<%BANNER%>

Agenda Setting and the BCS

Permanent Link: http://ufdc.ufl.edu/UFE0024652/00001

Material Information

Title: Agenda Setting and the BCS Agenda Setting Effects on Desired College Football Ranking in the Bowl Championship Series
Physical Description: 1 online resource (103 p.)
Language: english
Creator: Lawhorne, Todd
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: agenda, analysis, bcs, college, content, football, public, relations, setting
Journalism and Communications -- Dissertations, Academic -- UF
Genre: Mass Communication thesis, M.A.M.C.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Mass Communication AGENDA-SETTING AND THE BCS: AGENDA-SETTING EFFECTS ON DESIRED COLLEGE FOOTBALL RANKING IN THE BOWL CHAMPIONSHIP SERIES POLL By Todd Lawhorne August 2009 Chair: Michael Mitrook Major: Mass Communication The tradition rich institution of college football remains the only major U.S. sport which does not determine its annual champion through the process of tournament play. To determine a National Champion, the Bowl Championship Series (BCS) was created in 1998 to ensure a year end match up between the No. 1 and No. 2 ranked teams in the nation through a complex ranking system made up of one-third computer tabulations and two-third human voter. But the BCS has been riddled with controversy since its inception and arguments as to which two teams should play in the National Title game is an annual dispute. Due to this controversy, college football coaches have turned to the practice of campaigning through the media as an attempt to sway BCS poll voters into ranking their team higher in the Harris Interactive poll and USA Today/Coaches poll, the two polls which make up the human voting percentage of the BCS ranking formula. The following study attempts to determine whether this campaigning behavior displayed by college football coaches is effective in obtaining a higher ranking in the BCS through the generation of increased and positive media coverage justified by the agenda-setting theory. Agenda-setting theory was applied to this study based on its premise which states that the salience of a topic can be transferred from the media to the public based on amount of media coverage the topic receives. It could be theorized that as these campaigns are covered in the media, the salience of the topics addressed in the campaign are transferred from the media to the voting publics of the BCS. In order to measure amount and valence of media coverage, as well as mentions of campaigning behavior, a content analysis was employed which examined news articles from The New York Times (NYT), USA Today, The Sporting News (TSN), and the Associated Press (AP). This coded news content data was compared to the final BCS poll, final Harris Interactive poll, and final USA Today/Coaches poll to determine whether increased and favorable media coverage was correlated with a higher BCS ranking. The study found a strong correlation in both the relationships between increased and favorable media coverage and higher BCS ranking, as well as increased campaigning behavior coverage and higher BCS ranking. Both college football coaches and college football sports information directors may find the results of this study useful in future attempts at using campaigning as a means of generating more and favorable media coverage for their teams. This increased media coverage could help in persuading human voters, through topic salience transfer effects of agenda-setting theory, in the hopes of obtaining a higher BCS ranking, and ultimately, inclusion in the BCS National Title game.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Todd Lawhorne.
Thesis: Thesis (M.A.M.C.)--University of Florida, 2009.
Local: Adviser: Mitrook, Michael A.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-02-28

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024652:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024652/00001

Material Information

Title: Agenda Setting and the BCS Agenda Setting Effects on Desired College Football Ranking in the Bowl Championship Series
Physical Description: 1 online resource (103 p.)
Language: english
Creator: Lawhorne, Todd
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: agenda, analysis, bcs, college, content, football, public, relations, setting
Journalism and Communications -- Dissertations, Academic -- UF
Genre: Mass Communication thesis, M.A.M.C.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Mass Communication AGENDA-SETTING AND THE BCS: AGENDA-SETTING EFFECTS ON DESIRED COLLEGE FOOTBALL RANKING IN THE BOWL CHAMPIONSHIP SERIES POLL By Todd Lawhorne August 2009 Chair: Michael Mitrook Major: Mass Communication The tradition rich institution of college football remains the only major U.S. sport which does not determine its annual champion through the process of tournament play. To determine a National Champion, the Bowl Championship Series (BCS) was created in 1998 to ensure a year end match up between the No. 1 and No. 2 ranked teams in the nation through a complex ranking system made up of one-third computer tabulations and two-third human voter. But the BCS has been riddled with controversy since its inception and arguments as to which two teams should play in the National Title game is an annual dispute. Due to this controversy, college football coaches have turned to the practice of campaigning through the media as an attempt to sway BCS poll voters into ranking their team higher in the Harris Interactive poll and USA Today/Coaches poll, the two polls which make up the human voting percentage of the BCS ranking formula. The following study attempts to determine whether this campaigning behavior displayed by college football coaches is effective in obtaining a higher ranking in the BCS through the generation of increased and positive media coverage justified by the agenda-setting theory. Agenda-setting theory was applied to this study based on its premise which states that the salience of a topic can be transferred from the media to the public based on amount of media coverage the topic receives. It could be theorized that as these campaigns are covered in the media, the salience of the topics addressed in the campaign are transferred from the media to the voting publics of the BCS. In order to measure amount and valence of media coverage, as well as mentions of campaigning behavior, a content analysis was employed which examined news articles from The New York Times (NYT), USA Today, The Sporting News (TSN), and the Associated Press (AP). This coded news content data was compared to the final BCS poll, final Harris Interactive poll, and final USA Today/Coaches poll to determine whether increased and favorable media coverage was correlated with a higher BCS ranking. The study found a strong correlation in both the relationships between increased and favorable media coverage and higher BCS ranking, as well as increased campaigning behavior coverage and higher BCS ranking. Both college football coaches and college football sports information directors may find the results of this study useful in future attempts at using campaigning as a means of generating more and favorable media coverage for their teams. This increased media coverage could help in persuading human voters, through topic salience transfer effects of agenda-setting theory, in the hopes of obtaining a higher BCS ranking, and ultimately, inclusion in the BCS National Title game.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Todd Lawhorne.
Thesis: Thesis (M.A.M.C.)--University of Florida, 2009.
Local: Adviser: Mitrook, Michael A.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-02-28

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024652:00001


This item has the following downloads:


Full Text

PAGE 1

AGENDA-SETTING AND THE BCS: AGENDA-SETTING EFFE CTS ON DESIRED COLLEGE F OOTBALL RANKING IN THE BOWL CHAMPIONSHIP SERIES POLL By TODD LAWHORNE A THESIS PRESENTE D TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULLFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS IN MASS COMMUNICATIONS UNIVERSITY OF FLORIDA 2009 1

PAGE 2

2009 Todd Lawhorne 2

PAGE 3

To m y family and friends 3

PAGE 4

ACKNOWL EDGMENTS First, I would like extend my sincere gr atitude to the members of my supervisory committee Dr. Julie Dodd, Dr. Spiro Kiousis and Dr. Michael Mitrook for their hard work and understanding during this difficult process. I would like to extend a special thanks to my committee chair, Dr. Michael Mitrook, for without his patient and exper tise I would not have been able to complete this study. I would further like to thank Mr. Steve McCl ain of the University of Florida Athletic Association for the contribution to this study with his priceless insight into the practices of a college football sports information director. I would also like to thank my classmates a nd friends who helped me, inside the classroom and out, cope with the everyday st resses of graduate student life. I would like to give a much deserved thank you to Amanda Ehrlich, who wore many hats during this process, but none more important than loving girlfriend. She gave of herself mentally, physically, and emotionally in enumerable ways, and for that I will be forever grateful. You are my hero and I love you. I would also like to make a special mention of my brot her Christopher Lawhorne whose friendship and encouragement has meant the world to me. I hope that he is as proud of my accomplishments, as I am of his. I love you, bro. Most importantly, I would like to thank my parents, David and Mary Anne Lawhorne, for their unquestioning and undying support. They have taught me invaluable life lessons and built for me the solid foundation from which I have accomp lished the greatest feats in my life. I love you and thank you both, from the bottom of my heart. 4

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 CHAPTER 1 INTRODUCTION..................................................................................................................10 Rational/Significance of Study...............................................................................................10 Purpose...................................................................................................................................13 2 LITERATURE REVIEW......................................................................................................15 Agenda-Setting Theory.......................................................................................................... .15 First-Level Agenda-Setting.............................................................................................16 Second-Level Agenda-Setting.........................................................................................17 Agenda Building..............................................................................................................19 Agenda-Setting in Political Campaigns...........................................................................21 Agenda-Setting in Sports.................................................................................................25 Bowl Championship Series.....................................................................................................30 What is the BCS?.............................................................................................................30 History of the BCS..........................................................................................................31 Previous BCS Campaigns................................................................................................34 Hypothesis..............................................................................................................................39 3 METHODOLOGY.................................................................................................................4 2 Content Analysis............................................................................................................... ......42 Background......................................................................................................................42 Purpose............................................................................................................................43 Media Coverage/Sample.................................................................................................44 Coding.............................................................................................................................47 Reliability........................................................................................................................50 Public Agenda/Voter Opinion.........................................................................................51 Data Analysis Strategy....................................................................................................52 4 RESULTS...................................................................................................................... .........53 Total Media Coverage and BCS Outcome.............................................................................53 Media Coverage and Final Harris Poll Outcome....................................................................55 Media Coverage and Final USA Today/Coaches Poll Outcome............................................58 Campaigning Coverage and Final BCS Poll Outcome...........................................................60 Positive Campaigning Coverage and Final BCS Poll Outcome.............................................61 5 DISCUSSION................................................................................................................... ......63 5

PAGE 6

Hypothesis One ................................................................................................................. ......63 Hypothesis Two................................................................................................................. .....65 Hypothesis Three............................................................................................................... .....66 Hypothesis Four................................................................................................................ ......68 Theoretical Implications....................................................................................................... ..71 Practical Implications......................................................................................................... ....72 Limitations.................................................................................................................... ..........75 Future Research......................................................................................................................77 Conclusion..............................................................................................................................78 APPENDIX A BCS CAMPAIGNING STUDY CODE BOOK.....................................................................80 B BCS CAMPAINGING STUDY CODE SHEET....................................................................82 C ALPHABETICAL TEAM NUMBERING.............................................................................83 D TOP TEN TEAMS IN THE FINAL BCS POLL IN THE YEARS 2005 TO 2008...............84 E TOP TEN TEMAS IN THE FINAL HARRI S INTERACTIVE POLL IN THE YEARS 2005 TO 2008.........................................................................................................................85 F TOP TEN TEAMS IN THE FINAL USA TODAY/COACHES POLL IN THE YEARS 2005 TO 2008.........................................................................................................................86 G TOTAL MEDIA MENTIONS OVER ALL FOUR YEARS.................................................87 H TOTAL TEAM MEDIA COVERAGE..................................................................................88 I TOTAL CAMPAIGNING OVER ALL FOUR YEARS.......................................................90 J TEAM CAMPAIGNING YEAR BY YEAR.........................................................................91 K SCATTERPLOTS FOR PPMC..............................................................................................92 REFERENCES..............................................................................................................................96 BIOGRAPHICAL SKETCH.......................................................................................................103 6

PAGE 7

LIST OF TABLES Table page 4-1 Final BCS and total media coverage PPMC......................................................................53 4-2 Final BCS and positiv e media valence PPMC...................................................................54 4-3 Final BCS and negativ e media coverage PPMC................................................................55 4-4 Final Harris poll and total media coverage PPMC............................................................56 4-5 Final Harris poll and positive media coverage PPMC.......................................................57 4-6 Final Harris poll and ne gative media coverage PPMC......................................................57 4-7 Final USA Today/Coaches poll and total media coverage PPMC....................................58 4-8 Final USA/Today Coaches poll and positive media coverage PPMC...............................59 4-9 Final USA Today/Coaches poll and negative media coverage PPMC..............................60 4-10 Final BCS and total campaign coverage PPMC................................................................61 4-11 Final BCS and positive campaign coverage PPMC...........................................................62 7

PAGE 8

Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Mass Communication AGENDA-SETTING AND THE BCS: AGENDA-SETTING EFFE CTS ON DESIRED COLLEGE F OOTBALL RANKING IN THE BOWL CHAMPIONSHIP SERIES POLL By Todd Lawhorne August 2009 Chair: Michael Mitrook Major: Mass Communication The tradition rich institution of college foot ball remains the only major U.S. sport which does not determine its annual champion through the process of tournament play. To determine a National Champion, the Bowl Championship Series (BCS) was created in 1998 to ensure a year end match up between the No. 1 and No. 2 ranked teams in the nation through a complex ranking system made up of one-third computer tabulations and two-third human voter. But the BCS has been riddled with controversy since its incep tion and arguments as to which two teams should play in the National Title game is an annual dis pute. Due to this controversy, college football coaches have turned to the practice of campaigning through the media as an attempt to sway BCS poll voters into ranking their team higher in the Harris Interactive poll and USA Today/Coaches poll, the two polls which make up the human voting percentage of the BCS ranking formula. The following study attempts to determine wh ether this campaigning behavior displayed by college football coaches is effective in obtaining a higher ranking in the BCS through the generation of increased and positive media covera ge justified by the agenda-setting theory. Agenda-setting theory was applied to this study based on its premise which states that the 8

PAGE 9

salience of a topic can be transf erred from the media to the pu blic based on amount of media coverage the topic receives. It could be theorized that as these campaigns are covered in the media, the salience of the topics addressed in the campaign are transferred from the media to the voting publics of the BCS. In order to measure amount and valence of media coverage, as well as mentions of campaigning behavior, a content analysis was employed which examined news articles from The New York Times (NYT) USA Today, The Sporting News (TSN) and the Associated Press (AP). This coded news content data was compared to the final BCS poll, final Harris Interactive poll, and final USA Today/Coaches poll to determ ine whether increased and favorable media coverage was correlated with a higher BCS ranking. The study found a strong correlation in both the relationships between increased and favorable media coverage and higher BCS ranki ng, as well as increased campaigning behavior coverage and higher BCS ranking. Both college football coaches and college football sports information directors may find the results of th is study useful in futu re attempts at using campaigning as a means of generating more and favor able media coverage for their teams. This increased media coverage could help in persuadi ng human voters, through topic salience transfer effects of agenda-setting theory, in the hope s of obtaining a higher BCS ranking, and ultimately, inclusion in the BCS National Title game. 9

PAGE 10

10 CHAPTER 1 INTRODUCTION Rational/Significance of Study Division I college football is th e only major sports league, e ither amateur or professional, not to incorporate any form of a postseason playoff system in which to determine an annual national champion (Mandel, 2007). Instead, a syst em of polling, coined the Bowl Championship Series (BCS), is used to crea te a season ending match-up between a No. 1 and No. 2 ranked team in a National Title game. Since the BCS ranking is comprised of one-third computer rankings and two-thirds human voters (H arris Interactive Poll and the US A Today Coaches Poll), it is feasible to say that these huma n voters hold the potential be pers uaded by a number of factors in regards to how high to rank teams (bcsfootball.org, 2008). It is clear from a look at previous college football media coverage that BCS campaigning by college coaches does occur (e.g., Bla udshun, 2003, SI.com, 2005, Weiss, 2006). For example, in November of the 2008 college football season, the University of Oklahomas head football coach Bob Stoops and the University of Texass head coach Mack Brown were forced to engage in a debate as to whos team was the mo re deserving representati ve of the Big 12 South in their leagues championship game against th e University of Missouri of the Big 12 North. Both Texas and Oklahoma possessed identical conf erence records, but the tiebreaker, determined prior to the season, would come down to which team was ranked higher by the BCS (Maisel, 2008). Both Brown and Stoops argued, through the media to voters, why their team was more deserving of a higher BCS ranki ng. Stoops argued, if you're going to forgive a team for losing at hom e to an unranked team because theyre playing pretty well now. W ell were playing pretty well now too. If it's logical for so meone else, it's logical for us."( Yanda, 2008). In contrast Brown stated, we are a great foot ball team and we did beat Oklahoma on a neutral site. We lost

PAGE 11

on the last play of the gam e out at Tech (Texassports.com, 2008) It seemed the voters in the USA Today/Coaches and Harris Interactive polls took both arguments into consideration by evidence of an extremely close vote (six points difference in the Harris Poll, one point difference in the Coaches poll) (ESPN.com, 2008). In the end, Oklahoma was voted higher over Texas and went on to play in and win the Big 12 championshi p game against Missouri. In the aftermath of the debate both coaches expressed their op inion on campaigning. Stoops stated: it's unfortunate; no one likes to do it (Yanda, 2008). Brown added: [two years ago] I was criticized for saying I thought our team was good enough to be in a BCS game, and, my gosh, I was a politician and a whiner. Now, what the system's doing, it's making coaches talk about why their teams should be voted, and that's very unfair to the coaches, in my estimation. But it is mo re popular now than it was then, so I guess maybe I was a trendsetter (Yanda, 2008) In addition to coaches, college football sports information directors (SID) play a role in disseminating BCS campaign messages. In a persona l interview, University of Florida football sports information director Stev e McClain expressed his interest in a study which may prove or disprove the necessity of BCS campaigning. As opposed to Heisman Trophy campaigning for specific athletes, McClain argues that BCS campai gning is a relatively new practice which must still be explored. McClain expressed evidence of this when he stated, we have a PR plan in place for all of our high profile athletes which ma ximizes exposure for those players. But we do not have a plan in place for arguing our position in any polling situation. McClain said he believes because of a lack of experience with the subject of campaigning for higher poll ranking, that the collegiate sports communications fi eld may be divided on their opinion of its effectiveness. McClain believes that the collegia te football communicati ons field would largely value a study in determining BCS campaignings worth (effort, time, and money) (S. McClain, personal communication, November 23, 2008). 11

PAGE 12

In addition to winning a national title, there are other advantages garnered by a university with regard to inclusion in the BCS nationa l title gam e. Joseph Martinich in his 2002 Interfaces article noted that each teams participation in a BCS bowl game guarantees an approximate 10 million dollar payout for each university. Martinich de scribes in more detail what is at stake for a university whose football team may be included in the title game: selection to participate in a BCS bowl, especially the national championship game, is of great importance to US college s because of the immediate financial benefits and because of the increases in financial contributions and student applications that result from participation (p. 87) However, despite the evidence that campaigni ng in college football does exist, and may now be a necessary practice under the current BCS system, there is no empirical evidence that shows campaigning works. John Fortunatos work in agenda-setting (the tr ansfer of salience of issues from the media to the public) and sport, focuses mainly on two professional sports leagues (National Basketball Associati on [NBA] and the National Footba ll League [NFL]) and the agenda-setting effects of their multi-million dollar television contracts with media outlets. But, Fortunato does not offer any quantitative eviden ce of agenda-settings ef fects. Instead, he incorporates interviews of public relations officers to highlig ht best practices in message dissemination and media relations in the NBA ( 2000), and analysis of television ratings and schedules to make assumptions about NFL agenda-s etting practices (2008). Fortunatos work, though minimal, is a solid base for further invest igation into agenda-settings effects on sport media coverage and public salience of intended issues and attributes. Due to the deficiency of empirical studies of agenda-setting applicati on to sport, a content analysis which looks at specific media outlets fo cus on issues and attributes in a BCS college football campaign may break new ground in the area. Universitys sports information officers, 12

PAGE 13

as well as college football coaches, could use the findings of an em pirical study as evidence that campaigning for higher BCS ranking does or do es not accomplish th e desired result. Purpose To address the deficiency of empirical data on the effectiven ess of BCS campaign behavior, this study attempted to quantitatively inve stigate the argument that campaigning for higher ranking in the human polls of the final BCS standings is necessary. First, a literature review was conducted to analyze past campai gning behavior of Divi sion I college football coaches, university representative, or players along with classic communication and persuasion theory (agenda-setting), in order to identify best practices for engaging in campaign behavior for college football Bowl Championship Seri es (BCS) title game inclusion. McCombs and Shaws agenda-setting theory (1972) was applied to analyze the salience of campaign messages and subsequent effectiven ess of previous campaigns in this area. McCombs and Shaws research during the 1968 pr esidential campaign showed, voters tend to share the medias composite definition of what is important [and] str ongly suggests an agendasetting function of the mass media, (p. 185). Ag enda-setting theory ex plores the relationship between stories the media presents and how the public perceives these stories. In more specific terms, this review analyzed agenda-setting at its second-level thr ough numerous political campaigns where the salience of the candidate s attributes were studied (Lopez-Escobar, Llamas, & McCombs, 1998). John Fortanatos wo rk, which utilizes agenda-setting theory, was also examined to determine further application of the theory to sports. In addition, a history of the BCS was analy zed to determine college footballs current need for the organization, and its future in the sport of college football. Prior BCS campaigning by National Collegiate Athletic A ssociation (NCAA) football coaches were also analyzed to 13

PAGE 14

determ ine how extensively the practice of campa igning has been utilized in the past. The findings regarding the effectiveness of these campa igns can then be used to advise collegiate football coaches and university sports informati on directors who find themselves in contention for participation in a BCS game. Finally, a content analysis of messages disseminated through the media by college football coaches campaigning for higher BCS ra nking was conducted. The findings of this content analysis combined with analysis of the final BCS rankings could help support the argument that campaigning for higher ranking is effective. For instance, a high frequency of positive team messages in the media correlated with a high final BCS ranking may work in support of the argument. A high frequency of positive team messages in the media correlated with a low final BCS ranking may work against the argument. 14

PAGE 15

15 CHAPTER 2 LITERATURE REVIEW Agenda-Setting Theory We can see that the news of it comes to us now fast, now slowly; but that whatever we believe to be a true picture, we treat as if it were the environment itself. Walter Lippman, 1922 The World Outside and the Pictures In Our Heads The press may not be successful much of th e time in telling people what to think, but it is stunningly successful in telli ng its readers what to think about. -Bernard Cohen, 1963 The Press and Public Policy Background In 1972, Maxwell McCombs and Donald Shaw, fr om the University of North Carolina at Chapel Hill, used these statements on communication as a starting point to construct their groundbreaking Public Opinion Quarterly article The Agenda-S etting Function of Mass Media. Despite prior literature and studies in the fields of both mass media and communication (e.g., Lippman, 1922, Lasswell, 1948, Lang & Lang 1966, Cohen, 1963, Trenaman & McQuail, 1961), McCombs and Shaws (1972) study was the firs t to both coin the term agenda-setting and lay a solid foundation for similar continued res earch in the field of media effects (Dearing & Rogers, 1996). McCombs and Shaws study focused on the publics salience of key i ssues pertinent to the 1968 United States presidential election. Their seminal study hypothesized that, the mass media set the agenda for each political campaign, in fluencing the salience of attitudes toward the political issues (p. 177). The study combined an interview of undecided voters in the Chapel Hill area with a content analysis of nine local media outlets coverage of the presidential candidates (1972).

PAGE 16

Because, the media appear to have exerte d a considerable impact on voters judgments of what they considered the major issues of the campaign (1972, p.180), the findings of McCombs and Shaws study did not prove, but strongly suggest th e existence of an agendasetting function(p.184). The authors blame the lack of proof on minima l consistency between different forms (TV, newspaper, & magazines) a nd outlets (Central Broadcasting System [CBS], Time, Durham Sun) of media coupled with journalistic pol itical bias (p. 184). Yet, despite the lack of numerical scientific pr oof, their study did achieve the iden tification of the conditions for agenda-setting to occur. Over the last 30 year s, these findings have ope ned the door to countless studies focused on agenda-setti ng effects in political campa igning (e.g., Walters, Walters, & Grey, 1996, Lopez-Escobar, Llamas, & McCombs, 1998, Golan & Wanta, 2001) and other arenas such as, corporate reputa tions, organized religion, classroom learning, as well as colligate and professional sports (McCombs, 2005). The core theoretical idea behind agenda set ting is the transfer of salience between the media and the public (McCombs et al., 1997). Some have argued that a limitation of the original McCombs and Shaw study was its focus on medi a exposure alone (Fortunato, 2008a). Since that time, research has expanded extensively an d agenda-setting theory has evolved into five distinct stages: basic agenda-s etting effects, attribute agenda-setting, psychological effects, sources of media agenda, and consequences (McCombs, 2005). First-Level Agenda-Setting Basic agenda-setting effects include what is commonly referred to as first-level agendasetting, which focuses on the salience of objects (McCombs, 2005). For instance, in a political campaign, objects can consist of the main issu es debated, the candida tes parties, or the candidates themselves. Journalists choose to focus on certain objects over others, creating a perceived reality in the minds of the public (Turk & Franklin, 1987). This concept relates to the 16

PAGE 17

idea th at, activities and issues the media cover and include in their content make up the medias agenda of what the public should think about (p. 30). The more the public is exposed to a specific object, the more salient that object becomes on the publics agenda. Because of the abundance of potential news stories competing for attention, it is important to note that media space is limited. This limited amount of type space (newspapers, magazines) or air time (radio, te levision) creates the need for journalists to pick and choose which stories to cover, creating what is called the media agenda (McCombs, 2004). Protess and McCombs (1991) not e that there are a number contingent conditions which may either enhance or diminish the effects of agenda-setting (p. 98). One contingent condition is the duration in which the media covers a story. The authors note that objects are most salient in the first month of exposure, and some become more salient in the public agenda faster than others. Another condition is Geographic Proxim ity, which refers to research that suggests agenda-setting effects occur more frequently in national news stories rather than local. Yet another contingent condition is type of medium employed in the dispersal of news. The authors note that newspapers, as opposed to broadcast and digital mediums, seem to be the most effective in transfer of salience (p. 98). And finally another condition is audience attributes, where some people are more aware of the news and have a greater nee d for orientation (p. 99). The authors state other conditions such as type of methodology a nd when it is conducted may lead researchers to much different findi ngs on the same topic, while using the same conditions (Protess & McCombs, p. 98). Second-Level Agenda-Setting Second-level agenda-setting focuses on the ch aracteristics or attr ibutes of an object covered by the media. As stated earlier, age nda-setting is the relati onship between the topic which the media reports and how important that topic becomes to the audience. However, as 17

PAGE 18

Kiousis (2005) states, second-level agenda-setting theory has shifted the focus of research away from investigating what topics news media cover to how they cover them. This statement suggests that early agenda-set ting theories emphasized the object covered in the media, over that same objects properties, qualities, and characteristics (p. 4). In the traditional context of agenda-setting, the term transfe r of salience is just as important when speaking about the first-level of agenda-setting as it is when speaking about the second-level. Lopez-Escobar, Llamas, and McCombs (1998), revisite d an earlier study by McCombs and Evatt (1995) and noted: objects numerous attributes, those characteristics and prop erties that sketch out the picture of each object in our heads and in the news coverage. Just as objects vary in salience, so do the attributes of each object. When members of the public and journalists describe objects, public issues political leaders, or what ever, some attributes are emphasized, some mentioned only in pa ssing, and others no t at all (p. 337) Yet characteristics of an object need not be limited to a cl assic definition, for instance, when describing a political candidate as trustw orthy, qualified, or candid. Craft and Wanta (2004), in their study of 9/11 media coverage of terrorist attacks, saw an objects attributes as consequences. They argue that threats and risk s including future, possibly biological, attacks; war and attendant war protests abroad; airline safety and economic downturn (p. 461) may be used as characteristics or attributes of an obj ect or event when discussing second-level agendasetting. Expanding on second-level research, Kiousis, Bantimaroudis, and Ban (1999) argued that in political campaigns not all attributes hold the sa me value. Their research found that attributes such as corruptness and honesty were more eff ective in shifting public opinion of a candidate (p. 424). These attributes were said to be more salient, possibly because voters are more likely to identify with personality traits rather than polit ical stances, such as a to pic like education (1999). 18

PAGE 19

Choosing which attributes to emphasize and which ones to ignore is where the ideas of second-level agenda-setting and framing converg e (McCombs, 2005). Framing is the act of defining attributes of particular objects wh ich will act as the frame through which the public perceives the said object (McCombs, Lop ez-Escobar, & Llamas, 2000). In broader communication terms, Entmans (1993) wide ly noted definition of framing states: to frame is to select some aspects of a perc eived reality and make them more salient in a communicating text, in such a way as to pr omote a particular problem, definition, casual interpretation, moral evaluation and /or trea tment recommendation of the item described (p. 52) Agenda Building McCombs (1992) describes a fourth phase of agenda-setting resear ch whose focus can be summed up with the question, who sets th e news agenda? (p. 816). This theoretical movement, which is now more commonly referred to as agenda building, moves past the original McCombs and Shaw (1972) idea that the news media se ts the publics agenda. Since the development of this branch of agenda-setting research, there have been a number of studies conducted in multiple fields analyzing the effects of agenda building (e.g., Berger 2001, Cho & Benoit, 2005, Curtin, 1999, Math es & Pfetsh, 1991). Agenda building is said to occur when public relations or any comm unications practitioner attempts to influence the media agenda through the placement of messages. If the message is successfully picked up by the media outlet then in turn the message has the potential to influence th e public agenda (Curtin 1999). Gandy (1982) uses the term sources to mean anyone supplying the media with information when describing communicators w ho hope to influence the medias agenda: it is the goal of all sources to influence decisions by changing the stock of information upon which these decisions are based. Since the public is generally only marginally involved in the determination of public polic y, and because the costs of molding public 19

PAGE 20

opinion are often quite high, sour ces have greater incentives to use the press to define public op inion than to infl uence it directly (p. 272) Kiousis, Mitrook, Wu, and Seltzer (2006) reiterat e previous evidence that public relations plays an important role in setting the media s agenda noting that public relations impacts anywhere from 25% to 80% of news content (p. 267). Berger (2001) states that public relations practitioners can influence the medias agenda through a number of tactics: story suggestions, news releases, spokespersons, pseudoevents, satel lite feeds, and technical reports (p. 95). For instance Ohl, Pincus, Rimmer, and Harrison (19 95) found in their study on the effects of press releases in a corporate takeover situation that news re leases were effective in influencing the media in the agenda-setting process as long as the release reflects the s ources point of view and is accompanied by follow-up contact information a nd the nature of that contact between the company and the reporter (p. 100). However, some researchers believe that the origin al idea of agenda-setting is flawed in its position that the transfer of information and attitude salience from the media to public is a oneway, linear model. Walters, Walters, and Gray (1996), building on Severin and Tankard (1979), point out that it may be within reason to believ e and test the theory that issue salience can be transferred from public to media (p. 10). In agenda building, as it applies to sports and sports media, Denham (1997) looked at how Sports Illustrated (SI), a highly respected sports week ly magazine, may have influenced public policy through feature stor ies pertaining to excessive st eroid use among athletes. Denham notes that Sports Illustrateds agenda was, to demonstrate to r eaders the dangers associated with using the drugs for performance gain(p. 265). Denha m notes that the issue became so salient on the government agenda that the Sports Illustrated articles were used as evidence in the House and Senate Subcommittees war on dr ugs hearings in 1990 (p. 260). 20

PAGE 21

Denham continued his sports media agenda building research w ith his 2004 study of Sports Illustrateds expos featuring Major League Base ball star Ken Caminitis admission of steroid use. Denham argues that the agenda building process happened when other news sources, mostly newspapers, ran stories based on the original Caminiti SI article, then transferred the salience of that issue to th e public agenda. Denham discusses th e high credibility of SI when he states: Sports Illustrated can be considered the sporting publica tion of record and, consequently it has the capacity to ignite a firestorm among newspaper reporters, as well television and radio news broadcasters, when a salient, or highly charged issue arises. (p. 54) Denham found that the eventual tr ansfer of salience to the public occurred when Major League Baseball reacted to an excessive number of newspaper articles generated by the SI story by implementing a mandatory steroid drug testing po licy beginning the following season (p. 53). Agenda-Setting in Political Campaigns McCombs and Shaws (1972) seminal st udy on the 1968 United States Presidential election began a long tradition into the political research of agenda-setting theory (e.g., Cho & Benoit, 2005, Kiousis, Bantimaroudis, & Ban, 1999, Golan & Wanta, 2001, Iyengar & Kinder, 1987, Miller, Andsager, & Reichert, 1998). In a 1997 study on regional and municipal elections in the Spanish province of Navarra, McCombs, Ll amas, Lopez-Escobar, and Rey speculate that with the expansion of agenda-setting theory to the attribute level, the understanding of the dynamics of elections may change forever. They forecasted mass media assuming a more powerful role in the transfer of salience of political issues th rough the accentuation of candidate attributes. In addition, not onl y will the voting public be info rmed on issues the media finds relevant to the campaign, but these attributes will help the voter decide how to feel about those issues (1997, p. 706). 21

PAGE 22

In their experimental study, Kiousis, Bantimarroudis, and Ban (1999) examined both issue and attribution agenda-set ting by asking subjects to read ne ws articles about factitious candidates competing in a congressi onal campaign (p. 418). The arti cles featured attributes of each candidate (educated, uneducated, moral, corr upt), and subjects were then surveyed to measure if candidate attributes contained within the article had transferred and become salient in the subjects minds. Overall, the researcher s found that media attention toward certain attributes appears to shift publ ic opinion of political candidates (p. 424). These pro secondlevel agenda-setting findings were stronger for some attributes (c orrupt or moral) than others (educated or not educated) (1999). McCombs, Lopez-Escobar and Llamas (2000) examined the 1996 Spanish Presidential elections this time using both levels of agenda-setting noting that, objects defining media and public agendas also can be the political candidates competing in an election, [or] rival institutions vying for public attention (p. 78). McCombs et al examined candidates images under five categories: issues positions and political ideology, biographical information, perceived qualifications, personality, and integrity (p. 82). These five attributes were applied to all three presidential candidates, and then through a content analysis, media coverage was coded either positive (not corrupt) or negative (co rrupt) (p. 82). The researchers found a strong correlation between the transfer of salience between positive medi a coverage of a candidates attributes and the final results of the presidential campaign furt hering support on the effects of second-level agenda-setting (2000). Framing, mentioned as one of the four di mensions of second-level agenda-setting, was used as the basis for Golan and Wantas (2001) content analysis of three newspapers agenda22

PAGE 23

setting effects on public percep tions of presidential candidates in the 2000 New Ha mpshire primary elections. To the importan ce of framing Golan and Wanta noted: in the context of political campaigns, the mass medias coverage of candidates or campaigns can sometimes shape voter perceptio ns concerning the objects attributes. Thus, the implications of public affairs framing by the mass media can have a strong influence on voter perceptions (p. 249) In their content analysis, the researchers analyzed paragraphs from newspaper articles and coded them for issues (taxes, campaign reform, forei gn policy, education, etc.) and attributes (trust, leadership, patriotism, winner, etc.) focusi ng on only the two most prominent Republican candidates (McCain and Bush). They coupled this content analysis with the results of a Gallup Poll which gauged a positive or negative attitude toward candidate issues and attributes. The study found that the transfer of salience from newspaper to pub lic was prominent in factual information content but was not as effective in the transfer of attitudinal content such as positive or negative feelings towards a candida tes issues or at tributes (2001, p. 255). King (1997) focused on candidate images (as oppos ed to issues) as they relate to agendasetting. King argued that most researchers agree that candidate images, rather than issues, are more effective in swaying voters perceptions of a candidate. After showcasing many definitions of image King settled on the Nimmo and Savage (1976) definition of candidate image as a human construct imposed on an array of perceived at tributes projected by an object (p. 8). To study candidate image salience tr ansfer in the 1994 Taipei, Taiwan, mayoral election, King conducted a content analysis of articles from th ree different newspapers which he scanned for personal attributes of candidates, party affiliation, and policy stances. King combined the content analysis with a telephone survey c onducted by an outside source to evaluate the candidates image in the publics mind. Kings findings showed that t he press significantly contributed to the construction of candidate images in the he ads of the voters (p. 40). 23

PAGE 24

In the Kiousis and McCombs (2004) content an alysis study on political figures in the 1996 United States Presidential election, the au thors focused on the candidates and general political figures them selves as opposed to the political issu es of the campaign.1 By testing attitudinal dispersion and polarization, the authors hoped to disseminate some specific attitudinal consequences that can be ascribed to the agenda-setting process (social learning) (p. 39). Kiousis and McCombs (2004) conducted a c ontent analysis of 11 political figures by examining seven major media outlets to determine the congruence of news coverage, public salience, and public attitude stre ngth (p. 44). They combined th eir content analysis with an independent poll which analyzed pu blic salience, attitude dispersion, and attitude polarization of the political figures (p. 45). In the end, their resear ch found, strong correlat ion between the amount of attention that news media pay to political figures and both the public salience and the strength of public attitudes to ward these persons (p. 49). Other researchers (e.g., Ghorpade, 1986, Robe rts, 1997) have expanded agenda-setting effects research from news media into the realm of political advert ising. Building on the research of others (e.g., Atkins & Heald, 1976), Roberts (1997) found that much like the news medias messages transferring salience to the pub lic, political advertising was a factor in the voters decision making process when choosing a candidate (p. 86) However, Roberts also argues that political advertising and news media are unique in their issue salience transference methods. Golan, Kiousis, and Mc Daniel (2007) found that nega tive attributes of opposing candidates were included in negative political advertising as a direct attempt at influencing voters perceptions of said candidates (p. 438). In addition, Ghorpade (1986) combined the ideas of agenda-settings issue salience and advertisings behavior change models of top-of-mind 1 The researchers not only looked at th e transfer of salience of the political fi gures from media to public, but also examined the publics perceived attitudes toward the same figures. 24

PAGE 25

recall and f irst-brand awareness (p. 24). Ghorpade found that advertising can focus consumer attention on what attributes of th e product to think about, and that this transfer of salience can lead to intended behavioral outcome (p. 26). Agenda-Setting in Sports McCombs (2008) related his theory to th e sports industry when he discussed its relationship to the emergence of National Bask etball Association teams and the amount of television coverage they rece ived. In this study, McCombs found, agenda-setting links this broad media agenda to first and second-level ag enda-setting effects among the public and with subsequent attitudes and behavior especially viewing sports on television, becoming a fan and attending sports events, (p. 553). McCombs identifies, sports news and broadcasts of actual sporting events to be the media agenda in athletics (2005). McCombs (2005) credits John Fortunato fr om the Fordham University School of Business for engaging in groundbreaking work incor porating agenda-setting into the worlds of both colligate and professional sports. Fortunato (2001) used agenda-set ting theory to explain the symbiotic relationship between television br oadcasting and the NBA. Fortuanto combined past research in both first and second-level agen da-setting with uses and gratification theory, which examined how people satisfy their ne eds through mass communication and mass media dependency. Mass media dependency is when one s needs and wants are dependent on another entity satisfying those needs and wants. Combin ing these three views, Fortunato developed an integration model to explain the te levision broadcastNBA relationship: (a) the media can be successful in telling the audience which issues to think about (agendasetting level-one), (b) the media can be succe ssful in telling the audience how to think about that issue (agenda-setting level-two) (c) organizations ar e dependent on the mass media for communication links (organizationa l media dependency), (d) individuals are dependent on the mass media for informati on and social connection (individual media dependency), and (e) audiences actively select and interpret media to satisfy their own needs based on their own experien ce, attitude, and values (use s and gratifications) (p. 47) 25

PAGE 26

Just as ag enda-setting theory is used to describe the relationshi p between a political candidates message and the salience of that message in the publics mind, Fortuanto argues that sports organizations such as the NBA, also can possess agendas which bene fit from the help of the mass media to gain exposure among th eir potential audiences (2001, p. 37). Because most sports organizations or associations have multi-million dollar contracts with major media outlets such as Nati onal Broadcasting Company (NBC), American Broadcasting Company (ABC), or CBS (Fortunato, 2000), it may be in the best interest of both entities to promote one another. Because the medi a outlets are the gatekeep ers and the setters of the agenda, when a contract is signed, it coul d be inferred that the exposure of a sports organization should increase based on the increased coverage of the sports organization through the media outlet. Therefore, it can be said th at having a strong relations hip with a large media outlet may be extremely advantageous to a sports organization that wishes to increase its salience with an audience (Fortunato, 2001). Fortunato (2008a) applied McCombs and Shaws (1972) original theory of agenda-setting (increased public salience of object through fre quent exposure to a message) in his study of the football games which the NFL chose to televise He found that the NF L worked with their television sponsors to determine which teams were most likely to draw viewers and what were the most advantageous times that teams shoul d play to maximize exposure among their intended audiences. Fortunato also found that second-leve l or attribute agenda-setting is a factor in determining which games should be televised. He argues that players and coaches are attributes of the object (professional f ootball) which the media can focus on. Fortunato found that: this strategy of developing both programmi ng schedule components, placement of games (agenda-setting through exposure) and having the best teams and players participating in those games (agenda-setting thr ough selection of attributes ) gives the league and the networks the best opportunity to transfer th e salience of the NFL to the public and produce 26

PAGE 27

agenda-s etting effects. In this example of agenda-setting being app lied to the NFL, the proper exposure in the programming schedule and showing better teams in these games not only transferred the topic salience, but increased the behavior towa rd that topic, and increase in watching NFL ga mes on television (2008, p. 45) However, as Fortuanto points out, the media does not hold all the power when it comes to setting the agenda. Fortunato ( 2008) argues public relations practit ioners must operate from an assumption that they have the power to influe nce mass media content, acting as an advocate on behalf of the organization they represent through the public relation strategies they implement (p. 482). With this statement Fortunato delves into the agenda-setting to pics of who sets the media agenda?, second-level age nda setting, and framing. To s ee how this statement applies to sports, Fortunato examined the public relations strategies which the NBA employs to promote the league, its teams, and players and to frame the coverage the league receives in the most favorable fashion (p. 482). Through personal interviews with media relati ons practitioners working for the NBA and its teams, Fortunato found that by making the le ague and its teams easier to cover from a media/journalistic perspective, more coverage was generated. NBA public relations practitioners made statistics, box scores, and game notes fr om match-ups around the country instantaneously accessible through fax to any member of the medi a. In addition to information dispersal, practitioners proactively produced NBA, related stories in the form of video and radio news releases, which allowed media outlets to save time and money on self-produced story spots (2000, p. 485). Another strategy used by the NBA to cultiv ate its relationship with the mass media was media access to the leagues players and coaches. Both players and coaches were made available to the media before and after all games for pers onal interviews. The NBA demonstrated their understanding of the important ro le that agenda-setting and fram ing played in their everyday 27

PAGE 28

operations by im plementing a mandatory player media relations class that all rookie players in the league must complete (Fortunato, 2000). Player and coach media relations were also revisited and emphasized during th e playoffs. The NBA was fully aware that their television viewer-ship increased during the pl ayoffs, and while there were more viewers, the opportunity to grow and frame their organization wa s at its height (Fortunato, 2000). From a framing perspective, the NBA not only allowed easy access to information and stories, but the league made a consorted and consci ous effort to illustrate, tell, or present stories and information in a pro-NBA li ght (Fortunato, 2000). What ma kes the NBA a credible source to the media is the NBAs close da ily interactions with its player s and personnel. Because there is such a close working atmosphere within the league, the understanding between the two parties is that the NBA would be privy to more accurate inside information than would an outside source (2000, p. 486). When dealing with collegiate sports, it has b een said that the athletic activities of the university link that school to the outside world (Shulma n & Bowen, 2001, Fortunato, 2008b). The media plays an undeniable role in the transf er of information about a universitys activities to the public. In Goffs (2000) study of a Mi dwest university, he f ound that % of all newspaper articles written about the univers ity pertained to athletics (2008b). Additionally, Mitrook an d Seltzers (2005) conference pa per, hopes to break ground in the area of agenda-setting as it applies to college football in a paper that is still in progress and is set to be published in a 2009 edition of the Journal of Sports and Media Their paper entitled, The Agenda Setting Influence of Expert Op inion on Media Coverage of the Heisman Trophy Race uses a quantitative content analysis to meas ure the relationship between amount and type of media coverage and the annual Heisman trophy race. Their findings suggest that players 28

PAGE 29

who win the Heism an benefit from an agenda set ting effect due to favorable media coverage and higher rankings (2005, p. 15). The communications theory of agenda-setting has previously been applied to the sports industry, but not extensively. Besides the Mitrook and Seltzer (2005) conf erence paper, there is somewhat of a deficiency of quantitative agenda-setting research in the area of collegiate athletics, especially in college football. College football is an ever growing, multimillion dollar industry which relies heavily on the media to gene rate and sustain its pr ofitability. The sport may benefit greatly from the findings of a study on the effect s of agenda-setting on Bowl Championship Series (BCS) voting, seeing as the voters a large determinant of team participation. The BCS is the pinnacle of econom ic and social status for all 119 NCAA Division I football teams, with inclusion and partic ipation being the seas ons ultimate goal. Sports Information Director Practices It has been widely publized that universit y athletic departments and college football sports information directors ac tively engage in raising awaren ess and campaign for specific athletes who are in contention for college footballs most prestigious individual award, the Heisman trophy (Cohen, 1985). The Heisman is awarded annually based on a vote of the Heisman trust which is made up of a group of past Heisman winners and college football writers (The Heisman, 2003). Tactics employed by college football sports information directors for increased player exposure in the eyes of the voting public have ranged from the subtle to the extreme. A subtle example would include the maili ng of statistical data a nd highlight clips of a Heisman hopeful to voting members of the Heisma n trust. An extreme example would be when, the University of Oregon erected a ten story hi gh billboard of senior quarterback and Heisman hopeful Joey Harrington in Times Square, New York (Wahl, 2003). 29

PAGE 30

These examples are documented and admitted tactics used by sports information directors to increase exposure for their team or player in the Heisman race. Yet, tactics employed by college SIDs for acquiring higher ranking in the BCS are not as widely understood or admitted to. Other than the evidence of past campaigni ng employed by coaches through the media, it is not clear how SIDs and coaches work to craft th ese campaigning messages or even if they work on the development at all. Bowl Championship Series What is the BCS? The Bowl Championship Series purpose is to match the nations No. 1 and No. 2 teams in a de-facto national championship game as decided through a multi-factor mathematical formula. The formula consists of three e qually weighted components. These are the USA Today Coaches Poll the Harris Interactive College Football Poll and an average of six computer rankings (Anderson & Hester, Rich ard Billingsley, Colley Matrix, Kenneth Massey, Jeff Sagarin, and Peter Wolfe). In addition to the BCS national title game, four other BCS bowl games are played. These are the Tostitos Fiesta Bowl, th e FedEx Orange Bowl, the Rose Bowl, and the Allstate Sugar Bowl (Bowl Champi onships Series FAQ, 2003). The BCS is managed by the commissioners of the 11 NCAA Division I-A conferences, the director of athletics at the University of Notre Dame, and representatives of the bowl organizations. The six major conferences with automatic bids to BCS bowl games are the Atlantic Coast, the Big East, the Big Ten, the Big 12, the Pacific 10, and the Southeastern (The BCS is, 2003). Despite common perceptions, the BCS is not affiliated with the NCAA, and, in fact, the BCS is not an actual organization at all. The BCS is merely an agreement by the aforementioned participants who govern the pa rameters of the agreement: everything pertaining to the BCS and its national champi onship game, from payouts to entry rules to 30

PAGE 31

uniform colors, is determined by administrators from the nations majo r conferences(Mandel, 2007). History of the BCS The Bowl Championship Series was created to end years of controversy that surrounded the lack of a true Division I college football championship. The BCS aimed to match the No. 1 and No. 2 teams in the nation while ensuring inte grity and excitement to the important regular season and the historic bowl game tradition (Mandel, 2007). Prior to the 1992 NCAA Division I college f ootball season, no plan was in place which ensured a season ending match up of the two t op ranked teams in the nation. The Division I college football post-season consisted of conference affiliated bowl games which did not take into consideration a teams Associated Press or United Press Internat ional (UPI) ranking, only their final conference rankings. For instance, the winners of the Pacific 10 and Big 10 conferences would meet annually in the Rose Bowl in Pasadena California regardless of national rankings. Other bowl games also had single conferen ce affiliations such as The Sugar Bowl with the Southeastern Conference and the Orange Bowl with the Big Ea st Conference (Bowl Championship ,2008). Under these circumstan ces, controversies as to who was the true national champion arose. In some instances this system led to two or more teams remaining undefeated or having similar record s at the end of the post season, all claiming their right to be national champions. Split national titles after the year 1964 (the first year teams in a split national title both played in bowl games) in clude: 1965, Alabama (AP) 9-1-1, Michigan State (UPI) 10-0-1, 1970, Nebraska (AP) 11-0-1, Te xas (UPI) 10-1-0, 1973 Notre Dame 11-0-0, Alabama (UPI) 11-1-0, 1974, Oklahoma (AP) 11-00, Southern California (USC) (UPI) 10-1-0, 1978, Alabama (AP) 11-1-0, USC (UPI) 12-1-0, 1990, Colorado (AP) 11-1-1, Georgia Tech (UPI) 11-0-1, 1991, Miami (FL) (AP) 12-0-0, Wa shington (Coaches) 12-0-0, 1997, Michigan 31

PAGE 32

(AP) 12-0, Nebraska (C oaches) 13-0, 2003, Louisiana State (LSU) (BCS) 13-1, USC (AP) 12-1 (CollegeFootball.com, 2009). In 1992, the Bowl Coalition was formed by th e majority of commissioners from major conferences, Notre Dame (an independent), and four of the major bowl committees. Similar to the Rose Bowl, the Coalitions aim was to match the champions of the Big East and Atlantic Coast Conference (ACC) against SEC, Big Eight or Southwest conference champions under their respective bowl affiliated games. This coal ition succeeded in pairing up the two top ranked teams, two out of the three years that it was as sembled. The downfall of the coalition was the lack of participation by the Pac 10 and Big 10 c onferences in conjunction with their exclusive tie-in with the non-participating Rose Bowl. W ithout the complete participation of the major conferences there would be no tr ue match-up of No. 1 and No. 2. The next step in the evolution of the BCS came in 1995 when the Bowl Alliance was formed. The Alliance aimed to pair the APs No. 1 and No. 2 ranked Division I college football teams regardless of previous conference/bowl affiliations. Three major bowl venues (Sugar, Orange, Fiesta) where incorporated and would ro tate the hosting of the annual one-two match up. However, again the lack of par ticipation of both the PAC 10 and the Big 10 was a weakness to a true one-two match-up (Bowl Championship ., 2008). In 1997, the Pac 10 and the Big 10, in associatio n with the Rose Bowl, agreed to become members of the Bowl Alliance and the BCS was born. The Rose Bowl became one of four bowl venues which would host a season ending match-up of the top two ranked teams, regardless of conference affiliation. After the creation of the BCS, the system us ed to determine the top two teams underwent a number of modifications. At its inception, the BCSs ranking formula was made up of four 32

PAGE 33

com ponents: the AP Writers pol l, the ESPN/USA Today Coaches Poll, three computer ranking systems, and a strength of sche dule calculation. The first modi fication to this formula came in 1999 when five computer ranking systems were adde d to bring the total to eight. Then in 2001, the strength of schedule component was modified to include a quality win criteria. In 2002 one of the computer rankings was eliminated becau se it was decided that a margin of victory component of the system was not appropriate. That left seven computer ranking systems, where the lowest ranking would be dropped and the rema ining six would be averaged together. In 2004, the BCS formula had a substantial overhaul which included eliminating the strength of schedule component, and the weighting the computer rankings, AP Writers poll, and the USA Today Coaches poll equally at one third. In ad dition, one of the comput er ranking systems was dropped for a new total of six a nd the overall computer rank would be an average of the teams four highest computer results (Bowl Championship ., 2008). Controversy erupted in 2005 with the AP wr iters poll declining furt her participation in the BCS, citing that the BCS, damaged and cont inues to damage AP's reputation for honesty and integrity in its news accounts through the forced association of the A.P. poll with the B.C.S. rankings" (Thamel, 2004). As a result of th is development, the BCS created the Harris Interactive Poll to replace the AP. The Harris Poll is a panel of randomly selected former players, coaches and administrato rs, and former and current media (Harris Interactive ., 2008) Finally, in 2006, the last of the substantial modifications took effect with the creation of a fifth BCS bowl game, the BCS National Champions hip Game. This game would allow the four traditional BCS bowls (Fiesta, Orange, Rose, a nd Sugar) to operate simply as bowl games, matching highly ranked BCS teams. The BCS Na tional Championship Game would be played 33

PAGE 34

last, in addition to these four bowl gam es, at one of the four bowl venues and would match the No. 1 and No. 2 ranked teams (Bowl Championship ., 2008). Because the BCS is made up of two-thirds hum an voters, it is speculated that these voters can be influenced by a number of outside sources in their ranking decisions. The media is one such outside source of influence. Martinich (2002) hypothesized that: because of biases, conflicts of interest, a nd lack of knowledge (esp ecially by coaches who do not see many other teams play during the season), the USA Today /ESPN Coaches Poll and to a lesser degree the AP Writers poll, would be inferior to more objective computer rankings (p. 85) Previous BCS Campaigns Roy Kramer, former Southeastern Conferen ce (SEC) commissioner and co-creator of the BCS, said, there were three major objectives when the BCS started: to expand in terest nationally in the sport, preserve the bowl structure becau se of the postseason opportunities it provides and bring together the top two teams at the end of the year (Carey, 2007) Although the creation of the BCS has led to the accomplishment of thes e objectives since its most complete version in the 1998 season, the BCS ha s also created a number of controversies. The question of bringing together the top two team s has become a topic of debate in the court of public opinion as well as in the traditional polls that make up the BCS ranking system. In three different years, a debate emerged over which teams would be matching up as the No. 1 and No. 2 teams in the nation. The que stion was not over who was No.1 but over who were the No. 2 and No. 3 teams. Prior to th e 1998 bowl season, BCS teams were determined by a combination of human polls, including the traditional Associated Press and the college Coaches. As the system operates currently, mem bers of the Bowl Championship Series decided that, in addition to the traditional polls, they w ould use three computer rankings and a strengthof-schedule formula to make their selection (UMTERP.CSTV.com, 1998). Because of the BCS 34

PAGE 35

com puter rankings, the fight for the No. 2 spot in the championship game became extremely heated in 2004 bowl season. The conflict resided in the discrepancy between the Coaches and AP writers polls and the BCS computer rankings: the coaches and writers agreed that No. l was the University of Southern California (USC), which finished a strong 11-1 season with a 5228 romp past Oregon State. The final BCS standings-which determine who will play in the title game in the Sugar Bowl disagreed, placing Oklahoma, a 35-7 loser to Kansas State in the Big 12 championship game, against LSU, a 34-13 winner over Georgia in the SEC title game (Blaudschun, 2003) When the 11-1 USC team was not included in the BCS championship game in lieu of the University of Oklahoma, USC head coach Pe te Carroll conveyed messages through the media such as, were number one in the country and were going to do everything in our power to hold on to that spot. If we win that football game (2004 Rose Bowl against Michigan), we feel like were the number one team in the country regardle ss of what that other game (BCS title game) is called (Weiss, 2003). The statements of USC coach Pete Carroll had the potential to influence the AP writers vote, yet they did not have any effect on the outcome of the heavily computer weighted final BCS rankings. LSU went on to be named National Champion by the BCS, defeating Oklahoma in the title game. The BCS final rankings would put USC as number two, but the AP writers would go on to declare USC their champion. After the undesired outcome of a split national championship, the BCS decided to completely revamp the formula. It was obvious to most college football fans that the BCS formula relied too heavily on the computer as pect of the polling as opposed to the human elements of the AP and Coaches poll. For the 2005 bowl season, the BCS unveiled a revision to the formula that weighted the human polls tw ice as heavily as computer rankings, making it nearly impossible for the top team in the human polls to be denied a spot in the BCS title game. 35

PAGE 36

Regardless of the BCS r evisions, the result of the 2004 bowl season will remain a USC and LSU split of the national championship. In the 2005 bowl season, a similar controvers y occurred with a dissimilar outcome. The University of Auburn Tigers took issue with how teams were chosen to participate in the BCS National Championship game. After finishing the regular season with a perfect 11-0 record in the Southeastern Conference, the Tigers woul d go on to win the SEC championship game in Atlanta against SEC eastern opponent the Univ ersity of Tennessee on December 4, 2004 (SEC title likely ., 2004). Later that same day, the University of Oklahoma, after completing a perfect 11-0 regular season in Big 12 conference play, also won its conference title game against Colorado, 42-3. USC finished the regular season 12-0 in the Pacific Ten Conference and had been ranked first in the AP poll since the fi rst week of polling (Augus t 22nd) and had been leading the BCS standings since th e first edition of the poll. USC was the consensus No. 1 team in both the traditional polls a nd the BCS standings going into th e National Championship game. The other available slot in th e title game would have to be decided upon between Auburn and Oklahoma (ESPN.com, 2008). After winning the SEC championship game, Au burn head coach Tommy Tuberville was aware that his team could face the possibility of exclusion from the National Championship game. He made a plea to the AP writers and Coaches to vote his team No. 2 when he said, this is a true team from top to bottom. I just hope everyone is fair when they vote tonight. I know well get at least one first-plac e vote (in the coaches poll), (SEC title likely ., 2004). In the end Tubervilles plea to voters would not be en ough to leapfrog his team over Oklahoma to take on USC in the title game. However, Tuberville was not finished campa igning. There was an outside chance that if his team won by a gr eat enough margin in th eir bowl match-up with 36

PAGE 37

Virginia Tec h, and USC and Oklahoma battled to a close finish in the Orange Bowl, his team could be voted number one in the AP poll, forcing another split national championship. Tubervilles Auburn Tigers struggled against Virginia Tech but pulled out a close 16-13 victory in the January 4 Sugar Bowl, while USC went on to defeat Okla homa convincingly 55-19 in the Orange Bowl National Championship game (SI.com, 2005). The outcomes of the games did not stop Tuberville from camp aigning for his team, stating, were 13-0. We won the SEC. Somebody is going to pick us national champions, and thats all we want. We want to be recognized as a good football team. Neither (USC or Oklahoma) is better than us. Well play them anytime, anywhere (Gardiner, 2005). In the end, it seems Tubervilles campaigning did not work as USC was named BCS national champion as well as voted No. 1 in the final AP poll. (ESPN.com, 2008) After the controversies in 2004 and 2005, the AP poll decided to no longer affiliate themselves with the BCS. A statement released by the Associated Press football writers claimed, by stating that the AP poll is one of the thr ee components used by BCS to establish its rankings, BCS conveys the impression that AP condones or otherwise participates in the BCS system (Klemz, 2005). For the 2006 bowl season, the AP poll would be replaced in the BCS formula with the Harris Interactive College Football Poll, a panel of former coaches, athletes and media who rank the top 25 teams in college football eac h week. The Harris poll would account for one third of the weight in the fina l BCS standings, which was the same amount that the AP poll had accounted for. (The Bowl Championship ., 2005). The inclusion of the new poll would not prevent further cont roversies. In the 2007 bowl season, the University of Florida believed th ey deserved the No. 2 spot in the National Championship game after a strong 11-1 regular season in the SEC. Going into the SEC 37

PAGE 38

cham pionship game, the BCS and AP poll No. 4 ranke d Florida would need help in the form of No. 3 ranked USC losing to the University of California at Los Angeles (UCLA). The University of Florida went on to beat the Univer sity of Arkansas in the SEC championship game in Atlanta, 38-28 on December 2, 2006. Florida also received the help they needed from UCLA earlier that afternoon when th e Bruins went onto shock USC 13-9 in both teams regular season finales (Weiss, 2006). Ohio State University (OSU) had spent the entire season as both the BCS and APs No. 1 team and had cemented their spot at the top of college football with a regular season ending victory over then No. 2 University of Michiga n. After losing to OSU, Michigan found itself clinging to the No. 3 slot in the BCS, behi nd USC and ahead of Florida (ESPN.com, 2008). With USCs loss to UCLA, coupled with Fl oridas SEC championship game win against Arkansas, it looked as though the fi ght for the last spot in the national championship game would come down to Florida and Michigan in the final BCS poll (Romano, 2006). To the dismay of many Michigan fans, in cluding Wolverines h ead coach Lloyd Carr, Florida head coach Urban Meyer began to cam paign for his teams inclusion in the BCS National Championship game. Meyer stated, I think that'd be unfair to Oh io State (to rematch against Michigan), and I think it'd be unfair to the country. (I) Just don't believe that's the right thing to do. You're going to tell Ohio State they have to go beat the same team twice, which is extremely difficult? If that does happen, all the [university] presidents need to get together immediately and put together a playoff system. I mean like now, January or whenever, to get that done. I do believe we deserve a shot. I think the country will feel the same way. The SEC against the Big Ten will be a great title game Michigan has a great team, but I think they had their shot (Kilgore, 2006) Michigan Head Coach Lloyd Carr also took a dvantage of the opportunity to join the conversation about his teams inclusion in the BCS national title game saying, "I hope that the voters will not penalize our team because we didn't play the last two week s," Carr said. "I don't 38

PAGE 39

want to get into a cam paign. That's not what's be st for the game. The BCS is set in order to put the two best teams together in the championship games. We all have our views"(Kilgore, 2006). After both coaches had their say, the decision came down from the BCS on December 3, 2006, that Florida had leapfrogged Michigan to th e No. 2 spot, entitling them to a match up with OSU in the BCS National Title game (Prisbell, 2006).2 Coaches and Harris poll voters did not admit th at the campaigning of Florida coach Urban Meyer made a difference in their ballots, but it could be argued that they were listening. As Harris poll voter and former University of Iowa head coach Jim Walden stated, if you look at the Big Ten conference, it is a j oke. I voted my heart and I vote d my strength of what I believe in. In my opinion, Florida is the number one te am in the nation.'' Another Harris Poll voter, George Lapides, clearly reiterated Meyers rematch sentiment when he stated, ''I liked the idea of a conference champion playing a conference cham pion. I think that's more appealing than a rematch. Lapides said that if Florida and Mi chigan played each other he thought Michigan would win, yet still moved Florida past Mi chigan in his final poll (Thamel, 2006). Even though Michigan head coach Lloyd Ca rr would go on to call Meyers campaigning comments inappropriate, Florida would end up facing OSU in the national championship game on January 8, 2007 (Weiss, 2006). Florida would go on to win the BCS title by beating OSU 4114 in Tempe, Arizona, and securing the No. 1 spot in the AP poll with a Michigan loss to USC in the Rose Bowl on January 1, 2007 (Blaushun, 2007). Hypothesis Agenda-setting is the transfer of issue or issue attribute salience from the media to the public. It could be argued that college football coaches cam paign messages could transfer 2 Florida finished a slim 26 points ahead of Michigan in the Coaches poll and 38 points ahead in the Harris poll. 39

PAGE 40

salience of the issue of team rank from the media to human voters in both the USA Today Coaches poll and the Harris Interactive poll. The tr ansfer of issue salience of team ranking could be achieved through the accentuation of a teams attributes by the coach while engaging in campaigning behavior. This transfer of salienc e may persuade a human voter to rank a team higher if the media coverage of that team is favorable. Based on the aforementioned assumption deri ved from a review of agenda-setting literature as well as a history of the BCS and its past controversies the following hypotheses are proposed: H1: Teams exhibiting more media coverage will be ranked higher in the final overall BCS poll than teams e xhibiting less media coverage. H1a: Teams exhibiting more positive media coverage will be ranked higher in the final BCS poll than teams exhibiting less positive media coverage. Teams exhibiting more media coverage will be ranked higher in the final overall Harris Interactive poll than teams exhibiting less media coverage. H2: H2a: Teams exhibiting more positive media coverage will be ranked higher in the final Harris Interactive poll than teams exhibiting less positive media coverage. Teams exhibiting more media coverage will be ranked higher in the final overall USA Today/Coaches poll than teams exhibiting less media coverage. H3: H3a: Teams exhibiting more positive media coverage will be ranked higher in the final USA Today/Coaches poll than teams exhibiting less positive media coverage. H4: Teams engaging in campaigning beha vior will be ranked higher in the final BCS poll than teams not engaged in campaigning behavior. 40

PAGE 41

H4a: Teams engaging in positive campaigning behavior will be ranked higher in the final BCS poll than teams not engaged in positive campaigning behavior. For the purposes of this study, campaign behavior will be operationally defined as bolstering, promoting and framing with th e goal of achieving desired poll ranking. Additionally, favorable will be operationally de fined as frequently, justifiably, and positively mentioned. 41

PAGE 42

CHAP TER 3 METHODOLOGY Content Analysis Background Though other scholars have attempted to defi ne the practice of content analysis (e.g., Waples & Berelson, 1941, Leites & Pool, 1942, Kaplan & Goldsen, 1943, Janis, 1943), Berelsons 1952 book Content Analysis has been called the standard codification of the field (Pool, 1959). Berelson defines content analysis as a research technique for the objective, systematic, and quantitative description of the manifest content of communication (1952, p. 18). Berelsons definition characterizes four requir ements of a proper conten t analysis. The first requirement he called syntactic and semantic, whic h refers to the need for the content of the analysis to be manifest as opposed to latent. To reiterate, the words and the meanings of those words must actually occur in the text and cannot be the inferred intentions hidden by the author. The content must be manifest, argues Berelson, for the reasons of valid ity and reliability (1952, p. 16). The second requirement, objec tivity, is included for scientific justification and states that the categories of analysis should be defined so precisely that different analysts can apply them to the same body of content and secure the same results (p. 16). The third requirement is system, which is included to rule out research ers bias, by requiring the inclusion of all occurrences of the category (p. 17). Finally, Berelson includes the requirement which separates content analysis from reading. Quantification is, the extent to which the analytic categories appear in the conten t, that is, the relative empha ses and omissions (p. 17). In addition to Berelsons requirements, Ho lsti (1969) builds off previous work (e.g., Cartwright, 1953, Barcus, 1959) when he describes his three requirements of a proper content analysis as objectivity, systematic, and generality. Holstis first requirement, objectivity, puts a 42

PAGE 43

great em phasis on the analysts pred etermined rules. Holsti states that if these predetermined rules or guidelines cannot be followed by research ers reexamining the content, then the rules may be biased. Holstis second requirement, syst ematic, states that all content must be analyzed whether that content supports th e analysts hypothesis or not. Fi nally, Holstis third requirement, generality, says that it is not enough to look at only one datum, for example how many times a word appears in text. Holsti explains that compar ison is an important fact or in content analysis when he states, a datum about communication cont ent is meaningless until it is related to at least one other datum (p. 5). Constant themes that reoccu r throughout content analysis literature are the ideas of reliability and validity. Krippendor ff (1980) states that reliability assesses the extent to which any research design and any data resulting fr om them represent variations in real phenomena rather than the extraneous circumstances of measur ement (p. 129). In other words, reliability is determined when the data that is measured su stains consistency over numerous measurements. On the other hand, the concept of validity is a mo re subjective and complicated subject due to the many uses and definitions (Weber, 1990). In general, the term validity refers to the use of fact to reinforce a statement therefore making that stat ement logic filled (Riffe, Lacy, & Fico, 1998). Purpose This study employed a content analysis to j udge the positive or negative impact that media coverage has on the ranking of teams vying for higher BCS poll position. Because campaigning for higher BCS ranking is a new and relatively controversial topic, it can be surmised that campaigning teams will receive heightened media coverage. This content analysis of media coverage did not attempt to identify be st forms of campaigning by coaches, university representatives, or players to e ngage in, but rather it attempted to determine if heightened media coverage and the appearance of campaigning affected the human po lling elements of the BCS. 43

PAGE 44

A content an alysis was employed in this st udy due the rich tradition of its use as a methodology while measuring the media effects of agenda-setting. Most studies have used a combination of content analysis and surv ey (e.g., McCombs & Shaw, 1972, Shaw & Martin, 1992, Walters, Walters, & Grey, 1996, McCombs, Llamas, Lopez-Escobar, & Rey, 1997, Golan & Wanta, 2001, Kiousis & McCombs, 2004, and Wu & Seltzer, 2006) to measure the media agenda (content analysis) and it s impact on the public agenda (survey), while other studies have relied on content analysis alone (e.g., Ohl, Pi ncus, Rimmer, & Harrison, 1995, Roberts, Wanta, & Dzwo, 2002, and Kiousis & Sheilds, 2008). Content analysis has been a favorite among agenda-setting researchers for its proven abil ity to link large amounts of media content on a specific topic, with the salience of that specific to pic to the public (McCombs, 1992). Kosicki (1993), explained content analysis s function in agenda-setting research when he said, the amount of space or time devoted to particular issues should be measured, and this measurement should relate to either the amount of attention people pay to issues or to their judgments of the issues importance (p. 105). Media Coverage/Sample The media outlets which the content was drawn from were the newspapers The New York Times and USA Today the magazine The Sporting News and from the Associated Press newswire. The New York Times was analyzed based on its use in past agenda-setting content analyses (e.g., Miller, Andsager, & Riechert, 1998, Hester & Gibson, 2003, Yioutas & Segvic, 2003, and Kiousis, 2005) and due the fact that it is generally regarded as the most respected U.S. news medium (Dearing & Rogers, 1996). USA Today a national American daily newspaper, was included in this content analysis based on its circ ulation, which is the widest of any newspaper in the U.S. averaging 2.27 m illion copies every weekday (USAToday.com, 2008). In addition, USA Todays sponsorship of the Coaches Poll (one of the two human 44

PAGE 45

elem ents of the BCS) could have increased the am ount of media coverage of the BCS in its own publication. The bi-weekly sports magazine The Sporting News (TSN) was included in this content analysis due to its reput ation as a well established and hi ghly respected sports magazine. TSN established in 1886 is widely known as th e bible of baseball but in 1942 expanded its content to include a wide range of other sports including college football (Funk, 2009). The Sporting News the oldest sports public ation in the United States couples its 715,767 bi-weekly circulation with a online vers ion at www.SportingNews.com and is considered by many in the sports industry a publisher of reli able sports data (Bosman, 2006). Finally, the Associated Press (AP) was used a source in this content an alysis for two reasons. First, the AP is internationally known as the leader in unbiased news information. Evidence of this can be found on the APs own Website; founded in 1846, AP today is the largest and most trusted source of independent news and information. On any given day, more than half the world's population sees news from AP (AP.org, 2009). Secondly, AP sports stories were selected for inclusion in this content analysis due to their accessibility through LexisNexis. The AP yielded, by far, the greatest amount of articles containing the three search topics (2309). The time period from which the media conten t was drawn included the later half of the four most recent college football seasons. The la st eight weeks of four regular college football seasons from 2005 to 2008 were analyzed due to the emphasis on the BCS standings. In addition, 2005 was the first year to include both polls which were analyzed in this study ( USA Today Coaches poll and the Harris In teractive poll). The first annual installm ent of the BCS poll is traditionally released in mi d October, which is approximately seven weeks after the beginning of the college football season (15 week season) in the last week of Augus t. In the 2005 season, the time period ran from October 15 to Decembe r 3. Again, for the 2006 season, the time period 45

PAGE 46

analyzed was from October 15 to December 3. For the 2007 season, the analyzed time period was from October 14 to December 2. Finally, for the 2008 season, the time period analyzed was from October 19 to December 7. Articles from all four sources the Associated Press New York Times The Sporting News and USA Today were located using LexisNexis online data base. LexisNexis has been used in a number of recent content analysis to study the ef fects of news coverage on the public agenda (e.g., Tedesco 2001, Hester & Gibson, 2003, and Kiousis, 2005). Academically, LexisNexis is a highly respected online search engine which, accord ing to the website, will allow the researcher access to over 6,000 news, business, and legal sources. In addition, the search engine accesses, outstanding news coverage (w hich) includes deep b ack-files and up-to-the-minute stories in national and regional newspapers, wire services broadcast transcripts, international news (Lexisnexis.com, 2008). In the LexisNexis search, articles were accessed by using the search terms Bowl Championship Series, USA Today Coaches poll, and Harri s Interactive poll. These terms were used to define the focus of th e search and identify all articles which pertain only to the BCS race, and the two pertinent human polls. These search results were reviewed to ensure that the most appropriate content was incor porated into the final analysis. Only articles within the predetermined time pe riod which pertain to college f ootball were included in the analysis. All duplicate articles were eliminated along with other such articles which did not fit the extraneous search criteria (ex. did not pertain to college football). In the end a sample size of 500 articles was agreed upon by the researchers as an acceptable amount. Using the three search terms on LexisNexis Academic between the specified dates, and discarding duplicates and extraneous articles, the sear ch yielded 216 articles from the Associated Press, 105 articles from the New York Times 33 articles from The Sporting News 46

PAGE 47

and 146 articles from USA Today The low disproportional amount of articles used from The Sporting News was due to LexisNexiss inability to acce ss articles from that publication in both the years of 2007 and 2008. Coding All 500 articles were then coded to determine type and amount of media coverage individual teams received during the predetermi ned time period. Individual teams included the top ten ranked teams in each years final BCS po ll. The final BCS poll is released the Monday after the end of the regular season and will dete rmine the two teams which will match up in the BCS championship game. The top ten teams in each years final BCS poll were examined in each of the last four seasons (2005 to 2008). Thes e final top ten teams were tracked due to their play on the field (final record). In the final weeks of the BCS polling, only teams which had records good enough to compete for a place in the title game were ranked in the top ten (zero, one, or two losses). It was highly unlikely that teams outside the top ten would be eligible for title game participation due to record. For the 2005 season, the top ten teams in the final BCS poll included, in order of ranking, were, Southern California, Texas, Penn State, Ohio State, Oregon, Notre Dame, Georgia, Miami (FL), Aubur n, and Virginia Tech. For the 2006 season the final top teams, in order of ranking, included, Ohio State, Florida, Michigan, Louisiana State, Southern California, Louisville, Wisconsin, Bois e State, Auburn, and Ok lahoma. The final top ten teams in the BCS poll in 2007, in order of ranking were, Ohio State, Louisiana State, Virginia Tech, Oklahoma, Georgia, Missouri, So uthern California, Kansas, West Virginia, and Hawaii. For the 2008 season the top ten teams in the final BCS poll, in order of ranking were Oklahoma, Florida, Texas, Alabama, Southern California, Utah, Texas Tech, Penn State, Boise State, and Ohio State. 47

PAGE 48

The nam e of the team being coded was iden tified by a predetermine numbering system. Of all fifty possible positions in the top te n of the final BCS polls between 2005 and 2008, only 24 teams filled those 40 positions. Those teams were assigned a number in alphabetical order 1 through 24. The teams were coded as follows: 1 Al abama, 2 Auburn, 3 Boise State, 4 Florida, 5 Georgia, 6 Hawaii, 7 Louisiana State, 8 Loui sville, 9 Kansas, 10 Miami (FL), 11 Michigan, 12 Missouri, 13 Notre Dame, 14 Ohio State, 15 Or egon, 16 Oklahoma, 17 Penn State, 18 Southern California, 19 Texas, 20 Texas Tech, 21 Ut ah, 22 Virginia Tech, 23 West Virginia, 24 Wisconsin. Each article was first analyzed to determin e word count. The length of an article was used to determine prominence and importance to r eaders. The length of the article may indicate importance. Length of the article was coded based on word count; Short being under 300 words, medium being 300 to 600 words, and long being ove r 600 words. (1=short, 2=medium, 3=long). Next, each article was examined for the pres ence of each team ranked as one of the top ten teams in the final BCS poll by recording th e present teams corresponding number (1-24). The presence of one or more of the predetermi ned teams acted as a base calculation for the of number of media references a team had. It could be theorized that the more articles a particular team was mentioned in the higher ranked a team was (weekly or overall ). In addition, the articles were examined to determine what team was the focus of each article and subsequently coded by recording that teams predetermined number (1-24). Examples of being the focus included if the teams name was me ntioned in the title of the arti cle, if the teams players and coaches were quoted, or if one t eam was mentioned more times than another team. It must be noted that more than one team may have been th e focus of articles. For example if the article talked of a future game between two teams, if the article argued for more than two teams 48

PAGE 49

inclus ion in the BCS title game, or if the article put relatively equal emphasis on more than one team. If a teams position in the BCS was mentioned, then that position was recorded on a scale of 1 to 10. If the team was mentioned as leadi ng the race, the number one team, or in first place then that justified the team being recorded as 1. If the team was mentioned as a close second the other BCS title game team, or the runner up, then that wa rranted that team being coded as 2, and so on and so forth for teams thre e through 10. If the team was mentioned as a BCS contender without position in the BCS sta nding, then that instance was coded as 11. If the team was mentioned but not as a BCS contende r, then this was coded as 0. A weekly score was tabulated for each team based on a point sy stem which awards 10 points to a team every time it was coded as 1, 9 points for 2, 8 points fo r 3, 7 points for 4, 6 points for 5, 5 points for 6, 4 points for 7, 3 points for 8, 2 points for 9, 1 point for 10, and 0 points for 11. This way a weekly tabulation could be kept to determine wh ich team is perceived to be leading the BCS race. In addition, a teams valence attitude measure was conducted where each story was coded as either positive, negative, or neutral (1 = positive, 2 = negative, 3 = neutral). As before, a weekly score was tallied to determine direction and degree of attitude shift in week-to-week media coverage of the final top 10 teams. A pos itive attitude conveyed in the article was valued at 1, a neutral attitude at 2, and a neutral attitude was valued at 3. In an attempt to quantify campaign behavi or, the presence of campaign language was tabulated. For this study, the term campaign language was operationally defined as any attempt by a coach, university represen tative, or player to argue, state, or reinforce why his or her team should have higher ranking in the BCS poll. This definition was developed through the 49

PAGE 50

analysis of past instances of college football coaches or team repr esentatives disseminating statements through the media regarding their teams ranking. These instances were then described by the media as campaigning or possessing political campaigning characteristics. Behavior and language displayed in these campai gning instances ranged from general umbrella statements about team rank to specific acknowle dgment of team attributes such as win-loss record, quality wins, head-to-head match-ups, and individual player accomplishments. Any article containing campaign la nguage was coded by recording the team which was being campaigned for by that teams corresponding number (1 -24). In addition, when a team had been identified as being campaigned for using the defined campaign language, then the attitude or tone in the article toward campaigning language was recorded and subsequently coded as 1=positive, 2=negative, 3=neutral. Reliability Reliably was tested by using two separate met hods of inter-coder reliability. The first method was Scotts Pi.3 Scotts Pi was used based on Schi ff and Reiters (2004) account that for almost 50 years, Scotts Pi has served as the accepted standard for inter-coder reliability for nominal data in communication studies (Sco tt 1955; Krippendorff 1980; Wimmer & Dominick 1997; Keyton 2001) (p. 3). Scot ts Pi method was employed because of its past use in numerous agenda-setting cont ent analyses (e.g., Wanta & Wu, 1993; Kiousis & Shields, 2008; Kiousis & Wu, 2008). The second method of inte r-coder reliability used was Krippendorffs Alpha. Krippendorffs Alpha was employed because of its past use in numerous agenda-setting content analyses (e.g., Tafati 2003; Craig 2004; Stroud & Kenski 2007). To reduce the volume 3 Scotts Pi = (P o P e ) / (1 P e ) where P o = Percent of Observed Agreement = Number of Observed Agreements / Total Number of Cases Where P e = Percent of Expected Agr eement = (Number of Cases where the Category is Absent/ (Number of Cases x 2)) 2 + (Number of Cases where the Category is Present/ (Number of Cases x 2)) 2 = (% of 0s) 2 + (% of 1s) 2 Where (% of 0s) = (Total # of 0s / (Total # of Cases x 2 Coders) (% of 1s) = (Total # of 1s / (Total # of Cases x 2 Coders) 50

PAGE 51

of data to a m anageable proportion the two coders examined a sub-sample of approximately 10% of the total articles sampled (H olsti, 1969). The sub-sample of 10% of the total number of articles coded was 50 (n=50). Using Scotts Pi, coder agreement ranged fr om a low of 96% (.86) to a high of 100% (1.00) and had a mean of 98.6% (.994) for all 83 variables. Using Krippendorffs Alpha, coder agreement again range d from a low of 96% (.861) to a high of 100% (1.00) and also had a mean of 98.6% (.994) for all 83 variables. A code book outlining coding instructions, expectations, guidelines, and definitions was created and used by all coders in an attempt to unify analysis understanding. A code sheet containi ng coding categories and valences was supplied to coders to ensure pr oper recording and catalogi ng of coded articles. Public Agenda/Voter Opinion Just as McCombs and Shaw (1972) used a combination of content analysis and public opinion poll to study the 1968 United States Presiden tial election, this stu dy combined a content analysis with college football ranking polls to study the 2005 to 2008 colle ge football seasons. The USA Today Coaches poll and Harris Interactive poll ac ted as voter opini on indicators as they are the two human polls which determine two-thirds of the BCS tabulation. The two college polls were examined on a week-by-week basis to determine a teams position in the polls in a week-by-week comparison with that same team s media coverage. If a particular team rose in the polls during week number one, the media c overage of that same team during week number one was compared to see if the c overage increased or decreased. In addition to a week-by-week look at the USA Today Coaches poll and the Harris Interactive poll, the final week of both polls was examined to assess overall correlation between a teams overall media coverage and final outcome For example, if a team received more and favorable media coverage than another throughout the entire season, then it could be theorized that the first team would be ranked highe r than the latter in the final polls. 51

PAGE 52

Data Analysis Strategy Based on the accessibility of the sample and the data recovered from the coding process, a correlation coefficient analysis technique was employed. Based on the bivariate nature of the data it was possible to conduct, a Pers on Product Moment correlation (PPMC) for each hypothesis to determine the relationships betwee n the media agenda (amount and valence of coverage) and voter opinion (a teams final standing in the thr ee polls). In addition, a singletailed test was applied due to the directional nature of the hypotheses asked in the study. A PPMC was used in lieu of a Spearman Rank-Order correlation due to the intervaled data. A PPMC is believed to be a more accurate measur e of the linear relations hip between variables using intervaled data than a Spearman Rank-Orde r correlation (Field, 20 00). Furthermore, for data analysis purposes, all four seasons were analyzed as one composite season. team one represented all four teams which finished No. 1 in the final polls over the four years analyzed (2005 to 2008), team two represented all four te ams which finish No. 2 in the final polls, and so and so forth through 10, which represented the t eams that finished in the top 10 of the final BCS poll (Appendix A). 52

PAGE 53

CHAP TER 4 RESULTS Total Media Coverage and BCS Outcome H1: Teams exhibiting more media coverage will be ranked higher in the final BCS poll than teams exhibiting less media coverage. To analyze H1 (hypothesis one), a Pears on Product Moment Correlation (PPMC) was computed to assess the relationship between a teams final BCS standing and total media coverage. Bivariate comparisons were made be tween a number of variables including final BCS poll and total media coverage and media coverage va lence. Due to the directional nature of the hypothesis a one-tailed te st was conducted. In regards to H1, there was a significant positive correlation between the two variables at the 0.01 level, r = 0.583, n = 40, p = 0.000 (Table 41). A scatterplot su mmarizes the results (Appendix K, Figure 1). Overall, there was a strong, positive co rrelation betwee n the final BCS poll and total amount of media coverage. A highe r ranking in the final BCS poll was correlated with an increase in total amount of media coverage. Table 4-1 Final BCS and total media coverage PPMC Variables Final BCS Media coverage Pearson correlation 1 .583** Sig. (1-tailed) .000 Final BCS N 40 40 Pearson correlation .583** 1 Sig. (1-tailed) .000 Media coverage N 40 40 Note. **. Correlation is significant at the 0.01 level (1-tailed). In summation, there was a strong correlation between total media coverage and final BCS standing. 53

PAGE 54

Positive Media Coverage and Final BCS Outcome H1a: Team s exhibiting more positive media covera ge will be ranked higher in the final BCS poll than teams exhibiting less positive media coverage. A correlation between a teams final BCS standing and the valence of total media coverage was conducted using a PPMC. In rega rds to a teams final BCS standing and positive media coverage, there was a significant positiv e correlation at the 0.01 le vel, r = 0.426, n = 40, p = 0.003 (Table 4-2). A scatterplot summarizes the results (Appendix K, Figure 2). Overall, there was a strong, positive corr elation between the final BCS standings and positive media coverage. The higher a team was ranked the more positive the media coverage was which the team received. Table 4-2. Final BCS and positive media valence PPMC Variables Final BCS Media coverage Media val. pos. Pearson correlation 1 .583** .426** Sig. (1-tailed) .000 .003 Final BCS N 4040 40 Pearson correlation .583**1 .786** Sig. (1-tailed) .000 .000 Media coverage N 4040 40 Pearson correlation .426**.786** 1 Sig. (1-tailed) .003 .000 Media val. pos. N 40 40 40 Note. **. Correlation is significant at the 0.01 level (1-tailed). To further analyze the strong correlation be tween a teams higher final BCS ranking and positive media coverage, a PPMC was conducted to test if there was a correlation between a teams final BCS ranking and negative media cove rage. After running a PPMC the statistics did not show a significant correlation between the two variables, r = 0.173, n = 40, p = 0.143 (Table 54

PAGE 55

4-3). Overall there was no significant correlation between fina l BCS ranking and negative m edia coverage. Table 4-3. Final BCS and ne gative media coverage PPMC Variables Final BCS Media coverage Media val. neg. Pearson correlation 1 .583** .173 Sig. (1-tailed) .000 .143 Final BCS N 4040 40 Pearson correlation .583**1 .309* Sig. (1-tailed) .000 .026 Media coverage N 40 40 40 Pearson correlation .173 .309* 1 Sig. (1-tailed) .143 .026 Media val. neg. N 40 40 40 Note. **. Correlation is significant at the 0.01 level (1-tailed). In summation, there was a strong correlation between positive media coverage and final BCS standing. In addition, no significant correlation was found be tween final BCS standing and negative media coverage. Media Coverage and Final Harris Poll Outcome H2: Teams exhibiting more media coverage will be ranked higher in the final Harris Interactive poll than teams exhibiting less media coverage. To analyze H2, once again, a PPMC was comput ed to assess the re lationship between a teams final Harris Interactive poll standing and total media cove rage. Bivariate comparisons were made between a number of variables including final Harris poll and total media coverage and media coverage valence. Due to the direct ional nature of the hypothe sis, a one-tailed test was conducted. In regards to H2 there was a significant pos itive correlation between the two variables at the 0.01 level, r = 0.516, n = 40, P = 0.000 (Table 44). A scatterplot su mmarizes the results 55

PAGE 56

(Appendix K, Figure 3). Overall, there was a strong, positive correla tion between the final Harris poll and total am ount of media coverage. A higher ranki ng in the final Harris poll was correlated with an increase in to tal amount of media coverage. Table 4-4. Final Harris poll and Total Media Coverage PPMC Variables Final Harris Media coverage Pearson correlation 1 .516** Sig. (1-tailed) .000 Final Harris N 40 40 Pearson correlation .516**1 Sig. (1-tailed) .000 Media coverage N 40 40 Note. **. Correlation is significant at the 0.01 level (1-tailed). In summation, there was a strong correlati on between total media coverage and final Harris Interactiv e poll standing. Positive Media Coverage and Final Harris Poll Outcome H2a: Teams exhibiting more positive media covera ge will be ranked higher in the final Harris Interactive poll than teams exhibiting less positive media coverage. A correlation between a teams final Harris In teractive poll standings and the valence of total media coverage was conducted using a PPMC. In regards to final Harris poll standing and positive media coverage, there was a significant positive correlation at the 0.01 level, r = 0.374, n = 40, p = 0.009 (Table 4-5). A scatterplot summa rizes the results (Appendix K, Figure 4). Overall, there was a strong, pos itive correlation between the fi nal Harris poll standings and positive media coverage. The higher a team was ranked the more positive the media coverage was for that team. 56

PAGE 57

Table 4-5. Final Harr is poll and positive media coverage PPMC Variables Final Harris Media coverage Media val. pos. Pearson correlation 1 .516** .374** Sig. (1-tailed) .000 .009 Final Harris N 40 40 40 Pearson correlation .516**1 .786** Sig. (1-tailed) .000 .000 Media coverage N 40 40 40 Pearson correlation .374**.786** 1 Sig. (1-tailed) .009 .000 Media val. pos. N 40 40 40 Note. **. Correlation is significant at the 0.01 level (1-tailed). To further analyze the strong correlation between a teams higher final BCS ranking and positive media coverage, a PPMC was conducted to test for any correlation between final BCS ranking and negative media covera ge. After running the PPMC, th e statistics did not show a correlation between the two variables, r = 0.143, n = 40, p = 0.190 (Table 4-6). Overall there was no significant correlation between final Harris poll ranking and negative media coverage. Table 4-6. Final Harris poll a nd negative media coverage PPMC Variables Final Harris Media coverage Media val. Neg. Pearson correlation 1 .516** .143 Sig. (1-tailed) .000 .190 Final Harris N 40 40 40 Pearson correlation .516**1 .309* Sig. (1-tailed) .000 .026 Media coverage N 40 40 40 Pearson correlation .143 .309* 1 Sig. (1-tailed) .190 .026 Media val. neg. N 40 40 40 Note. **. Correlation is significant at the 0.01 level (1-tailed). 57

PAGE 58

In summ ation, there was a strong correlation between positive media coverage and final Harris poll standings. In addi tion, no significant correlation wa s found between final Harris poll standings and negative media coverage. Media Coverage and Final USA Today/Coaches Poll Outcome H3: Teams exhibiting more media coverage will be ranked higher in the final USA Today/Coaches poll than teams exhibiting less media coverage. To analyze H3, again a PPMC was computed to assess the relationship between a teams final USA Today/Coaches poll standing and total media coverage. Bivariate comparisons were made between a number of variables including final USA Today/Coaches poll and total media coverage and media coverage valence. Due to the directional nature of the hypothesis a onetailed test was conducted. In regards to H3, there was a significant positive correlation between the two variables at the 0.01 level, r = 0.489, n = 40, P = 0.001 (Table 47). A scatterplot su mmarizes the results (Appendix K, Figure 5). Overall, there was a strong, positive corr elation between the final USA Today poll and total amount of media cove rage. A higher ranking in the final USA Today/Coaches poll was correlated with an incr ease in total amount of media coverage. Table 4-7. Final USA Today/Coaches poll and total media coverage PPMC Variables Final USA Media coverage Pearson correlation 1 .489** Sig. (1-tailed) .001 Final USA N 40 40 Pearson correlation .489** 1 Sig. (1-tailed) .001 Media coverage N 40 40 Note. **. Correlation is significant at the 0.01 level (1-tailed). 58

PAGE 59

In summation, there was a strong correl ation between total media coverage and final USA Today/Coaches poll standings. Positive Media Coverage and Fina l USA Today/Coaches Poll Outcome H3a: Teams exhibiting more positive media cove rage will be ranked higher in the final USA Today/Coaches poll than teams exhibiting less positive media coverage. A correlation between a teams final USA Today/Coaches poll standing and the valence of total media coverage was conducted using a PP MC. In regards to final USA Today/Coaches poll standing and positive media coverage, there wa s a significant positive correlation at the 0.05 level, r = 0.329, n = 40, p = 0.019 (Table 4-8). A scatterplot summarizes the results (Appendix K, Figure 6). Overall, there was a moderately strong positive correlati on between the final USA Today/Coaches poll standings and positive media coverage. The higher a team was ranked, the more positive the media coverage was which that team received. Table 4-8. Final USA Today/Coaches poll and positive media coverage PPMC Variables Final USA Media coverage Media val. pos. Pearson correlation 1 .489** .329* Sig. (1-tailed) .001 .019 Final USA N 4040 40 Pearson correlation .489**1 .786** Sig. (1-tailed) .001 .000 Media coverage N 40 40 40 Pearson correlation .329*.786** 1 Sig. (1-tailed) .019 .000 Media val. pos. N 40 40 40Note. **. Correlation is significant at the 0.01 level (1-tailed). *. Correlation is significant at the 0.05 level (1-tailed). To further analyze the correlation between a teams higher final BCS ranking and positive media coverage, a PPMC was conducted to test if there was any correlation between 59

PAGE 60

final USA Today/Coaches poll ranking and negati ve m edia coverage. After running the PPMC the statistics did not show a si gnificant correlation between the two variables, r = 0.125, n = 40, p = 0.245 (Table 4-9). Overall there was no significant correlation between final USA Today/Coaches poll ranking and negative media coverage. Table 4-9. Final USA Today/Coaches poll and negative media coverage PPMC Variables Final USA Media coverage Media val. neg. Pearson correlation 1 .489** .125 Sig. (1-tailed) .001 .245 Final USA N 40 40 40 Pearson correlation .470**1 .309* Sig. (1-tailed) .001 .026 Media coverage N 40 40 40 Pearson correlation .125 .309* 1 Sig. (1-tailed) .245 .026 Media val. neg. N 40 40 40 Note. **. Correlation is significant at the 0.01 level (1 -tailed). *. Correlation is significant at the 0.05 level (1-tailed). In summation, there was a moderately st rong correlation betw een positive media coverage and final USA Today/Coaches poll stan dings. In addition, no significant correlation was found between final USA Today/Coaches poll standings and negative media coverage. Campaigning Coverage and Final BCS Poll Outcome H4: Teams engaging in campaigning behavior will be ranked higher in the final BCS poll than teams not engaged in campaigning behavior. Finally, to analyze H4, again a PPMC was comp uted to assess the relationship between a teams final BCS standing and total campaigning coverage. Bivariate comparisons were made between a number of variables including fi nal BCS poll and total campaign coverage and 60

PAGE 61

cam paign coverage valence. Due to the direc tional nature of the hypothesis, one-tailed tests were conducted. In regards to hypothesis four there was a si gnificant positive corre lation between the two variables at the 0.01 level, r = 0.371, n = 40, P = 0.009 (Table 4-10). A sc atterplot summarizes the results (Appendix K, Figure 7). Overall, there was a strong, positive correlation between the final BCS poll and total amount of campaign cove rage. The higher a team ended up ranked in the final BCS poll was correlated with an in crease in total amount of campaign coverage. Table 4-10. Final BCS and total campaign coverage PPMC Variables Final BCS Campaign coverage Pearson correlation 1 .371** Sig. (1-tailed) .009 Final BCS N 40 40 Pearson correlation .371** 1 Sig. (1-tailed) .009 Campaign coverage N 40 40 Note. **. Correlation is significant at the 0.01 level (1-tailed). In summation, there was a strong correlation between total campaign coverage and final BCS poll standings. Positive Campaigning Coverage and Final BCS Poll Outcome H4a: Teams engaging in positive campaigning beha vior will be ranked higher in the final BCS poll than teams not engaged in positive campaigning behavior. A correlation between a teams final BCS sta ndings and the valence of total campaigning coverage was conducted using a PPMC. In regards to final BCS standing and positive campaigning coverage there was a significant positive correlation at the 0.05 level, r = 0.281, n = 40, p = 0.040, (Table 4-11). A scatterplot summa rizes the results (Appendix K, Figure 8). Overall, there was a moderately strong, positive correlation between the final BCS standings and 61

PAGE 62

positiv e campaigning coverage. The higher a team was ranked, the more positive the campaigning coverage was that that team received. Table 4-11. Final BCS and positive campaign coverage PPMC Variables Final BCS Campaign coverage Campaign val. pos. Pearson correlation 1 .371** .281* Sig. (1-tailed) .009 .040 Final BCS N 4040 40 Pearson correlation.371**1 .214 Sig. (1-tailed) .009 .092 Campaign coverage N 4040 40 Pearson correlation.281*.214 1 Sig. (1-tailed) .040 .092 Campaign val. pos. N 40 40 40 Note. **. Correlation is significant at the 0.01 level (1-tailed). *. Correlation is significant at the 0.05 level (1-tailed). In summation, there was a moderately st rong correlation betw een positive campaign coverage and final BCS poll standings. 62

PAGE 63

63 CHAPTER 5 DISCUSSION Overall, the results of the study show strong su pport for all eight hypotheses. In general, for the first six hypotheses, it can be stated that teams which received bo th more and favorable media coverage were ranked higher in the fi nal BCS, Harris Interactive, and USA/Today Coaches polls, respectively. A strong correlation between positive media coverage and higher ranking in all polls, coupled w ith a week correlation between negative media coverage and higher ranking in all polls, added support that va lence plays a large roll in team ranking. In addition, a strong correlation between campai gning coverage and hi gher BCS ranking added support that teams engaged in campaigning behavior will exhibit higher BCS ranking than teams not engaged in campaigning behavior. Hypothesis One H1: Teams exhibiting more media coverage will be ranked higher in the final BCS poll than teams exhibiting less media coverage. Out of the eight hypotheses, this hypothesis was supported by the st rongest relationship observed, with a strong positive correlation betwee n the variables. This strong correlation was backed up with further examination of individual t eam media data. In thr ee out of the four years examined, the team which was ranked the highes t in the final BCS poll received the most media attention. In 2005, No. 1 USC was mentioned 71 times, 4 more times than No. 2 ranked Texas (Appendix H, Figure 1). In 2006, No. 1 Ohio Stat e was mentioned 60 times, 5 more times than No. 2 ranked Florida (Appendix H, Figure 2). In 2008, No. 1 ranked Oklahoma was mentioned 61 times, 19 more times than No. 2 ranked Florid a, and 3 more times than No. 3 ranked Texas (Appendix H, Figure 4). In 2007, No. 2 ranked LSU garnered the most media attention with 52 mentions while No. 1 Ohio State was only four behind with 48 mentions (Appendix H, Figure 3).

PAGE 64

Hypothesis One A H1a: Team s exhibiting more positive media c overage will be ranked higher in the final BCS poll than teams exhibiting less positive media coverage. Positive media coverage valence played a large roll in both 2005 and 2006. In 2005, No. 1 ranked USC received the largest amount of pos itive media coverage with 22 positive mentions while runner up Texas received th e second largest amount with 17 (Appendix H, Figure 1). In 2006, BCS No. 1 Ohio State lead all teams with 18 positive mentions, while No. 2 Florida was second with 13 positive mentions (Appendix H, Figure 2). In 2007, BCS No. 1 LSU finished third with 7 positive mentions while No. 2 Ohio State finished first with 9 positive mentions (Appendix H, Figure 3). Finally, in 2008, BCS No. 1 Oklahoma was third on the list with 6 positive mentions while No. 2 Florida was second with 7 positive mentions (Appendix H, Figure 4). These numbers are consistent with the fact that the BCS No. 1 and No. 2 teams for each year finished in the top five in both positive media c overage and total media coverage for each year. In addition, adding support to the first hypothesis, no team finishing in the BCS top three finished lower than sixth out of ten in total media coverage, except for Virginia Tech in 2007 who finished eighth (Appendix H, Figures 1-4). Since the BCS is made up of two-thirds hum an voters and one-third computer rankings the fact that this hypothesis was so strongly supported can only te ll two-thirds of the story. The BCS is influenced by nonhuman factors and cannot be a complete determinant of agenda-setting effects of topic salience transfer from media to the voting pu blic. After analyzing the final BCS poll against media coverage, further analysis of the two human polls by themselves against total media coverage and valence of coverage must be conducted to determine if the correlation continues to be strong. 64

PAGE 65

Hypothesis Tw o H2: Teams exhibiting more media coverage will be ranked higher in the final Harris Interactive poll than teams exhibiting less media coverage. This hypothesis was also str ongly supported. To test this hypothesis, the correlation between total media coverage of teams finishing in the final top ten of the BCS poll and the voter opinion of the final Harris Interactive poll were examined for each of the four years studied. After analysis, it was determined that this hypothesis was supported by the second strongest correlation between variables only behind H1. Si nce the final Harris Interactive poll and the final BCS poll were very similar in their team rankings, it could be presumed that the strong correlations found in H1 w ould carryover to H2. In agenda-setting terms, th e strong correlation between me dia coverage and final poll standings was extremely significant. The str ong correlation between the two variables suggests the possible presence of agenda-setting effects in the form of topic salience transfer from the media to the voting public. The strongest evid ence of this was in the correlation between individual teams media coverage data and that teams final Harris poll standings. Once again, as in the first hypothesis in both 2005 and 2006, teams ranked No. 1 (USC, 2005; Ohio State, 2006) and No. 2 (Texas, 2005; Florida, 2006) had the most and second most article mentions, respectively (Appendix E, & Appendix H, Figure 1 & 2). In 2007, the No. 1 ranked team, Ohio State, had the second most article mentions wh ile the No. 2 ranked team, LSU, had the most mentions (Appendix E, & appendix H, Figure 3). In 2008, the No. 1 ranked team, Oklahoma, had the most mentions followed by the No. 3 ranked team, Texas, with the second most mentions. The No. 2 ranked team, Florida, had the fifth most mentions in 2008 (Appendix E, & appendix H, Figure 4). 65

PAGE 66

Hypothesis Tw o A H2a: Teams exhibiting more positive media c overage will be ranked higher in the final Harris Interactive poll than teams exhibiting less positive media coverage. The voter opinion of the Harri s Interactive poll proved to be an adequate measure of public topic salience as seen th rough the shared high media covera ge correlation with the final BCS poll. In addition, teams which received more positive media coverage ended up ranked higher in the final Harris Interactive poll as shown by a strong negative correlation between the two variables. This was also evident with an analysis of individual team positive media coverage. Just as with both the BCS in 2005, No. 1 ranked USC received the most positive media coverage, followed by No. 2 ranked Texas in the Harris poll (Appendix E, & appendix H, Figure 1). The same can be said, when speaki ng about the Harris Interactive poll for both the 2006 and 2007 seasons (Appendix E, & Appendix H, Figures 2 & 3). These correlations are also si gnificant in supporting an agen da-setting affect. Since the correlation is high between media coverage a nd Harris Interactive po ll ranking, it could be theorized that college football team ranking salie nce was transferred from the media to the voting public of the Harris poll. The amount of media c overage of a specific team may have influenced a voter into believing that a spec ific team deserved a higher ranking. The same can be said for valence. Since there was a high correlation be tween positive media coverage of a team and higher ranking of that team, it could be theorized that a voter in the Harris poll may have been influenced to rank a team higher based on the medi as positive representa tion of that team. Hypothesis Three H3: Teams exhibiting more media coverage will be ranked higher in the final USA Today/Coaches poll than teams exhibiting less media coverage. 66

PAGE 67

This hypothesis was also str ongly supported. T his was the third strongest correlation between media coverage and voter opinion poll behind the BCS poll and the Harris Interactive poll. Once again, evidence of this was seen in an examination of individual team data from the years in question. Since the final Harris Inte ractive polls and the USA Today/Coaches polls from all four years did not contain any signifi cant differences in ranking order data, these comparisons were somewhat similar. In this instance, the final top two ranked teams in 2005, 2006, and 2007 received the most media mentions in ranking order, respectively (Appendix F, & appendix H, Figures 1,2 & 3). In 2008, the No. 1 ranked team, Oklahoma, had the most article mentions, followed by the No. 2 ranked team, Texas, and with the No. 2 ranked team, Florida, finishing fifth in article mentions (Appendix F, & appendix H, Figure 4). Hypothesis Three A H3a: Teams exhibiting more positive media c overage will be ranked higher in the final USA Today/Coaches poll than teams e xhibiting less positive media coverage. As with the Harris poll, teams which receiv ed more positive media coverage ended up ranked higher in the final USA Today/Coaches poll as shown by a strong positive correlation between the two variables. Agai n, this was also evident with an analysis of individual team positive media coverage. Just as with both the BCS and Harris polls in 2005 the USA Today/Coaches poll No. 1 ranked USC received th e most positive media coverage, followed by No. 2 ranked Texas (Appendix F, & appendix H, Figure 1). The same can be said when speaking about the USA Today/Coaches poll for both the 2006 and 2007 seasons (Appendix F, & appendix H, Figures 2 & 3). These correlations, as in H2, are significant in su pporting an agenda-set ting affect. Since the correlation is high between media coverage and USA Today/Coaches poll ranking it could be 67

PAGE 68

theorized th at the topic of colle ge football team ranking salience was transferred from the media to the voting public of the coaches. The amount of media coverage of a specific team may have influenced a voter into believing that specific team deserved a hi gher ranking. Again, the same could be said for valence. Since there was a high correlation between positive media coverage of a team and higher ranking it could be theorized th at a voter in the Coaches poll may have been influenced to rank a team higher based on the me dias positive represen tation of that team. Hypothesis Four H4: Teams engaging in campaigning behavior will be ranked higher in the final BCS poll than teams not engaged in campaigning behavior. This hypothesis was also str ongly supported. Evidence of th is was seen in a strong positive correlation between the amount of campaigning language used by a team and that teams final BCS standing. Additional evidence of this correlation could also be found in individual team campaigning data from three of the four years studied. In 2005, six teams were determined to have engaged in campaigni ng behavior through the use of campaigning language. USC, who ended the season ranked No. 1 in the BCS, was tallied as using the most campaigning language with 6 instances, followed by BCS No. 2 ranked Texas, who used campaigning language 4 times. The other four teams, Virginia Tech, Auburn, Georgia, and Oregon, all were found to be campaigning 2 or le ss times and ended the season ranked no higher than fifth in the final BCS poll (Appendix J, Figure 1). In 2006, six teams we re determined to be engaged in campaigning behavior with the top three teams ranked in the final BCS poll all present in the list including N o. 2 ranked Florida, who lead the way with 6 campaigning examples, followed by Michigan with 3, and Ohio State with 2. The other three teams engaged in campaigning, Boise State, Auburn, and Oklahoma, were mentioned as campaigning 1 time a 68

PAGE 69

piece and finish ranked 8, 9, and 10, respectivel y (Appendix J, Figure 2). In 2008, only three teams engaged in campaigning, with BCS N o. 1 ranked Oklahoma, engaging 10 times in campaigning followed by No. 3 Texas, contribut ing 7 times; No. 6 Utah was found campaigning only once (Appendix J, Figure 4). The campaigning data supported evidence of documented instances of college football coaches campaigning in both 2006 and 2008. In 2006, Florida was engaged in a campaign battle with Michigan for inclusi on in the BCS title game and the data showed that Florida with 6 and Michigan with 3 lead the way in campaigning behavior for that year (Appendix J, Figure 2). In 2008, Texas and Oklahoma engaged in a campaign battle for inclusion into the Big 12 title game and again the data reflected their campai gning behavior with Oklahoma leading the way with 10 mentions, and Texas a close second with 7 (Appendix J, Figure 4) In 2007, five teams engaged in campaigning behavior with No. 10 ra nked Hawaii, leading the way with 4 (Appendix J, Figure 3). That year Hawaii went undefeated and felt it necessary to campaign for inclusion into a BCS bowl game. Campaigning may have b een necessary because Hawaii played in a nonBCS conference (Western Athle tic Conference) and needed an at-large bid from the BCS commissioners to gain inclusion in a BCS game. It could be sa id that the campaigning worked as Hawaii was given an at large bid to the 2008 Sugar bowl where they lost to Georgia 41-10 (ESPN.com, 2008). In regards to agenda-setting, the strong correlation between higher BCS ranking and campaigning may support the agenda-setting effect at the attribute or second-level in the transfer of topic attribute salience from the media to the voter. A coach, player, or team representative campaigning for his or her team may bring the a ttributes of why a team should be ranked higher to the attention of the voters through the media. If the media includes this campaigning, which 69

PAGE 70

contains team ranking attributes, the salience of these at tributes m ay be tran sferred to the voting publics contained within the BCS formula (Harris and Coaches poll). Because a strong positive correlation between campaigning coverage a nd higher team ranking was found, it could be theorized that campaigning may affect a voters ranking of a team. Since teams that campaigned ended the season ranked higher than teams that di d not campaign, it may be theorized that teams attribute salience was transferred from the media to the voting public of the BCS through campaigning and was reflected in the polls as a higher ranking. As stated above, these team attributes covered in the media may have been taken into consideration by the voting public and used to make a determination of team rank. Team attributes, which may persuade a voter to rank a team higher, would include a teams representative mentioning why his or her team deserves to be ranked higher. Some of these attributes ma y include strength of schedule, win-loss record, quality wins or loses, key play er injuries, team work ethic or desire to win, and BCS system flaws. Hypothesis Four A H4a: Teams engaging in positive campaigning behavior will be ranked higher in the final BCS poll than teams not engaged in positive campaigning behavior. The moderately strong correlation between positive campaigning coverage and higher BCS ranking shows the act of campaigning for hi gher ranking through the media by coaches or team representatives was perceived as a mos tly positive endeavor by the journalists and media outlets which chose to include cam paigning behavior in their articl es. The significance of this positive correlation between positive campaigning coverage and higher BCS ranking was further supported through the findings of no strong correlation between ne gative campaigning coverage and higher BCS ranking. The fact that campaigning was seen as an overall positive activity to be 70

PAGE 71

engaged in by college coaches or team represen tative shows that the pr actice of cam paigning is becoming a more widely accepted tactic for obtaining higher ranking. Theoretical Implications This study expands on agenda-setting theory in two ways. First, this study expands on the very limited research on agenda -setting as it is applied to sport. Other than the limited amount of previous studies (Fortunado, 2000, 2001, & 2008, Mitrook & Seltzer, 2005) the theory of agenda-setting has not been applied to sporti ng events and sports media coverage. Whereas Fortunado (2000, 2001, & 2008) looked at agenda-s etting effects on the National Basketball Association and the National Football League from a case study perspective, this study conducted quantitative research through the use of a content analysis in sport, much like Mitrook and Seltzer (2005). This quantitat ive approach is similar to th e research conducted in agendasetting studies of political campaigning (e .g., Kiousis, Bantimaroudis, & Ban, 1999; Golan & Wanta, 2001; Cho & Benoit, 2005). This application of a quantitative conten t analysis to college football communications holds implications for research in other areas of sport such as professional football, professional and college ba sketball, professional soccer, or any other high profile sport. But this is true only if the sport is high prof ile enough to generate enough news coverage to support a statistically significant content analysis. Second, this study expands agenda-setting th eory by incorporati ng a voter opinion source (Harris Interactive poll; USA Today/Coac hes poll) as a replacement for the traditionally used survey of the voting public in numerous prio r studies dealing with ag enda-setting effects in political campaigning (e.g., McCombs & Shaw, 1972; King, 1997; Kiousis & McCombs, 2004). Analogizing the BCS ranking system to a politi cal race, the Harris In teractive poll and USA Today/Coaches poll, act as a voter opinion poll due to the fact th at they are both ranking systems 71

PAGE 72

com prised of human voters. These voting publics have the potential to be influenced in their voting by media coverage of college football game s, similar to the voting public in a political race. In addition, even though this study did not examine the question of who sets the media agenda?, voter opinion polls may have the poten tial to influence media content. This study maybe reserved for future agenda building research as it applies to sport. Practical Implications In order to enlighten sports communications pr actitioners to the significance of the data retrieved in this study, a number of recommendations are presented. Due to the lack of previous empirical studies linking increased media cove rage to higher team ranking, positive media coverage to higher team ranking, and campaigning coverage to higher team ranking, this studys findings offer an initial view in to defining best practices for co llege football sports information directors and other college sports media relations practitioners. In the college sports information field it is no t an official position of a sports information director (SID) to lobby for a team s higher ranking. Yet, a college SID should be interested in this studys findings where increase d media coverage is strongly corr elated with higher ranking. If a SID wishes to contribute to his or her t eams success they should c onsider an attempt at generating additional media attentio n for their team through their rela tionships with journalists. This is especially true if the team in question is a football team who is in a close race for higher ranking and inclusion into a BCS bowl game or ultimately the BCS National Championship game. Additional media coverage of a team has the potential to attract the attention of active voters in either the Harris Interactive poll or th e USA Today/Coaches poll. If the football SID can generate more media coverage than an opposi ng SID, then this studys findings show that the team with the most media attention has a greate r chance of attaining the higher ranking. This 72

PAGE 73

technique could be em ployed by SIDs representing other college spor ts teams other than football, but in no other sport is the question of higher ranking as crucial to competing for a National Championship as it is in college football. College basketball SI Ds should use this technique of generating increased media coverage to produ ce higher ranking in seeding teams for the NCAA basketball tournament. Higher s eeded teams, in any tournament, theoretically, have an easier road to the Final Four and National Championshi p game being pitted ag ainst teams of a lesser quality. Since teams are seeded by human voter s, it could be argued, based on these findings, that teams could receive a higher seeding th rough increased media coverage which may influence the seeding committee. In terms of higher ranking being strongly correlated with positive media coverage, college football SIDs should be interested in mani pulating the valence of me dia coverage of their team. Through their continuing re lationships with national and lo cal journalists, a college SID has the potential to influence the valence of me dia coverage of their team. Through second-level agenda-setting effects, this positive media coverage has the potential to in fluence voters in both the Harris and Coaches polls by convincing voters of a teams worthiness of a higher ranking. Positive media coverage of a team may help to transfer the salience of positive attributes of a team from the media to the voting public. If th e media chooses to focus on the positive facets of a team, such as win-loss record, qu ality wins, or team drive, then the transfer of salience of these attributes to the voting public may influence th em to rank that team hi gher than a team not receiving as much positive media attention. One of a SIDs principle media relations responsibilities is the active attempt to positively fr ame his or her teams image. This framing is accomplished through the dissemination of specia lly selected story ideas and by properly handled crisis situations. The findings of this study should encourage a SID to value the impact 73

PAGE 74

with which his or her po sitive relationships w ith the media may have on team ranking. Strong positive media relations can result in a positiv e media perception of a team. These strong positive media relationships should be fostered by a college football SID through exceptional cooperation with media which woul d include an increase in appr oved interview requests, more readily available statistical records, and mo re complete information transparency. The practical implications of the data, wh ich show increased campaigning coverage being correlated with higher team ranking, is useful to both college football SIDs and college football coaches alike. This strong correlation is eviden ce that campaigning for ones team does hold the possibility of influencing voters to rank a team higher. Even though this evidence does not definitively prove that campaigning works, it does not show that campaigning hurts. The practice of campaigning by college football coaches, players, or other team representatives has become more acceptable in recent years, as seen in its increased employment by respected head coaches such as Floridas Urban Meyer, Texass Mack Brown, and Oklahomas Bob Stoops. The recent increased use of campaigning may cas t the perception that not only is campaigning effective but it may also be a necessity. In fact, there seems to be an underlying tone among college coaches that campaigning has now become a necessary evil in the war for higher BCS ranking. The results of this study should be used by college football coaches, or any team representative, includ ing the teams SID, to further support th is perception. For coaches, players, and SIDs, this study has determined that campaigning is a very useful tool necessary to winning a football National Title. In addition, these campaigning techni ques should be carried over to college basketball, where head coaches and othe r team representatives may employee them in attempts at garnering higher s eeding in the NCAA tournament 74

PAGE 75

Limita tions While this study produced some insight into agenda-settings aff ect on college football ranking, the research was subject to a few of limitations. Most of the limitations in this study were related to the design and execution of th e chosen methodology. The first limitation deals with the abbreviated time frame of the study. The study looked at the only four college football seasons in which the BCS determined the college football National Championship through the use of the Harris Interactive poll, a human poll ca pable of being persuaded by outside influences such as news reports, expert speculation, and eq uivalent polls. During these four years, the Harris Interactive poll and the USA Today/Coaches poll were used as tw o-thirds of the BCS determinant. The time frame of the study could have been expanded to include the years of 1998 to 2004, in which time the Associated Press Writers poll, in lieu of the Harris Interactive poll, was in use by the BCS. Despite the years between 1998 and 2004, where the AP and USA Today/Coaches polls were counted only as a 50 percent factor of the BCS calculation (BCSfootball.org, 2008), the expanded time fram e may have added more reliability and credibility to the study a nd possibly changed the signi ficance of the findings. The second limitation concerns the sample si ze in regards to the medium analyzed for content. The content analysis used was lim ited to only four sources which included the Associated Press, The New York Times The Sporting News and the USA Today. Only four sources were selected to keep the sample size ma nageable for the coders. The researchers would have preferred to use the s ources of ESPN.com and SportsIllustrated.com due to their comprehensive college football content and industry respectabili ty, instead of the Associated Press (AP) and The Sporting News (TSN). The AP and TSN were used because of their accessibility through LexisNexis which supplie d both a word count and a unique tracking 75

PAGE 76

num ber for each article referenced, whereas ESPN .com and SportsIllustrated.com did not. In addition, the content analyzed was limited to prin t articles. News coverage concerning college football was not tracked using television or ra dio shows and most importantly not web-only based articles. Again, even though a large amount of the articles included in th is study may have appeared in the print and online versions of the publications, a la rge amount of information and news coverage concerning college football appears on web-only based sports news cites, especially ESPN.com and SportsIllustrated.com, and these were not included in the study. Finally, the last limitation of this study concerns media bias. It could be argued that some college teams receive more coverage based on who they are. Schools which have a long tradition of winning in college f ootball may receive more coverage from the media based on the laws of supply and demand. When a team is succ essful over along amount of time, a fan base is built which turns into a demand for information about that team. Teams such as USC, Oklahoma, and especially Notre Dame have large fan bases and may generate more total and possibly more positive media coverage than othe r less traditionally successful college football schools. This media bias may skew the numbers to reflect more media coverage for teams ranked lower in the BCS final poll. In additi on, schools which may have been successful in a recent prior year may garner more media covera ge based on a holdover of media expectation for that team to succeed again, such as Miami in 2005, and Auburn in 2005 and 2006. Another cause of increased total media coverage and positive media coverage may deal with a media underdog bias. It could be argued that the media is fascinated with surprise success stories. These stories are interesting to the media beca use they are interesting to the public. The increased media coverage of a team, which is not a regular finisher in the final BCS top ten, may receive more total and favorable media coverage simply because it is an underdog, as in the 76

PAGE 77

cases of Louisville and B oise State in 2006, Missouri and Kansas in 2007, and Texas Tech in 2008. Future Research Future research should include a larger samp le size in a number of different respects. First, future researchers may consider incorp orating other forms of media in their content analysis. The addition of televi sion and radio transcripts may be useful in broadening the scope of opinion and influence on pollste r voting. In addition, and most importantly, for a sports news media content analysis to be more legitimized, future researchers shoul d include web-only based sources such as ESPN.com and SI.c om. It could be theorized that voters in the Harris Interactive and USA Today/Coaches poll obtain a good deal of their information about teams which they vote on from Internet sources, whether that info rmation is presented in statistical, news, or feature form. Future research should also include a surv ey of voters in the Harris Interactive and USA/Today Coaches polls. This survey combined with the results of the polls themselves could act as a more informative voter opinion component of the experiment. Voters could supply the researchers with valuable information on why th ey voted for one team over another. Factors such as sports news media influence and sim ilar polling influence could be measured based on voters survey responses. Next, future re search could also build off Fourtnado (2000) in the area of sports agenda building research by survey ing and interviewing college football sports information directors. For example, SIDs could be asked about their practices for generating news content specifically for higher team ranking. Additionally, SIDs could be surveyed on their direct voter communication prac tices with intent to influe nce pollsters team rankings. Information on whether or not campaigning or incr eased favorable media coverage could have a 77

PAGE 78

positiv e effect on team ranking could be valuable information to a college football SID as Steve McClain, sports information director for the Univer sity of Florida football team, confirmed in the aforementioned personal interview conducted specifically for this study (McClain, 2008). Finally, future research could include teams wh ich did not finish in the top ten BCS poll. This study looked at teams which were ranked only in the top ten of th e final BCS poll in the years between 2005 and 2008. In total, only 24 team s were tracked during th at period of time. The use of these 24 teams brought into question whether or not more team should have been included. For example, a team may have started the year in the BCS Top Ten and maintained a top ten ranking all season until the final one or two editions of the poll (Arkansas 2006, Arizona State 2007), but because this team did not finish in the top ten in the final poll it was not included. The statistical data c oncerning a team in this situati on could add important information in regards to whether or not an increase or decr ease in media coverage co uld have affected their drop in rank. Future studies could include track ing all teams included in both the first and last editions of the poll and not lim iting the study to tracking teams th at generally moved up in the polls or maintained a top ten ranking, and not leavi ng out teams which dropped out of the final BCS ranking. Conclusion This study attempted to expand on previous agenda-setting research, bo th at the first and second-level, by conducting a quantitative content an alysis in the previous ly un-researched area of college football ranking a nd sports information communicati ons. By measuring the amount and valence of college football media coverage this study showed that a strong correlation does exist between increased positive media coverage and higher college football poll ranking. Furthermore, by measuring amount and valence of campaigning behavior this study also found 78

PAGE 79

a strong co rrelation between increased positiv e campaigning coverage and higher college football poll ranking. In the tradition of political campaign agenda -setting research, this study employed the use of a content analysis combined with a poll of the voting public. Yet, unlike previous agendasetting research this study expa nded on past research by employing a voter opinion poll in lieu of a public opinion poll in the form of the BCS poll, Harris Interactive poll, and USA/Today Coaches poll. In addition, this study broke new ground into agenda-setting research by conducting a quantitative content an alysis of media coverage and campaigning as it pertains to BCS college football ranking. The findings of this study will benefit sports communications practitioners, especially college sports information directors, in develo ping a wider spectrum of understanding of their field and the relationships which they foster with media personnel. The evidence in this study of higher team ranking through the ag enda-setting effects of increased positive media coverage should be examined by college sports communicati ons practitioners and taken into consideration during the construction of con tingency models for increasing team poll rankings and overall team image. 79

PAGE 80

80 APPENDIX A BCS CAMPAIGNING STUDY CODE BOOK Source: All articles from the four sources will be recorded to k eep track of publication and for future reference. The four source s will be recorded with the initials Associated Press (AP) New York Times (NYT) The Sporting News (TSN) and USA Today (USA) Date range: 10/15/2005 12/3/2005, 10/15/2006 12/3/2006, 10/14/2007 12/2/2007, 10/19/2008 12/7/2008. Article ID number: Every article in the sample should receive a unique identification number that consists of the date and the first three le tters in the authors last name in this format: monthdayyearXXX. For example, an article that was published on October 19, 2008 and was written by John Smith would have the following identifier: 10192008SMI. If more than one article in the sample is published on the same date and written by the same author, the ID number will be followed by a, b, c and so on in chronological order. Coder: Coders name will be recorded to keep track of individuals work and the ability to reference the coder for further information. Date: Record the date in the mm/dd/yyyy format. Length: Number of words contained in article. Headline: The name of the story, which will help to identify the article at a later time. Teams: Of all 40 possible positions in the top te n of the final BCS poll in the years 2005 2008, only 24 teams filled those 40 positions with some teams appearing multiple times. Those teams will be assigned a number in alphabetical order 1 through 24. Those teams in alphabetical order and their corresponding numbers are, 1 Alabama, 2 Auburn, 3 Boise State, 4 Florida, 5 Georgia, 6 Hawaii, 7 Louisiana State, 8 Louisville, 9 Kansas, 10 Miami (FL), 11 Michigan, 12 Missouri, 13 Notre Dame, 14 Ohio State, 15 Oklahoma, 16 Oregon, 17 Penn State, 18 Southern California, 19 Texas, 20 Texas Tech, 21 Utah, 22 Virginia Tech, 23 West Virginia, 24 Wisconsin (Page 3). I. Word Count Number of Words in each article is coded as either 1. short (0-300 words), 2. medium (301-600 words), or 3. long (601 words or more). II. Article Type Articles are coded for Type If the article is of an editori al nature or expresses an opinion then the article will be considered a Feature article and coded as 1. If the article only contains facts then that ar ticle will be considered a News article and coded as 2. If the article is neither a feature or a news article o r a it is a combination of both then that article will be considered N/A and be coded as 3. (Feature = 1, News = 2, N/A = 3). III. Team Presence

PAGE 81

Artic les are coded for the Presence of a team. The team must be mentioned by name only once. Teams will be determined by the top ten teams in each final BCS poll (2005, 2006, 2007, & 2008). Team will be indicated by number assignment and presence of team will be coded by recording the corresponding number of team in appropriate spaces on code sheet. IV. Article Attitude Toward Team Each article is coded for its general Attitude Toward a Team If a pertinent team is present in the article then the tone toward each team mentioned in the article is coded as being either 1. Positive, 2. Negative, or 3. Neutral V. Position in Race A teams Position in the BCS will be recorded. That position will be recorded on a scale of 0 to 26. If the team is mentioned as leading the race, the number one team, or in first place then that would justify the team be ing recorded as 1. If the team is mentioned as a close second, the other BCS title game team, or the runner up, then that would warrant the that team being coded as two, a nd soon and so forth of teams three through 25. If the team is mentioned as a BCS contender with out position in the BCS standing, then it will be coded as 26. If the team is men tioned, but not as a BCS contender, then that will be coded as 0. Teams 1 through 25 correspond to the rankings of the teams in each week, in any of the th ree polls (BCS, Harris Interactive, and USA Today/Coaches), in each indivi dual years race (Page 4). VI. Article Focus Articles are coded for conten t. Content is coded by Team Focus (teams name is mentioned in the headline, the teams players and coaches were interviewed, or if a team is mentioned more often than another). If a t eam is the main focus of the article then that team is record by assigned number (1-24) Note : It is possible for more than one team to be the focus of the article. VII. Presence of Campaign Language References to campaign langua ge are coded as reflecting Presence. The term campaign language will be operationally defined as any attempt by a coach or university representative to argu e, state, or reinforce why his or her team should have higher ranking in the BCS poll. If campaign language appear s in the text then the appearance of the language in the article will be coded by recording the team corresponding number (1-24) which is being campaigned for. VIII. Article Attitude Toward Campaigning Each article is coded for its general A ttitude T oward Campaigning Behavior If a pertinent team is identified as being campai gned for (Presence) then the tone of that article towards campaigning will be coded as being either 1. Positive, 2. Negative, or 3. Neutral 81

PAGE 82

APPENDIX B BCS C AMPAINGING STUDY CODE SHEET Article ID #: ___________________ Headline:_____________________________ Coder: ___________________ _____________________________ Source: ___________________ _____________________________ Date: ___________________ Length: _____________________________ I. Word Count Circle One 1. Short 2. Medium 3. Long II. Article Type Circle One 1. Feature 2. News 3. Blog 4. N/A III. Team Presence, IV. Attitude toward team, & V. Position in Race Team/Presence Attitude Towards Team Position No. 1-24 Circle one 0-26 BCS Harris Coaches _______ 1. Positive 2. Negative 3. Neutral ______ ______ ______ _______ 1. Positive 2. Negative 3. Neutral ______ ______ _______ _______ 1. Positive 2. Negative 3. Neutral ______ ______ _______ _______ 1. Positive 2. Negative 3. Neutral ______ ______ _______ _______ 1. Positive 2. Negative 3. Neutral ______ ______ _______ _______ 1. Positive 2. Negative 3. Neutral ______ ______ _______ _______ 1. Positive 2. Negative 3. Neutral ______ ______ _______ _______ 1. Positive 2. Negative 3. Neutral ______ ______ _______ _______ 1. Positive 2. Negative 3. Neutral ______ ______ _______ _______ 1. Positive 2. Negative 3. Neutral ______ ______ _______ VI. Article Focus Circle As Many As Apply 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 VII. Presence of Campaign Language & VIII. Attitude Toward Campaigning Team/Presence Article Attitude Toward Campaigning No. 1-24 Circle one _________ 1. Positive 2. Negative 3. Neutral _________ 1. Positive 2. Negative 3. Neutral _________ 1. Positive 2. Negative 3. Neutral _________ 1. Positive 2. Negative 3. Neutral _________ 1. Positive 2. Negative 3. Neutral _________ 1. Positive 2. Negative 3. Neutral _________ 1. Positive 2. Negative 3. Neutral _________ 1. Positive 2. Negative 3. Neutral _________ 1. Positive 2. Negative 3. Neutral _________ 1. Positive 2. Negative 3. Neutral 82

PAGE 83

APPENDIX C ALPHABE TICAL TEAM NUMBERING 1. Alabama 2. Auburn 3. Boise State 4. Florida 5. Georgia 6. Hawaii 7. Louisiana State 8. Louisville 9. Kansas 10. Miami (FL) 11. Michigan 12. Missouri 13. Notre Dame 14. Ohio State 15. Oklahoma 16. Oregon 17. Penn State 18. Southern California 19. Texas 20. Texas Tech 21. Utah 22. Virginia Tech 23. West Virginia 24. Wisconsin 83

PAGE 84

APPENDIX D TOP TEN TEAMS IN THE FINAL BCS POLL IN THE YEARS 2005 TO 2008 YEAR 2005 2006 2007 2008 1. Southern Cal Ohio State Ohio State Oklahoma 2. Texas Florida LSU Florida 3. Penn State Michigan Virginia Tech Texas 4. Ohio State LSU Oklahoma Alabama 5. Oregon Southern Cal Georgia Southern Cal 6. Notre Dame Louisville Missouri Utah 7. Georgia Wisconsin Southern Cal Texas Tech 8. Miami (FL) Boise State Kansas Penn State 9. Auburn Auburn West Virginia Boise State 10. Virginia Tech Oklahoma Hawaii Ohio State 84

PAGE 85

APPENDIX E TOP TEN TEMAS IN THE FINAL HARRIS IN TERACTIVE POLL IN THE YEARS 2005 TO 2008 YEAR 2005 2006 2007 2008 1. Southern Cal Ohio State Ohio State Florida 2. Texas Florida LSU Oklahoma 3. Penn State Michigan Oklahoma Texas 4. Ohio State LSU Georgia Alabama 5. Notre Dame Louisville USC Southern Cal 6. Oregon Wisconsin Virginia Tech Penn State 7. Auburn USC Missouri Utah 8. Georgia Oklahoma Kansas Texas Tech 9. Miami (FL) Boise State West Virginia Boise State 10. LSU Auburn Hawaii Ohio State 85

PAGE 86

APPENDIX F TOP TEN TEAMS IN THE FINAL USA TODAY /COACHES POLL IN THE YEARS 2005 TO 2008 YEAR 2005 2006 2007 2008 1. Southern Cal Ohio State Ohio State Oklahoma 2. Texas Florida LSU Florida 3. Penn State Michigan Oklahoma Texas 4. Ohio State LSU Georgia USC 5. Oregon Wisconsin Virginia Tech Alabama 6. Notre Dame Louisville USC Penn State 7. Auburn USC Missouri Utah 8. Georgia Oklahoma Kansas Texas Tech 9. Miami (FL) Boise State West Virginia Boise State 10. LSU Auburn Hawaii Ohio State 86

PAGE 87

APPENDIX G TOTAL ME DIA MENTIONS OVER ALL FOUR YEARS Team Total Media Mentions Years USC 171 4 Ohio State 157 4 Texas 125 2 Oklahoma 116 3 Florida 97 2 Virginia Tech 78 2 LSU 66 2 Penn State 65 2 Michigan 51 1 Texas Tech 51 1 Auburn 49 2 Alabama 44 1 Notre Dame 41 1 Georgia 40 2 Miami 40 1 Kansas 37 1 Missouri 37 1 West Virginia 32 1 Louisville 30 1 Boise State 25 2 Utah 25 1 Hawaii 20 1 Oregon 16 1 Wisconsin 5 1 87

PAGE 88

APPENDIX H TOTAL TE AM MEDIA COVERAGE Figure 1: 2005 Team Coverage Team BCS Rank Total Mentions Positive Negative USC 1 71 22 2 Texas 2 67 17 1 Va. Tech 10 58 11 4 ND 6 41 12 1 Miami 8 40 6 1 Penn State 3 34 6 1 Ohio State 4 29 3 1 Georgia 7 29 7 2 Auburn 9 25 3 2 Oregon 5 16 2 0 Figure 2: 2006 Team Coverage Team BCS Rank Total Mentions Positive Negative Ohio State 1 60 18 2 Florida 2 55 13 3 Michigan 3 51 10 2 USC 5 46 11 9 Louisville 6 30 11 1 Auburn 9 24 5 1 Boise St. 8 18 7 2 LSU 4 14 4 0 Oklahoma 10 12 2 0 Wisconsin 7 5 1 1 Figure 3: 2007 Team Coverage Team BCS Rank Total Mentions Positive Negative LSU 2 52 7 3 Ohio State 1 48 9 8 Oklahoma 4 43 7 1 Missouri 6 37 9 1 Kansas 8 37 9 0 W. Va. 9 32 8 1 USC 7 26 3 5 Va. Tech 3 20 2 0 Hawaii 10 20 5 0 Georgia 5 11 1 1 88

PAGE 89

Figure 4: 2008 Team Coverage Team BCS Rank Total Mentions Positive Negative Oklahoma 1 61 6 2 Texas 3 58 7 0 T. Tech 7 51 8 0 Alabama 4 44 5 0 Florida 2 42 7 0 Penn State 8 31 3 1 USC 5 28 2 0 Utah 6 25 4 0 OSU 10 20 3 1 Boise St. 9 7 1 1 89

PAGE 90

APPENDIX I TOTAL CAMPAIGNING OVE R ALL FOUR YEARS Teams Total Oklahoma 13 Texas 10 Florida 6 USC 6 Hawaii 4 Ohio State 4 Michigan 3 Auburn 2 Georgia 2 Va. Tech 2 Boise St. 1 Oregon 1 Utah 1 W. Va. 1 Total 56 90

PAGE 91

APPENDIX J TEAM CAMPAIGNING YEAR B Y YEAR Figure 1 Figure 2 2005 BCS Team Cmpn. 1 USC 4 2 Texas 3 10 Va. Tech 2 9 Auburn 1 7 Georgia 1 5 Oregon 1 Total 12 2006 BCS Team Cmpn. 2 Florida 6 3 Michigan 3 1 Ohio State 2 8 Boise St. 1 9 Auburn 1 10 Oklahoma 1 Total 14 Figure 3 Figure 4 2007 BCS Team Cmpn. 10 Hawaii 4 1 Ohio State 2 4 Oklahoma 2 7 USC 2 9 W. Va. 1 Total 11 2008 BCS Team Cmpn. 1 Oklahoma 10 3 Texas 7 6 Utah 1 Total 18 91

PAGE 92

APPENDIX K SCATT ERPLOTS FOR PPMC Figure 1 Figure 2 92

PAGE 93

Figure 3 Figure 4 93

PAGE 94

Figure 5 Figure 6 94

PAGE 95

Figure 7 Figure 8 95

PAGE 96

96 REFERENCES Berger, B. (2001). Private Issues and Public Po licy: Locating the Corporate Agenda in AgendaSetting Theory. Journal of Public Relations Research 13(2), 91-126. Berelson, B. (1952). Content Analysis In Communication Research. New York, NY: Straford Press, Inc. Blaudshun, M. (2003, Decem ber, 8). On College Football Mark Blaudschun; BCS Match-up SetWith One Big Problem. The Boston Globe, p. D1. Blaudshun, M. (2007, January, 9). Ohio St. trou nced in title game. The Boston Globe. p. E1. Bosman, J. (2006). Sporting News is Formally Seeking a Buyer. Retrieved Jan. 10, 2009, from http://www.nytim es.com/2006/02/21/ business/media/21mag.html?_r=1 Bowl Championship Series. (2003 ). Bowl Championship Series FAQ. Retrieved Sept. 2, 2008, from http://www.bcsfootball.org/bcsfb/faq Bowl Cha mpionship Series. (2003). Th e BCS is. Retrieved Apr. 10, 2008, from http://www.bcsfootball.org/bcsfb/definition Bowl Cha mpionship Series in Association with Fox. (2008). Bcsfootball.org. Retrieved Nov. 25, 2008, fro m http://www.bcsfoot ball.org/bcsfb/history. Business Network. (2000). ESPN.com Celebrates Five Years as the Most Popular Sports Web Site On April 1; Since 1995, ESPN.com Ha s Pioneered The Way Fans Consume Sports Information and Entertainment. Retrieved Dec. 20, 2008, from http://findarticles.com /p/articl es/mi_m0EIN/is_2000_March_30/ai_60951313 Carey, J. (2007). Man behind creation of BC S pleased with results. Retrieved Jun. 4, 2008, from http://www.usatoday.com /sports/college /football/bowls/2007-12-06bcs2 football_N.htm?csp=34 Cohen, B. (1963). The Press and Foreign Policy. Princeton, NJ: Princeton University Press. Cohen, S. S. (1985, Decem ber 9). Happy 50th birthday to the Heisman Trophy, college footballs Oscar. Sports Illustrated 63, 6-7. CollegeFootball.com. (2009). College Football National Champions. Retrieved Jan. 13, 2009, fro m http://www.collegefootballpoll.com/champions_national.html Craft, S. & Wanta, W. (2004). U.S. Public Concer ns in the Aftermath of 9-11: A Test of Second Level Agenda-Setting. International Journal of Public Opinion Research 16(4), 456-463.

PAGE 97

Curtin, P. A. (1999). Reevaluating Public Rela tions Inform ation Subsidies: Market-Driven Journalism and Agenda-Building Theory and Practice. Journal of Public Relations Research 11(1), 53-90. Gandy, O. (1991). Beyond Agenda-Setting. In D. L. Protess & M. McC ombs (Eds.), Agenda Setting: Readings on Media, P ublic Opinion, and Policymaking (pp. 263-275). Hillsdale, NJ: Lawrence Erlbaum Associates. Dearing, J. W., & Rogers, E. M. (1996). Agenda Setting. Thousand Oaks, CA: Sage Publishing. Denha m, B. E. (1997). Sports Illustrated, The "W ar on Drugs," and the Anabolic Steroid Control Act of 1990. Journal of Sport & Social Issues 21(3), 260-273. Denha m, B. E. (2004). Sports Illustrated, the ma instream press and the enactment of drug policy in Major League Baseball. Journalism 5(1), 51-68. Entm an, R. M. (1993). Framing: Toward Clarification of a Fractured Paradigm. Journal of Communication, 43, 51-55. ESPN.com (2008). Hawaii's nearly $4.4M payout sets record for non-BCS school. Retrieved Mar. 25, 20 09, from http://sports.espn.go.com/ncf/news/story?id=3350323 ESPN.com.. (n.d.). 2003 NCAA Football Rankings Final (Jan. 4). Retrieved Apr. 10, 2008, from http://sports.espn.go.com/n cf/rankingsindex?seasonYear=2003 ESPN.com.. (n.d.). 2004 NCAA Football Rankings Final (Jan. 4). Retrieved Apr. 10, 2008, from http://sports.espn.go.com/n cf/rankingsindex?seasonYear=2004 ESPN.com.. (n.d.). 2006 NCAA Football Rankings Final (Jan. 9). Retrieved Apr. 10, 2008, from http://sports.espn.go.com/n cf/rankingsindex?seasonYear=2006 ESPN.com.. (n.d.). 2008 BCS Standings November 30, 2008. Retrieved Dec. 10, 2008, from http://sports.espn.go.com/n cf/BCSStandings?week=7 Field, A. (2000). Discovering statistics using SPSS for windows Newbury Park, CA: Sage. Fortunato, J. A. (2000). Public Relations Stra tegies for Creating Mass Media Content: A Case Study of the National Basketball Association. Public Relations Review 26(4), 481-497. Fortunato, J. A. (2001). The Ultimate Assist: The Relationship and Broadcast Strategies of the NBA and Te levision Networks. Cresskill, NJ: Hampton Press, Inc. Fortunato, J. A. (2008a). NFL Agenda-setting: The NFL Programm ing Schedule; A Study of Agenda-setting. Journal of Sports Media 3(1), 27-48. Fortunato, J. A. (2008). Restor ing a reputation: The Duke University lacrosse scandal. Public Relation s Review 34, 116-123. 97

PAGE 98

Funk, D. (2007). History of the Sporting News: The Early Years. Retrieved Jan. 10, 2009, from http://www.associatedcontent.com /article/ 264801/the_history_of_sporting_news_the_early .html Gardiner, A. (2005, January, 5). 13-0 season, Sugar Bowl win hard to top, Auburns Tuberville says. USA Today p. 6C. Ghorpade, S. (1986). Agenda Setting: A Te st of Advertising's Neglected Function. Journal of Advertising Research 23-27. Golan, G. L., Kiousis, S. K., & McDaniel, M. L. (2007). Second-Level Agenda Setting and Political Ad vertising: Investigating the transf er of issue and attri bute saliency during the 2004 US presidential election. Journalism Studies 8(3), 432-443. Golan, G. & W anta, W. (2001). S econd-Level Agenda Setting in the New Hampshire Primary: A Comparison of Coverage in Three Newspapers and Public Perceptions of Candidates. Journalism and Mass Communication Quarterly 78(2), 247-259. Harris Interactive Inc. (2008). Harrisinteractive.com. Retrieved Nov. 25, 2008, from http://www.harrisin teractive.com/news/bcspoll.asp Holsti, O. R. (1969). Content Analysis for the Social Sciences and Humanities. Philippines: Addison-W esley Publishing Company Inc. Kilgore, A. (2006, December, 4). A topsey-turve y picture; USCs loss, Floridas win muddles BCS title match-up. The Washington Post p. E01. Kiousis, S. & McCombs, M. (2004). Agenda-Setti ng Effects and Attitude Strength: Political Figures During the 1996 Presidential Election. Communication Research 31(1), 36-57. Kiousis, S. (2005). Compelling Argum ents and A ttitude Strength: Exploring the Impact of Second-Level Agenda Setting on Public Opin ion of Presidential Candidate Images. The Harvard International Journal of Press/Politics 10(2), 3-27. Kiousis, S., Bantim aroudis, P., & Hyun, B. (1999). Candidate Image Attributes: Experiments on the Substantive Dimension of Second Level Agenda Setting. Communication Research 26(4), 414-428. Kiousis, S., Mitrook, M., W u, X., & Seltzer, T. (2006). Firstand Second-Level AgendaBuilding and Agenda-Setting Effects: Exploring the Linkages Among Candidate News Releases, Media Coverage, and Public Opin ion During the 2002 Florida Gubernatorial Election. Journal of Public Relations Research 18(3), 265-285. King, P. (1997). The Press, Candidate Im ages, a nd Voter Perceptions. In M. McCombs, D. L. Shaw, & D. Weaver (Eds.), Communication and Democracy (pp. 29-40). Mahwah, New Jersey : Lawrence Erlbaum Associates. 98

PAGE 99

Kosicki, G. M. (1993). P roblems and O pportunities in Agenda-Setting Research. Journal of Communication, 43(2), 100-127. Klem z, P. (2005). AP removes poll from BCS formula. Retrieved Sept. 21, 2008, from http://badgerherald.com /sports /2005/01/17/ap_removes_poll_from.php Krippendorff, K. (1980). Content Analysis: An Intr oduction to its Methodology. Beverly Hill, CA: Sage Publications Inc. LexisNexis. (2008). LexisNexis Academic & Library Solutions. Retrieved Dec. 18, 2008, from http://a cademic.lexisnexis.com/contact.aspx Lieberman, D. (2006). Newspaper sales dip, but websites gain. Retrieved Dec. 20, 2008, from http://www.usatoday.com /money/media/ 2006-05-08-newspaper-circulation_x.htm Lippmann, W. (1991). The World Outs ide and the Pictures in Our H eads. In D. L. Protess & M. Maxwell (Eds.), Agenda Setting: Readings on Media, Public Opinion, and Policymaking (pp. 5-15). Hillsdale, N J: La wrence Erlbaum Associates. Lopez-Escobar, E., Llamas, J. P., & McComb s, M. (1998). Agenda Setting and Community Consensus: First and Second Level Effects. International Journal of Public Opinion Research 10(4), 335-348. Maisel, I. (2008). Breaking the huddle on the po litics of the B ig 12 South. Retrieved December 10, 2008, from http://sports.espn.go.com/ncf/notebook?page=iform0812 Mandel, S. (2007). Bowls, Polls, & Tattered Souls: Tackling the Chaos and Controversy That Reign over College Football. Hoboken, NJ: John Wiley & Sons, Inc. Martinich, J. (2002). College Football Rankings: Do the Computers Know Best? Interfaces 32(5), 85-94. McCom bs M. (1992). Explorers and Surveyor s: Expanding Strategies for Agenda-Setting Research. Journalism Quarterly 69(4), 813-824. McCom bs, M., & Evatt, D. (1995). Los temas y los aspectos: Explorando una nueva dimension de la agenda setting. Comunicacion y Sociedad, 8, 7-32. McCombs, M., Llamas, J. P., Lopez-Escobar, E ., & Rey, F. (1997). Candidate Images in Spanish Elections: Second-Level Agenda-Setting Effects. Journalism and Mass Communication Quarterly 74(4), 703-717. McCom bs M., Lopez-Escobar, E., & Llamas, J. P. (2000). Setting the Agenda of Attributes in the 1996 Spanish General Election. Journal of Communication, 77-92. McCom bs, M. (2004). Setting the Agenda: The Mass Media and Public Opinion. Malden, MA: Polity Pre ss. 99

PAGE 100

McCom bs, M. (2005). A look at ag enda-setting: Past, present a nd future. Journalism Studies, 6(4), 543-557. McCombs, M.E., & Shaw, D.L. (1972). The ag enda-setting function of mass media. Public Opinion Quarterly, 36(2), 176-187. Mitrook, M., & Seltzer. (2005, May). The Agenda-Setting Influence of Expert Opinion on Media Coverage of the Heisman Trophy race. Paper presented at the J ournalism Studies Division for the International Communication Associ ation annual conference, New York, NY. Nimmo, D., & Savage, R. L. (1976). Candidate s and their Images: Concepts, methods, and findings. Santa Monica, CA: Goodyear. Ohl, C. M., Pincus, J. D., Rimmer, T., & Harri son, D. (1995). Agenda Building Role of News Releases in Corporate Takeovers. Public Relations Review 21(2), 89-101. Pool, I. D. (1959). Trends in Content Analys is T oday: A summary. In I. D. Pool (Ed.), Trends in Content Ana lysis (pp. 189-233). : Univers ity of Illinois P ress. Prisbell, E. (2006, December 4). Florida gets nod for the title game; SEC champion leapfrogs Michigan. The Washington Post p. E01. Protess, D. L., & McCombs, M. (1991). Agenda Setting: Readings on Media, Public Opinion, and Policym aking. Hillsdale, NJ: Lawrence Erlbaum Associates. Riffe, D., Lacy, S., & Fico, F. (1998). Analyzing Media Messages: Using Quantitative Content Analysis in Research. Mahwah, New Jersey: Lawrence Erlbaum Associates. Roberts, M. (1997). The Press, Candidate Im ages, and Voter Perceptions. In M. McCombs, D. L. Shaw, & D. Weaver (Eds.), Communication and Democracy: Exploring the Intellectual Frontiers in Agenda-Setting Theory (pp. 85-96). Mahwah, New Jersey: Lawrence Erlbaum Associates. Rom ano, J. (2006, December 3). Entitled to more than laptops whim. St. Petersburg Times. p. 1C. SEC title likely no help for Aubur ns BCS bid. (2004, December 5 ). St. Petersburg Times p. 8C. Retrived from, http://www.sptimes.com/2004/12/05/Sports/SEC_title_likely_no_h.shtml Severin, W., & Tankard, J. (1979). Communications Theories. New York, NY: Hastings House. Schiff, F. & Reiter, G. (2004). "Mea suring Intercoder Reli ability: Scott' s Pi Reconsidered" Paper presented at the annua l meeting of the American Sociological Association, Hilton San Francisco & Renai ssance Parc 55 Hotel, San Francisco, CA, Online <.PDF>. 2009-03-04 from http://www.allacademic.com/meta/p110405_index.html 100

PAGE 101

SI.com (2005). No. 1? There's no doubt: It's USC Trojans rout S ooners 55-19 in Orange Bowl to win second straight title Retrieved Apr. 10, 2008, from http://sportsillustrated.cnn.com/2005/football/n caa/specials/bowls/2004/01/04/orange.bowl /index.html Texassports.com. (2008). Football: Texas A&M Quotes/Mack Brown. Retrieved Dec. 10, 2008, fro m http://www.texassports.com/s ports/m-footbl/recaps/112808aac.html Thamel, P. (2004, December 22). Associated Pre ss Football Poll Is Pulled From B.C.S. Equation. The New York Times. Retrieved, (2008, November 25) http://www.nytimes.com/2004/12/22/sports/ncaa football/22bcs.html?scp=1&sq=ap%20bcs %202004&st=cse Thamel, P, (2006, December 4). Florida pa sses Michigan for spot in title game. The New York Times p. D1. The Bowl Championship Series. (2005, July 11). BCS announces standings formula (press release). Retrieved Aug. 1, 2008, from http://harrisinteractive.com/ne ws/newsletters/clientnews/BCS_2005.pdf The Heisman Memorial Trophy. (Accessed April 21, 2009). Heisman balloting-how it works [Online]. Available: www.heisman.com/balloting.html Turk, J. V. & Franklin, B. (1987). Informati on Subsidies: Agenda-Setting Traditions. Public Relations Review 29-41. UMTERPS.CSTV.com. (1998). New Formula for Football C hampionship Announced Top bowl game to be chosen by polls, computers, formulas. Retrieved Apr. 10, 2008, from http://umterps.cstv.com/sports /m-footbl/specrel/061098aaa.html Wahl, G. (2001). Frenzied State: Oregon and Oregon St ate both have Heisman candidates and national title hopes, so their cute little rivalry has sudde nly turned very, very serious. Retrieved Mar. 25, 2009, from http://vault.sportsillustrated.cnn.com/vault/article/magazine/MAG1023180/index.htm Walters, T. N., Walters, L. M., & Gray, R. (1196). Agenda Building in the 1992 Presidential Campaign. Public Relations Review 22(1), 9-24. Weber, R. P. (1990). Basic Content Analys is: Second Addition. Beverly Hill, CA: Sage Publications Inc. Weiss, D. (2003, December, 8). USC gets life o ut in computer cold Olka. (2) vs. LSU (3) to settle whos No. 1. Daily News p. 60. Weiss, D. (2006, December 4). As usual, its BC -Mess. Meyer campaign gets Florida shot. Daily News p. 66. 101

PAGE 102

Yanda, S. (2008, Nove mber 27). Time for the Real Campaign: the BCS; Sooners' Stoops Lobbying Hard, But Does It Help? The Washington Post, p E03. 102

PAGE 103

103 BIOGRAPHICAL SKETCH Todd David Lawhorne was in 1975, in Ft. Lauderd ale, Florida. After completing his high school training at Gainesville Hi gh School, Todd went on to earn his Associate of Arts degree from Santa Fe Community College in 1996. In 2000, Todd earned his Bachelor of Arts degree from the Theatre Conservatory at the University of Central Florida in the discipline of theatre performance. After working in both New York and Los Angeles as a professional actor for eight years, Todd attended the University of Fl orida and will receive his Master of Mass Communication degree in the summer of 2009.