VDUFDVP FXUVLQJ H[FODPDWLRQV PXOWLSOH TXHVWLRQ PDUNV HPRWLFRQV OLNH ZKLFK FRQQRWH DQJHU DQG EORFN FDSLWDO ZRUGV 7KHVH VRFLDOO\ XQGHVLUDEOH DFWV PDLQO\ FRPH IURP ODFN RI VWDQGDUG UXOHV LQ WKH HOHFWURQLF FRPPXQLFDWLRQ ZRUOG QDPHO\ WKH ,QWHUQHW 0F/DXJKOLQ HW DK f :KLOH UDGLR WRRN \HDUV WR UHDFK DQ DXGLHQFH RI PLOOLRQ DQG WHOHYLVLRQ WRRN WKH ,QWHUQHW WRRN MXVW IRXU \HDUV :HLVH f ,Q D UHFHQW UHSRUW RQ LWV ILUVW PDMRU VWXG\ RI WKH HFRQRPLF LPSDFW RI WKH ,QWHUQHW WKH &RPPHUFH 'HSDUWPHQW RI WKH 8QLWHG 6WDWHV GLVFORVHG WKDW 1HW WUDIILF LV GRXEOLQJ HYHU\ KXQGUHG GD\V DQG HOHFWURQLF FRPPHUFH VKRXOG UHDFK ELOOLRQ E\ 0RUH WKDQ PLOOLRQ SHRSOH DUH QRZ RQ OLQH :HLVH f +RZ PDQ\ XVHUV DUH VXUILQJ RQ WKH ,QWHUQHW WRGD\" $FFRUGLQJ WR D VXUYH\ FRQGXFWHG E\ WKH 5HOHYDQW.QRZOHGJH FRPSDQ\ LQ ODWH )HEUXDU\ RI WKH :HE XVHUV LQ WKH 8QLWHG 6WDWHV KDYH JURZQ WR D JURZWK RI RYHU PLOOLRQ XVHUV VLQFH WKH FRPSDQ\nV ODVW VWXG\ LQ -DQXDU\ -HII /HY\ &(2 RI WKH 5HOHYDQW.QRZOHGJH $ SXEOLF SRVW RU HPDLO PHVVDJH WKDW H[SUHVVHV D VWURQJ RSLQLRQ RU FULWLFLVP )ODPHV FDQ EH IXQ ZKHQ WKH\ DOORZ SHRSOH WR YHQW WKHLU IHHOLQJV WKHQ UHWXUQ WR WKH WRSLF DW KDQG 2WKHUV DUH VLPSO\ LQVXOWLQJ DQG FDQ OHDG WR IODPH ZDUV

PAGE 30

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fV DUWLFOH DSSHDULQJ LQ WKH 5HXWHUV 1RUWK $PHULFDQ :LUH )HE f RQ SXEOLF UHODWLRQV ILUPV KDQGOLQJ WKH ,QWHUQHW KH UHPDUNHG WKDW SXEOLF UHODWLRQV YHWHUDQV

PAGE 31

DJUHHG WKDW DERXW WKH ZRUVW WKLQJ D FRPSDQ\ FRXOG GR LV LJQRUH SHUVRQV ZKR EDGPRXWK WKHP LQ F\EHUVSDFH 6SHDNLQJ RI WKH DGYDQWDJHV RI WKH ,QWHUQHW IRU SXEOLF UHODWLRQV 6WUHQVNL f SRLQWHG RXW WKDW fSURIHVVLRQDO SXEOLF UHODWLRQV FRPPXQLFDWRUV FDQ FKRRVH IURP D EORVVRPLQJ DUUD\ RI F\EHUVSDFH UHVRXUFHV WKURXJK ZKLFK WR FKDQQHO WKHLU PHVVDJHVf S f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f EHOLHYHV WKDW WKH SRSXODULW\ VWHPV IURP WKH ORVV RI FUHGLELOLW\ RI WUDGLWLRQDO PHGLD E\ WKH JHQHUDO SXEOLF +RZHYHU LW LV DOVR UHFRJQL]HG E\ PRVW XVHUV WKDW fPXFK RI WKH GLVFXVVLRQ LV WULYLDO DQG WKDW PRUH RIWHQ WKDQ QRW GHEDWH UDSLGO\ GLVLQWHJUDWHV LQWR YHUEDO DVVDXOWV DQG LGOH EDQWHUf 5HYDK S f 7KH GDUN VLGH RI WKH ,QWHUQHW LV WKDW fDQ\RQH FDQ VD\ DQ\WKLQJ WR DQ\ERG\ DW DQ\WLPHf FRPPHQWHG 'RQ 0LGGOHEHUJ SUHVLGHQW RI D SXEOLF

PAGE 32

UHODWLRQV ILUP .DOLVK f $V D UHVXOW VRPH SHRSOH DUH SDUWLFLSDWLQJ LQ D YDULHW\ RI QHZVJURXSV WKDW GDPDJH FRPSDQLHVf UHSXWDWLRQV 7R HPSKDVL]H WKH LPSRUWDQFH RI PRQLWRULQJ QHZVJURXSV GLVFXVVLRQ IRU UHIHUHQFHV WR WKH EXVLQHVV FRUSRUDWLRQV .DOLVK FLWHG WKH VXJJHVWLRQ RI $GDP &RRSHU FUHDWLYH PDQDJHU DW WKH ,QWHUDFWLYH 6ROXWLRQV *URXS RI (GHOPDQ 3XEOLF 5HODWLRQV :RUOGZLGH VWDWLQJ WKDW f:H ILQG WKDW WKH QHZVJURXSV DUH PXFK PRUH LQWHUDFWLYH WKDQ WKH ZHE VR WKH QHZVJURXSV DUH UHDOO\ ZKDW ZH ZRXOG FRQVLGHU WKH NH\ SODFH WR OLVWHQ WR SHRSOHf 5HXWHUV 1RUWK $PHULFDQ :LUH )HE f 7ZR H[DPSOHV LQYROYLQJ FRUSRUDWH FULVHV RQ WKH ,QWHUQHWUHODWHG GLVFXVVLRQ EROVWHU &RRSHUfV YLHZSRLQW 2QH H[DPSOH ZDV WKH FULVLV SHUWLQHQW WR WKH LQWURGXFWLRQ RI ,QWHOfV QHZ 3HQWLXP FRPSXWHU SURFHVVRU 7KH ,QWHO RUJDQL]DWLRQ GLVFRYHUHG WKH 3HQWLXP IODZ LQ WKH VXPPHU RI EXW GHFOLQHG WR DFNQRZOHGJH WKH SUREOHP SXEOLFO\ %\ IDLOLQJ WR UHFRJQL]H WKH LPSRUWDQFH RI WKH PHVVDJH SRVWHG RQ D &RPSX6HUYH IRUXP ,QWHO SDLG RXW RYHU PLOOLRQ WR UHSODFH WKH IODZHG FKLS 7KLV LV RQH H[DPSOH RI LJQRULQJ WKH SXEOLFfV FRQFHUQV DERXW LWV RZQ SURGXFWV DV ZHOO DV LWV RZQ LPDJH $QRWKHU UXPRUOLNH LQFLGHQW RFFXUUHG LQ 1RYHPEHU ZKHQ SHRSOH RQ WKH ,QWHUQHW ODEHOHG IDVKLRQ GHVLJQHU 7RPP\ +LOILJHU D UDFLVW DQG FDOOHG IRU D ER\FRWW RI KLV SURGXFWV ,Q FRQWUDVW WR WKH ,QWHOfV UHDFWLRQ WKH IDVKLRQ GHVLJQHU 7RPP\ +LOILJHU SDLG DWWHQWLRQ WR WKH RIIHQGLQJ ,QWHUQHW VLWHV DQG UHVSRQGHG E\ ODXQFKLQJ DQ HPDLO FDPSDLJQ WKDW DVVXUHG PLQRULWLHV RQ WKH ,QWHUQHW WKDW WKH\ ZHUH YDOXHG FXVWRPHUV DQG WKDW WKH UDFLVW UXPRUV ZHUH WRWDOO\ JURXQGOHVV &RPSX6HUYH KDV EHHQ RQH RI PDMRU FRPPHUFLDO RQOLQH VHUYLFHV LQ WKH 8QLWHG 6WDWHV IRU RYHU ILIWHHQ \HDUV

PAGE 33

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fRQOLQH FULVLVf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f FLWHG 9DOHQWHfV LGHD DERXW WKH FULWLFDO PDVV fWKH FULWLFDO PDVV LV DFKLHYHG ZKHQ DERXW WR SHUFHQW RI WKH SRSXODWLRQ KDV DGRSWHG WKH LQQRYDWLRQ :KHQ WKLV OHYHO KDV EHHQ UHDFKHG WKH LQQRYDWLRQ FDQ EH VSUHDG WR WKH UHVW RI WKH VRFLDO V\VWHPf S f %HIRUH GHOYLQJ PRUH LQWR WKH LPSDFW RI QHZ FRPPXQLFDWLRQ WHFKQRORJLHV LW PD\ EH ZRUWKZKLOH WR UHHYDOXDWH WKH GHILQLWLRQ RI PDVV PHGLD 'RHV WKH ,QWHUQHW SOD\ WKH UROH RI DQ LQWHUDFWLYH PHGLXP" 5DIDHOL f UHJDUGHG WKLV QHZ W\SH RI FRPPXQLFDWLRQ

PAGE 34

PHGLXP DV VLPLODU WR RWKHU PHGLD KRZHYHU LW LQGHHG LQWHJUDWHV WZR LQWHUHVWV LQ FRPPXQLFDWLRQ ZLWK WKH WHFKQRORJLFDO LQQRYDWLRQ RI FRPSXWHU QHWZRUNLQJ fLQWHUSHUVRQDO FRQYHUVDWLRQ DQG PDVV SURSDJDWLRQ RI PHVVDJHVf S f ,Q %XF\fV f DUWLFOH GLVFXVVLQJ DQ REMHFWLYH PHDVXUHPHQW RI LQWHUDFWLYLW\ KH VXJJHVWHG WKDW WKH LQWHUDFWLYH HYHQW VKRXOG QRW RQO\ DFWLYDWH DQ LQWHUSHUVRQDO FRQYHUVDWLRQ EXW LW KDV WR EH fD IRUP RI PDVV FRPPXQLFDWLRQ LQ ZKLFK WKH VHQGHU EURDGFDVWV D PHVVDJH WR D PDVV DXGLHQFH LQ WKH SURFHVV RI UHVSRQGLQJ WR DQ LQGLYLGXDOf S f +LHEHUW f FRQVLGHUHG PHVVDJH WUDQVPLVVLRQ WKURXJK D PDVV PHGLXP DV DQ HIIHFWLYH FRPPXQLFDWLRQ WKDW HOLFLWV IHHGEDFN IURP WKH DXGLHQFH 5DIDHOLnV f GHILQLWLRQ RI LQWHUDFWLYLW\ EDVHG RQ WKH LVVXH RI UHVSRQVLYHQHVV UHFRJQL]HV WKUHH SHUWLQHQW OHYHOV QRQLQWHUDFWLYH FRPPXQLFDWLRQ VHTXHQFHV TXDVLn LQWHUDFWLYH UHDFWLYHf FRPPXQLFDWLRQ DQG IXOO\ LQWHUDFWLYH FRPPXQLFDWLRQ 7KH GLVWLQFWLRQ EHWZHHQ TXDVL DQG IXOO LQWHUDFWLYLW\ GHSHQGV RQ WKH QDWXUH RI WKH FRPPXQLFDWLRQ UHVSRQVHV ,Q WKH IXOO\ LQWHUDFWLYH FRPPXQLFDWLRQ VHTXHQWLDO PHVVDJHV fGHSHQG XSRQ WKH UHDFWLRQ LQ HDUOLHU WUDQVDFWLRQV DV ZHOO DV RQ WKH FRQWHQW H[FKDQJHGf S f 'HULYLQJ IURP 5DIDHOLfV PRGHO LQWHUDFWLYLW\ LV fIHHGEDFN WKDW UHODWHV ERWK WR SUHYLRXV PHVVDJHV DQG WR WKH ZD\ SUHYLRXV PHVVDJHV UHODWHG WR WKRVH SUHFHGLQJ WKHPf S f 5DIDHOL DQG /D5RVHfV f VWXG\ RQ HOHFWURQLF EXOOHWLQ ERDUGV WUHDWHG HOHFWURQLF EXOOHWLQ ERDUGV DV D NLQG RI FROODERUDWLYH PDVV PHGLXP fD QHZ W\SH RI PDVV FRPPXQLFDWLRQV PHGLXP LQ ZKLFK WKH DXGLHQFH DFWV ERWK DV WKH VRXUFH DQG WKH UHFHLYHU RI WKH PHVVDJHf S f $Q LQWHUDFWLYH PHGLXP LV D fPDQ\WRPDQ\f FRPPXQLFDWLRQ FKDQQHO XQOLNH FRQYHQWLRQDO PDVV PHGLD VXFK DV WHOHYLVLRQ UDGLR DQG VWDWLF SULQW PHGLD

PAGE 35

ZKLFK DUH D fRQHWRPDVVf FKDQQHO $OVR HOHFWURQLF EXOOHWLQ ERDUGV UHSUHVHQW DQ LQWHUDFWLYH PHGLXP WKDW GHPDQGV XVHUVf FRQWULEXWLRQ DV ZHOO DV FRQVXPSWLRQ *DUUDPRQH +DUULV DQG $QGHUVRQ f H[SORUHG WKH PRWLYDWLRQV IRU SDUWLFLSDWLRQ RQ D SROLWLFDO EXOOHWLQ ERDUG 7KH\ REVHUYHG QR VWDWLVWLFDOO\ VLJQLILFDQW UHODWLRQVKLSV EHWZHHQ WKH XVH RI %%6 IHDWXUHV DQG GHPRJUDSKLFV RU SROLWLFDO SDUWLVDQ SUHGLVSRVLWLRQV $OVR LQ WKHLU VWXG\ VXUYHLOODQFH ZDV UDQNHG DV WKH QXPEHU RQH PRWLYDWLRQ IRU %%6 XVH IROORZHG E\ FXULRVLW\ DQG NQRZOHGJH RI RWKHUVf RSLQLRQV ,Q KLV DUWLFOH GHILQLQJ HOHFWURQLF EXOOHWLQ ERDUG XVHUV 5DIDHOL f FRQFOXGHG WKDW WKH XVHUV RI ERDUGV FRXOG EH FKDUDFWHUL]HG DV ERWK D IDLWKIXO DQG DFWLYH DXGLHQFH 7R IXUWKHU SUREH WKH VXFFHVV RI HOHFWURQLF EXOOHWLQ ERDUG V\VWHPV 5DIDHOL HW DO f IRXQG WKDW WKH GLYHUVLW\ RI ERDUG FRQWHQWV KDG SRVLWLYH LPSDFW RQ ERDUG XVHUVf FRQWULEXWLRQ OHYHOV DQG VPDOO JURXS VL]H FRQWULEXWHG WR DGRSWLRQ OHYHO RI WKH %%6V $IWHU FRQWHQW DQDO\]LQJ SXEOLF UHODWLRQV SUDFWLWLRQHUVf XVH RI DQ HOHFWURQLF QHZVJURXS FDOOHG 35)RUXP 7KRPVHQ f GLVWLQJXLVKHG WKUHH PDLQ IXQFWLRQV LQ XVLQJ WKH 35)RUXP H[FKDQJH RI LQIRUPDWLRQ DQG DGYLFH GHEDWH RYHU LVVXHV DIIHFWLQJ WKH SURIHVVLRQ DQG FXOWLYDWLRQ RI D VHQVH RI VHOIYDOLGDWLRQ DQG HQKDQFHG HIILFDF\ ERWK DW D SHUVRQDO DQG SURIHVVLRQDO OHYHO $PRQJ WRROV RQ WKH ,QWHUQHW WKDW DUH LQIOXHQWLDO LQ ERWK SXEOLF UHODWLRQV DQG DGYHUWLVLQJ ILHOGV %REELWW f UHFRPPHQGHG WKH 86(1(7 QHZVJURXS DV D SXEOLF UHODWLRQV WRRO IRU ERWK LVVXHV DQG DXGLHQFH UHVHDUFK %HFDXVH VWXG\ RI LVVXHV DQG $V QRWHG LQ 7KRPVHQfV DUWLFOH 35)RUXP IRXQGHG LQ LV DQ RQOLQH GLVFXVVLRQ JURXS IRU SXEOLF UHODWLRQV SURIHVVLRQDOV 8SGDWHG LQ /LVW VHDUFK ZHEVLWH KWWSZZZOVRIWFRPOLVWVOLVWBTKWPOf WKHUH ZHUH SHRSOH VXEVFULELQJ WR WKLV JURXS DV RI 0D\

PAGE 36

DXGLHQFH DQDO\VLV LV GHSHQGHQW RQ WKH FUHGLELOLW\ RI WKH UHVSRQGHQWV WKH UHVHDUFKHU PXVW DVVXPH WKDW H[SUHVVHG RSLQLRQV RU FRPPHQWV DUH LQGHHG WUXWKIXO 7UHDWLQJ WKH 86(1(7 DV D SROLWLFDO GLVFXVVLRQ IRUXP *URSHU f FRQFOXGHG WKDW WKH SROLWLFDO OHDGHUVKLS GLG QRW H[LVW LQ WKH WKUHH VDPSOHG QHZVJURXSV 1R VLQJOH SDUWLFLSDQW LQ WKH QHZVJURXSV GLVFXVVLRQ ZDV DEOH WR GRPLQDWH WKH GLDORJXHV DQG SURYRNH SDVVLRQ DPRQJ KLV RU KHU IROORZHUV +LOO DQG +XJKHV f DQDO\]HG PHVVDJHV LQ SROLWLFDO QHZVJURXSV DQG IRXQG WKDW PRVW SROLWLFDO 86(1(7 JURXSV H[KLELWHG WKH WUDGLWLRQDO FKDUDFWHULVWLFV RI D VRFLDOO\ FRKHVLYH JURXS 7KH DXWKRUV IXUWKHU DUJXHG WKDW DOWKRXJK WKHVH JURXSV EHFDPH D QHZ IRUP RI SROLWLFDO FRPPXQLW\ IRU WKH SXUSRVH RI SROLWLFDO GLDORJXH LW GLG QRW VLJQDOL]H DQ\ SDUDGLJP VKLIW ,QVWHDG fSHRSOH DUH PHUHO\ PRYLQJ WKHLU DJHROG SDWWHUQV RI LQWHUDFWLRQ LQWR D QHZ UHDOPf S f 7R ZKDW H[WHQW VKRXOG WKH SXEOLF UHODWLRQV SHRSOH HYDOXDWH WKLV QHZ LQWHUDFWLYH PHGLXP WKH ,QWHUQHW" $FFRUGLQJ WR D FRPSUHKHQVLYH GHPRJUDSKLF VXUYH\ UHOHDVHG LQ 0DUFK RI E\ &RPPHUFH1HW DQG 1LHOVHQ 0HGLD 5HVHDUFK D ODUJH PDMRULW\ RI :HE XVHUVf§ SHUFHQWf§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

PAGE 37

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f DQG 0DF.XHQnV f PRGHO RI SXEOLF GLDORJXH $W WKH HQG RI WKH WKLUG FKDSWHU UHVHDUFK TXHVWLRQV DQG K\SRWKHVHV DUH SUHVHQWHG

PAGE 38

&+$37(5 ,,, &5,6,6 &20081,&$7,21 /,7(5$785( 2YHUYLHZ ,Q UHFHQW GHFDGHV LQ $PHULFDQ EXVLQHVV WKHUH KDV RIWHQ EHHQ D WHQVLRQ EHWZHHQ FRUSRUDWH SXEOLF UHODWLRQV SURIHVVLRQDOV DQG PHGLD MRXUQDOLVWV :KHQ WKLQJV DUH JRLQJ ZHOO SXEOLF UHODWLRQV PDQDJHUV GHVSHUDWHO\ ZDQW PHGLD FRYHUDJH EXW fJRRG QHZVf LV VHOGRP D KLJK SULRULW\ IRU SHRSOH DVVLJQLQJ UHSRUWHUV DQG ZULWLQJ KHDGOLQHV +RZHYHU GXULQJ D EXVLQHVV FULVLV WKH QHZV PHGLD DSSHDU IDU PRUH LQVSLUHG WR LQYHVWLJDWH WKH LVVXH ,Q WKHVH fGDPDJH FRQWUROf VLWXDWLRQV SXEOLF UHODWLRQV PDQDJHUV PXVW EH DZDUH WRWDOO\ RI ZKDW LV EHLQJ SULQWHG DQG VDLG LQ DOO PHGLD 2IWHQ WKLV LQIRUPDWLRQ LV LQDFFXUDWH RU ELDVHG DQG WKH FRUSRUDWLRQ PXVW VRPHKRZ FRUUHFW WKH PLVFRQFHSWLRQ DQG UHKDELOLWDWH WKH FRUSRUDWLRQfV SXEOLF UHSXWDWLRQ %LUFK 7UDYHUVR f $V ZLWK WUHDWLQJ D VSUHDGLQJ GLVHDVH HDUO\ LQWHUYHQWLRQ LV HVVHQWLDO WR VXFFHVVIXO WUHDWPHQW RI D FRUSRUDWH FULVLV 3UHYHQWLRQ RI VRPHWKLQJ GHWULPHQWDO WR WKH RUJDQL]DWLRQ UHPDLQV D YLWDO SDUW RI FRUSRUDWH SXEOLF UHODWLRQV ZRUN $V QR RUJDQL]DWLRQ LV H[HPSW IURP D FULVLV WKH DUW DQG VFLHQFH RI FULVLV PDQDJHPHQW LV QR ORQJHU DQ XQXVXDO VSHFLDOW\ EXW DQ LPSHUDWLYH MRE IRU SXEOLF UHODWLRQV SUDFWLWLRQHUV 0DJJDUW f 5DWKHU WKDQ FRSLQJ ZLWK D FULVLV DIWHU LW KDV UHDFKHG RYHUZKHOPLQJ SURSRUWLRQV LQ WKH PHGLD LQWHOOLJHQW SXEOLF UHODWLRQV SURIHVVLRQDOV PXVW OHDP WR GHWHFW SRWHQWLDO FULVHV DW WKH HPEU\RQLF VWDJH 'HDOLQJ ZLWK DQ HPHUJLQJ FULVLV EHIRUH LW UHDFKHV WKH PDVV SXEOLF LV WKH NH\ WR VXFFHVV

PAGE 39

0HGLD SOD\ DQ HVVHQWLDO UROH LQ KHOSLQJ SHRSOH XQGHUVWDQG KRZ WKH FRUSRUDWLRQ UHDFWV WR WKH FULVLV 0HGLD DOVR SURYLGH SXEOLF UHODWLRQV SURIHVVLRQDOV ZLWK LQIRUPDWLRQ DV WR KRZ WKH JHQHUDO SXEOLF IHHOV DERXW WKH FULVLV HYHQW VR WKH FRUSRUDWLRQ FDQ DGRSW DSSURSULDWH PDQDJHPHQW WDFWLFV WR DPHOLRUDWH WKH FULVLV VLWXDWLRQ (IIHFWLYH FRPPXQLFDWLRQ ZLWK NH\ SXEOLFV LQ WKH VRFLHW\ LV WKH SXUSRVH RI SXEOLF UHODWLRQV SUDFWLFH LQ D EXVLQHVV FRUSRUDWLRQ HVSHFLDOO\ ZKHQ LW HQFRXQWHUV DQ XQH[SHFWHG FULVLV $PRQJ DOO NH\ SXEOLFV WKH PHGLD VWDQG RXW DV WKH PRVW VLJQLILFDQW LQ WKH SXEOLF UHODWLRQV ILHOG EHFDXVH WKH\ GLVVHPLQDWH LQIRUPDWLRQ DQG VHUYH DV LQWHUSUHWHUV RI VRFLDO SKHQRPHQD +RZHYHU WKH PHGLD GR QRW RSHUDWH LQ LVRODWLRQ 3XEOLF UHODWLRQV SURIHVVLRQDOV FDQ RIWHQ DIIHFW WKH EHKDYLRU RI WKH PDVV PHGLD 1RW RQO\ VKRXOG DQ RUJDQL]DWLRQ SD\ DWWHQWLRQ WR PHGLD FRYHUDJH ZKLFK VRPHWLPHV JHQHUDWHV QHJDWLYH HIIHFWV RQ WKH SXEOLF DJHQGD LI QR SURPSW DFWLRQ KDV EHHQ LPSOHPHQWHG E\ WKH FRUSRUDWLRQV EXW SXEOLF SHUFHSWLRQ RI DQG RSLQLRQ DERXW WKH FULVLV PXVW EH HYDOXDWHG %HFDXVH PRVW SHRSOH SHUFHLYH WUXWK WR EH ZKDWHYHU SXEOLF RSLQLRQ H[SUHVVHV DW WKH FULVLV WLPH WKH HVVHQWLDO SXUSRVH RI FULVLV FRPPXQLFDWLRQV LV WR DIIHFW WKH DWPRVSKHUH RI SXEOLF RSLQLRQ KRSLQJ WKDW DQ\WKLQJ JRRG DERXW WKH RUJDQL]DWLRQ ZLOO SHUVLVW DQG EDG FRQQRWDWLRQV ZLOO VRRQ GLVDSSHDU )UHTXHQWO\ DVNHG TXHVWLRQV LQ D FULVLV E\ D SURIHVVLRQDO SXEOLF UHODWLRQV SHUVRQ DUH f'RHV WKH JHQHUDO SXEOLF IHHO WKH VDPH ZD\ DV WKH PHGLD MRXUQDOLVWV"f DQG LQ WHUPV RI WKH SHUFHSWLRQ RI D FULVLV f7R ZKDW GHJUHH DUH SHRSOH DIIHFWHG E\ PHGLD FRYHUDJH RI WKH FULVLV DQG WKH RUJDQL]DWLRQ"f 7KH DVVXPSWLRQ EHKLQG WKHVH TXHVWLRQV SRVLWV WKDW WKH JHQHUDO SXEOLF LV RIWHQ LQIOXHQFHG E\ WKH PHGLD FRYHUDJH

PAGE 40

,Q UHFHQW \HDUV D QXPEHU RI QHZ PHGLD WHFKQRORJLHV KDYH PDGH WKH SXEOLF UHODWLRQVn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fV JUHDWHVW DGYDQWDJH LV LWV LQWHUDFWLYLW\ EHWZHHQ LQIRUPDWLRQ SURYLGHUV DQG UHFHLYHUV $Q DXGLHQFH PHPEHU PD\ EHFRPH DQ DFWLYH PHVVDJH SURGXFHU QRW MXVW D SDVVLYH UHFHLYHU )URP WKH FULVLV PDQDJHPHQW SHUVSHFWLYH WKH ,QWHUQHW DOVR RIIHUV DQ H[FHOOHQW RSSRUWXQLW\ IRU SXEOLF UHODWLRQV SURIHVVLRQDOV WR JDXJH WKH LPSDFW RI WKH FULVLV RQ WKH

PAGE 41

&RUSRUDWLRQ FRYHUHG E\ WKH PHGLD DQG WR PRQLWRU WKH SXEOLFfV UHDFWLRQ WR LW 7KH H[FKDQJH EHWZHHQ WKH PHGLD SXEOLF UHODWLRQV SURIHVVLRQDOV DQG WKH SXEOLF LV LPSHUDWLYH IRU WKH PRGHP FRUSRUDWLRQ HVSHFLDOO\ LQ D FULVLV VLWXDWLRQ %\ XQGHUVWDQGLQJ WKHVH LQWHUDFWLRQV WKH ,QWHUQHW FDQ IXQFWLRQ DV D PRUH XVHIXO WRRO IRU SXEOLF UHODWLRQV SUDFWLWLRQHUV :KHQ D FRUSRUDWH FULVLV RFFXUV WKH QHZV PHGLD DUH WHPSWHG RIWHQ WR UHO\ RQ LQIRUPDWLRQ QRW RQO\ IURP FRQYHQWLRQDO MRXUQDOLVWLF VRXUFHV VXFK DV ZLUH VHUYLFHV EXW RQ LQIRUPDWLRQ GHULYHG IURP WKH ,QWHUQHW ZHEVLWHV %LUFK f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nV RSLQLRQ DERXW WKH FULVLV &ULVLV 0DQDJHPHQW 5HVHDUFK DQG 3XEOLF 2SLQLRQ 1RWKLQJ LV PRUH FKDOOHQJLQJ DQG XQQHUYLQJ WR SXEOLF UHODWLRQV SURIHVVLRQDOV WKDQ WKH XQSUHGLFWDEOH FULVLV $ FULVHV FDQ FDXVH D WHPSRUDU\ GLVUXSWLRQ RI DFWLYLW\ ZLWK QR ORQJODVWLQJ FRQVHTXHQFHV WR WKH ERWWRP OLQH RU LW FDQ SHUPDQHQWO\ GDPDJH D FRUSRUDWLRQfV UHSXWDWLRQ UHVXOWLQJ LQ UHGXFHG SURILWDELOLW\ 8QGHUVWDQGLQJ SURSHUO\ KRZ D FULVLV LV FRQWDLQHG DQG PDQDJHG PD\ LQ H[WUHPH FDVHV GHWHUPLQH WKH YHU\ VXUYLYDO RI D EXVLQHVV RUJDQL]DWLRQ

PAGE 42

'HILQLWLRQ RI &ULVLV :KDW LV D FULVLV" 1R XQLYHUVDOO\ DFFHSWHG GHILQLWLRQ KDV EHHQ DGRSWHG EXW LQ JHQHUDO D FULVLV LV D VLWXDWLRQ WKDW WKUHDWHQV WKH QRUPDO DFWLYLW\ RI DQ RUJDQL]DWLRQ $ FULVLV FDQ SURFHHG IURP D PHUH GLVUXSWLRQ RI DFWLYLW\ WR GDPDJLQJ WKH FRUSRUDWH UHSXWDWLRQ DQG UHGXFLQJ SURILWDELOLW\ +D\HV f EHOLHYHV WKDW D FULVLV UHVXOWV IURP D PDMRU LQFRQJUXHQFH EHWZHHQ WKH H[SHFWDWLRQV RI D FRLSRUDWLRQ DQG ZKDW KDSSHQV LQ WKH HQYLURQPHQW :HLFN f WKLQNV WKDW fFULVHV DUH FKDUDFWHUL]HG E\ ORZ SUREDELOLW\KLJK FRQVHTXHQFH HYHQWV WKDW WKUHDWHQ WKH PRVW IXQGDPHQWDO JRDOV RI DQ RUJDQL]DWLRQf S f 'XWWRQ f LGHQWLILHV WKUHH HVVHQWLDO GLPHQVLRQV LQYROYHG LQ D FULVLV LQ WKH VWXG\ RI WKH GHFLVLRQPDNLQJ SURFHVVHV EHWZHHQ FULVLV DQG QRQFULVLV LVVXHV LQ DQ RUJDQL]DWLRQ LPSRUWDQFH LPPHGLDF\ DQG XQFHUWDLQW\ /HUELQJHU f WUHDWV D FULVLV DV DQ HYHQW WKDW HQGDQJHUV DQ RUJDQL]DWLRQfV IXWXUH SURILWDELOLW\ JURZWK DQG HYHQ LWV VXUYLYDO )HDP %DQNV f GHILQHV D FULVLV EURDGO\ DV f D PDMRU RFFXUUHQFH ZLWK D SRWHQWLDOO\ QHJDWLYH RXWFRPH DIIHFWLQJ DQ RUJDQL]DWLRQ FRPSDQ\ RU LQGXVWU\ DV ZHOO DV LWV SXEOLFVf S f 7KH ,QVWLWXWH IRU &ULVLV 0DQDJHPHQW UHIHUV WR FULVLV DV D VLJQLILFDQW EXVLQHVV GLVUXSWLRQ ZKLFK UHVXOWV LQ H[WHQVLYH QHZV PHGLD FRYHUDJH DQG SXEOLF VFUXWLQ\ ,UYLQH t 0LOODU f 7R VXP XS D FULVLV FDQ EH VRPHWKLQJ XQH[SHFWHG WKDW RFFXUV E\ VXUSULVH LQ DQ\ W\SH RI FRLSRUDWLRQ :RRGV f VWHPV IURP WKH LQWHUDFWLRQ RI IDLOXUHV EHWZHHQ WKH FRUSRUDWLRQ DQG WKH H[WHUQDO HQYLURQPHQW 'n$YHQL HW DK f UHTXLUHV IDVW DQG DFFXUDWH UHDFWLRQ WR QHXWUDOL]H KLJK WKUHDW WR FRUSRUDWH YDOXHV *RQ]£OH]+HUUHUR t 3UDWW f FUHDWHV XQFHUWDLQW\ 0LWFKHOO f WKUHDWHQV WKH UHSXWDWLRQ DQG DVVHWV RI WKH

PAGE 43

RUJDQL]DWLRQ %DUWRQ f DQG FDXVHV SXEOLFV HVSHFLDOO\ PHGLD WR VFUXWLQL]H WKH RUJDQL]DWLRQ ,UYLQH t 0LOODU f 7\SHV RI &ULVHV $OWKRXJK FRUSRUDWH FULVLV FDQ RFFXU VXGGHQO\ ZLWKRXW IRUHZDUQLQJ LW LV ZLVH WR LGHQWLI\ WKH W\SHV RI FULVHV ZKHQ WKH\ WDNH SODFH LQ RUGHU WR DGRSW WKH SURSHU VWUDWHJLF SODQ 7KH ,QVWLWXWH IRU &ULVLV 0DQDJHPHQW SXWV WKH FULVLV LQWR FDWHJRULHV DQG LGHQWLILHV WKUHH RULJLQV RI FULVHV PDQDJHPHQW HPSOR\HHV DQG RWKHU ,UYLQH t 0LOODU f 7KH FDWHJRULHV DUH EXVLQHVV FDWDVWURSKH HQYLURQPHQWDO GDPDJH FRQVXPHU DFWLRQ GLVFULPLQDWLRQ ILQDQFLDO GDPDJHV ODERU GLVSXWHV VH[XDO KDUDVVPHQW ZKLWH FROODU FULPH FDVXDOW\ DFFLGHQW FODVV DFWLRQ VXLWV GHIHFWVUHFDOOV H[HFXWLYH GLVPLVVDO KRVWLOH WDNHRYHU PLVPDQDJHPHQW ZKLVWOH EORZLQJ DQG ZRUNSODFH YLROHQFH 0LWURII HW DO f FDWHJRUL]HV W\SHV RI FULVLV ZKLFK LQFOXGH FULPLQDO DWWDFNV HFRQRPLF DWWDFNV ORVV RI SURSULHWDU\ LQIRUPDWLRQ LQGXVWULDO GLVDVWHU QDWXUDO GLVDVWHU HTXLSPHQWSODQW PDOIXQFWLRQ OHJDO SUREOHP SHUFHSWXDOUHSXWDWLRQDO KXPDQ UHVRXUFHVRFFXSDWLRQDO HQYLURQPHQWDOKHDOWK DQG UHJXODWRU\ ,Q KHU FDVH VWXG\ ERRN .DWKOHHQ )HDP%DQNV f OLVWV WKH IROORZLQJ ILYH W\SHV RI FULVLV SURGXFW WDPSHULQJ HQYLURQPHQWDO QDWXUDO GLVDVWHUV YLROHQFH DQG FHOHEULWLHV DQG FULVHV $FFRUGLQJ WR /HUELQJHU f WKHUH DUH VHYHQ FULVLV W\SHV QDWXUDO WHFKQRORJLFDO FRQIURQWDWLRQ PDOHYROHQFH VNHZHG PDQDJHPHQW YDOXHV GHFHSWLRQ DQG PDQDJHPHQW PLVFRQGXFW (QYLURQPHQWDO 0RQLWRULQJ LQ 3XEOLF 5HODWLRQV %XVLQHVV FULVHV SXW FRUSRUDWH UHSXWDWLRQV DQG HFRQRPLF VXUYLYDO WR WKH WHVW (IIHFWLYH FULVLV FRPPXQLFDWLRQ SODQQLQJ QRW RQO\ UHVXOWV LQ JRRG LVVXH PDQDJHPHQW EXW LQ IDYRUDEOH SXEOLF SHUFHSWLRQ RI WKH FRUSRUDWLRQ 2QH RI +HDULWfV f VXJJHVWLRQV WR

PAGE 44

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f VHYHUDO UHDVRQV IRU PRQLWRULQJ PHGLD FRYHUDJH DUH LGHQWLILHG 7KH HIIHFWLYHQHVV RI JHQHUDO LVVXH PDQDJHPHQW FDQ EH GHWHUPLQHG E\ DQDO\]LQJ WKH WRQH RI PHGLD FRYHUDJH 7KH LQIRUPDWLRQ REWDLQHG IURP WKH PHGLD FRQWHQW DOVR KHOSV SXEOLF UHODWLRQV SHRSOH H[DPLQH SXEOLF SHUFHSWLRQ RI DQ RUJDQL]DWLRQfV UHVSRQVH WR D FULVLV 0RQLWRULQJ DQG DQDO\]LQJ PHGLD FRYHUDJH KHOS SXEOLF UHODWLRQV SUDFWLWLRQHUV DXGLW WKH UHSXWDWLRQ RI WKH RUJDQL]DWLRQ 6FUXWLQL]LQJ PHGLD FRYHUDJH RQ LVVXHV SURYLGHV SXEOLF UHODWLRQV SUDFWLWLRQHUV LGHDV DERXW KRZ WR SRVLWLRQ DQ RUJDQL]DWLRQ LQ IURQW RI SXEOLFV DQG KRZ WR PRGLI\ WKH VWUDWHJLHV RU WDFWLFV ZKHQ KDQGOLQJ D FULVLV 8QGHUVWDQGLQJ PHGLD FRYHUDJH LV RQH SDUW RI IRUPDWLYH UHVHDUFK LQ WKH LPSOHPHQWDWLRQ RI SXEOLF UHODWLRQV VWUDWHJLHV

PAGE 45

3XEOLF 2SLQLRQ LQ 3XEOLF 5HODWLRQV 3XEOLF RSLQLRQ UHVHDUFK KDV DOZD\V EHHQ D YLWDO WRSLF DFURVV D GLYHUVLW\ RI VRFLDO VFLHQFH ILHOGV VXFK DV SROLWLFDO VFLHQFH SV\FKRORJ\ DQG PDVV FRPPXQLFDWLRQ 7R XQGHUVWDQG SXEOLF DWWLWXGHV SROOV DUH RIWHQ HPSOR\HG DV DQ HIILFLHQW ZD\ RI PHDVXULQJ WKH SXEOLF DJHQGD RQ PDMRU LVVXHV :KHWKHU SXEOLF RSLQLRQ UHSUHVHQWV DQ DJJUHJDWLRQ RI LQGLYLGXDO YLHZV RU D FROOHFWLYH PRYHPHQW UHPDLQV GHEDWDEOH 3ULFH f 3XEOLF RSLQLRQ PRUH RIWHQ WKDQ QRW FDQ GHWHUPLQH ZKDW JRYHUQPHQWV DQG FRUSRUDWLRQV GR ZKHQ HQFRXQWHULQJ WKH XQH[SHFWHG FULVLV ,Q D W\SLFDO FULVLV VLWXDWLRQ WKH SXEOLF SHUFHLYHV fWKH WUXWKf WR EH ZKDWHYHU SXEOLF RSLQLRQ VD\V WKDW LW LV 3XEOLF UHODWLRQV SUDFWLWLRQHUV DUH UHVSRQVLEOH IRU VXUYH\LQJ TXHVWLRQV RI SXEOLF RSLQLRQ 7KHUHIRUH LW LV YLWDO IRU DQ RUJDQL]DWLRQ WR SURYH WR LWV VWDNHKROGHUV WKDW WKH SUHYDLOLQJ QHJDWLYH SXEOLF RSLQLRQ LV QRW DFFXUDWH 3XEOLF RSLQLRQ FDQ DQG PXVW EH VKLIWHG WR WKH FRUSRUDWLRQfV DGYDQWDJH LI DQ RUJDQL]DWLRQ GHVLUHV WR HPHUJH IURP D FULVLV VLWXDWLRQ ZLWK D PLQLPXP RI GDPDJH WR LWV UHSXWDWLRQ ,W LV GLIILFXOW HQRXJK WR GHDO ZLWK D VLWXDWLRQ LQ ZKLFK WKH FRUSRUDWLRQ LV LQGHHG JXLOW\ EXW HYHQ PRUH IUXVWUDWLQJ LV FRSLQJ ZLWK EDG LQIRUPDWLRQ DQG IDOVH DFFXVDWLRQV /RRNLQJ EDFN WR SDVW FULVLV HYHQWV XQGHU YDULRXV FLUFXPVWDQFHV PDQ\ FRPSDQLHVf§VXFK DV ([[RQ 1$6$ 'HQQ\nV DQG ,QWHOf§IDLOHG WR FRPPXQLFDWH HIIHFWLYHO\ ZLWK WKH JHQHUDO SXEOLF LQFOXGLQJ WKHLU OR\DO FXVWRPHUV ZKHQ WKH\ IDFHG D FULVLV RU SUHn FULVLV VLWXDWLRQ &RQYHQWLRQDO SUHVV FRQIHUHQFHV RU JHQHUDO SUHVV VWDWHPHQWV PD\ QRW VXEGXH WKH DQ[LHWLHV RI SHRSOH ZKR DUH GLUHFWO\ LQIOXHQFHG E\ D FULVLV &RUSRUDWLRQV QHHG WR JR GLUHFWO\ WR WKH SXEOLF WR JDXJH KRZ ZHOO SXEOLF UHODWLRQV HIIRUWV DUH ZRUNLQJ WKLV LV ZKHUH SURDFWLYH WDFWLFV EHFRPH HVVHQWLDO

PAGE 46

'DPDJH FRQWURO LV WKH NH\ REMHFWLYH RI FULVLV PDQDJHPHQW 7KH PRUH D SXEOLF UHODWLRQV SUDFWLWLRQHU NQRZV DERXW KRZ D FULVLV LV EHLQJ LQWHUSUHWHG E\ WKH PHGLD DQG WKH JHQHUDO SXEOLF WKH PRUH HIIHFWLYH WKH GDPDJH FRQWURO ZLOO EH 7KH FRPPXQLFDWLRQ FKDOOHQJH LV WR FRQYLQFH YDULRXV DXGLHQFHV RI WKH FRPSDQ\n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f 0HGLD SOD\ D FUXFLDO UROH LQ XQGHUVWDQGLQJ SXEOLF RSLQLRQ WUHQGV RQ LVVXHV $OWKRXJK WKH LPSDFW RI PDVV PHGLD RQ DWWLWXGH IRUPDWLRQ DQG FKDQJH KDV SURYHQ WR EH OHVV LQIOXHQWLDO WKDQ VRPH H[SHFWHG .ODSSHU f PDVV PHGLD VWLOO IXQFWLRQ DV D SULPDU\ UHVRXUFH IRU LQIRUPDWLRQ DERXW PDQ\ WRSLFV 7KHUHIRUH LW EHFRPHV LPSRVVLEOH WR LJQRUH ZKDW PDVV PHGLD FRYHU RU VD\ DERXW DQ LVVXH SHUWLQHQW WR DQ RUJDQL]DWLRQ ZKLOH VLPSO\ SURELQJ SXEOLF RSLQLRQ 0HGLD LPSDFW RQ VSUHDGLQJ D FULVLV LV RIWHQ WLPHGHSHQGHQW ,W LV EHWWHU WR FRQWURO DQ LVVXH LQ WKH HDUO\ VWDJH WKDQ ZDLW XQWLO LW EHFRPHV D FULVLV PDQDJHPHQW LVVXH $V UHFRPPHQGHG E\ %LUFK f WKH XVH RI WUDFNLQJ UHVHDUFK HYHQ GDLO\ LQ WKH HDUO\

PAGE 47

VWDJHV DOORZV FRUSRUDWH PDQDJHPHQW WR NQRZ ZKDW WKH JHQHUDO SXEOLF RU NH\ VWDNHKROGHUV DUH UHDOO\ WKLQNLQJ DQG VD\LQJ 3HUPDQHQW PRQLWRULQJ RI DFWLYH SXEOLFV JHWWLQJ FRQWLQXRXV IHHGEDFN WKURXJK GLIIHUHQW FKDQQHOV IURP WKHVH SXEOLFV DOVR IDFLOLWDWHV LVVXH WUDFNLQJ 8PDQVN\ 9HQGUHOO f ,Q VXPPDU\ ZKHQ GHDOLQJ ZLWK DQ XQH[SHFWHG FULVLV WKH WZRZD\ UHVHDUFK PRQLWRULQJ ERWK WKH PHGLD DQG SXEOLF DJHQGD LV SLYRWDO 'XULQJ D FRUSRUDWH FULVLV FRPPXQLFDWLQJ ZLWK WKH JHQHUDO SXEOLF LV YHU\ LPSRUWDQW &ULVLV FRPPXQLFDWLRQ WHQGV WR HYROYH WKURXJK WKUHH VWDJHV PDVV PHGLD SXEOLFLW\ SXEOLF RSLQLRQ DURXVDO DQG WKHQ SXEOLF SROLF\ PDNHUVf UHVSRQVH 0D\HU f 3XEOLF RSLQLRQ VWHPV IURP WKH PHGLD LVVXH FRYHUDJH DQG WKHQ LPSDFWV JRYHUQPHQW SROLF\ PDNHUV $W WKH LQLWLDO SKDVH RI D FULVLV DQ RUJDQL]DWLRQ PD\ KDYH WKH SRZHU WR LQIOXHQFH HYHQWV $IWHU WKH FULVLV KDV UHFHLYHG ODUJH DPRXQWV RI DWWHQWLRQ IURP ERWK PHGLD DQG WKH SXEOLF WKH DELOLW\ RI WKH FRPSDQ\ WR FRQWURO LWV SXEOLF UHSXWDWLRQ GLPLQLVKHV ,Q KHU ERRN RQ FULVLV PDQDJHPHQW )HDP%DQNV f WHUPV fFRQWDLQPHQWf DV HIIRUWV H[HUWHG E\ WKH RUJDQL]DWLRQ WR OLPLW WKH GXUDWLRQ RI WKH FULVLV RU NHHS LW IURP VSUHDGLQJ WR RWKHU DUHDV DIIHFWLQJ WKH RUJDQL]DWLRQ $Q HIIHFWLYH FULVLV FRPPXQLFDWLRQV SODQ PXVW LQYROYH D VRXQG PHGLD PRQLWRULQJ SURJUDP WR HQVXUH DFFXUDF\ RI PHGLD FRYHUDJH 7UDYHUVR f XUJHG WKDW D PHGLD PRQLWRULQJ SURJUDP LV QHFHVVDU\ ZKHQ IDFLQJ WKH FULVLV DV PHGLD SURYLGH D IDVW FKDQQHO WR GHOLYHU LQIRUPDWLRQ WR WKH SXEOLF 6LWXDWHG LQ VXFK D G\QDPLF LQIRUPDWLRQ HUD 6WUHQVNL f SRLQWHG RXW WKDW D NH\ DUHD LQ ZKLFK SXEOLF UHODWLRQV SURIHVVLRQDOV PXVW H[FHO LQ FULVLV FRQWURO LV WR WDNH UHVSRQVLELOLW\ IRU PRQLWRULQJ WKH RQOLQH VHUYLFHV VXFK DV %%6V DQG DOO NLQGV RI QHZVJURXSV WKDW H[LVW LQ F\EHUVSDFH +H ZDUQHG WKDW LJQRULQJ WKHVH

PAGE 48

JURXSV LV PLVVLQJ D PDMRU SRUWLRQ RI WKH SRWHQWLDO DXGLHQFH .RWFKHU f RIIHUHG UHDVRQLQJ ZK\ QHZ FRPPXQLFDWLRQ PHGLD PXVW QRW EH QHJOHFWHG E\ SXEOLF UHODWLRQV SUDFWLWLRQHUV +H VWDWHV &RPPXQLFDWRUV FDQ QR ORQJHU RSHUDWH XQGHU WKH ROG DVVXPSWLRQ WKDW PDVV FRPPXQLFDWLRQ LV SDVVLYH LQ QDWXUH DQG WKDW LQWHUDFWLYH SHUVRQDO FRPPXQLFDWLRQ FDQ RQO\ UHDOO\ RFFXU LQ VPDOO JURXSV RU IDFHWRIDFH ,Q DGGLWLRQ FRPPXQLFDWRUV VKRXOG QRW DVVXPH WKHLU SUHVHQW XQGHUVWDQGLQJ RI WHFKQRORJ\ DQG LWV DSSOLFDWLRQV ZRXOG EH VXIILFLHQW DW D WLPH RI FULVLV S f &ULVHV FDQ VWDUW ZLWK UXPRUV ZKLFK FDQ EH DEVROXWHO\ IDOVH RU SDUWO\ IDOVH ZLWKRXW VXEVWDQWLDO YHULILFDWLRQ DQG FUHGLEOH VRXUFHV ,Q KHU WHQ JXLGHOLQHV IRU UHGXFLQJ OHJDO ULVNV LQ FULVLV PDQDJHPHQW )LW]SDWULFN f VWUHVVHG WKH XUJHQW QHHG WR UHVSRQG WR UXPRUV VSUHDGLQJ DOO RYHU WKH YDULHW\ RI LQIRUPDWLRQ FKDQQHOV WKDW PD\ WXUQ RXW WR FDXVH D ORQJn WHUP VXVWDLQHG FULVLV IRU D FRPSDQ\ 6WXUJHVf 3XEOLF 2SLQLRQ 0RGHO RI &ULVLV 0DQDJHPHQW &RPPXQLFDWLRQ LV DQ LPSRUWDQW WRRO LQ KDQGOLQJ WKH LPSUHVVLRQ RI NH\ VWDNHKROGHUV DW WKH WLPH RI FULVLV 'f$YHQL HW DO f 6WXUJHV f HPSKDVL]HV WKH YDOXH RI FRPPXQLFDWLRQ GXULQJ WKH FULVLV WR WKH SXEOLFV 7KH FULVLV GRHV QRW KDSSHQ LQ LVRODWLRQ VR WKH FRPPXQLFDWLRQ SODQ VKRXOG FRQVLGHU KRZ SXEOLF RSLQLRQ LQWHUDFWLQJ ZLWK RWKHU VRFLDO HYHQWV DQG HQYLURQPHQW LV IRUPHG EHIRUH GXULQJ DQG DIWHU WKH FULVLV +H LOOXVWUDWHV WKH HYROXWLRQ RI JURXS RSLQLRQ WKURXJK DQ HLJKWVWHS SURFHVV DV VKRZQ LQ )LJXUH

PAGE 49

6RFLDO 1RUP 6RFLDO $FWLRQ 3XEOLFfV 2SLQLRQ )RUPV /DWHQW ,VVXH (YHQW 2FFXUV ? 'HEDWH 2FFXUV ),*85( *URXS RSLQLRQ IRUPDWLRQ SURFHVV ,Q WKH ILUVW VWDJH DQ LVVXH PXVW EH VDOLHQW EXW ODWHQW IRU WKH SXEOLF PHPEHUV $V SXEOLF DZDUHQHVV LV WKH QHFHVVDU\ FRQGLWLRQ LQ WKH IRUPXODWLRQ RI SXEOLF RSLQLRQ DQ HYHQW RFFXUV DQG FRPHV LQWR SHRSOHfV PLQG LQ WKH QH[W VWHS DIWHU WKH\ EHFRPH DZDUH RI WKDW HYHQW 7KHQ WKH SUR DQG FRQ IDFWLRQV DPRQJ WKH PHPEHUV VROLGLI\ LQ WKH WKLUG VWDJH IROORZHG E\ SXEOLF GHEDWH DPRQJ WKH PHPEHUV VWHS f $IWHU D WLPH ODSVH RFFXUV VWHS f ZKHQ HQRXJK KDV EHHQ VDLG DQG GRQH SXEOLF RSLQLRQ WKXV HPHUJHV LQ VWHS ,Q VWHS VRPH DFWLRQ ZLOO EH XVHG WR UHLQIRUFH WKH SXEOLF RSLQLRQ IRUPHG LQ WKH SUHYLRXV VWHS )LQDOO\ GXH WR WKH SHUIRUPDQFH RI DFWLRQ D VRFLDO QRUP LV JRLQJ WR EH HVWDEOLVKHG

PAGE 50

6WXUJHV DSSOLHV WKLV FRQFHSW RI RSLQLRQ IRUPDWLRQ WR WKH GLUHFWLRQ RI SXEOLF RSLQLRQ LQ WKH FULVLV VLWXDWLRQ DV LOOXVWUDWHG LQ )LJXUH 7KH SXEOLF RSLQLRQ HLWKHU QHJDWLYH RU SRVLWLYH H[SDQGV DV WKH FULVLV HUXSWV DQG IDGHV DZD\ ZKLOH WKH FULVLV DEDWHV DQG HQGV 7KHUHIRUH WKH REMHFWLYH RI FULVLV FRPPXQLFDWLRQ LV fWR LQIOXHQFH WKH SXEOLF RSLQLRQ GHYHORSPHQW WR WKH SRLQW WKDW RSLQLRQV KHOG LQ WKH SRVWFULVLV SHULRG DUH DW WKH VDPH OHYHO RU JUHDWHU LQ SRVLWLYH RSLQLRQV RU ORZHU LQ QHJDWLYH RSLQLRQf S f DPRQJ WKH PHPEHUV RI WKH SXEOLF f§ f§1HJDWLYH 2SLQLRQ nrr AAAf3RVLWLYH 2SLQLRQ 7HUPLQDWLRQ ),*85( 6WXUJHVn PRGHO RI SXEOLF RSLQLRQ ZLWK FULVLV PDQDJHPHQW SODQ 0DF.XHQfV 6WUDWHJ\ 0RGHO RI 3XEOLF 'LDORJXH 5HJDUGOHVV RI WKH IRUP IDFHWRIDFH HOHFWURQLF EXOOHWLQ ERDUGV QHZVJURXSV HWFf RQH FKRRVHV WR WDNH SDUW LQ WKH SROLWLFDO FRQYHUVDWLRQ 0DF.XHQfV f GLVFXVVLRQ DERXW SHUVRQDO VWUDWHJLHV LQ WKH SXEOLF GLDORJXH WR SURPRWH GHOLEHUDWLYH GHPRFUDF\ LV HPSOR\HG WR H[DPLQH KRZ RQOLQH XVHUV JHW LQYROYHG LQ WKH GLVFXVVLRQ JURXSV 7KH EDVLF LGHD EHKLQG 0DF.XHQfV VWUDWHJ\ PRGHO IRU SXEOLF GLDORJXH LV WKDW GHOLEHUDWLYH GHPRFUDF\ LV D IXQFWLRQ RI SROLWLFDO SDUWLFLSDQWV ZKR LQWHUDFW ZLWK RQH DQRWKHU 7KHVH LQWHUDFWLRQV VHW WKH SROLWLFDO DQG LQWHOOHFWXDO FRQWH[WV IRU GHPRFUDWLF GHFLVLRQV DQG PD\ WDNH WZR IRUPV 2QH Z ? A ? 3UHFULVLV %XLOG 8S %UHDN 2XW $EDWHPHQW

PAGE 51

RFFXUV EHWZHHQ WKRVH ZKR KROG VLPLODU YLHZV DQG DQRWKHU VWHPV IURP LQGLYLGXDOV RI GLIIHUHQW YLHZV +RZHYHU RQO\ WKH ODWWHU DOORZV JHQXLQH GHEDWH DQG DQ H[FKDQJH RI LGHDV $V D UHVXOW WKH FRQYHUVDWLRQ ZKLFK HDFK LQGLYLGXDO FUHDWHV EHFRPHV FUXFLDO LQ VKDSLQJ SXEOLF RSLQLRQ DQG WKHQ GHYHORSV LQWR GHOLEHUDWLYH GHPRFUDF\ 5HJDUGLQJ HQJDJHPHQW LQ SROLWLFDO GLVFRXUVH 0DF.XHQ KROGV WKDW RQHfV ZLOOLQJQHVV WR HQJDJH LQ SXEOLF GLVFRXUVH GHSHQGV fQRW RQO\ RQ KRZ PXFK RQH HQMR\V GLVFXVVLRQ RU KDWHV FRQWURYHUV\ EXW DOVR RQ WKH OLNHOLKRRG RI UXQQLQJ LQWR V\PSDWKHWLF DQG FRQWUDU\ YLHZSRLQWVf S f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f WKH UDWLR RI SOHDVXUH IURP UHZDUGLQJ FRQYHUVDWLRQ WR SDLQ RI GLVDJUHHPHQWV WR GHSLFW KRZ DQ LQGLYLGXDO IHHOV DERXW WKH FRQYHUVDWLRQ DQG KLV RU KHU RZQ WROHUDQFH RI RSSRVLQJ YLHZSRLQWV ,W DOVR UHSUHVHQWV D SRVLWLYH LQFHQWLYH WR HQJDJH LQ WKH SROLWLFDO FRQYHUVDWLRQ /RZ OHYHOV RI ([SUHVVLYLW\ UHTXLUH IULHQGO\ HQYLURQPHQWV WR PDNH FRQYHUVDWLRQ ZRUWKZKLOH %\ FRQWUDVW KLJK OHYHOV RI ([SUHVVLYLW\ UHQGHU SROLWLFDO FRQYHUVDWLRQ DWWUDFWLYH HYHQ ZKHQ IDFLQJ RSSRVLQJ PDMRULWLHV

PAGE 52

$SSOLFDWLRQV RI )RXU 7\SHV RI 6WUDWHJLHV 7KH RXWFRPH RI XVH RI 7$/.&/$0 VWUDWHJLHV LQ WKH SROLWLFDO GLDORJXH LV VKRZQ LQ 7DEOH DQG 7$%/( 3D\RII PDWUL[ IRU SOD\HU HQFRXQWHULQJ IULHQG )ULHQGfV 6WUDWHJ\ 7$/. &/$0 3OD\HUfV 6WUDWHJ\ 7$/. 5HLQIRUFHPHQW %OXVWHU &/$0 0XVLF 3DVV 7$%/( 3D\RII PDWUL[ IRU SOD\HU HQFRXQWHULQJ IULHQG )ULHQGfV 6WUDWHJ\ 7$/. 5($&7 &/$0 7$/. 5HLQIRUFHPHQW 5HLQIRUFHPHQW %OXVWHU 3OD\HUfV 5($&7 5HLQIRUFHPHQW 3DVV 3DVV 6WUDWHJ\ &/$0 0XVLF 3DVV 3DVV

PAGE 53

$V VHHQ LQ 7DEOH WKH FRQYHUVDWLRQ WKDW UHLQIRUFHV GHPRFUDWLF GLDORJXH EHWZHHQ WZR SDUWLHV ZKR UHZDUG VHOIH[SUHVVLRQ ZLOOLQJQHVV WR VSHDN RXW KDSSHQV RQO\ LI ERWK LQGLYLGXDOV FKRRVH 7$/. VWUDWHJ\ %\ WKH VDPH WRNHQ LI ERWK LQGLYLGXDOV FKRRVH &/$0 VWUDWHJ\ D QRQSROLWLFDO FRQYHUVDWLRQ 3DVVf ZLOO IROORZ 7KH WKHPH LQ WKLV PDWUL[ RI HPSOR\PHQW RI ERWK VWUDWHJLHV LV WKDW DOO RXWFRPHV EXW 5HLQIRUFHPHQW DQG 'LVDJUHHPHQW UHVXOW LQ OLWWOH RU WHPSRUDU\ YDOXH WKURXJK WKH LQWHUDFWLYH FRQYHUVDWLRQ 2WKHU WKDQ WKH DIRUHPHQWLRQHG WZR VWUDWHJLHV DQ LQGLYLGXDO PLJKW RSW IRU WKH WKLUG RQHf§ 5($&7 $V VKRZQ LQ 7DEOH WKH XQLTXH DGYDQWDJH RI WKLV VWUDWHJ\ LV WR DYRLG GLVDJUHHPHQWV ZKHQ ERWK SOD\HU DQG IULHQG DGRSW HLWKHU 7$/. RU 5($&7 VWUDWHJ\ LQ WKH FRQYHUVDWLRQ +RZHYHU LW DOVR OLPLWV WKH DWWUDFWLYHQHVV RI WKH 7$/. RSWLRQ LI DQ RSSRQHQW LV LQYROYHG %\ XVLQJ WKLV VWUDWHJ\ D SHUVRQ FDQ FKHHUIXOO\ UHVSRQG WR IULHQGO\ 7$/.(56 DQG UHPDLQLQJ VLOHQW EHIRUH RSSRVLQJ YLHZSRLQWV 7KH 5($&725 QHYHU LQLWLDWHV D SROLWLFDO FRQYHUVDWLRQ EXW VLPSO\ IROORZV WKH SDUWQHUV ZLWK ZKRP VKH RU KH DJUHHV $V WKH QXPEHU RI 7$/.LQJ IULHQGV LQFUHDVHV WKH DWWUDFWLYHQHVV RI 5($&7 ZLOO DOVR LQFUHDVH OLNHZLVH 7KH ODVW VWUDWHJ\ 0DF.XHQ PHQWLRQV LV FDOOHG 6,*1$/ 7KH FRQVHTXHQFH RI VLJQDOLQJ IRU LQGLYLGXDO FKRLFH SURGXFHV D FRPPXQLW\ RI LVRODWHG JURXSV DQG HOLPLQDWHV PHDQLQJIXO SXEOLF GLDORJXH 6R WR VSHDN VLJQDOLQJ QRW RQO\ SUHYHQWV D FRQVLGHUDWLRQ RI DOWHUQDWLYH YLHZV EXW DOVR JHQHUDWHV D FRQVLVWHQF\ RI RSLQLRQV DURXQG WKH V\PEROV HPERGLHG LQ WKH VLJQDO LWVHOI S f %HFDXVH WKH OHYHO RI ([SUHVVLYLW\ LQ WKH SXEOLF HQYLURPQHQW GHWHUPLQHV DQ LQGLYLGXDOfV ZLOOLQJQHVV WR HQJDJH LQ WKH SROLWLFDO GLDORJXH 0DF.XHQfV LQWHJUDWHG

PAGE 54

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£OH]+HUUHUR HW DK +DLQVZRUWK 0HQJ f :LWK WKH RSLQLRQV LGHQWLILHG DV SRVLWLYH RU QHJDWLYH RU FRPELQDWLRQ RI ERWK SHRSOH WHQG WR UHDFW SRVLWLYHO\ WR JRRG QHZV DQG QHJDWLYHO\ WR EDG QHZV 6LJHOPDQ HW DK f %DVHG XSRQ 6WXUJHVf f SXEOLF RSLQLRQ IRUPDWLRQ PRGHO GXULQJ FULVLV QHJDWLYH SXEOLF RSLQLRQ DERXW WKH FRUSRUDWLRQ DW WKH EUHDNRXW RI WKH FULVLV UHDFKHV LWV KLJKHVW SRLQW DQG WKHQ HEEV WR WKH SUHn FULVLV VWDJH DIWHU WKH FULVLV $V 6WXUJHV GHPRQVWUDWHV HDUOLHU LQ )LJXUH LW LV DQWLFLSDWHG WKDW WKH KLJKHVW QXPEHU RI QHJDWLYH PHVVDJHV DERXW DQ RUJDQL]DWLRQ ZLOO EH SRVWHG DV WKH FULVLV HUXSWV IROORZLQJ WKH LQLWLDO PHGLD FRYHUDJH )ORZ ZLOO SXEOLF RSLQLRQ RQ DQ LQWHUDFWLYH FRPPXQLFDWLRQ FKDQQHO VXFK DV WKH ,QWHUQHW GLIIHU IURP 6WXUJHVf PRGHO RQ SXEOLF RSLQLRQ GXULQJ D FULVLV" $V FRPPHUFLDO SURYLGHUV LQFUHDVH RQ WKH ,QWHUQHW DQG PRUH SROLWLFDO LQIRUPDWLRQ LV SURYLGHG WKH SUREOHP RI ZKR VHWV WKH DJHQGD IRU WKH QHZ PHGLXP DOVR EHFRPHV D FRQFHUQ %HKU HW DK f PDLQWDLQHG WKDW SXEOLF FRQFHUQV DERXW HQHUJ\ DQG LQIODWLRQ ZHUH GHWHUPLQHG E\ WHOHYLVLRQ FRYHUDJH ZKLFK LQ WXUQ ZDV GHSHQGHQW XSRQ ZRUOG

PAGE 55

FRQGLWLRQV %\ FRPSDULQJ WKH FRQWHQWV RI WKUHH ORFDO DQG UHJLRQDO GDLO\ QHZVSDSHUV ZLWK WKH UHVXOWV VXUYH\LQJ WKUHH ZDYHV RI UHVSRQGHQWV GUDZQ IURP WKH VDPH VWXG\ SRSXODWLRQ DFURVV HLJKW PRQWKV 6DOZHQ f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f 7R ZKDW GHJUHH LV WKH IUHTXHQF\ RI RQOLQH GLDORJXHV RQ WKH 86(1(7 DIIHFWHG E\ DQ RUJDQL]DWLRQfV VWUDWHJLF FULVLV LQWHUYHQWLRQ" 0DF.XHQ f GHPRQVWUDWHV WKDW SHRSOH ZKR IHHO KLJK OHYHOV RI ([SUHVVLYLW\ ZLOO PRUH OLNHO\ DGRSW D 7$/. VWUDWHJ\ WKDQ WKRVH ZKR GR QRW $OWKRXJK &/$0 LV RQH RI WKH VWUDWHJLHV XVHG E\ LQGLYLGXDOV WR HQJDJH LQ WKH SXEOLF GLDORJXH LW LV LPSRVVLEOH WR GHWHFW ZKR UHPDLQV VLOHQW E\ DGRSWLQJ WKLV VWUDWHJ\ LQ WKH QHZVJURXSV GLVFXVVLRQ &/$0 XVHUV DUH UHJDUGHG DV WKRVH ZKR UHDG WKH PHVVDJHV ZLWKRXW SRVWLQJ WKHLU RZQ RSLQLRQV RQ WKH QHZVJURXSV

PAGE 56

+RZ ZLOO WKH IUHTXHQF\ DQG WRQH RI RQOLQH GLDORJXHV RQ WKH 86(1(7 SRVWHG E\ 7$/.(56 DQG 5($&7256 YDU\ LQ WKH OLJKW RI 0DF.XHQfV VWUDWHJ\ PRGHO" 5HVHDUFK +\SRWKHVHV *LYHQ WKDW WKH ,QWHUQHW SURYLGHV SXEOLF UHODWLRQV SURIHVVLRQDOV DFFHVV WR OLVWHQLQJ WR ZKDW SHRSOH UHDOO\ WKLQN DQG VD\ DERXW WKH FULVLV RU WKH WDUJHWHG FRUSRUDWLRQ RQ WKH 86(1(7 K\SRWKHVHV LQ WKLV VWXG\ VWUHVV WKH FKDQJH RI QXPEHU RI RQOLQH GLVFXVVLRQ PHVVDJHV EHIRUH GXULQJ DQG DIWHU WKH FRUSRUDWH FULVLV $QDO\VLV RI WKHVH PHVVDJHV ZLOO DOVR KHOS DGYDQFH WKH PHDVXUHPHQW RI WKH SXEOLF DJHQGD LQ DJHQGDVHWWLQJ WKHRU\ 6WXUJHV f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n PRGHO LV WKDW SXEOLF QHJDWLYH RSLQLRQ ZLOO FRPSDUDWLYHO\ LQFUHDVH PRUH WKDQ SXEOLF SRVLWLYH RSLQLRQ VLQFH WKH FULVLV FDQ KDYH QHJDWLYH LPSDFW RQ WKH VRFLHW\ DV ZHOO DV WKH JHQHUDO SXEOLF 7KHUHIRUH E\ XVLQJ DOO WKH VDPSOLQJ GD\V DIWHU WKH FULVLV HUXSWLRQ DV D WLPH IUDPH D VHFRQG K\SRWKHVLV LV IRUPHG

PAGE 57

+\SRWKHVLV 7KH DYHUDJH QXPEHU RI GDLO\ QHJDWLYH 86(1(7 GLDORJXHV ZLOO EH JUHDWHU WKDQ WKH DYHUDJH QXPEHU RI GDLO\ SRVLWLYH 86(1(7 GLDORJXHV DIWHU WKH HUXSWLRQ RI WKH FULVLV 1XPHURXV DJHQGDVHWWLQJ VWXGLHV OHDG WR D FRPPRQ FRQFOXVLRQ WKDW QHZV PHGLD FRYHUDJH RI DQ LVVXH KDV FRQVLGHUDEOH LPSDFW RQ WKH SXEOLF DJHQGD &RQFHUQLQJ WKH ILUVW OHYHO RI DJHQGDVHWWLQJ WKHRU\ WKH PRUH PHGLD FRYHUDJH RI DQ LVVXH WKH PRUH VDOLHQW WKDW LVVXH ZLOO EHFRPH LQ SHRSOHnV PLQGV 7KH QH[W K\SRWKHVLV LV KHQFH PDGH EDVHG XSRQ WKH DJHQGDVHWWLQJ FRQFHSW DV GLVSOD\HG LQ )LJXUH ),*85( &DXVDO UHODWLRQVKLS EHWZHHQ WKH PHGLD DJHQGD DQG WKH SXEOLF DJHQGD +\SRWKHVLV 7KH GDLO\ QXPEHU RI WKH 86(1(7 GLDORJXHV ZLOO EH UHODWHG WR WKH GDLO\ QXPEHU RI QHZV VWRULHV DERXW D FULVLV DIWHU LWV HUXSWLRQ ,Q 6WXUJHVn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cf§\ 5HSRUWV

PAGE 58

'HULYHG IURP WKH DERYH K\SRWKHWLFDO DVVXPSWLRQ LW LV H[SHFWHG WKDW WKH LQWHUYHQWLRQ VWUDWHJLHV ZLOO KDYH SRVLWLYH LQIOXHQFHV RQ SHRSOHnV RSLQLRQ DERXW WKH FRUSRUDWLRQ ,Q 6WDJHVn PRGHO WKH DPRXQW RI SXEOLF SRVLWLYH RSLQLRQ DERXW WKH FRPSDQ\ UHPDLQV VLPLODU DW ERWK SUHFULVLV DQG DEDWHPHQW VWDJHV VR WKH QH[W K\SRWKHVLV ZLOO EH DV IROORZV +\SRWKHVLV )ROORZLQJ D FULVLV LQWHUYHQWLRQ VWUDWHJ\ WKH DYHUDJH QXPEHU RI GDLO\ SRVLWLYH RQOLQH GLDORJXHV RQ WKH 86(1(7 FRQFHUQLQJ WKH FRPSDQ\ ZLOO EH HTXDO WR WKH DYHUDJH QXPEHU RI GDLO\ SRVLWLYH RQOLQH GLDORJXHV EHIRUH WKH FULVLV HUXSWLRQ $IWHU WKH LQWHUYHQWLRQ VWUDWHJLHV DPRXQW RI ERWK QHJDWLYH DQG SRVLWLYH RSLQLRQ IURP WKH SXEOLF GHFUHDVHV LQ FRPSDULVRQ ZLWK WKRVH ZKHQ WKH FULVLV HUXSWV DFFRUGLQJ WR 6WDJHVn PRGHO 7KXV WKH K\SRWKHVLV LV GHYHORSHG DV IROORZV +\SRWKHVLV 7KH QXPEHU RI WKH 86(1(7 GLDORJXHV IROORZLQJ D FULVLV LQWHUYHQWLRQ ZLOO EH OHVV WKDQ WKH QXPEHU RI WKH 86(1(7 GLDORJXHV EHIRUH WKH LQWHUYHQWLRQ DQG DIWHU WKH FULVLV 1R VWXG\ KDV DSSOLHG 0DF.XHQn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

PAGE 59

%\ XVLQJ FRPSXWHU VRIWZDUH WR DQDO\]H WKH WRQH RI PHGLD FRYHUDJH WKH K\SRWKHVLV LV IRUPHG WR WHVW WKH UHODWLRQVKLS EHWZHHQ WKH WRQH RI WKH PHGLD DJHQGD DQG WKH SXEOLF RQn OLQH GLDORJXHV +\SRWKHVLV 7KH 2SWLPLVP 6FRUH RI PHGLD PHVVDJHV ZLOO EH SRVLWLYHO\ UHODWHG WR WKH QXPEHU RI QHJDWLYHO\ WRQHG 86(1(7 PHVVDJHV SRVWHG DIWHU WKH PHGLD VWRULHV ZHUH UHOHDVHG 7KH QH[W FKDSWHU ZLOO GHVFULEH LQ GHWDLO WKH PHWKRG XVHG WR WHVW WKH K\SRWKHVHV DQG WR DQVZHU WKH UHVHDUFK TXHVWLRQV

PAGE 60

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

PAGE 61

WKH QH[W VHFWLRQ 7KH ILQDO VHFWLRQ LQWURGXFHV RI D FRPSXWHU VRIWZDUH FDOOHG ',&7,21 IRU WKH SXUSRVH RI WHVWLQJ WKH ODVW K\SRWKHVLV 'HVFULSWLRQ RI &RQWHQW $QDO\VLV 7KH PRVW SRSXODU DQG IUHTXHQWO\ FLWHG GHILQLWLRQ RI FRQWHQW DQDO\VLV LV RIIHUHG E\ %HUHOVRQ f f&RQWHQW DQDO\VLV LV D UHVHDUFK WHFKQLTXH IRU WKH REMHFWLYH V\VWHPDWLF DQG TXDQWLWDWLYH GHVFULSWLRQ RI WKH PDQLIHVW FRQWHQW RI FRPPXQLFDWLRQf S f .ULSSHQGRUII f DOVR GHILQHV FRQWHQW DQDO\VLV DV D UHVHDUFK WHFKQLTXH IRU PDNLQJ UHSOLFDEOH DQG YDOLG LQIHUHQFHV IURP GDWD WR WKHLU FRQWH[W 7KH LQIHUHQFHV LQYROYH WKH VHQGHU UHFHLYHU RI WKH PHVVDJH WKH PHVVDJH LWVHOI DV ZHOO DV WKH FRPPXQLFDWLRQ FKDQQHOV 7KLV XQREWUXVLYH TXDQWLWDWLYH PHDVXUH RI WH[WV VHUYHV PDQ\ SXUSRVHV VXFK DV LGHQWLILFDWLRQ RI LQWHQWLRQV RI WKH FRPPXQLFDWRUV XQGHUVWDQGLQJ RI DWWLWXGLQDO DQG EHKDYLRUDO UHVSRQVHV WR FRPPXQLFDWLRQV DQG GHVFULSWLRQ RI WUHQGV LQ FRPPXQLFDWLRQ FRQWHQW :HEEHU f &RUSRUDWH &ULVLV 6HOHFWLRQ 7KLV VWXG\ LV LQWHQGHG WR SUHGLFW WKH WUHQG RI SXEOLF GLDORJXHV RQ WKH 86(1(7 DERXW FRUSRUDWH FULVHV 7KH LPSDFW RI QHZV PHGLD FRYHUDJH DQG WKH FRPSDQ\fV LQWHUYHQWLRQ VWUDWHJ\ ZLOO EH LQWHJUDWHG LQWR WKH PRGHO 7ZR H[DPSOHV RI WKH VDPH W\SH RI FULVLV ZLOO EH UDQGRPO\ VHOHFWHG 7KH ILUVW FULVLV ZLOO EH DQDO\]HG DQG XVHG WR PRGLI\ WKH FRGLQJ FDWHJRULHV DQG WR LPSURYH WKH ZD\V RI UHWULHYLQJ D VXIILFLHQW QXPEHU RI PHGLD VWRULHV DQG RQOLQH GLVFXVVLRQ PHVVDJHV )RU WKH SXUSRVHV RI JHQHUDOL]DELOLW\ RI ILQGLQJV DQG WR XQGHUVWDQG WKH LQIOXHQFHV RI WKH FULVLV LQ WKH RQOLQH 86(1(7 GLDORJXHV SURGXFW WDPSHULQJ ZDV VHOHFWHG DV WKH WDUJHW FULVLV )HDP%DQNV f FRQVLGHUV SURGXFW WDPSHULQJ FULVHV DV D UHVXOW RI FODLPV

PAGE 62

H PDGH DJDLQVW PDQXIDFWXUHUVf SURGXFWV $IWHU DQDO\]LQJ QHZV VWRULHV DERXW W\SHV RI EXVLQHVV FULVLV ,UYLQH DQG 0LOODU f IRXQG WKDW FULVHV RI GHIHFWV t UHFDOOV RI SURGXFWV KDV LQFUHDVHG SHUFHQW EHWZHHQ DQG %HFDXVH SURGXFW WDPSHULQJ KDV WKH SXEOLF VDIHW\ DQG KHDOWK DW KLJK ULVN /HUELQJHU f FODVVLILHV LW DV D FULVLV RI PDOHYROHQFH ZKLFK RIWHQ QHFHVVLWDWHV SURGXFW UHFDOOV 3DVW H[DPSOHV LQFOXGH -RKQVRQ t -RKQVRQfV 7\OHQRO PXUGHUV LQ *HUEHUfV SURGXFW WDPSHULQJ LQ 3HUULHUfV UHFDOOLQJ RI ZDWHU ERWWOHV LQ DQG 3HSVLfV V\ULQJH WKUHDW LQ 7KH WLPH SHULRG IURP ZKLFK WKH WZR FULVLV FDVHV ZHUH VHOHFWHG ZDV IURP 0DUFK WR 'HFHPEHU 7KH FULVLV HYHQW ZDV FKRVHQ IRU VWXG\ EHFDXVH LW PHW WKH FRQGLWLRQV WKDW DUH SUHGHILQHG LQ WKH ODWWHU SDUW RI WKLV FKDSWHU 1HZV &RYHUDJH $QDO\VLV $ FRUSRUDWH FULVLV KDV HPHUJHG ZKHQ WKH QHZV PHGLD UHSRUW DQ LVVXH LQ D QHJDWLYH RU DGYHUVDU\ WRQH &RQWHQW DQDO\VLV RI FULVLV FRYHUDJH E\ WKH PHGLD KDV XVXDOO\ EHHQ XVHG WR JXLGH DQ RUJDQL]DWLRQf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t 0F&RPEV :DQWD t +X :LQWHU HW DO f :LWK WKH FRQVLGHUDEOH QXPEHU RI SXEOLF GLDORJXHV RQ WKH 86(1(7 WKH WLPH IUDPH IRU DQDO\VLV RI PHGLD FRYHUDJH LV VL[ ZHHNV GD\V EHIRUH RFFXUUHQFH RI D FULVLV

PAGE 63

DQG GD\V IROORZLQJ WKH FULVLV HUXSWLRQ $OWKRXJK RQOLQH GLVFXVVLRQ PD\ RU PD\ QRW HQG RQH PRQWK DIWHU RFFXUUHQFH RI WKH FULVLV WKH UDQGRP HUURU EHFRPHV ODUJHU DV PRUH GD\V RI PHVVDJHV DUH UHWULHYHG ZKHQ DSSO\LQJ VWDWLVWLFDO DQDO\VLV WR WHVW WKH HIIHFWV RI QHZV PHGLD FRYHUDJH LQ UHODWLRQ WR WKH QXPEHU RI RQOLQH GLDORJXHV 7KH RQOLQH /H[LV1H[LV GDWD EDVH VHUYHG DV D IRXQGDWLRQ RI PHGLD QHZV FRYHUDJH 7KH NH\ZRUGV IRU VHDUFKLQJ QHZV DUWLFOHV LQ WKH /H[LV1H[LV GDWD EDVH YDULHG GHSHQGLQJ RQ HDFK FULVLV 0HGLD 6HOHFWLRQ IRU $QDO\VLV RI 1HZV 6WRULHV 7HOHYLVLRQ SURYLGHV SROLWLFDO LQIRUPDWLRQ WR PRVW FLWL]HQV LQ :HVWHUQ LQGXVWULDO VRFLHWLHV DQG KDV EHFRPH WKH PRVW FUHGLEOH VRXUFH RI LQIRUPDWLRQ %URVLXV t .HSSOLQJHU f =XFNHU f PDLQWDLQHG WKDW DW WKH QDWLRQDO OHYHO WKH SXEOLF PD\ EH PRUH LQIOXHQFHG E\ WKH WKUHH QHWZRUNVf QHZVFDVWV WKDQ E\ QHZVSDSHUV EHFDXVH RI WHOHYLVLRQfV DFFHVVLELOLW\ 6KDZ DQG 0F&RPEV f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
PAGE 64

WKH VWXG\ EHFDXVH EHVLGHV KDYLQJ D QDWLRQDO FLUFXODWLRQ LW LV WKRXJKW WR KDYH DQ DJHQGDn VHWWLQJ HIIHFW RQ VPDOOHU PHGLD WKDW FRXOG EH VHHQ LQ VXUYH\ GDWD RI D FURVV VHFWLRQ RI WKH SRSXODWLRQ 7KH 1HZ
PAGE 65

$SSHQGL[ $f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fUHVSRQVLYH LQWHUDFWLYLW\f SURSRVHG E\ 5DIDHOL f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

PAGE 66

6WHSV WR 'RZQORDG 2QOLQH 0HVVDJHV %HFDXVH RQOLQH PHVVDJHV RI GD\ GXUDWLRQ ZRXOG EH GRZQORDGHG IRU DQDO\VLV WKH HIILFLHQW ZD\ LV WR XVH WKH SRZHU VHDUFK PHWKRG LQ WKH 'HMD 1HZV VHDUFK HQJLQH 6HH $SSHQGL[ %f ZKLFK HQDEOHV D XVHU WR VSHFLI\ D WLPH IUDPH ZLWKRXW ZRUU\LQJ DERXW KDYLQJ WRR PDQ\ PHVVDJHV EH\RQG WKH SDUWLFXODU WLPH VSDQ RI GD\V 7KH IROORZLQJ GHVFULEHV KRZ WR VHDUFK IRU WKH PHVVDJHV LQ WKH 'HMD 1HZV VHDUFK HQJLQH VWHS E\ VWHS 7\SH LQ WKH NH\ ZRUGV LQ WKH 6HDUFK IRU EODQN URZ 8VH fWKUHDGHGf DV WKH UHVXOWV IRUPDW DQG VRUWHG E\ f'DWHf /HDYH *URXSVf $XWKRUVf DQG 6XEMHFWVf EODQN 6SHFLI\ WKH EHJLQQLQJ DQG HQGLQJ GDWH LQ 'DWH IURP DQG 7R EODQN URZV &OLFN RQ )LQG EXWWRQ $IWHU WKH VHDUFK UHVXOWV VKRZHG RQ WKH VFUHHQ HYHU\ PHVVDJH ZDV GRZQORDGHG IRU IXUWKHU DQDO\VLV :KLOH GRZQORDGLQJ DOO SHUWLQHQW PHVVDJHV EDVHG RQ DERYH PHQWLRQHG FULWHULD WKH UHVHDUFKHU KDV DOVR FOLFNHG RQ fYLHZ WKUHDGf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

PAGE 67

GLVFXVV WKH FULVLV DV UHVSRQGLQJ WR RWKHU XVHUVn PHVVDJHV ZLWKRXW PHQWLRQLQJ DQ\ NH\ZRUGV WKDW DUH XVHG WR VHDUFK IRU FULVLVUHODWHG PHVVDJHV 'HILQLWLRQV RI &RGLQJ &DWHJRULHV 7KH FDWHJRULHV RI GLVFXVVLRQ PHVVDJHV QHHGHG WR EH GHILQHG LQ RUGHU WR FRPSDUH WKH SURJUHVV RI PHGLD FRYHUDJH RQ WKH FRUSRUDWH FULVLV 6WHPSHO ,,, f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£OH] +HUUHUR HWDO +HDULW /HUELQJHU 0LWURII 6WDQWRQ 7KRPVHQ 7UDYHUVR :LOFR[ HW DO :RRGV f $V QRWHG LQ WKH VHFRQG FKDSWHU LW LV QRW XQXVXDO WR UHFHLYH VRFDOOHG fIODPLQJf PHVVDJHV IURP WKH QHZVJURXSV GLVFXVVLRQV %XH\ f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

PAGE 68

DQG ER\FRWW RI WKH SURGXFW RU VHUYLFH WKLV PHVVDJH ZRXOG EH FRGHG DV ff ZKLFK LQGLFDWHV D QHJDWLYH PHVVDJH 7KLV UXOH ZDV DOVR DSSOLHG WR SRVLWLYH PHVVDJHV 7KH FRGLQJ VFKHPHV IRU QHZV VWRULHV DQG RQOLQH PHVVDJHV DUH VKRZQ LQ $SSHQGL[ & DQG $SSHQGL[ 1HJDWLYH SRVWHG PHVVDJHV ZLOO EH FRGHG DV WKRVH WKDW Df UHTXHVW OHJDO DFWLRQ DJDLQVW WKH FRPSDQ\ Ef UHTXHVW FRPSHQVDWLRQ IURP WKH FRPSDQ\ Ff PHQWLRQ RI SHUVRQDO RU RWKHUVf H[SHULHQFHV VXIIHULQJ IURP WKH FULVLV HYHQW Gf XUJH WKH ER\FRWW RI WKH FRPSDQ\fV SURGXFW RU VHUYLFH Hf JLYH VDUFDVWLF MRNHV RU KDUVK FRPPHQWV DERXW WKH FRPSDQ\ RU If EODPH RU DFFXVH WKH FRPSDQ\ IRU LWV ZURQJGRLQJ RU LWV GHIHFWLYH SURGXFWVf 3RVLWLYH SRVWHG PHVVDJHV ZLOO EH FRGHG DV WKRVH WKDW Df XUJH HQG RI D ER\FRWW Ef FRPSOLPHQW WKH FRPSDQ\ Ff PHQWLRQ WKH ODZV DV XQIDLU WR WKH RUJDQL]DWLRQ Gf SUDLVH FRPSDQ\fV LQWHUYHQLQJ DFWLRQV RU Hf GHQ\ WKDW WKH FULVLV HYHQW LV VROHO\ GXH WR WKH FRPSDQ\ 1HXWUDO SRVWHG PHVVDJHV ZLOO EH FRGHG DV WKRVH LQFOXGLQJ Df IRUZDUGHG PHGLD UHSRUWV ZLWKRXW H[SUHVVLQJ RSLQLRQ Ef IRUZDUGHG ZHEVLWH LQIRUPDWLRQ ZLWKRXW H[SUHVVLQJ RSLQLRQ

PAGE 69

Ff SHUVRQDO TXHVWLRQV RU UHVSRQVHV WR RWKHUVf PHVVDJHV ZLWKRXW H[SUHVVLQJ DQ\ RSLQLRQV RU Gf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f SRLQWV RXW WKDW WKHVH QHZVJURXS SRVWLQJV KDYH WKH TXDOLW\ RI VSRQWDQHLW\ DQG DUH XQFHQVRUHG IXQFWLRQLQJ PRUH OLNH WUXH GLDORJXHV 8QOLNH WUDGLWLRQDO FRQYHUVDWLRQ LQ WKH UHDO ZRUOG WKH IHHOLQJV H[SUHVVHG LQ WKH SRVWLQJV DUH QRW PLUURUHG HLWKHU SK\VLFDOO\ RU V\PEROLFDOO\ DQG WKH GLDORJXH LV fDEVHQW IURP SK\VLFDO SUR[LPLW\ IDFHWRIDFH LQWHUDFWLRQ DQG QRQYHUEDO FXHVf S f 0RUULV DQG 2JDQ f UHJDUG WKH 86(1(7 DV D PDQ\WR PDQ\ DV\QFKURQRXV FRPPXQLFDWLRQ 6R WR VSHDN WZR RU PRUH RQOLQH XVHUV FRPPXQLFDWLQJ ZLWK HDFK RWKHU GR QRW KDYH WR EH RQOLQH DW WKH VDPH WLPH (DFK FRXOG OHDYH PHVVDJHV WR RWKHUV DQG UHWULHYH UHSOLHV DW KLV RU KHU FRQYHQLHQFH *URSHU f 7R FRQWUDVW ZLWK GLDORJLF FRPPXQLFDWLRQ PDLQO\ DSSOLHG WR LQWHUSHUVRQDO FRPPXQLFDWLRQ 7KRPVHQ f FLWHG %DOO5RNHDFK DQG 5HDUGRQfV FRQFHSW RI fWHOHORJLF FRPPXQLFDWLRQf IRU WKH DQDO\VLV RI EXOOHWLQ ERDUG V\VWHPV $FFRUGLQJ WR %DOO5RNHDFK

PAGE 70

DQG 5HDUGRQ f WHOHORJLF FRPPXQLFDWLRQ fLQYROYHV DOWHUQDWLQJ GLDORJXH EHWZHHQ SHRSOH DW D GLVWDQFH ZKR XVH ERWK FRQYHQWLRQDO DQG XQFRQYHQWLRQDO ODQJXDJH DQG HOHFWURQLF RU RSWLFDO FKDQQHOVf S f 7KLV W\SH RI FRPPXQLFDWLRQ LV RIWHQ JHRJUDSKLFDOO\ DQG WHPSRUDOO\ XQERXQGHG %XH\ f FRQVLGHUV PHGLDWHG H[FKDQJHV RI PHVVDJHV DV D IRUP RI PDVV FRPPXQLFDWLRQ LQ ZKLFK WKH VHQGHU EURDGFDVWV D PHVVDJH WR D PDVV DXGLHQFH LQ WKH SURFHVV RI UHVSRQGLQJ WR DQ LQGLYLGXDO $OO PHVVDJHV RQ WKH 86(1(7 KDYH D VXEMHFW KHDGLQJ DQG WKRVH ZKLFK FDUU\ WKH VDPH VXEMHFW OLQH DUH FDOOHG D fWKUHDGf +LOO DQG +XJKHV f GHILQH WKUHDG DV UHVSRQVHV WR WKH RULJLQDO PHVVDJH RU WR VRPHWKLQJ VDLG LQ VRPHRQH HOVHfV UHSO\ 2QFH D WKUHDG KDV EHHQ VWDUWHG WKRXJK LW LV QR ORQJHU XQGHU WKH VHQGHUfV FRQWURO )RU WKLV VWXG\ RQOLQH GLDORJXHV RQ WKH 86(1(7 DUH GHILQHG DVf§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nV SHUFHSWLRQ RI WKH RUJDQL]DWLRQ $LPLQJ WR H[DPLQH WKH UHODWLRQVKLS EHWZHHQ WKH PHGLD FRYHUDJH RI WKH FULVLV DQG RQOLQH

PAGE 71

SXEOLF RSLQLRQ DERXW LW D FULVLV LQ WKLV VWXG\ LV GHILQHG DV VLPLODU WR WKDW E\ ,UYLQH DQG 0LOODU ff§ $ VLJQLILFDQW EXVLQHVV GLVUXSWLRQ ZKLFK QHJDWLYHO\ LPSDFWV LWV VWDNHKROGHUV VXFK DV HPSOR\HHV VWRFNKROGHUV FXVWRPHUV RU FRPPXQLW\ PHPEHUV DQG UHVXOWV LQ H[WHQVLYH QHZV PHGLD FRYHUDJH DQG SXEOLF VFUXWLQ\ &ULVLV LQWHUYHQWLRQ &ULVLV FRPPXQLFDWLRQ UHVXOWV IURP D QHHG WR RIIVHW SRWHQWLDO QHJDWLYH FRQVHTXHQFHV RI QRW FRPPXQLFDWLQJ 6WXUJHV f +HDULW f VXJJHVWV WKDW WR LQWHUYHQH LQ WKH FULVLV LV WR GLIIXVH KRVWLOLW\ GLUHFWHG WRZDUG WKH FRUSRUDWLRQ VXIIHULQJ IURP LW 5HVHPEOLQJ WKH LGHD RI FULVLV LQWHUYHQWLRQ )HDP%DQNV f XVHV fFRQWDLQPHQWf WR UHIHU WR fWKH HIIRUW WR OLPLW WKH GXUDWLRQ RI WKH FULVLV RU NHHS LW IURP VSUHDGLQJ WR RWKHU DUHDV DIIHFWLQJ WKH FRUSRUDWLRQf S f /HUELQJHU f UHJDUGV FULVLV LQWHUYHQWLRQ DV DQ DFWLRQ WKDW KHOSV UHPRYH DPELJXLW\ GXULQJ D FULVLV DQG LPSURYHV WKH XQGHUVWDQGLQJ RI WKH SXEOLF DQG WKH PHGLD )RU WKLV VWXG\ FULVLV LQWHUYHQWLRQ LV GHILQHG DVf§ :KHQ WKH RUJDQL]DWLRQ DFWV VWUDWHJLFDOO\ WR UHGXFH WKH QHJDWLYH LPSDFW RI WKH FULVLV RQ VWDNHKROGHUV DIWHU WKH RFFXUUHQFH RI WKH FULVLV 2SHUDWLRQDO 'HILQLWLRQV 7KH RSHUDWLRQDO GHILQLWLRQV RI NH\ WHUPV DUH OLVWHG DV IROORZV f 2QOLQH 'LDORJXHV RQ WKH 86(1(7 fQXPEHU RI PHVVDJHV IURP GLIIHUHQW LQGLYLGXDOV ZLWKLQ RQH VXEMHFW WKUHDGf +HUH WKH WKUHDG LV GHILQHG DV PHVVDJHV UHVSRQGLQJ WR WKH RULJLQDO PHVVDJH RU WR RWKHU PHVVDJHV XQGHU WKH VDPH GLVFXVVLRQ WRSLF RU VXEMHFW

PAGE 72

f &ULVLV WKH HYHQWV WKDW f QHJDWLYHO\ LPSDFW WKH WDUJHW VXEMHFWV HJ WKH JHQHUDO SXEOLF FXVWRPHUV DQG HPSOR\HHV f DUH FRYHUHG E\ ERWK D QDWLRQDO HOLWH QHZVSDSHU DQG D QDWLRQDO WHOHYLVLRQ QHWZRUN f FRQWLQXH WR EH UHSRUWHG E\ PDMRU QDWLRQDO QHZV PHGLD IRU DW OHDVW D RQHZHHN GXUDWLRQ DQG FRYHUHG E\ VRPH QHZV VRXUFHV IRU D WRWDO RI VHYHQ GD\V GXULQJ WKH PRQWK DQG f DUH GLVFXVVHG E\ DQ ,QWHUQHW QHZVJURXS SHULRGLFDOO\ IRU DW OHDVW WZR ZHHNV VLQFH WKH LQLWLDO QHZV PHGLD FRYHUDJH 7KH FULVLV LV FRQVLGHUHG WR KDYH HUXSWHG RQ WKH ILUVW GD\ WKDW DQ\ RI WKH VHOHFWHG QHZV PHGLD UHSRUWHG WKH FULVLV HYHQW f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

PAGE 73

SURGXFW UHFDOOV RU GHVWUXFWLRQ GHFLVLRQ WR HQIRUFH ODZV WR FRPEDW WKH FULVLV RIIHU RI PRQHWDU\ RU SURGXFW FRPSHQVDWLRQ PDNLQJ D SXEOLF DSRORJ\ DFNQRZOHGJHPHQW RI ZURQJGRLQJ UHSODFHPHQW RI SHUVRQV ZKR FDXVHG WKH FULVLV RU SURPLVH WR SUHYHQW D VLPLODU FULVLV LQ WKH IXWXUH f 7$/.(5 WKH RQOLQH XVHUV ZKR HLWKHU UHVSRQG WR ZLWKRXW H[SUHVVLQJ DJUHHPHQW RU H[SUHVV GLIIHUHQW RSLQLRQV RU LQLWLDWH WKH VXEMHFW PHVVDJH ZLWKLQ RQH WKUHDG r 5($&725 WKH RQOLQH XVHUV ZKR H[SUHVV WKHLU DJUHHPHQW WR WKH LQLWLDO PHVVDJH RU PHUHO\ UHSO\ WR RWKHU XVHUV ZLWKRXW YRLFLQJ WKHLU RSLQLRQV DERXW RWKHUVn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f WDNHQ E\ WKH FRUSRUDWLRQ GXULQJ WKH FULVLV SHULRG 7KH ODVW LQGHSHQGHQW YDULDEOH LV WKH QXPEHU RI RQn OLQH PHVVDJHV SRVWHG E\ ERWK 7$/.(56 DQG 5($&7256

PAGE 74

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f XUJHG WKDW WKLV W\SH RI UHOLDELOLW\ EH D PLQLPXP VWDQGDUG IRU FRQWHQW DQDO\VLV 7R WHVW WKH LQWHUFRGHU UHOLDELOLW\ PHVVDJHV UDQGRPO\ VHOHFWHG IURP WKH ILUVW FDVH ZHUH JLYHQ WR WZR FRGHUV ZKR DUH QDWLYH (QJOLVK VSHDNHUV 7KH\ ZHUH LQVWUXFWHG LQ GHWDLO RQ KRZ WR FRGH D QRPLQDO YDULDEOHf§WRQH RI PHVVDJH 1HZ LWHPV VXJJHVWHG E\ WKH FRGHUV ZHUH DGGHG WR WKH FRGLQJ VFKHPH $OWKRXJK WKHUH LV QR DJUHHPHQW RQ D PLQLPXP LQWHUFRGHU UHOLDELOLW\ FRHIILFLHQW PRVW FRQWHQW DQDO\VWV VXJJHVW WKDW WKH UHVHDUFKHU VKRXOG ORRN DW ZKHWKHU RU QRW WKH H[WHQW RI DJUHHPHQW H[FHHGV FKDQFH 6WHPSHO ,,, f .ULSSHQGRUII f IRXQG WKDW DQ DJUHHPHQW FRUUHODWLRQ RI OHVV WKDQ WHQGHG WR EH VWDWLVWLFDOO\ LQVLJQLILFDQW DQG WKDW SHUFHQW RI DJUHHPHQW LV QRW SODXVLEOH LQ PRVW FDVHV )RU WKLV VWXG\ WKH SUHWHVW LQWHUFRGHU UHOLDELOLW\ FRHIILFLHQW ZDV H[SHFWHG WR UHDFK DW OHDVW SHUFHQW EHIRUH WKH UHVHDUFKHU VWDUWHG WR FRGH PHVVDJHV LQ WKH VHFRQG FDVH $FFRUGLQJ WR 6FRWWnV f IRUPXOD UHOLDELOLW\ UHDFKHG WKDW LV ERWK FRGHUV KDG FRQVLVWHQW DJUHHPHQWV RQ RXW RI PHVVDJHV LQ WKH VDPH FDWHJRU\ %HFDXVH WKH\

PAGE 75

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f IRU WH[WXDO DQDO\VLV ZDV HPSOR\HG WR FDOFXODWH WKH QXPEHU RI QHJDWLYH ZRUGV DSSHDULQJ LQ HDFK QHZV VWRU\ $IWHU VHOHFWHG DUWLFOHV DUH DQDO\]HG WKLV WH[WXDO DQDO\VLV SURJUDP ZLOO FDOFXODWH VFRUHV RI ILYH f0DVWHU 9DULDEOHVf &HUWDLQW\ 2SWLPLVP $FWLYLW\ 5HDOLVP DQG &RPPRQDOLW\ (YHU\ 0DVWHU 9DULDEOH VFRUH KDV LWV RZQ IRUPXOD 7DNH 2SWLPLVP DV DQ H[DPSOH 7KH GHILQLWLRQ RI WKH 2SWLPLVP 6FRUH LV ODQJXDJH HQGRUVLQJ RU KLJKOLJKWLQJ WKH SRVLWLYH HQWDLOPHQWV RI VRPH SHUVRQ JURXS FRQFHSW RU HYHQW +DUW S f 7KH IRUPXOD IRU WKLV YDULDEOH LV 3UDLVH 6DWLVIDFWLRQ ,QVSLUDWLRQf %ODPH +DUGVKLS 'HQLDOf 7HUPV RI 3UDLVH 6DWLVIDFWLRQ ,QVSLUDWLRQ %ODPH +DUGVKLS DQG 'HQLDO DUH FRQVLVWHG RI GLIIHUHQW ZRUGV VWRUHG LQ WKH GLFWLRQDULHV LQ WKH VRIWZDUH 7R VHH WR ZKDW GHJUHH D QHZV VWRU\ FRYHUHG WKH FULVLV HYHQW LQ D QHJDWLYH WRQH WKH 2SWLPLVP 6FRUH ZDV XVHG +RZHYHU WR VLPSOLI\ WKH IRUPXOD RQO\ QXPEHUV RI ZRUGV UHSUHVHQWLQJ DERYH VL[ WHUPV ZHUH SXW LQWR WKH IRUPXOD LQVWHDG RI WKH VWDQGDUGL]HG VFRUH GLVSOD\HG LQ WKH UHSRUW ILOH 6HH $SSHQGL[ (f 2QFH WKH 2SWLPLVP 6FRUH IRU HDFK QHZV

PAGE 76

VWRU\ ZDV REWDLQHG LW ZDV FRPSDUHG ZLWK WKH QXPEHU RI GDLO\ QHJDWLYH 86(1(7 PHVVDJHV WR WHVW K\SRWKHVLV 7KH QH[W FKDSWHU ZLOO ILUVW SUHVHQW GHVFULSWLYH DQDO\VLV 7KHQ GHVFULSWLRQ RI HDFK FULVLV FDVH DORQJ ZLWK UHVXOWV RI GRZQORDGLQJ PHGLD VWRULHV DQG 86(1(7 PHVVDJHV LV SURYLGHG IROORZHG E\ WKH GLVFXVVLRQ RI RXWFRPHV RI K\SRWKHVHV WHVWLQJ 3RVWKRF DQDO\VLV ZLOO HQG WKH FKDSWHU

PAGE 77

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

PAGE 78

UHVXOWV SDJHV +RZHYHU EHFDXVH QRW DOO WKH PHVVDJHV GLVSOD\LQJ RQ WKH UHVXOWV SDJHV ZHUH UHODWHG WR WKH FULVLV PDQ\ PHVVDJHV ZHUH HOLPLQDWHG )LQDOO\ DQG RQOLQH PHVVDJHV IURP WKH 0DWWHO DQG WKH +XGVRQ )RRGV FDVHV ZHUH VDPSOHG WKURXJK GHWDLOHG H[DPLQDWLRQ 7KH\ ZHUH XVHG DV WKH EDVLV IRU WKH SXEOLF DJHQGD WR WHVW WKH K\SRWKHVHV 7RQH RI 2QOLQH 0HVVDJHV 7RQH RI RQOLQH PHVVDJHV ZDV FRGHG DV HLWKHU SRVLWLYH QHJDWLYH QHXWUDO RU XQLGHQWLILHG E\ FRGHUV ,Q WKH 0DWWHO FDVH QHJDWLYH bf DQG SRVLWLYH PHVVDJHV bf ZHUH SRVWHG GLVFXVVLQJ WKH FULVLV ,Q WKH +XGVRQ )RRGV FDVH PHVVDJHV bf ZHUH FRQVLGHUHG QHJDWLYH DQG PHVVDJHV SRVLWLYH bf 7KH UHPDLQLQJ PHVVDJHV LQ ERWK FDVHV ZHUH HLWKHU QHXWUDO RU XQLGHQWLILHG ,QWHUHVWLQJO\ SHUFHQWDJHV RI QHJDWLYH DQG SRVLWLYH PHVVDJHV LQ ERWK FDVHV ZHUH YHU\ FORVH )UHTXHQF\ $SSHDUDQFH RI 8VHU 1DPHV LQ WKH 'LVFXVVLRQ 0HVVDJHV $FFRUGLQJ WR *URSHUfV f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

PAGE 79

8VHUVn 6WUDWHJLHV 7ZR W\SHV RI XVHUVn VWUDWHJ\ ZHUH FDWHJRUL]HG 2QH LV 7$/.(5 VWUDWHJ\ DSSOLHG WR XVHUV ZKR LQLWLDWH WKH GLVFXVVLRQ WKUHDG RU H[SUHVV GLVDJUHHPHQW ZLWK RWKHU XVHUV $QRWKHU LV 5($&725 VWUDWHJ\ DGRSWHG E\ XVHUV ZKR WHQG WR DJUHH ZLWK RWKHU SHRSOHnV RSLQLRQ RU GR QRW H[SUHVV WKHLU RZQ RSLQLRQV ,Q WKH 0DWWHO FDVH XVHUV DGRSWHG 7$/.(5 VWUDWHJ\ ZKLOH 5($&725 VWUDWHJ\ ZDV HPSOR\HG E\ XVHUV KL WKH +XGVRQ )RRGV FDVH XVHUV ZHUH LGHQWLILHG WR KDYH D 7$/.(5 VWUDWHJ\ DQG HPSOR\HG 5($&725 VWUDWHJ\ %HFDXVH VRPH RQn OLQH XVHUV KDG SRVWHG PRUH WKDQ RQH PHVVDJH WKH\ FRXOG DGRSW ERWK 7$/.(5 DQG 5($&725 VWUDWHJLHV LQ YDULRXV SRVWLQJV ,Q WRWDO SHUFHQW RI XVHUV IRU 0DWWHO DQG IRU +XGVRQ )RRGV ZHUH FODVVLILHG DV HLWKHU 7$/.(56 RU 5($&7256 )UHTXHQFLHV RI )ODPLQJ 0HVVDJHV $V GLVFXVVHG HDUOLHU LQ &KDSWHU WZR IODPLQJ PHVVDJHV DUH RIWHQ SRVWHG DQG VHHQ LQ WKH QHZVJURXSV D SXEOLF GLVFXVVLRQ IRUXP ZLWKRXW D FHQWUDO DXWKRULW\ WR UHJXODWH XVHUVf EHKDYLRU &RQWUDU\ WR H[SHFWDWLRQ WKH DQDO\VLV RI WKH RQOLQH PHVVDJHV LQ WKH WZR FDVHV UHYHDOHG WKDW YHU\ IHZ PHVVDJHV FRXOG EH FRGHG DV IODPLQJ ,Q WKH 0DWWHO FDVH IODPLQJ PHVVDJHV bf ZHUH DQDO\]HG WDUJHWHG RQ WKH 0DWWHO FRPSDQ\ RQ RWKHU XVHUV DQG RQ ERWK WKH FRPSDQ\ DQG XVHUV $V WR WKH +XGVRQ )RRGV FDVH IODPLQJ PHVVDJHV ZHUH HYHQ IHZHU 2QO\ ZHUH UHWULHYHG WKH JRYHUQPHQW ZDV WKH IODPLQJ WDUJHW LQ PHVVDJHV WDUJHWHG RQ RWKHU XVHUV DQG WDUJHWHG RQ ERWK WKH JRYHUQPHQW DQG XVHUV

PAGE 80

0DWWHO &DEEDJH 3DWFK 'ROO 'HIHFW &DVH &DVH 'HVFULSWLRQ 7KH 0DWWHO WR\ FRPSDQ\ HQFRXQWHUHG D SURGXFW GHIHFW FULVLV LQ WKH 8QLWHG 6WDWHV DW WKH HQG RI 7KH IROORZLQJ VWRU\ LV EDVHG RQ WKH PHGLD VWRULHV UHWULHYHG IURP WKH /H[LV1H[LV GDWD EDVH 2Q 'HFHPEHU WKH 1HZ
PAGE 81

5HWULHYDO RI 0HGLD 6WRULHV IURP WKH /H[LV1H[LV 'DWD %DVH $V WKLV FULVLV ZDV VROHO\ UHODWHG WR WKH 0DWWHO FRPSDQ\ WKH NH\ZRUG IRU VHDUFKLQJ WKH QHZV VWRULHV DERXW WKH FULVLV ZDV PDWWHO $IWHU LGHQWLI\LQJ WKH GDWH ZKHQ DQ\ RI VL[ PDMRU PHGLD ILUVW UHSRUWHG WKH FULVLV HYHQW WKH WLPH IDPH ZDV VHW XS WR UHWULHYH QHZV VWRULHV ,Q WKLV FDVH WKH GDWH ZDV 'HFHPEHU ZKHQ WKH 1HZ
PAGE 82

,Q WKLV FDVH PRVW PHGLD VWRULHV FDPH IURP WZR PDMRU QHZV VRXUFHV WKH 1HZ
PAGE 83

UHVSHFWLYHO\ 7DEOH VKRZV D VLJQLILFDQW GLIIHUHQFH EHWZHHQ WKH WZR JURXSV 7KHUHIRUH WKLV K\SRWKHVLV ZDV VXSSRUWHG 7$%/( )UHTXHQFLHV RI PHVVDJHV DSSHDULQJ RQ HDFK VDPSOLQJ GD\ LQ WKH 0DWWHO FDVH 9DOLG &XPXODWLYH )UHTXHQF\ 3HUFHQW 3HUFHQW 3HUFHQW 7RWDO 7RWDO

PAGE 84

7$%/( 5HVXOWV RI +\SRWKHVLV WHVW LQ WKH 0DWWHO FDVH 1 0HDQ 6WG 'HYLDWLRQ 6WG (UURU WB YDOXH GD\V EHIRUH FULVLV r GD\V DIWHU FULVLV r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rS OHYHO +\SRWKHVLV 7KH GDLO\ QXPEHU RI WKH 86(1(7 GLDORJXHV ZLOO EH UHODWHG WR WKH GDLO\ QXPEHU RI QHZV VWRULHV DERXW D FULVLV DIWHU LWV HUXSWLRQ &RQVLVWHQW ZLWK WKH FRQFHSW RI DJHQGDVHWWLQJ WKHRU\ WKH PRUH QHZV VWRULHV DUH DERXW D FULVLV WKH PRUH RQOLQH GLVFXVVLRQ PHVVDJHV VKRXOG EH SRVWHG RQ WKH 86(1(7

PAGE 85

6LPSOH UHJUHVVLRQ DQDO\VLV ZDV XWLOL]HG WR WHVW WKLV K\SRWKHVLV 7KH QXPEHU RI QHZV VWRULHV RQ HDFK GD\ DIWHU WKH FULVLV HUXSWLRQ ZDV WKH LQGHSHQGHQW YDULDEOH ZKHUHDV WKDW RI WKH RQOLQH GLDORJXHV ZDV WKH GHSHQGHQW YDULDEOH $V U ZDV U f DQG VLJQLILFDQW ) f S f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b GD\V DIWHU LQWHUYHQWLRQ rS OHYHO

PAGE 86

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f YHUVXV GD\V DIWHU WKH LQWHUYHQWLRQ DGRSWLRQ Q f WKH K\SRWKHVLV ZDV VXSSRUWHG DV GHPRQVWUDWHG LQ 7DEOH

PAGE 87

7$%/( 5HVXOWV RI +\SRWKHVLV WHVW LQ WKH 0DWWHO FDVH 1 0HDQ 6WG 'HYLDWLRQ 6WG (UURU WB YDOXH GD\V DIWHU FULVLV DQG EHIRUH LQWHUYHQWLRQ GD\V DIWHU LQWHUYHQWLRQ rS OHYHO +\SRWKHVLV 7KH QXPEHU RI GDLO\ QHJDWLYH 86(1(7 GLDORJXHV SRVWHG E\ 7$/.(56 ZLOO EH JUHDWHU WKDQ WKDW E\ 5($&7256 7$/.(56 SRVWHG QHJDWLYH RQOLQH PHVVDJHV ZKHUHDV QHJDWLYH RQOLQH PHVVDJHV ZHUH SRVWHG E\ 5($&7256 %\ XVLQJ D &KLVTXDUH WHVW WKLV K\SRWKHVLV ZDV QRW VXSSRUWHG DV VKRZQ LQ 7DEOH b 1 f S 7$%/( 5HVXOWV RI +\SRWKHVLV WHVW LQ WKH 0DWWHO FDVH 1HJDWLYH 3RVLWLYH 7RWDO GI &KLVTXDUH YDOXH 7$/.(5 16 5($&725 7RWDO +\SRWKHVLV 7KH 2SWLPLVP 6FRUH RI PHGLD PHVVDJHV ZLOO EH SRVLWLYHO\ UHODWHG WR WKH QXPEHU RI QHJDWLYHO\ WRQHG 86(1(7 PHVVDJHV SRVWHG DIWHU WKH PHGLD VWRULHV ZHUH UHOHDVHG

PAGE 88

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 S $OWKRXJK WKH YDULDQFH H[SODLQHG ZDV VPDOO WKLV K\SRWKHVLV ZDV VXSSRUWHG R DV 8f 2f 8f f &' ’ ’ V D! f ; ( ’ ’ ’ ’ ’ ’ ’ ’ ’ ’ ’’ ’’’’’ ’ ’ ’ 2SWLPLVP 6FRUH IRU $OO 1HZV 6WRULHV (DFK 'D\ ),*85( %LYDULDWH FRUUHODWLRQ EHWZHHQ WKH QXPEHU RI QHJDWLYH RQOLQH PHVVDJHV DQG WKH 2SWLPLVP 6FRUH LQ WKH 0DWWHO FDVH

PAGE 89

+XGVRQ )RRGV 0HDW &RQWDPLQDWLRQ &DVH &DVH 'HVFULSWLRQ 7KH +XGVRQ )RRGV FRPSDQ\ HQFRXQWHUHG D VHULRXV PHDW FRQWDPLQDWLRQ FULVLV DFURVV WKH 8QLWHG 6WDWHV DIWHU -XO\ RI 7KH IROORZLQJ VWRU\ LV EDVHG XSRQ WKH PHGLD VWRULHV UHWULHYHG IURP WKH /H[LV1H[LV GDWD EDVH 2Q $XJXVW WKH $%& :RUOG 1HZV 7RQLJKW KDG WKH IROORZLQJ FRYHUDJH DQ $UNDQVDV PHDW SURFHVVRU LV UHFDOOLQJ SRXQGV RI IUR]HQ KDPEXUJHU SDWWLHV ZKLFK PD\ EH FRQWDPLQDWHG ZLWK WKH ( FROL EDFWHULD +XGVRQ )RRGV GLVWULEXWHG WKH KDPEXUJHU QDWLRQZLGH 7KH SUREOHP ZDV ILUVW VSRWWHG LQ &RORUDGR ZKHQ VHYHUDO SHRSOH ZKR KDG HDWHQ WKH KDPEXUJHU EHFDPH LOO 7KLV QHZV DUWLFOH LV WKH HDUOLHVW RQH WKDW VKRZHG +XGVRQ )RRGV LQ DOO QHZV GRFXPHQWV RI WKH /H[LV1H[LV GDWD EDVH LQ ERWK -XO\ DQG $XJXVW RI 7KHUHIRUH WKH EHJLQQLQJ GD\ RI WKLV FULVLV ZDV FRQVLGHUHG WR EH $XJXVW $IWHU WKH JRYHUQPHQW XUJHG PRUH UHFDOO RI WKH FRQWDPLQDWHG PHDW +XGVRQ )RRGV UHFDOOHG DERXW PLOOLRQ EHHI SURGXFWV IURP WKH PDUNHW RQ $XJXVW FORVHG LWV SODQW LQ 1HEUDVND DQG UHFDOOHG PLOOLRQ EHHI SDWWLHV RQ $XJXVW DFFRUGLQJ WR WKH $%& QHZV UHSRUWV ,Q WKH EHJLQQLQJ RI 6HSWHPEHU +XGVRQ )RRGV DJUHHG WR PHUJH ZLWK LWV ULYDO FRPSDQ\ 7\VRQ )RRGV $V WR WKLV FULVLV FDVH WKH HUXSWLRQ GD\ ZDV $XJXVW DQG WKH DGRSWLRQ RI WKH LQWHUYHQWLRQ VWUDWHJLHV RFFXUUHG WZLFH 7KH ILUVW ZDV RQ $XJXVW ZKHQ WKH FRPSDQ\ UHFDOOHG RQH PLOOLRQ SRXQGV RI EHHI SDWWLHV DQG WKH VHFRQG ZDV RQ $XJXVW

PAGE 90

ZKHQ DERXW PLOOLRQ EHHI SDWWLHV ZHUH UHFDOOHG DQG D IRRG SODQW LQ 1HEUDVND ZDV VKXW GRZQ 7KH LQWHQVLW\ RI WKLV FULVLV H[WHQGHG EH\RQG WKH GD\V IUDPHG LQ WKLV VWXG\ 7KH 86 1HZV t :RUOG 5HSRUW DQG WKH 1HZVZHHN LQ WKHLU 6HSWHPEHU HGLWLRQV KDG VWRULHV IHDWXULQJ WKH IRRG FRQWDPLQDWLRQ FDVHV LQ WKH SDVW GHFDGH WR GHPRQVWUDWH WKH LPSRUWDQFH RI IRRG VDIHW\ 7KH 1HZ
PAGE 91

KHQFH ZDV IURP $XJXVW WR 6HSWHPEHU DQG QHZV VWRULHV ZHUH GRZQORDGHG DQG VHUYHG DV WKH EDVLV RI WKH PHGLD DJHQGD 7$%/( )UHTXHQFLHV RI QHZV VWRULHV RQ HDFK QHZV PHGLD LQ +XGVRQ )RRGV FDVH )UHTXHQF\ 3HUFHQW 9DOLG 3HUFHQW &XPXODWLYH 3HUFHQW 9DOLG $%& 1HZV &11 1%& 1HZV 1HZ
PAGE 92

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

PAGE 93

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r GD\V DIWHU FULVLV rS OHYHO +\SRWKHVLV 7KH DYHUDJH QXPEHU RI GDLO\ QHJDWLYH 86(1(7 GLDORJXHV ZLOO EH JUHDWHU WKDQ WKH DYHUDJH QXPEHU RI GDLO\ SRVLWLYH 86(1(7 GLDORJXHV DIWHU WKH HUXSWLRQ RI WKH FULVLV 7KH QXPEHUV RI SRVLWLYH DQG QHJDWLYH PHVVDJHV GXULQJ GD\V DIWHU WKH FULVLV HUXSWLRQ ZHUH FRPSDUHG 7KH WWHVW UHYHDOV D VLJQLILFDQW GLIIHUHQFH EHWZHHQ WKH DYHUDJH QXPEHUV RI GDLO\ QHJDWLYH DQG SRVLWLYH RQOLQH GLDORJXHV LQ 7DEOH &RQVHTXHQWO\ WKLV K\SRWKHVLV ZDV VXSSRUWHG

PAGE 94

7$%/( 5HVXOWV RI +\SRWKHVLV WHVW LQ WKH +XGVRQ )RRGV FDVH 1 0HDQ 6WG 'HYLDWLRQ 6WG (UURU WB YDOXH 3RVLWLYH 0HVVDJHV LS f§ 1HJDWLYH 0HVVDJHV rS OHYHO +\SRWKHVLV 7KH GDLO\ QXPEHU RI WKH 86(1(7 GLDORJXHV ZLOO EH UHODWHG WR WKH GDLO\ QXPEHU RI QHZV VWRULHV DERXW D FULVLV DIWHU LWV HUXSWLRQ 'LIIHUHQW IURP WKH ILQGLQJ LQ WKH 0DWWHO FDVH WKLV UHVXOW U S ) f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f S

PAGE 95

7$%/( 5HVXOWV RI FRPSDULVRQ RI WKUHH VWDJHV LQ +\SRWKHVLV LQ WKH +XGVRQ )RRGV &DVH 6XP RI 6TXDUHV GI 0HDQ 6TXDUH ) 6LJ 1XPEHU RI %HWZHHQ QHJDWLYH *URXSV PHVVDJH :LWKLQ *URXSV 7RWDO D 1RQ6LJQLILFDQW 7KH VHFRQG WHVW ZDV WR FRPSDUH WKH QXPEHU RI QHJDWLYH 86(1(7 PHVVDJHV EHIRUH DQG DIWHU WKH ILUVW LQWHUYHQWLRQ VWDJHV 7DEOH UHYHDOV D VLJQLILFDQW UHODWLRQVKLS ) f S 7$%/( 5HVXOWV RI FRPSDULVRQ RI WZR VWDJHV LQ +\SRWKHVLV LQ WKH +XGVRQ )RRGV &DVH 6XP RI 0HDQ 6TXDUHV GI 6TXDUH ) 6LJ 1XPEHU RI QHJDWLYH %HWZHHQ *URXSV D PHVVDJH :LWKLQ *URXSV 7RWDO D S OHYHO 7KH ILQDO WHVW ZDV WR FRPSDUH WKH QXPEHU RI QHJDWLYH 86(1(7 PHVVDJHV EHIRUH DQG DIWHU WKH VHFRQG LQWHUYHQWLRQ VWDJHV $ VLJQLILFDQW UHODWLRQVKLS ZDV IRXQG LQ 7DEOH ) f S

PAGE 96

7$%/( 5HVXOWV RI FRPSDULVRQ RI WZR VWDJHV LQ +\SRWKHVLV LQ WKH +XGVRQ )RRGV &DVH 6XP RI 0HDQ 6TXDUHV GI 6TXDUH ) 6LJ 1XPEHU RI QHJDWLYH %HWZHHQ *URXSV D PHVVDJH :LWKLQ *URXSV 7RWDO D S OHYHO +\SRWKHVLV )ROORZLQJ D FULVLV LQWHUYHQWLRQ VWUDWHJ\ WKH DYHUDJH QXPEHU RI GDLO\ SRVLWLYH RQOLQH GLDORJXHV RQ WKH 86(1(7 FRQFHUQLQJ WKH FRPSDQ\ ZLOO EH HTXDO WR WKH DYHUDJH QXPEHU RI GDLO\ SRVLWLYH RQOLQH GLDORJXHV EHIRUH WKH FULVLV HUXSWLRQ 7KH $129$ UHVXOWV LQ 7DEOH VXSSRUWHG WKLV K\SRWKHVLV P [? ) f S 7KDW LV ZKHWKHU WKH FRPSDQ\ DGRSWHG WKH LQWHUYHQWLRQ RU QRW QR VLJQLILFDQW GLIIHUHQFH ZDV IRXQG LQ WKH DYHUDJH QXPEHU RI GDLO\ SRVLWLYH RQOLQH GLDORJXHV DIWHU WKH LQWHUYHQWLRQ DQG EHIRUH WKH FULVLV UHJDUGOHVV RI WKH IHZ SRVLWLYH RQOLQH PHVVDJHV 7$%/( 5HVXOWV RI +\SRWKHVLV WHVW LQ WKH +XGVRQ )RRGV FDVH 6XP RI 6TXDUHV GI 0HDQ 6TXDUH ) 6LJ $YHUDJH RI 3RVLWLYH %HWZHHQ 2QOLQH 0HVVDJHV *URXSV (DFK 'D\ ZLWKLQ *URXSV 7RWDO B B D 1RQ6LJQLILFDQW

PAGE 97

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r GD\V DIWHU VW LQWHUYHQWLRQ r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f S

PAGE 98

7$%/( 5HVXOWV RI FRPSDULVRQ RI WZR VWDJHV LQ +\SRWKHVLV LQ WKH +XGVRQ )RRGV &DVH 1 0HDQ 6WG 'HYLDWLRQ 6WG (UURU WB YDOXH GD\V DIWHU FULVLV DQG EHIRUH QG LQWHUYHQWLRQ r GD\V DIWHU QG LQWHUYHQWLRQ rS OHYHO 7$%/( 5HVXOWV RI +\SRWKHVLV WHVW LQ WKH +XGVRQ )RRGV FDVH 1HJDWLYH 3RVLWLYH 7RWDO GI &KLVTXDUH YDOXH 7$/.(5 16 5($&725 7RWDO +\SRWKHVLV 7KH 2SWLPLVP 6FRUH RI PHGLD PHVVDJHV ZLOO EH SRVLWLYHO\ UHODWHG WR WKH QXPEHU RI QHJDWLYHO\ WRQHG 86(1(7 PHVVDJHV SRVWHG DIWHU WKH PHGLD VWRULHV ZHUH UHOHDVHG 7R WHVW WKLV K\SRWKHVLV ELYDULDWH FRUUHODWLRQ EHWZHHQ WKH DYHUDJH QXPEHU RI QHJDWLYH RQOLQH PHVVDJHV HDFK GD\ DQG WKH 2SWLPLVP 6FRUH ZDV H[DPLQHG DV VKRZQ LQ )LJXUH

PAGE 99

D !r WR F R IIO /8 F! &' FQ f UR %! ( ] Â’ R Â’ RQ Â’ Â’ Â’ &' Â’ Â’ Â’ Â’ Â’ 2SWLPLVP 6FRUH IRU $OO 1HZV 6WRULHV (DFK 'D\ ),*85( %LYDULDWH FRUUHODWLRQ EHWZHHQ WKH QXPEHU RI QHJDWLYH RQOLQH PHVVDJHV DQG WKH 2SWLPLVP 6FRUH LQ WKH +XGVRQ )RRGV FDVH 7KH DERYH ILJXUH VXJJHVWHG WKDW WKH 2SWLPLVP 6FRUHV LQ QHZV VWRULHV ZHUH QRW UHODWHG WR WKH QXPEHU RI QHJDWLYH PHVVDJHV SRVWHG HDFK GD\ DV U U f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

PAGE 100

),*85( 5HODWLRQVKLS EHWZHHQ WKH PHGLD DJHQGD DQG WKH SXEOLF DJHQGD LQ WKH 0DWWHO FDVH ),*85( 5HODWLRQVKLS EHWZHHQ WKH PHGLD DJHQGD DQG WKH SXEOLF DJHQGD LQ WKH +XGVRQ )RRGV FDVH 7KH ODFN RI D VLJQLILFDQW UHODWLRQVKLS EHWZHHQ WKH PHGLD DJHQGD DQG WKH SXEOLF DJHQGD LQ WKH +XGVRQ )RRGV FDVH LV HYLGHQW U f 8VLQJ 6366 FXUYHILWWLQJ D FXELF UHODWLRQVKLS ZDV IRXQG WR LPSURYH WKH H[SODQDWRU\ SRZHU RI WKH PHGLD DJHQGD

PAGE 101

LPSDFWLQJ WKH SXEOLF DJHQGD IURP WR DV GHPRQVWUDWHG LQ )LJXUH 7KH 0DWWHO GDWD LOOXVWUDWH WKH H[SHFWHG OLQHDU SDWWHUQ RI QHZV VWRULHV DQG 86(1(7 PHVVDJHV 5VT 1XPEHU RI 1HZV 6WRULHV RQ (DFK 'D\ ),*85( &XELF UHODWLRQVKLS EHWZHHQ WKH PHGLD DJHQGD DQG WKH SXEOLF DJHQGD LQ WKH +XGVRQ )RRGV FDVH $QDO\VLV RI 0HVVDJH 7\SH ,Q RUGHU WR EHWWHU XQGHUVWDQG WKH FXUYLOLQHDU UHODWLRQVKLS EHWZHHQ WKH SXEOLF DJHQGD DQG WKH PHGLD DJHQGD FRQWHQWV RI DOO RQOLQH PHVVDJHV IRU ERWK FDVHV ZHUH UHn H[DPLQHG DQG FDWHJRUL]HG 7KH FRGLQJ FDWHJRULHV ZHUH FUHDWHG EDVHG RQ 7KRPVHQfV f DUWLFOH DQDO\]LQJ WKH GLVFXVVLRQ PHVVDJHV LQ WKH 35)RUXP )LYH FDWHJRULHV ZHUH IRUPHG 4XHVWLRQ$QVZHU H[FKDQJH RI LQIRUPDWLRQ DERXW WKH FULVLV RU WKH WDUJHW FRPSDQ\ ([DPSOH

PAGE 102

6XEMHFW 5H +RZ PXFK 7XQD LV VDIH" )URP DWFVKDQH#DROFRP $7& 6KDQHf 'DWH 1HZVJURXSV DOWVSRUWZHLJKWOLIWLQJ ZDV UHFHQWO\ WROG WKDW HDWLQJ WXQD PRUH WKDQ WLPHV D ZHHN FRXOG FDXVH !PHUFXU\ SRLVRQLQJ LV WKHUH DQ\ WUXWK WR WKLV" +PPP 6RXQGV OLNH VRPHRQH LV KLWWLQJ \RX ZLWK VRPH SDUDQRLD WKLQN LI LW ZDV WKDW HDV\ WR JHW SRLVRQHG IURP WXQD WKH )'$ ZRXOGQnW DOORZ LW RQ WKH PDUNHW +HFN ORRN DW +XGVRQ %HHI ,WnV OLNH RQH JLUO JRW (&ROL SRLVRQLQJ DQG WKH JRY SXOOHG RYHU D PLOOLRQ SRXQGV RI EHHI RII WKH PDUNHW 2SLQLRQ GLVFXVVLRQ GHEDWHV RU DUJXPHQWV FRQFHUQLQJ WKH FULVLV H[SUHVVLRQ RI SHUVRQDO H[SHULHQFHV ([DPSOH 6XEMHFW 5H >27@ 86 1HZV )LQDOO\ )LQGV D n&OLQWRQ :DWFKn 6FDQGDO )URP JFUXVH#ZRUOGQHWQRVSDPDWWQHW JDU\ FUXVHf 'DWH 1HZVJURXSV DOWFXUUHQWHYHQWVFOLQWRQZKLWHZDWHU 2Q 6DW $XJ %LOO1DOW\ &%DVKHU#ZRUOGQHWDWWQHW! ZURWH !KWWSZZZXVQHZVFRPXVQHZVLVVXHOIHHGKWP !86 1(:6 t :25/' 5(3257 !6HSWHPEHU ! 7KH QH[W EDG EHHI VFDQGDO" ! &DWWOH IHHG QRZ FRQWDLQV WKLQJV OLNH PDQXUH DQG GHDG FDWV ! ,nP QRW VXUH ZDQWHG WR NQRZ DERXW WKLV :HOO PD\EH %XW LW VXUH SXWV D QHZ OLJKW RQ KRZ UHPRYLQJ DJULFXOWXUDO VXEVLGLHV ZLOO SXW WKH VPDOO IDUPHUV RXW RI EXVLQHVV 6LQFH DFFRUGLQJ WR WKH DUWLFOH WKLV PDQXUH UHF\FOLQJ LV SUDFWLFHG E\ VPDOO

PAGE 103

IDUPHUV DQG QRW DJULEXVLQHVV DP HYHQ OHVV FRQYLQFHG WKDW LQGLYLGXDO IDUPLQJ VKRXOG EH VXEVLGL]HG 0HGLD 5HSRUW IRUZDUGLQJ PHGLD VWRU\ UHODWHG WR WKH FULVLV XSGDWHV RI PHGLD UHSRUWV ([DPSOH 6XEMHFW )URP $0 1HZV $EXVH IRU $XJXVW 6HS )URP 'DLV\ GDLV\#JHRFLWLHVFRP! 'DWH f 1HZVJURXSV DOWFXOWXUHIDEXORXV 'HDU ,V 7KH ,UUDGLDWRU 5HDG\ 2QH UHVXOW RI WKH +XGVRQ )RRGV KDPEXUJHU FRQWDPLQDWLRQ ILDVFR LV UHQHZHG KWWSZZZSPDFQHWIRRGLUUKWP GLVFXVVLRQ DERXW ZK\ WKH EHHI LQGXVWU\ GRHVQn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nW PDNH VHQVH WR PH XQOHVV WKH JRYHUQPHQW ZDV MXVW WDNLQJ PRUH SRZHU (YHU\GD\ $PHULFDQnV ULJKWV DUH EHLQJ UHPRYHG E\ DQ RSSUHVVLYH JRYHUQPHQW DQG $PHULFD VOHHSV (YHU\GD\ ZH WDON DERXW KRZ EDG LW LV 7DON LV FKHDS DQG RXU ULJKWV VOLS DZD\ GDLO\ &RPELQDWLRQ RI FDWHJRU\ RU ([DPSOH

PAGE 104

6XEMHFW IRRG LUUDGLDWLRQ )URP -RKQ 0F&DUWK\ MPF#VWHDPVWDQIRUGHGX! 'DWH 1HZVJURXSV VFLHQYLURQPHQW +HUHnV D 5HXWHUV VWRU\ ,W IROORZHG WKH 1<7LPHV 2S(G E\ 5LFKDUG 5KRGHV E\ RQO\ D IHZ GD\V %HLQJ D QHZV VWRU\ DQG QRW DQ 2S(G LW KDV GLUHFW TXRWHV IURP YDULRXV SHRSOH 7KH 1DGHU PLQLRQ &DUROLQH 6PLWK ZDV QHJDWLYH EXW ZDV UDWKHU PLOG IRU D 1DGHULWH 6KH ZDV VXIILFLHQWO\ PLOG LW PD\ UHOLHYH IHDUV DPRQJ SROLWLFLDQV RI DQ HQYLURQPHQWDOLVW DWWDFN 7KHUH ZDV DOVR WKH XVXDO H[SHUW *DU\ 6PLWK ZKR VDLG WKDW LI RQO\ WKH FRQVXPHUV ZRXOG FRRN WKH IRRG HQRXJK LW ZRXOG EH VDIH >, ZRQnW GR WKDW EHFDXVH OLNH PHDW UDUH DQG WKLQN WKH QXPEHU RI IRRG SRLVRQLQJ LQFLGHQWV LV ORZ HQRXJK VR WKDW HDWLQJ UDUH PHDW RU HYHQ VWHDN WDUWDUH LV RQH RI WKH OHVVHU ULVNV IDFH@ 7KH 1DGHU PLQLRQ VD\V LUUDGLDWLRQ FKDQJHV WKH WDVWH RI PHDW ZKHUHDV 5KRGHV VD\V LW GRHVQnW ,W ZRXOG EH LQWHUHVWLQJ WR NQRZ LI DQ\RQH KDV HYHU GRQH DQ DFWXDO WHVW :$6+,1*721 5HXWHUf ,UUDGLDWLRQ ,W PD\ VRXQG VLQLVWHU EXW IRRG VDIHW\ H[SHUWV VD\ LW FRXOG EHFRPH D NH\ WRRO WR NHHS VLFNQHVVFDXVLQJ EDFWHULD RXW RI WKH 86 PHDW VXSSO\ 0DWWHO &DVH $V VKRZQ LQ )LJXUH EHORZ RXW RI WKH WRWDO PHVVDJHV bf ZHUH FRGHG LQ WKH FDWHJRU\ RI HLWKHU TXHVWLRQDQVZHU RU RSLQLRQ :KLOH QXPEHUV RI WKH TXHVWLRQDQVZHU W\SH PHVVDJHV ZHUH HTXDOO\ GLVWULEXWHG XQWLO -DQXDU\ WKH QXPEHU RI RSLQLRQ W\SH PHVVDJHV UHDFKHG LWV SHDN EHIRUH -DQXDU\ 7KLV VXJJHVWV WKDW WKH RQOLQH XVHUVn FRQFHUQV DERXW WKH FULVLV GLPLQLVKHG DIWHU LW KDV SURJUHVVHG WR WKH DEDWH VWDJH 2YHUDOO WKH GLVWULEXWLRQ RI IUHTXHQF\ RI WKH PHVVDJHV FRUUHVSRQGHG ZLWK PHGLD FRYHUDJH RI WKH FULVLV DV SUHYLRXVO\ VHHQ LQ )LJXUH

PAGE 105

4 t $ 2 SLQLRQ 6DPSOLQJ 'DWH ),*85( 'LVWULEXWLRQ RI QXPEHUV RI RQOLQH PHVVDJHV E\ W\SHV LQ WKH 0DWWHO FDVH +XGVRQ )RRGV &DVH $Q LQWHUHVWLQJ ILQGLQJ DURVH IURP WKH +XGVRQ )RRGV FDVH ZKHQ FRQWUDVWLQJ WKH SXEOLF DJHQGD ZLWK WKH PHGLD DJHQGD WKHUH DUH PRUH PHGLD VWRULHV WKDQ 86(1(7 PHVVDJHV 7KH GLVWULEXWLRQ RI QXPEHUV RI PHVVDJHV E\ W\SHV GXULQJ WKH VDPSOLQJ SHULRG LV GHPRQVWUDWHG LQ )LJXUH 4 t $ 2 SLQLRQV !? ),*85( 'LVWULEXWLRQ RI QXPEHUV RI RQOLQH PHVVDJHV E\ W\SHV LQ WKH +XGVRQ )RRGV FDVH $QDO\VLV RI )LJXUH VXJJHVWHG WKDW WKH RFFXUUHQFH RI WKH SXEOLF RQOLQH PHVVDJHV FDQ EH GLYLGHG LQWR WKUHH VWDJHV 7KH ILUVW VWDJH UXQV IURP WKH ILUVW VDPSOLQJ

PAGE 106

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f DQG KDQGOLQJ RI WKH FRQWDPLQDWHG PHDW PDLQ WRSLF RI TXHVWLRQV LQ WKH ILQDO VWDJHf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

PAGE 107

DJJUHJDWLQJ GDWD LQ RQH YDULDEOH ZDV WR DGG WKHP WRJHWKHU IURP WZR FDVHV DQG WKHQ WR GLYLGH WKH VXP E\ WZR $ORQJ ZLWK WKRVH IRXU YDULDEOHV WRQH RI PHVVDJHV DQG XVHUVn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f U 7KHUH ZDV D VWURQJ VLJQLILFDQW FRUUHODWLRQ EHWZHHQ WKH QXPEHU RI RQOLQH PHVVDJHV DQG WKH QXPEHU RI PHGLD VWRULHV )LJXUH VKRZV WKDW PRVW RQOLQH PHVVDJHV ZHUH SRVWHG EHWZHHQ GD\ WR GD\ WKDW LV RQOLQH XVHUV SDLG DWWHQWLRQ WR WKH FRUSRUDWH FULVLV LQ WKH ILUVW GD\V DQG WKHQ WKHLU LQWHUHVWV ZDQHG

PAGE 108

),*85( )UHTXHQFLHV RI FRPELQHG RQOLQH PHVVDJHV DQG PHGLD VWRULHV )UHTXHQFLHV RI FRPELQHG QHJDWLYH RQOLQH PHVVDJHV DQG WKH 2SWLPLVP 6FRUHV LQ PHGLD VWRULHV DUH LOOXVWUDWHG LQ )LJXUH 7KH FRUUHODWLRQ FRHIILFLHQW LQGLFDWHG D PRGHUDWHO\ VLJQLILFDQW UHODWLRQVKLS EHWZHHQ WKH WZR YDULDEOHV U S ) f U &RQFHUQLQJ WKH WKLUG TXHVWLRQ D &KLVTXDUH WHVW LQ 7DEOH GLVSOD\HG D QRQn VLJQLILFDQW GLIIHUHQFH EHWZHHQ 7$/.(56 DQG 5($&7256 LQ SRVWLQJ WKH QHJDWLYH DV ZHOO DV SRVLWLYH RQOLQH PHVVDJHV RQ WKH 86(1(7 [ 8 1 f S 7KLV UHVXOW ZDV FRQVLVWHQW ZLWK ZKDW KDV EHHQ IRXQG LQ HDFK FULVLV FDVH DV 7$/.(56 GLG QRW SDUWLFXODUO\ SRVW PRUH QHJDWLYH PHVVDJHV WKDQ 5($&7256

PAGE 109

f§ff§ RI &RPELQHG 1HJDWLYH 2Qn OLQH 0HVVDJHV f§Dpf§&RPELQHG 2SWLPLVP 6FRUHV b D V = D 6HTXHQFH RI 'D\V ),*85( )UHTXHQFLHV RI FRPELQHG QHJDWLYH RQOLQH PHVVDJHV DQG WKH 2SWLPLVP 6FRUHV LQ PHGLD VWRULHV 7$%/( 5HVXOWV RI &KLVTXDUH WHVW ZLWK FRPELQHG GDWD 1HJDWLYH 3RVLWLYH 7RWDO GI &KLVTXDUH YDOXH 7$/.(5 16 5($&725 7RWDO

PAGE 110

7KH VXPPDWLRQ RI ILQGLQJV DQG DQVZHUV WR WKH UHVHDUFK TXHVWLRQV ZLOO EH SUHVHQWHG LQ WKH QH[W FKDSWHU 7KH FULVLV FRPPXQLFDWLRQ PRGHO EDVHG XSRQ WKH SXEOLF DJHQGD H[DPLQHG LQ ERWK FDVHV ZLOO DOVR EH GLVFXVVHG IROORZHG E\ OLPLWDWLRQV RI WKH VWXG\ ,PSOLFDWLRQV DQG VXJJHVWLRQV IRU SXEOLF UHODWLRQV SURIHVVLRQDOV DV ZHOO DV DFDGHPLF UHVHDUFKHUV FRQFOXGH WKH ILQDO FKDSWHU

PAGE 111

&+$37(5 9, ',6&866,21 &RQWUDU\ WR WKH WUDGLWLRQDO ZD\ RI VXUYH\LQJ SHRSOH DERXW WKHLU RSLQLRQV WKLV UHVHDUFK HPSOR\HG FRQWHQW DQDO\VLV WR LQYHVWLJDWH ZKDW SHRSOH UHDOO\ WKLQN DQG VD\ DERXW D FRUSRUDWLRQ LQYROYHG LQ DQ XQH[SHFWHG FULVLV $JHQGDVHWWLQJ WKHRU\ IXQFWLRQHG DV D WKHRUHWLFDO IUDPHZRUN IRU GHYHORSLQJ UHVHDUFK TXHVWLRQV ,Q DGGLWLRQ 6WXUJHVf SXEOLF RSLQLRQ PRGHO DQG 0DF.XHQf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

PAGE 112

6XPPDWLRQ RI )LQGLQJV 0DWWHO &DVH 6HYHQ K\SRWKHVHV RXW RI LQ WKH 0DWWHO FULVLV FDVH ZHUH VXSSRUWHG 7KH QXPEHU RI QHJDWLYH 86(1(7 PHVVDJHV LQFUHDVHG DIWHU WKH FULVLV HUXSWHG 7KHUH ZHUH PHVVDJHV GLVFXVVLQJ WKH FULVLV GD\V EHIRUH WKH FRPSDQ\ GHFLGHG WR WDNH LQWHUYHQWLRQ DFWLRQV $PRQJ WKHP bf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 S 7KH RQO\ K\SRWKHVLV QRW VXSSRUWHG ZDV WKH WHVW RI ELYDULDWH FRUUHODWLRQ EHWZHHQ XVHUVn VWUDWHJLHV DQG WKH WRQH RI RQOLQH PHVVDJHV 7KHUH ZDV QR VLJQLILFDQW FRUUHODWLRQ EHWZHHQ HPSOR\PHQW RI XVHUVn VWUDWHJLHV DQG WKH QXPEHU RI QHJDWLYH RQOLQH PHVVDJHV ;1 f S +XGVRQ )RRGV &DVH 2QO\ WKUHH K\SRWKHVHV ZHUH VXSSRUWHG LQ WKLV PHDW FRQWDPLQDWLRQ FDVH 7KH WRWDO QXPEHU RI RQOLQH PHVVDJHV DERXW WKH +XGVRQ )RRGV FRPSDQ\ DQG QHJDWLYH RQOLQH PHVVDJHV LQFUHDVHG DIWHU WKH FULVLV HUXSWHG :LWK D IHZ SRVLWLYH RQOLQH PHVVDJHV DQDO\]HG WKHLU QXPEHUV ZHUH VLPLODU DIWHU WKH LQWHUYHQWLRQ DFWLRQV ZHUH WDNHQ ) f J

PAGE 113

8QOLNH WKH ILQGLQJV LQ WKH 0DWWHO FDVH WKH QXPEHU RI PHGLD VWRULHV GLG QRW FDXVDOO\ LPSDFW WKH QXPEHU RI RQOLQH PHVVDJHV U S ) f W U ZLWK PHGLD VWRULHV DQG RQOLQH PHVVDJHV DQDO\]HG 1HLWKHU GLG WKH WRQH RI PHGLD VWRULHV KDYH D VLJQLILFDQW FRUUHODWLRQ ZLWK WKH QXPEHU RI QHJDWLYH RQOLQH PHVVDJHV U U S $V LQ WKH 0DWWHO FDVH QR VLJQLILFDQW FRUUHODWLRQ EHWZHHQ XVHUVn VWUDWHJ\ DQG WKH QXPEHU RI QHJDWLYH RQOLQH PHVVDJHV ZDV IRXQG [ 7 1 f S 'DWD $JJUHJDWLRQ LQ %RWK &DVHV 5HVXOWV IURP DQDO\VLV RI WKH DJJUHJDWHG GDWD FRUUHVSRQGHG ZLWK FRQFHSWV RI DJHQGDVHWWLQJ WKHRU\ 7KH ILQGLQJ U S ) f U VXSSRUWHG WKH DVVXPSWLRQ WKDW WKH FRQYHQWLRQDO PHGLD FRYHUDJH RI D FRUSRUDWH FULVLV RQH VLGHGO\ DIIHFWV WKH SXEOLF GLVFXVVLRQ DERXW WKH FULVLV RU WKH WDUJHW FRPSDQ\ 7KH WRQH WKDW WKH PHGLD XVHG WR FRYHU WKH FULVLV HYHQWV DOVR KDG D PRGHUDWHO\ VLJQLILFDQW FRUUHODWLRQ ZLWK WKH QXPEHU RI QHJDWLYH RQOLQH PHVVDJHV U S ) f U $JDLQ QR VLJQLILFDQW FRUUHODWLRQ EHWZHHQ XVHUVf VWUDWHJ\ DQG WKH QXPEHU RI QHJDWLYH RQOLQH PHVVDJHV ZDV IRXQG b 1 f S 5HVHDUFK 4XHVWLRQV 5HVHDUFK TXHVWLRQV LQ WKLV VWXG\ ZHUH EDVHG RQ WKH 6WXUJHVf SXEOLF RSLQLRQ PRGHO SDVW DJHQGDVHWWLQJ VWXGLHV DQG 0DF.XHQfV PRGHO RI SXEOLF GLDORJXH 7KH\ DUH VXPPDUL]HG IURP &KDSWHU ,,, DV IROORZV 'RHV WKH WUHQG RI RQOLQH SXEOLF RSLQLRQ GXULQJ WKH FRUSRUDWH FULVLV IROORZ 6WXUJHVf PRGHO"

PAGE 114

'RHV WKH PHGLD DJHQGD DIIHFW WKH SXEOLF DJHQGD LQ WHUPV RI WKH ILUVW DQG VHFRQG OHYHO RI DJHQGDVHWWLQJ WKHRU\" 'R RQOLQH XVHUV DSSOLHG WR 0DF.XHQfV FRQFHSW RI XVHUVf VWUDWHJLHV SRVW GLIIHUHQW QXPEHUV RI QHJDWLYH PHVVDJHV" 7KH WUHQG RI SXEOLF RSLQLRQ GXULQJ WKH WLPH RI WKH 0DWWHO FULVLV PDWFKHG 6WXUJHVf SXEOLF RSLQLRQ PRGHO RI FULVLV FRPPXQLFDWLRQ 7KH QXPEHU RI RQOLQH GLVFXVVLRQ PHVVDJHV PRYHG XS DQG GRZQ ZLWK WKH DGYHQW RI WKH FULVLV DQG DGRSWLRQ RI WKH LQWHUYHQWLRQV 7KH GHYHORSPHQW RI SXEOLF RSLQLRQ ZDV VLPLODU WR ZKDW ZDV SUHGLFWHG E\ 6WXUJHV f IURP SUHFULVLV WR WKH WHUPLQDWLRQ VWDJH 6HH )LJXUH RQ S f 7R FRPSDUH WKH QXPEHUV RI QHJDWLYH DQG WRWDO RQOLQH PHVVDJHV ZLWK WKH PHGLD UHSRUWV WKH PHGLD LPSDFW RQ WKH SXEOLF DJHQGD ZDV FRQVLVWHQW ZLWK WKH DJHQGDVHWWLQJ VWXGLHV 7KHUHIRUH DQVZHUV WR WKH ILUVW WZR TXHVWLRQV DUH SRVLWLYH ,Q WKH +XGVRQ )RRGV FDVH WKH WUHQG RI SXEOLF RSLQLRQ GLG QRW ORRN OLNH WKDW LQ WKH 0DWWHO FDVH DOWKRXJK WKH QXPEHU RI QHJDWLYH RU WRWDO RQOLQH PHVVDJHV LQGHHG JUHZ LQ WKH EHJLQQLQJ RI WKH FULVLV 7KH QXPEHU RI RQOLQH PHVVDJHV GLG QRW GHFOLQH HYHQ DIWHU +XGVRQ )RRGV WRRN LQWHUYHQWLRQ DFWLRQV 7KH SXEOLF VWLOO IHOW YHU\ FRQFHUQHG DERXW WKLV FULVLV 7KH UHDVRQ IRU WKLV FRXOG UHVXOW IURP WKH SURORQJHG PHGLD UHSRUWV DERXW IRRG VDIHW\ )RU LQVWDQFH QHZV DUWLFOHV EHWZHHQ 6HSWHPEHU DQG 2FWREHU DERXW WKH +XGVRQ )RRGV FRPSDQ\ RU IRRG VDIHW\ LQ PDMRU QHZVSDSHUV ZHUH IRXQG WKURXJK D VHDUFK LQ WKH /H[LV1H[LV GDWD EDVH 7KH PHGLD DJHQGD KDV JRQH EH\RQG WKH GD\ WLPH IUDPH LQ WKLV FDVH VWXG\ $V D PDWWHU RI IDFW WKH PHDW FRQWDPLQDWLRQ DFFLGHQW KDG EHHQ UHSRUWHG GDWLQJ EDFN WR -XQH

PAGE 115

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bf RI PHGLD VWRULHV DSSHDUHG IURP $XJXVW WR $XJXVW ,Q IDFW LI FRPSDULQJ ERWK YDULDEOHV ZLWKLQ DQRWKHU WLPH IUDPH WKH ILQGLQJ GLIIHUV )RU H[DPSOH LI WKH DYHUDJH QXPEHU RI QHZV VWRULHV LV FRPSDUHG ZLWK WKDW RI RQOLQH PHVVDJHV HDFK GD\ EHIRUH WKH ILUVW LQWHUYHQWLRQ WKH FRUUHODWLRQ LV VLJQLILFDQW U S ) f U $V GLVFXVVHG HDUOLHU DQVZHUV WR WKH WKLUG UHVHDUFK TXHVWLRQ LQ ERWK FDVHV LV QR 1R VLJQLILFDQW FRUUHODWLRQ EHWZHHQ XVHUVf VWUDWHJLHV 7$/.(5 DQG 5($&725 DQG WKH QXPEHU RI QHJDWLYH RQOLQH PHVVDJHV ZDV IRXQG HYHQ ZLWK GDWD FRPELQHG $V D UHVXOW WKHUH ZDV QR GLIIHUHQFH DPRQJ RQOLQH XVHUV EHLQJ LGHQWLILHG DV 7$/.(56 RU 5($&7256 LQ SRVWLQJ QHJDWLYH PHVVDJHV DERXW WKH FRPSDQ\ )DLOXUH WR GHWHFW WKH VLJQLILFDQW FRUUHODWLRQ EHWZHHQ XVHUVf VWUDWHJLHV DQG WKH QXPEHU RI QHJDWLYHO\ WRQHG RQOLQH PHVVDJHV FRXOG UHVW RQ WKH GLIIHUHQFH RI WKH VHWWLQJV ZKHUH WKH SXEOLF GLDORJXHV WDNH SODFH RQ D GDLO\ EDVLV 0DF.XHQ XUJHG WKDW GHPRFUDF\ ZLOO SURJUHVV ZKHQ SHRSOH XVH 7$/.&/$0 VWUDWHJLHV LQ D SXEOLF GLVFXVVLRQ IRUXP WR EHFRPH LQYROYHG LQ PHDQLQJIXO GLDORJXHV +RZHYHU LQ RUGHU WR IUDPH D IULHQGO\ VLWXDWLRQ SHRSOH ZLOO OLNHO\ DGRSW 5($&76,*1$/ VWUDWHJLHV ZLWKRXW H[SUHVVLQJ GLVDJUHHPHQW

PAGE 116

6LWXDWHG LQ WKH VRFDOOHG fYLUWXDO UHDOLW\f RI WKH ,QWHUQHW DQRQ\PRXV XVHUV E\ XVLQJ 7$/.(5 RU 5($&725 VWUDWHJLHV DUH DOORZHG WR IUHHO\ SRVW WKHLU RSLQLRQV HLWKHU DJUHHLQJ RU GLVDJUHHLQJ ZLWK RWKHU SHRSOH ZLWKRXW ZRUU\LQJ DERXW JHWWLQJ LQWR DQ XQIULHQGO\ HQYLURQPHQW WKDW LPSHGHV SXEOLF GLVFXVVLRQ RU PXWXDO FRPPXQLFDWLRQ 3URSRVHG &ULVLV &RPPXQLFDWLRQ 0RGHO ,Q WLPHV RI FULVLV PRQLWRULQJ WKH PHGLD DJHQGD DV ZHOO DV WKH SXEOLF DJHQGD LV WKH NH\ WR VXFFHVVIXO FULVLV PDQDJHPHQW &RQVLGHULQJ WKH XUJHQF\ RI WKH FULVLV DQG SUHVVXUH RQ WKH RUJDQL]DWLRQ D FRPPXQLFDWLRQ PRGHO WKDW IDFLOLWDWHV FRQWURO RYHU WUHQGV RI SXEOLF RSLQLRQ LV LQGLVSHQVDEOH $Q\ FULVLV FRPPXQLFDWLRQ PRGHO EDVHG RQ WKLV H[DPLQDWLRQ RI SXEOLF GLVFXVVLRQ PHVVDJHV RQ WKH 86(1(7 PXVW LQWHJUDWH VHYHUDO IDFWRUV 7KH IRXUWK K\SRWKHVLV WHVW UHYHDOHG WKH LPSRUWDQFH RI DGRSWLQJ LQWHUYHQWLRQ VWUDWHJLHV WKDW FDQ UHGXFH WKH QXPEHU RI QHJDWLYH RQOLQH PHVVDJHV DERXW WKH FRUSRUDWLRQ &RQVHTXHQWO\ WKH HDUOLHU WKH LQWHUYHQWLRQV DUH HPSOR\HG WKH TXLFNHU WKH QHJDWLYH SXEOLF RSLQLRQ DERXW WKH FRPSDQ\ LV QHXWUDOL]HG ,Q ERWK FDVHV H[DPLQHG WKH FRQWHQW RI RQOLQH PHVVDJHV ZDV UHH[DPLQHG IRU WKH TXDOLWDWLYH DQDO\VLV 7KH UHVXOWV VKRZHG WKDW QRW DOO PHVVDJHV SRVWHG RQ WKH QHZVJURXSV H[SUHVVHG RSLQLRQ 6RPH RI WKHP ZHUH LQIRUPDWLRQ H[FKDQJH RU TXHVWLRQVDQVZHUV SRVWLQJ 7KH FRUSRUDWLRQ VKRXOG QRW RQO\ IHHO FRQFHUQHG DERXW ZKDW SHRSOH VD\ DERXW WKH FULVLV EXW DERXW ZKDW SHRSOH ZDQW WR NQRZ DERXW WKH FULVLV VLWXDWLRQ ,W WKHQ FDQ UHVSRQG WR WKH TXHVWLRQV SRVWHG RQ WKH ,QWHUQHW 7DNH +XGVRQ )RRGV DV DQ H[DPSOH 6RPH RQOLQH XVHUV DVNHG DERXW KRZ PXFK KDUP WKH (FROL EDFWHULD FRXOG FDXVH WR KXPDQ EHLQJV 6RPH IHOW DQ[LRXV DERXW WKH

PAGE 117

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

PAGE 118

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

PAGE 119

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

PAGE 120

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fV RSLQLRQ DERXW DOO NLQGV RI LVVXHV ,I WLPH DQG FRVW SHUPLW LW LV VXJJHVWHG WKDW ERWK SHUVRQDO VXUYH\V DQG FRQWHQW DQDO\VLV RI WKH SXEOLF PHVVDJHV SRVWHG RQ WKH RQOLQH GLVFXVVLRQ IRUXPV EH XWLOL]HG WR JUDVS WUHQGV RI SXEOLF RSLQLRQ PRUH REMHFWLYHO\ DQG DFFXUDWHO\

PAGE 121

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fV FRQFHUQV 7KH H[WHQW WR ZKLFK D FULVLV ZLOO LPSDFW WKH JHQHUDO SXEOLF DQG VRFLHW\ ZLOO DOVR SURORQJ WKH WLPH OLQH IRU WKH SXEOLF RQOLQH GLVFXVVLRQ $OWKRXJK RQO\ SXEOLF RQOLQH GLDORJXHV ZHUH DQDO\]HG DV D VXUURJDWH RI WKH SXEOLF DJHQGD LW ZDV WUXH WKDW DGRSWLRQ RI FRUSRUDWH LQWHUYHQWLRQ VWUDWHJLHV LV WKH NH\ IDFWRU WR VXFFHHG LQ QHXWUDOL]LQJ RU UHYHUVLQJ SHRSOHf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

PAGE 122

,QWHUQHW EXW PRUH SURDFWLYH VWHSV QHHG WR EH WDNHQ WR FODULI\ DQ\ GRXEWV RU PLVXQGHUVWDQGLQJV DSSHDULQJ RQ WKH ,QWHUQHW DPRQJ RQOLQH XVHUV $OWKRXJK RQOLQH XVHUV GLVFXVVLQJ ERWK FULVHV UHSUHVHQW RQO\ D WLQ\ VHJPHQW RI WKH JHQHUDO SXEOLF WKHLU YRLFHV VKRXOG QRW EH LJQRUHG E\ SXEOLF UHODWLRQV SUDFWLWLRQHUV 2QH UHDVRQ LV EHFDXVH LW LV YHU\ OLNHO\ WKDW RQH SHUVRQf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n VHWWLQJ WKHRU\ DQG VKHGV OLJKW RQ XQGHUVWDQGLQJ WKH LPSDFW RI WKLV G\QDPLF QHZ PHGLD WKH ,QWHUQHW 7KLV UHVHDUFK GHPRQVWUDWHV WKDW WKH WUDGLWLRQDO FRQFHSW RI DJHQGDVHWWLQJ WKHRU\ FDQ EH DSSOLHG WR WKH QHZ PHGLD HQYLURQPHQW ZLWK VWDWLVWLFDOO\ VLJQLILFDQW UHVXOWV DV LQ WKH 0DWWHO FDVH DQG LQ WKH DQDO\VLV RI FRPELQHG GDWD IURP ERWK FDVHV 3ULRU WR WKLV LW ZDV RQO\ DVVXPHG WKDW LWV DSSOLFDELOLW\ ZDV WUDQVIHUDEOH DPRQJ WKH PDLQVWUHDP PHGLD $V IDU DV FRPPXQLFDWLRQ WKHRULHV DUH FRQFHUQHG DJHQGDVHWWLQJ WKHRU\ ZLOO VHUYH DV D

PAGE 123

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

PAGE 124

6HFRQG EHFDXVH WKH LPSDFW RI HDFK FRUSRUDWH FULVLV RQ WKH VRFLHW\ DQG WKH SXEOLF YDULHV GXUDWLRQ RI WKH SXEOLFf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

PAGE 125

DJHQGDVHWWLQJ WKHRU\ UHVHDUFKHUV FDQ LQFRUSRUDWH LQIRUPDWLRQ DERXW WKH FULVLV IURP WKH QHZV ZLUH VHUYLFHV FRUSRUDWH UHDFWLRQV SXEOLFL]HG LQ WKH QHZV UHOHDVHV DQG WKH JRYHUQPHQWDO UHDFWLRQV WR WKH SXEOLF DJHQGD WR EHWWHU XQGHUVWDQG WKH LQWHUUHODWLRQVKLSV EHWZHHQ WKHVH LQGHSHQGHQW YDULDEOHV DQG WUHQGV RI SXEOLF RSLQLRQ $IWHU DOO SXEOLF RSLQLRQ GRHV QRW FKDQJH RQO\ EHFDXVH RI WKH VLQJOH IDFWRU RI PHGLD FRYHUDJH $JHQGDVHWWLQJ WKHRU\ SURYHG XVHIXO LQ SUHGLFWLQJ GHYHORSPHQW RI SXEOLF RSLQLRQ DERXW WKH FRUSRUDWH FULVLV H[SUHVVHG RQ WKH 86(1(7 WKURXJK H[DPLQDWLRQ RI WKH QHZV PHGLD FRYHUDJH &RQWHQW DQDO\VLV RI WKH SXEOLF DJHQGD SURYLGHV DQRWKHU RXWORRN RI SXEOLF RSLQLRQ RQ WKH ,QWHUQHW )RU SXEOLF UHODWLRQV SURIHVVLRQDOV PRVW FULVHV KDSSHQ XQH[SHFWHGO\ EXW KRSHIXOO\ WKLV FRQWHQW DQDO\VLV RI RQOLQH PHVVDJHV ZLOO RIIHU LQVLJKW LQWR KRZ WKH\ VKRXOG KDQGOH WKH FULVLV LQ WHUPV RI PRQLWRULQJ WKH LQWHUDFWLYH QHZ PHGLD )RU WKH DFDGHPLF UHVHDUFKHUV FRQWHQW DQDO\VLV RI WKH SXEOLF DJHQGD FDQ LPSURYH WKH DJHQGDVHWWLQJ WKHRUHWLFDO IUDPHZRUN LQ YLHZ RI WKH ILUVW DQG VHFRQG OHYHO UHVHDUFK %HFDXVH PRVW ILQGLQJV FRUUHVSRQGHG ZLWK WKH LGHDV RI DJHQGDVHWWLQJ WKHRU\ WKLV VWXG\ VKRXOG FRQWULEXWH WR HQULFKLQJ WKH HPSLULFDO DV ZHOO DV WKHRUHWLFDO DSSOLFDELOLW\ RI WKLV WKHRU\ DW ERWK WKH ILUVW DQG VHFRQG OHYHOV WR WKH QHZ PHGLD UHVHDUFK

PAGE 126

$33(1',; $ +20( 3$*( 2) '(-$ 1(:6 6($5&+ (1*,1( ,Q LQWHUQDO 'L.8MVORQ 97t P\ g‘n Kr£ =sL .LWH f & -ULGnM rr LI ,N n-U£UL 5 WLYR LMLIH0nnf rO ,frr c9n1 n ,_ N\LIMLLQLLLL5V :LQGRZV ,V ,W 8SJUDGHZRUWK\" m!ff!} r .YDrUY0Q£ n9IW£6:: OR ,r EnQYMYUIFFPUD FDM ED!\ 2LRQ \IW£\Yn (n} ODPD 0HWW! _X XYr}L} Q LLIF LU} R ‹OLLOLLL 'OUJ0.OJD c IVLO9\O c eEZLeLLV L ;IHL0D.DLU( 1m:8m.f $OLW:JWIWW;FI0L f $0fLLL} IWLOZOFPD PV f LmK!!"JDZ}MeHDD ‘OnLnI n .OLWL,O&6D2r 6", f : LLLe \M==t;; &WS-OIL & OWI0 ULM n,OQnWKL URLUYLGnA+LYQ2*Vr

PAGE 127

$33(1',; % 32:(5 6($5&+ 3$*( 2) '(-$ 1(:6 6($5&+ (1*,1( M2 t P\ I LXGOL 8.6} 6P:P. 8Lr 3RZHU 6VDUDLO 6OOOGW IOMU (LDDSOLV OOLL $1+ OWp72"0D 0O 3QG cnLXN/LLOL6OM ,IIL6.6/(06 +LIR $UFKLYH &.LLS6:% __ ,SrWLmO L_ VtMOWIFLG 1XPEHU0 UQUWW&I-IF" IHVXLV (DUP£I IEVDWt$ (rDPSOn6"rXYW,} RU [IILHVr $XOL2UW"B I Q )@ 6, &? FNt ,% XEMWFE8fcB ) W0,,,-OLU c,P LDLR&OLIFMDL&nZ6&&IOO .Y 6HHUV NO ,W 72L /YGULSOY I2UVLO $SU ORR 'LZUDQFL ,f§R WLSSOH :DDO UfWRNOL 9HWKP %OWHFV f INJL6OJ
PAGE 128

$33(1',; & &2',1* 6&+(0( )25 1(:6 6725,(6 1HZV VWRU\ LGHQWLILFDWLRQ 0HGLD VRXUFH Â’ $%& 1HZV Â’ &11 Â’ 1%& 1HZV Â’ 1HZ
PAGE 129

PDNLQJ SXEOLF DSRORJ\ DFNQRZOHGJHPHQW RI ZURQJGRLQJ UHSODFHPHQW RI SHUVRQV ZKR FDXVHV WKH FULVLV RU SURPLVH WR SUHYHQWLQJ IURP RFFXUUHQFH RI WKH VLPLODU FULVLV

PAGE 130

$33(1',; &2',1* 6&+(0( )25 21/,1( 0(66$*(6 2QOLQH PHVVDJH LGHQWLILFDWLRQ 2QOLQH PHVVDJH GDWHG 7DUJHW RI IODPLQJ PHVVDJH 8VHUfV VWUDWHJ\ ’ 7$/.(5 8VHUfV QDPH LQ WKH HPDLO 7RQH RI RQOLQH PHVVDJH ’ 1HJDWLYHf§ :K\" Df UHTXHVW OHJDO DFWLRQ DJDLQVW WKH FRPSDQ\ Ef UHTXHVW FRPSHQVDWLRQ IURP WKH FRPSDQ\ Ff PHQWLRQ RI SHUVRQDO RU RWKHUV H[SHULHQFHV VXIIHULQJ IURP WKH FULVLV HYHQW Gf XUJH WKH ER\FRWW RI WKH FRPSDQ\fV SURGXFW RU VHUYLFH Hf JLYH VDUFDVWLF MRNHV RU KDUVK FRPPHQWV DERXW WKH FRPSDQ\ RU If EODPH RU DFFXVH WKH FRPSDQ\ IRU LWV ZURQJ GRLQJ RU LWV GHIHFWHG SURGXFWVf 0RQWK 'D\
PAGE 131

’ 3RVLWLYHf§ :K\" Df XUJH HQG RI D ER\FRWW Ef FRPSOLPHQW WKH FRPSDQ\ Ff PHQWLRQ WKH ODZV DV XQIDLU WR WKH RUJDQL]DWLRQ Gf SUDLVH FRPSDQ\fV LQWHUYHQLQJ DFWLRQV RU Hf GHQ\ WKDW WKH FULVLV HYHQW LV VROHO\ GXH WR WKH FRPSDQ\ ’ 1HXWUDOf§ :K\" Df IRUZDUGHG PHGLD UHSRUWV ZLWKRXW H[SUHVVLQJ RSLQLRQ Ef IRUZDUGHG ZHEVLWH LQIRUPDWLRQ ZLWKRXW H[SUHVVLQJ RSLQLRQ Ff SHUVRQDO TXHVWLRQV RU UHVSRQVHV WR RWKHUVf PHVVDJHV ZLWKRXW H[SUHVVLQJ DQ\ RSLQLRQV RU Gf VWDWHPHQWV WKDW WDNH WKH FULVLV LVVXH IRU JUDQWHG ZLWKRXW EODPLQJ RU SUDLVLQJ WKH FRPSDQ\ ’ 1RQDSSOLFDEOH 8VHUfV LGHQWLILFDWLRQ ’ &RPSDQ\fV HPSOR\HH ’ &RPSDQ\fV PDQDJHPHQW ’ 2WKHU RFFXSDWLRQ ’ 8QDEOH WR LGHQWLI\

PAGE 132

$33(1',; ( 6$03/( 2) 5(3257 ),/( ,1 ',&7,21 62)7:$5( 727$/ 180%(5 2) :25'6 $1$/<=(' 180%(5 2) &+$5$&7(56 ,1 3$66$*( &+$5$&7(56 3(5 :25' 180%(5 2) ',))(5(17 :25'6 $/3+$180(5,& ,'(17,),(56 '(6&5,37,9( ,'(17,),(5 QRQH! ,1387 7(;7 *HH] &KXFNLH /,9(6 $FWXDOO\ FDQ \RX LPDJLQH KRZ IULJKWHQLQJ VRPHWKLQJ OLNH WKDW ZRXOG EH WR D OLWWOH NLG $WWDFNHG E\ D GROO WKDW HDWV LWnV RZQHUV" r 7KLV SDVVDJH FRQWDLQV IHZHU WKDQ ZRUGV r r %HFDXVH ',&7,21n6 QRUPV DUH EDVHG RQ r r ZRUG VDPSOHV WKH UDZ VFRUHV UHSRUWHG r r KHUH FDQQRW EH FRPSDUHG WR WKH QRQQDO UDQJH r r RI VFRUHV SURYLGHG r ',&7,21$5< 727$/6 9$5,$%/( )5(4 )5(4f 3(5 &(17 1250$/ 5$1*( 2) 6&25(6 f 67$1'$5'n ,=(' 6&25(6 1XPHULFDO 7HUPV r $PELYDOHQFH r 6HOIUHIHUHQFH

PAGE 133

7HQDFLW\ +H /HYHOLQJ r &ROOHFWLYHV 3UDLVH r 6DWLVIDFWLRQ ,QVSLUDWLRQ r %ODPH +DUGVKLS r $JJUHVVLRQ r $FFRPSOLVKPHQW r &RPPXQLFDWLRQ r &RJQLWLYH 7HUPV r 3DVVLYLW\ r 6SDWLDO $ZDUHQHVV r )DPLOLDULW\ r 7HPSRUDO $ZDUHQHVV r 3UHVHQW &RQFHUQ r +XPDQ ,QWHUHVW r &RQFUHWHQHVV r 3DVW &RQFHUQ &HQWUDOLW\ 5DSSRUW r

PAGE 134

&RRSHUDWLRQ r 'LYHUVLW\ r ([FOXVLRQ /LEHUDWLRQ 'HQLDO r 0RWLRQ r 86(5 ',&7,21$5< 727$/6 ',&7,21$5< 1$0( )5(48(1&< QRQH! :25'6 '(6,*1$7(' 72 &20326( ,16,67(1&( 6&25( 1RQH &$/&8/$7(' 9$5,$%/(6 6&25( 1250$/ 5$1*( 2) 6&25(6 f 67$1'$5'n ,=(' 6&25(6 ,QVLVWHQFH r (PEHOOLVKPHQW 9DULHW\ r &RPSOH[LW\ &20326,7( 9$5,$%/(6 6&25( $FWLYLW\ 2SWLPLVP &HUWDLQW\ r 5HDOLVP r &RPPRQDOLW\

PAGE 135

5()(5(1&(6 f +LOILJHU GHQLHV ,QWHUQHW DFFXVDWLRQV RI UDFLVP $VVRFLDWHG 3UHVV %DOO5RNHDFK 6 t 5HDUGRQ f 0RQRORJXH GLDORJXH DQG WHOHORJ ,Q 5 3 +DZNLQV 0 :LHPDQQ t 6 3LQJUHH (GVf $GYDQFLQJ &RPPXQLFDWLRQ 6FLHQFH 0HUJLQJ 0DVV DQG ,QWHUSHUVRQDO 3URFHVVHV SS f 1HZEXU\ 3DUN &$ 6DJH %DUWRQ / f &ULVLV LQ 2UJDQL]DWLRQV &LQFLQQDWL 2+ 6RXWK:HVWHUQ 3XEOLVKLQJ &R %HFNHU / t 0F&RPEV 0 f 7KH UROH RI WKH SUHVV LQ GHWHUPLQLQJ YRWHU UHDFWLRQV WR SUHVLGHQWLDO SULPDULHV +XPDQ &RPPXQLFDWLRQ 5HVHDUFK %HKU 5 / t ,\HQJDU 6 f 7HOHYLVLRQ QHZV UHDOZRUOG FXHV DQG FKDQJHV LQ WKH SXEOLF DJHQGD 3XEOLF 2SLQLRQ 4XDUWHUO\ f %HQLJHU 5 f 0HGLD FRQWHQW DV VRFLDO LQGLFDWRUV 7KH *UHHQILHOG ,QGH[ RI DJHQGDVHWWLQJ &RPPXQLFDWLRQ 5HVHDUFK f %HUHOVRQ % f &RQWHQW $QDO\VLV LQ &RPPXQLFDWLRQV 5HVHDUFK 1HZ
PAGE 136

&DUQH\ $ t -RUGHQ $ f 3UHSDUH IRU EXVLQHVVUHODWHG FULVHV 3XEOLF 5HODWLRQV -RXUQDO f &RKHQ $XJXVWf $ UHSRUW RQ D QRQHOHFWLRQ DJHQGDVHWWLQJ VWXG\ 3DSHU SUHVHQWHG DW WKH $VVRFLDWLRQ IRU (GXFDWLRQ LQ -RXUQDOLVP 2WWDZD &DQDGD 'n$YHQL 5 $ t 0DF0LOODQ & f &ULVLV DQG WKH FRQWHQW RI PDQDJHULDO FRPPXQLFDWLRQV $ VWXG\ RI WKH IRFXV RI DWWHQWLRQ RI WRS PDQDJHUV LQ VXUYLYLQJ DQG IDLOLQJ ILUPV $GPLQLVWUDWLYH 6FLHQFH 4XDUWHUO\ 'HDULQJ : t 5RJHUV ( 0 f $JHQGD6HWWLQJ 7KRXVDQG 2DNV &$ 6DJH 3XEOLFDWLRQV 'XWWRQ ( f 7KH SURFHVVLQJ RI FULVLV DQG QRQFULVLV VWUDWHJLF LVVXHV -RXUQDO RI 0DQDJHPHQW 6WXGLHV nf '\HU 6 & -U 0LOOHU 0 0 t %RRQH f :LUH VHUYLFH FRYHUDJH RI WKH ([[RQ 9DOGH] FULVLV 3XEOLF 5HODWLRQV 5HYLHZ f (DWRQ + f $JHQGDVHWWLQJ ZLWK ELZHHNO\ GDWD RQ FRQWHQW RI WKUHH QDWLRQDO PHGLD -RXUQDOLVP 4XDUWHUO\ (JHOKRII : t 6HQ ) f $Q LQIRUPDWLRQSURFHVVLQJ PRGHO RI FULVLV PDQDJHPHQW 0DQDJHPHQW &RPPXQLFDWLRQ 4XDUWHUO\ f (QWPDQ 5 f )UDPLQJ 7RZDUG FODULILFDWLRQ RI D IUDFWXUHG SDUDGLJP -RXUQDO RI &RPPXQLFDWLRQ f (\DO & + f 7LPHIUDPH LQ DJHQGDVHWWLQJ UHVHDUFK $ VWXG\ RI WKH FRQFHSWXDO DQG PHWKRGRORJLFDO IDFWRUV DIIHFWLQJ WKH WLPH IUDPH FRQWH[W RI WKH DJHQGDVHWWLQJ SURFHVV 'RFWRUDO GLVVHUWDWLRQ 6\UDFXVH 8QLYHUVLW\ 6\UDFXVH 1< )HDP%DQNV f &ULVLV &RPPXQLFDWLRQV $ &DVHERRN $SSURDFK 0DKZDK 1/DZUHQFH (UOEDXP $VVRFLDWHV )LW]SDWULFN 5 f 7HQ JXLGHOLQHV IRU UHGXFLQJ OHJDO ULVNV LQ FULVLV PDQDJHPHQW 3XEOLF 5HODWLRQV 4XDUWHUO\ )RPEUXQ & f 7KH UHSXWDWLRQDO DXGLW 5HSXWDWLRQ 5HDOL]LQJ 9DOXH IURP WKH &RUSRUDWH ,PDJH SS f %RVWRQ 0$ +DUYDUG %XVLQHVV 6FKRRO 3UHVV )XQNKRXVHU 5 f 7KH LVVXHV RI WKH VL[WLHV $Q H[SORUDWRU\ VWXG\ LQ WKH G\QDPLFV RI SXEOLF RSLQLRQ 3XEOLF 2SLQLRQ 4XDUWHUO\

PAGE 137

*DUUDPRQH 0 +DUULV $ & t $QGHUVRQ 5 f 8VHV RI SROLWLFDO FRPSXWHU EXOOHWLQ ERDUGV -RXUQDO RI %URDGFDVWLQJ t (OHFWURQLF 0HGLD f *DXQW 3 t 2OOHQEXUJHU f ,VVXHV PDQDJHPHQW UHYLVLWHG $ WRRO WKDW GHVHUYHV DQRWKHU ORRN 3XEOLF 5HODWLRQV 5HYLHZ f *KDQHP 6 f )LOOLQJ LQ WKH WDSHVWU\ 7KH VHFRQG OHYHO RI DJHQGD VHWWLQJ ,Q 0 0F&RPEV / 6KDZ t :HDYHU (GVf &RPPXQLFDWLRQ DQG 'HPRFUDF\ ([SORULQJ WKH ,QWHOOHFWXDO )URQWLHUV LQ $JHQGD6HWWLQJ 7KHRU\ SS f 0DKZDK 1/DZUHQFH (UOEDXP $VVRFLDWHV ,QF *ROG t 6LPPRQV / f 1HZV VHOHFWLRQ SDWWHUQV DPRQJ ,RZD GDLOLHV 3XEOLF 2SLQLRQ 4XDUWHUO\ f *RQ]£OH]+HUUHUR $ t 3UDWW & % f +RZ WR PDQDJH D FULVLV EHIRUH RU ZKHQHYHU LW KLWV 3XEOLF 5HODWLRQV 4XDUWHUO\ f *URSHU 5 $XJXVW6HSWHPEHUf 3ROLWLFDO EHKDYLRU RI DFWRUV RQ HOHFWURQLF GLVFXVVLRQ JURXSV 3DSHU SUHVHQWHG DW WKH QG $QQXDO 0HHWLQJ RI $PHULFDQ 3ROLWLFDO 6FLHQFH $VVRFLDWLRQ 6DQ )UDQFLVFR &$ *UXQLJ ( f 6\PPHWULFDO 3UHVXSSRVLWLRQV DV D )UDPHZRUN IRU 3XEOLF 5HODWLRQV 7KHRU\ ,Q & + %RWDQ t 9 +D]OHWRQ -U (GVf 3XEOLF 5HODWLRQV 7KHRU\ SS f +LOOVGDOH 1/DZUHQFH (UOEDXP $VVRFLDWHV ,QF +DLQVZRUWK % f 7KH GLVWULEXWLRQ RI DGYDQWDJHV DQG GLVDGYDQWDJHV 3XEOLF 5HODWLRQV 5HYLHZ f +DUW 5 3 f ',&7,21 7KH 7H[W$QDO\VLV 3URJUDP 7KRXVDQG 2DNV &$ 6DJH 3XEOLFDWLRQV 6RIWZDUH +D\HV 5 ( f &RUSRUDWH FULVLV PDQDJHPHQW DV DGDSWLYH FRQWURO ,Q 6 $QGULROH (Gf &RUSRUDWH &ULVLV 0DQDJHPHQW SS f 3ULQFHWRQ 13HWURFHOOL +HDULW 0 f $SRORJLHV DQG SXEOLF UHODWLRQV FULVHV DW &KU\VOHU 7RVKLED DQG 9ROYR 3XEOLF 5HODWLRQV 5HYLHZ f +HUWRJ t )DQ 3 f 7KH LPSDFW RI SUHVV FRYHUDJH RQ VRFLDO EHOLHIV 7KH FDVH RI +,9 WUDQVPLVVLRQ &RPPXQLFDWLRQ 5HVHDUFK f +LHEHUW 5 ( f *OREDO SXEOLF UHODWLRQV LQ D SRVWFRPPXQLVW ZRUOG $ QHZ PRGHO 3XEOLF 5HODWLRQV 5HYLHZ f +LOO $ t +XJKHV ( f &RPSXWHUPHGLDWHG SROLWLFDO FRPPXQLFDWLRQ 7KH 86(1(7 DQG SROLWLFDO FRPPXQLFDWLRQ 3ROLWLFDO &RPPXQLFDWLRQ f

PAGE 138

+RIIPDQ 3 ( f ,QWHUQHW ,QVWDQW 5HIHUHQFH $ODPHGD &$ 6<%(; ,QF ,UYLQH 5 % t 0LOODU 3 $XJXVW f 'HEXQNLQJ WKH VWHUHRW\SHV RI FULVLV PDQDJHPHQW 7KH QDWXUH RI EXVLQHVV FULVHV LQ WKH nV 3DSHU SUHVHQWHG DW WKH WK $QQXDO &RQIHUHQFH RQ &ULVLV 0DQDJHPHQW /DV 9HJDV 19 ,WR < f 1HZ GLUHFWLRQV LQ FRPPXQLFDWLRQ UHVHDUFK IURP D -DSDQHVH SHUVSHFWLYH ,Q 3 *DXQW (Gf %H\RQG $JHQGDV 1HZ 'LUHFWLRQV LQ &RPPXQLFDWLRQ 5HVHDUFK SS f :HVWSRUW &7 *UHHQZRRG 3UHVV .DOLVK f 35 ILUPV VXUI WKH 1HW 5HXWHUV 1RUWK $PHULFDQ :LUH .LQJ 3W f 7KH SUHVV FDQGLGDWH LPDJHV DQG YRWHU SHUFHSWLRQV ,Q 0 0F&RPEV / 6KDZ t :HDYHU (GVf &RPPXQLFDWLRQ DQG 'HPRFUDF\ ([SORULQJ WKH ,QWHOOHFWXDO )URQWLHUV LQ $JHQGD6HWWLQJ 7KHRU\ SS f 0DKZDK 1/DZUHQFH (UOEDXP $VVRFLDWHV ,QF .ODSSHU 7 f 7KH (IIHFWV RI 0DVV &RPPXQLFDWLRQ 1HZ
PAGE 139

0D\HU 5 1 f *RQH \HVWHUGD\ KHUH WRGD\ &RQVXPHU LVVXHV LQ WKH DJHQGDVHWWLQJ SURFHVV -RXUQDO RI 6RFLDO ,VVXHV f 0D]XU $ f 3XWWLQJ UDGRQ RQ WKH SXEOLFnV ULVN DJHQGD 6FLHQFH 7HFKQRORJ\ DQG +XPDQ 9DOXHV f 0F&RPEV 0 f 1HZ IURQWLHUV LQ DJHQGD VHWWLQJ $JHQGDV RI DWWULEXWHV DQG IUDPHV 3DSHU SUHVHQWHG DW WKH $VVRFLDWLRQ IRU (GXFDWLRQ LQ -RXUQDOLVP DQG 0DVV &RPPXQLFDWLRQ &KLFDJR ,/ 0F&RPEV 0 t 6KDZ f 7KH DJHQGDVHWWLQJ IXQFWLRQ RI PDVV PHGLD 3XEOLF 2SLQLRQ 4XDUWHUO\ 0F&RPEV 0 'DQLHOLDQ / t :DQWD : f ,VVXHV LQ WKH QHZV DQG SXEOLF DJHQGD 7KH DJHQGDVHWWLQJ WUDGLWLRQ ,Q 7 / *ODVVHU t & 7 6DOPRQ (GVf 3XEOLF 2SLQLRQ DQG WKH &RPPXQLFDWLRQ RI &RQVHQW SS f 1HZ
PAGE 140

3ULFH 9 f &RPPXQLFDWLRQ &RQFHSWV 3XEOLF 2SLQLRQ 1HZEXU\ 3DUN &$ 6DJH 3ULHVW 6 + f 3XEOLF RSLQLRQ H[SHUW RSLQLRQ DQG WKH LOOXVLRQ RI FRQVHQVXV *OHDQLQJ SRLQWV RI YLHZ HOHFWURQLFDOO\ ,Q / %RUGHQ t +DUYH\ (GVf 7KH (OHFWURQLF *UDSHYLQH 5XPRU 5HSXWDWLRQ DQG 5HSRUWLQJ LQ WKH 1HZ 2Q/LQH (QYLURQPHQW SS f 0DKZDK 1/DZUHQFH (UOEDXP $VVRFLDWHV 5DIDHOL 6 f 7KH HOHFWURQLF EXOOHWLQ ERDUG $ FRPSXWHUGULYHQ PDVV PHGLXP &RPSXWHUV DQG WKH 6RFLDO 6FLHQFHV f 5DIDHOL 6 f ,QWHUDFWLYLW\ )URP QHZ PHGLD WR FRPPXQLFDWLRQ ,Q 5 +DZNLQV :LHPDQQ t 6 3LQJUHH (GVf $GYDQFLQJ &RPPXQLFDWLRQ 6FLHQFH 0HUJLQJ 0DVV DQG ,QWHUSHUVRQDO 3URFHVVHV SS f 1HZEXU\ 3DUN &$ 6DJH 5DIDHOL 6 t /D5RVH 5 f (OHFWURQLF EXOOHWLQ ERDUGV DQG SXEOLF JRRGV H[SODQDWLRQV RI FROODERUDWLYH PDVV PHGLD &RPPXQLFDWLRQ 5HVHDUFK f 5HHVH 6 t 'DQLHOLDQ / + f $ FORVHU ORRN DW LQWHUPHGLD LQIOXHQFHV RQ DJHQGD VHWWLQJ 7KH FRFDLQH LVVXH RI ,Q 3 6KRHPDNHU (Gf &RPPXQLFDWLRQ &DPSDLJQV DERXW 'UXJV *RYHUQPHQW 0HGLD DQG WKH 3XEOLF SS f +LOOVGDOH 1/DZUHQFH (UOEDXP $VVRFLDWHV 5HHVH 6 t 'DQLHOLDQ / + f ,QWHUPHGLD LQIOXHQFH DQG WKH GUXJ LVVXH &RQYHUJLQJ RQ FRFDLQH ,Q 3 6KRHPDNHU (Gf &RPPXQLFDWLRQ &DPSDLJQV DERXW 'UXJV *RYHUQPHQW 0HGLD DQG WKH 3XEOLF SS f +LOOVGDOH 1/DZUHQFH (UOEDXP $VVRFLDWHV 5HYDK 6 f ,WnV D MXQJOH RXW WKHUH LQ F\EHUVSDFH $PHULFDQ -RXUQDOLVP 5HYLHZ f 5REHUWV 0 f 3ROLWLFDO DGYHUWLVLQJnV LQIOXHQFH RQ QHZV WKH SXEOLF DQG WKHLU EHKDYLRU ,Q 0 0F&RPEV / 6KDZ t :HDYHU (GVf &RPPXQLFDWLRQ DQG 'HPRFUDF\ ([SORULQJ WKH ,QWHOOHFWXDO )URQWLHUV LQ $JHQGD6HWWLQJ 7KHRU\ SS f 0DKZDK 1/DZUHQFH (UOEDXP $VVRFLDWHV ,QF 5RJHUV ( 0 t 'HDULQJ : f $JHQGDVHWWLQJ UHVHDUFK :KHUH KDV LW EHHQ" :KHUH LV LW JRLQJ" ,Q $ $QGHUVRQ (Gf &RPPXQLFDWLRQ
PAGE 141

6KDZ & $SULO f 7KH $3 ,WnV HYHU\ZKHUH DQG SRZHUIXO /RV $QJHOHV 7LPHV SS 6KDZ / t 0F&RPEV 0 ( f 7KH (PHUJHQFH RI $PHULFDQ 3ROLWLFDO ,VVXHV 7KH $JHQGD6HWWLQJ )XQFWLRQ RI WKH 3UHVV 6W 3DXO 01 :HVW 3XEOLVKLQJ 6KRHPDNHU 3 :DQWD : t /HJJHWW f 'UXJ FRYHUDJH DQG SXEOLF RSLQLRQ ,Q 3 6KRHPDNHU (Gf &RPPXQLFDWLRQ &DPSDLJQV DERXW 'UXJV *RYHUQPHQW 0HGLD DQG WKH 3XEOLF SS f +LOOVGDOH 1/DZUHQFH (UOEDXP $VVRFLDWHV 6LJHOPDQ / /HERYLF :LOFR[ & t $OOVRS f $V WLPH JRHV E\ 'DLO\ RSLQLRQ FKDQJH GXULQJ WKH 3HUVLDQ *XOI &ULVLV 3ROLWLFDO &RPPXQLFDWLRQ 6RKQ $ % f $ ORQJLWXGLQDO DQDO\VLV RI ORFDO QRQSROLWLFDO DJHQGDVHWWLQJ HIIHFWV -RXUQDOLVP 4XDUWHUO\ 6WDQWRQ $ )HE f 3HQWLXP EURXKDKD D PDUNHWLQJ OHVVRQ %HWWHU FRPPXQLFDWLRQV ZRXOG KDYH KHOSHG $GYHUWLVLQJ $JH 6WDUEXFN : + *UHYH $ t +HGEHUJ % / 7 f 5HVSRQGLQJ WR FULVLV -RXUQDO RI %XVLQHVV $GPLQLVWUDWLRQ f 6WHPSHO P + f +RZ QHZVSDSHUV XVH WKH $VVRFLDWHG 3UHVV DIWHUQRRQ $:LUH -RXUQDOLVP 4XDUWHUO\ 6WHPSHO ,,, + t :HVWOH\ % + f 5HVHDUFK 0HWKRGV LQ 0DVV &RPPXQLFDWLRQ QG HGf (QJOHZRRG &OLIIV 13UHQWLFH +DOO 6WRQH & t 0F&RPEV 0 ( f 7UDFLQJ WKH WLPH ODJ LQ DJHQGDVHWWLQJ -RXUQDOLVP 4XDUWHUO\ 6WUHQVNL % f 7KH HWKLFV RI PDQLSXODWHG FRPPXQLFDWLRQ 3XEOLF 5HODWLRQV 4XDUWHUO\ f 6WXUJHV / f &RPPXQLFDWLQJ WKURXJK FULVLV $ VWUDWHJ\ IRU RUJDQL]DWLRQDO VXUYLYDO 0DQDJHPHQW &RPPXQLFDWLRQ 4XDUWHUO\ f 7DNHVKLWD 7 t 0LNDPL 6 f +RZ GLG WKH PDVV PHGLD LQIOXHQFH WKH YRWHUnV FKRLFH LQ WKH JHQHUDO HOHFWLRQ LQ -DSDQ" $ VWXG\ RI DJHQGDVHWWLQJ .HLR &RPPXQLFDWLRQ 5HYLHZ 7HUU\ ( f (GXFDWRU DQG SUDFWLWLRQHU GLIIHUHQFHV RQ WKH UROH RI WKHRU\ LQ SXEOLF UHODWLRQV ,Q & + %RWDQ t 9 +D]OHWRQ -U (GVf 3XEOLF 5HODWLRQV 7KHRU\ SS f +LOOVGDOH 1/DZUHQFH (UOEDXP $VVRFLDWHV

PAGE 142

7KRPVHQ 6 5 f #:RUN LQ F\EHUVSDFH ([SORULQJ SUDFWLWLRQHU XVH RI WKH 35)RUXP 3XEOLF 5HODWLRQV 5HYLHZ f 7UDYHUVR f 2SHQLQJ D FUHGLEOH GLDORJXH ZLWK \RXU FRPPXQLW\ 3XEOLF 5HODWLRQV -RXUQDO f 8PDQVN\ f +RZ WR VXUYLYH DQG SURVSHU ZKHQ LW KLWV WKH IDQ 3XEOLF 5HODWLRQV 4XDUWHUO\ f 9HQGUHOO % f 2LO VSLOOV VKRZ OHVVRQV VWLOO QRW OHDUQHG 3XEOLF 5HODWLRQV -RXUQDO f :DQWD : t :X <& f ,QWHUSHUVRQDO FRPPXQLFDWLRQ DQG WKH DJHQGDVHWWLQJ SURFHVV -RXUQDOLVP 4XDUWHUO\ f :DQWD : t +X <: f 7KH DJHQGDVHWWLQJ HIIHFWV RI LQWHUQDWLRQDO QHZV FRYHUDJH $Q H[DPLQDWLRQ RI GLIIHULQJ QHZV IUDPHV ,QWHUQDWLRQDO -RXUQDO RI 3XEOLF 2SLQLRQ 5HVHDUFK f :DQWD : t +X <: f 7LPHODJ GLIIHUHQFHV LQ WKH DJHQGDVHWWLQJ SURFHVV $Q H[DPLQDWLRQ RI ILYH QHZV PHGLD ,QWHUQDWLRQDO -RXUQDO RI 3XEOLF 2SLQLRQ 5HVHDUFK f :HDYHU *UDEHU 0F&RPEV 0 t (\DO & f 0HGLD $JHQGD 6HWWLQJ LQ D 3UHVLGHQWLDO (OHFWLRQ ,VVXHV ,PDJHV DQG ,QWHUHVW 1HZ
PAGE 143

:LOOQDW / t =KX -+ f 1HZVSDSHU FRYHUDJH DQG SXEOLF RSLQLRQ LQ +RQJ .RQJ $ WLPHVHULHV DQDO\VLV RI PHGLD SULPLQJ 3ROLWLFDO &RPPXQLFDWLRQ nf :LQWHU 3 t (\DO & + f $JHQGD VHWWLQJ IRU WKH FLYLO ULJKWV LVVXH 3XEOLF 2SLQLRQ 4XDUWHUO\ Af :RRGV % -XQH f ,W FRXOG KDSSHQ WR \RX $ IUDPHZRUN IRU XQGHUVWDQGLQJ DQG PDQDJLQJ FULVHV LQYROYLQJ UDFH 3DSHU SUHVHQWHG DW WKH 3XEOLF 5HODWLRQV 6RFLHW\ RI $PHULFD (GXFDWRUV 6HFWLRQ =KX -+ :DWW + 6Q\GHU / %
PAGE 144

%,2*5$3+,&$/ 6.(7&+ 7]RQJ+RPJ ']ZR ZDV ERP LQ D VRXWKHUQ FLW\ RI 7DLZDQ RQ 2FWREHU DQG KDV OLYHG LQ 7DLZDQ QHDUO\ \HDUV EHIRUH FRPLQJ WR WKH 8QLWHG 6WDWHV IRU JUDGXDWH VWXG\ +H UHFHLYHG D EDFKHORUfV GHJUHH LQ %ULWLVK DQG $PHULFDQ /LWHUDWXUH IURP 7XQJKDL 8QLYHUVLW\ LQ 7DL&KXQJ &LW\ RI 7DLZDQ $IWHU WZR\HDU PLOLWDU\ VHUYLFH IURP WR KH ZRUNHG DV D VDOHV SHUVRQ DQG DFFRXQW H[HFXWLYH IRU D GRPHVWLF VKLSSLQJ FRPSDQ\ IRU WZR DQG D KDOI \HDUV *DLQLQJ WKUHH \HDUV RI ZRUNLQJ H[SHULHQFHV LQ 7DLZDQ LQ ']ZR FDPH WR WKH 8QLWHG 6WDWHV WR SXUVXH KLV PDVWHUf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f /DQH &KXQJZX 5G 3LQJ7XQJ &LW\ 7DLZDQ

PAGE 145

, FHUWLI\ WKDW KDYH UHDG WKLV VWXG\ DQG WKDW LQ P\ RSLQLRQ LW FRQIRUPV WR DFFHSWDEOH VWDQGDUGV RI VFKRODUO\ SUHVHQWDWLRQ DQG LV IXOO\ DGHTXDWH LQ VFRSH DQG TXDOLW\ DV D GLVVHUWDWLRQ IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ & fKQ & 6XWKHUODQG &KDLU URIHVVRU RI -RXUQDOLVP DQG &RPPXQLFDWLRQV FHUWLI\ WKDW KDYH UHDG WKLV VWXG\ DQG WKDW LQ P\ RSLQLRQ LW FRQIRUPV WR DFFHSWDEOH VWDQGDUGV RI VFKRODUO\ SUHVHQWDWLRQ DQG LV IXOO\ DGHTXDWH LQ VFRSH DQG TXDOLW\ DV D GLVVHUWDWLRQ IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ 0DULO\Q 6 n5REHUWV $VVRFLDWH 3URIHVVRU RI -RXUQDOLVP DQG &RPPXQLFDWLRQV FHUWLI\ WKDW KDYH UHDG WKLV VWXG\ DQG WKDW LQ P\ RSLQLRQ LW FRQIRUPV WR DFFHSWDEOH VWDQGDUGV RI VFKRODUO\ SUHVHQWDWLRQ DQG LV IXOO\ DGHTXDWH LQ VFRSH DQG TXDOLW\ DV D GLVVHUWDWLRQ IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ *DLO %DNHU $VVRFLDWH 3URIHVVRU RI -RXUQDOLVP DQG &RPPXQLFDWLRQV FHUWLI\ WKDW KDYH UHDG WKLV VWXG\ DQG WKDW LQ P\ RSLQLRQ LW FRQIRUPV WR DFFHSWDEOH VWDQGDUGV RI VFKRODUO\ SUHVHQWDWLRQ DQG LV IXOO\ DGHTXDWH LQ VFRSH DQG TXDOLW\ DV D GLVVHUWDWLRQ IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ 0LFKDHO 0DUWLQH] $VVRFLDWH 3URIHVVRU RI 3ROLWLFDO 6FLHQFH

PAGE 146

7KLV GLVVHUWDWLRQ ZDV VXEPLWWHG WR WKH *UDGXDWH )DFXOW\ RI WKH &ROOHJH RI -RXUQDOLVP DQG &RPPXQLFDWLRQV DQG WR WKH *UDGXDWH 6FKRRO DQG ZDV DFFHSWHG DV SDUWLDO IXOILOOPHQW RI WKH UHTXLUHPHQWV IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ $XJXVW 'HDQ &ROOHJH RI -RXUQDOLVP DQG &RPPXQLFDWLRQV 'HDQ *UDGXDWH 6FKRRO