Citation
The Determinants of Word of Mouth Influence in Sport Viewership

Material Information

Title:
The Determinants of Word of Mouth Influence in Sport Viewership
Creator:
Asada, Akira
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (71 p.)

Thesis/Dissertation Information

Degree:
Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Sport Management
Tourism, Recreation, and Sport Management
Committee Chair:
KO,YONG JAE
Committee Co-Chair:
KAPLANIDOU,KYRIAKI
Committee Members:
STEPCHENKOVA,SVETLANA O
Graduation Date:
8/9/2014

Subjects

Subjects / Keywords:
Arithmetic mean ( jstor )
Consumer research ( jstor )
Marketing ( jstor )
Modeling ( jstor )
Persuasion ( jstor )
Recommendations ( jstor )
Social psychology ( jstor )
Sporting events ( jstor )
Statistical models ( jstor )
Syntactical antecedents ( jstor )
Tourism, Recreation, and Sport Management -- Dissertations, Academic -- UF
sport -- viewership -- word-of-mouth
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Sport Management thesis, M.S.

Notes

Abstract:
In the field of sport management, word of mouth (WOM) has been examined as an outcome of other focal concepts (e.g., game satisfaction, team identification, relationship quality). However, none of prior studies in the sport literature focused on the effectiveness of WOM from the information receivers' standpoint. The present study was conducted to examine the determinants of information receivers' perceived WOM influence in sport viewership. Specifically, the author investigated (a) the role of source (i.e., expertise and trustworthiness) and message characteristics (i.e., richness of message content and strength of message delivery) in predicting sports viewers' perceived WOM influence on their watching behavior and (b) the moderating roles of interpersonal factors (i.e., tie strength and homophily) and information receiver's psychological characteristics (i.e., involvement and susceptibility) on the relationships between perceived influence and its predictors. The data was collected from people who had received word of mouth recommendation about a sporting event in the past three months. The results showed that source's trustworthiness (p = .011), richness of message content (p = .001), and strength of message delivery (p < .001) have significant positive effects on perceived influence. The significant moderating effects of homophily and involvement were also discovered. Theoretical and practical implications were discussed based on the results. ( en )
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.
Thesis:
Thesis (M.S.)--University of Florida, 2014.
Local:
Adviser: KO,YONG JAE.
Local:
Co-adviser: KAPLANIDOU,KYRIAKI.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2016-08-31
Statement of Responsibility:
by Akira Asada.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Embargo Date:
8/31/2016
Resource Identifier:
968786009 ( OCLC )
Classification:
LD1780 2014 ( lcc )

Downloads

This item has the following downloads:


Full Text

PAGE 1

1 THE DETERMINANTS OF WORD OF MOUTH INFLUENCE IN SPORT VIEWER SHIP By AKIRA A S ADA A THE S IS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2014

PAGE 2

2 © 2014 Akira A sada

PAGE 3

3 To my dad (Tsuneho), mo m (Yuko), and brother (Wataru)

PAGE 4

4 ACKNOWLEDGMENTS I would like to express my deepest gratitude to my advisor Dr. Yong Jae Ko for his advice and encouragement . He kept saying your research idea is interesting. H e treated me as a research er not as a student. That motivates me to work hard. I could not complete this thesis research without his support and guidance . I would also like to extend my appreciation to Dr. Kyriaki Kaplanidou and Dr. Svetlana Stepchenkova . I believe that insights that they gave me are going to be the founda tion of mine as a researcher. I also thank my colleag ues, Dr. Jaewon Chang, Dr. Chanmin Park, Dr. Seunghoon Jeong, Dr. Hee Youn Kim , Tae H o, Akiko, Shintaro, Semih, Yonghwan, and Eric for their precious advice and suggestions . Finally and most importantly , I would like to thank my family for their supports .

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ............... 4 LIST OF TABLES ................................ ................................ ................................ ........................... 7 LIST OF FIGURES ................................ ................................ ................................ ......................... 8 ABSTRACT ................................ ................................ ................................ ................................ ..... 9 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .................. 10 Significance of Word of Mouth ................................ ................................ .............................. 10 Statement of Problem ................................ ................................ ................................ ............. 11 Purposes of the Study ................................ ................................ ................................ ............. 12 2 LITERATURE REVIEW ................................ ................................ ................................ ....... 14 The Pred icting Role of Source and Message Characteristics on Perceived Influence ........... 14 Source Characteristics ................................ ................................ ................................ ..... 14 Expertise ................................ ................................ ................................ ................... 14 Trustworthiness ................................ ................................ ................................ ........ 15 Message Characteristics ................................ ................................ ................................ .. 16 Richness of c ontent ................................ ................................ ................................ .. 17 Strength of d elivery ................................ ................................ ................................ .. 17 The Mode rating Roles of Tie Strength and Homophily ................................ ......................... 18 Tie Strength ................................ ................................ ................................ ..................... 19 Homophily ................................ ................................ ................................ ....................... 20 The Moderating Role of Involvement and Susceptibility ................................ ....................... 21 Involvement ................................ ................................ ................................ ..................... 21 Susceptibility ................................ ................................ ................................ ................... 23 3 METHODOLOGY ................................ ................................ ................................ ................. 26 Procedures ................................ ................................ ................................ ............................... 26 Instrumentation ................................ ................................ ................................ ....................... 27 Perceived Influence ................................ ................................ ................................ ......... 27 Expertise ................................ ................................ ................................ .......................... 28 Trustworthiness ................................ ................................ ................................ ............... 28 Richness of Content ................................ ................................ ................................ ......... 28 Strength of Delivery ................................ ................................ ................................ ........ 29 Tie Strength ................................ ................................ ................................ ..................... 29 Homophily ................................ ................................ ................................ ....................... 29 Involvement ................................ ................................ ................................ ..................... 29 Susceptibility ................................ ................................ ................................ ................... 30

PAGE 6

6 Data Analysis ................................ ................................ ................................ .......................... 30 Descriptive Statistics ................................ ................................ ................................ ....... 30 Measurement Model Test ................................ ................................ ................................ 30 Struct ural Equation Modeling Test ................................ ................................ ................. 31 Multiple Group Structural Equation Model Test ................................ ............................ 31 4 RESULTS ................................ ................................ ................................ ............................... 33 Descriptive Statistics ................................ ................................ ................................ .............. 33 Descri ption of the Sample ................................ ................................ ............................... 33 Independent Variables and Dependent Variable ................................ ............................. 35 Moderators ................................ ................................ ................................ ....................... 36 Measurement Models ................................ ................................ ................................ .............. 38 Structural Models ................................ ................................ ................................ .................... 41 Moderating Effects ................................ ................................ ................................ ................. 42 Tie S trength ................................ ................................ ................................ ..................... 43 Homophily ................................ ................................ ................................ ....................... 44 Involvement ................................ ................................ ................................ ..................... 46 Susceptibility ................................ ................................ ................................ ................... 48 5 DISCUSSION ................................ ................................ ................................ ......................... 50 General Discussi on ................................ ................................ ................................ ................. 50 The Predicting Role of Source and Message Characteristics ................................ .......... 50 The Moderating Role of Inte rpersonal Factors and Receiver C haracteristics ................. 52 Implication ................................ ................................ ................................ .............................. 54 Theoretical Implication ................................ ................................ ................................ ... 54 Practical Implication ................................ ................................ ................................ ........ 55 Limitations and Future Research ................................ ................................ ............................ 55 APPENDIX SURVEY QUESTIONNAIRE ................................ ................................ ............... 57 L IST OF REFERENCES ................................ ................................ ................................ ............... 64 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ......... 71

PAGE 7

7 LIST OF TABLES Table page 4 1 Demographic characteristics of participants ................................ ................................ ...... 33 4 2 Descriptive statistics for perceived influence and its antecedents ................................ ..... 35 4 3 Descriptive statistics for moderators ................................ ................................ .................. 37 4 4 Summary results for measurement model of perceived influence ................................ ..... 38 4 5 Correlation among items ................................ ................................ ................................ .... 40 4 6 Correlation among constructs ................................ ................................ ............................ 41 4 7 Standardized total effect (Dependent variable: Perceived influence) ................................ 41 4 8 Descriptive statistics for groups of moderators ................................ ................................ . 43 4 9 Comparing the models and structural invariance between the average and strong tie strength groups ................................ ................................ ................................ ................... 44 4 10 Comparing the models and structural invariance between the average and high homophily groups ................................ ................................ ................................ .............. 44 4 11 C ompar ing path coefficients and statistical significance between the average and high homophily groups ................................ ................................ ................................ ...... 45 4 1 2 Comparing the models and structural invariance between the average and high involvement groups ................................ ................................ ................................ ............ 47 4 1 3 C ompar ing path coefficients and statistical significance between the average and high involvement groups ................................ ................................ ................................ .... 47 4 14 Comparing the models and structural invariance between the low and high susceptibility groups ................................ ................................ ................................ .......... 49

PAGE 8

8 LIST OF FIGURES Figure page 2 1 A proposed model for hypotheses H1 ................................ ................................ ................ 18 2 2 A proposed model for hypotheses H2 ................................ ................................ ................ 20 2 3 A proposed model for hypotheses H3 ................................ ................................ ................ 21 2 4 A proposed model for hypotheses H 4 ................................ ................................ ................ 23 2 5 A proposed model for hypotheses H 5 ................................ ................................ ................ 24 2 6 A proposed research model ................................ ................................ ................................ 25 4 1 The direct effects of antecedents to perceived influence ................................ ................... 42 4 2 The moderating effect o f homophily on the relationship between perceived influence and its antecedents ................................ ................................ ................................ ............. 46 4 3 The moderating effect of involvement on the relationship between perceived influence and its antecedents ................................ ................................ ............................. 48

PAGE 9

9 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 Science THE DETERMINANTS OF WORD OF MOUTH INFLUENCE IN SPORT VI E WE R SHIP By Akira Asada August 2014 Chair: Yong Jae Ko Major: Sport Management In the field of sport management, word of mouth ( WOM ) has been examined as an outcome of other focal concepts (e.g., game satisfaction, team identification, relationship quality). However , none of prior studies in the sport literature fo cused on the effectiveness of WOM from the information receivers standpoint. T he present study was conducted to examine the determinant s of information receivers perceived WOM influence in sport viewership. Specifically, the author investigate d ( a ) the role of source (i.e., expertise and trustworthiness) and message characteristics (i.e., richness of message content and strength of message delivery) in predicting sport s viewer s perceived WOM influence on their watching behavior and (b ) the moderati ng role s of interpersonal factors (i.e., tie strength and homophily ) and information receiver s psychological characteristics (i.e., involvement and susceptibility ) on the relationship s between perceived influence and its predictors. The data was collected from people who ha d received word of mou th recommendation about a sporting event in the past three months. T he results showed that source s trustworthiness 180 , p = .011) , richness of message content = . 250 , p = .001) , and strength of message delivery = .2 0 9, p < .00 1 ) have significant positive effects on perceived influence. The significant moderating effects of homophily and involvement were also discovered. Theoretical and practical implications were discussed ba sed on the results .

PAGE 10

10 CHAPTER 1 INTRODUCTION Significance of Word of Mouth Word of m outh (WOM) is to person communication between a perceived noncommercial communicator and a receiver regarding a brand, a product, Walker, 2001, p. 63) . S ince the 1950s WOM has been known as a signifi cant consumer behavior and has drawn tremendous interest from both scholars and practitioners. Lazarsfeld and Katz (1955) found that WOM i s two times more effective than radio advertisement s , four times more effective than personal selling, and seven times more effective than print advertisement s . More recently , research ers ha ve confirmed that consumers perceive the information from WOM as more credible and trustworthy than conventional advertisements (Brown & Reingen , 1987; Liu , 2006; Murray , 1991). Specif ically , WOM referrals have a significant im pact on acquiring new customers ( Trusov, Bucklin, & Pauwels , 2009) and customers who get information through WOM are more likely to be recurrent buyers than other customers ( Villanueva, Yoo, & Hanssens , 2008) . Ultimately, WOM is considered to be one of the most important tools for enhancing the financial health of an organization (Harrison Walker, 2001; Villanueva et al, 2008) . WOM has received tremendous scholarly attention since online communities and social media became popular (Brown, Broderick, & Lee, 2007; Mangold & Faulds , 2009 ; Ye, Law, Gu, & Chen, 2011 ) . M ore than ever before, consumers m essages spread rapidly and extensively b ecause WO M communication on the Internet ha s no time or geographical constraints. T o further enhance WOM communication among consumers , c ompanies started using online communities and social media as their main marketing tools , known as WOM marketing or viral marketing , ( Kozinets , De Valck, Wojnicki, & Wilner , 2010 ) .

PAGE 11

11 Sport s org anizations are no exception to this trend. For example , the Arizona Diamondbacks , a professional baseball franchise, created a video clip titled D backs D o the Har l em Shake and uploaded it to YouTube. The video has been viewed more than 240,000 times and generate s a high volume of conversation among interested fans . Even though t he video does not give any information about the team , the team surveys showed that the video affected consumers intention to purchase tickets and merchandise (Fisher, 2013) . Numerous o ther sports organization s conduct this type of marketing activit y to enhance WOM communication among sports consumers because it has great influence on message recipients consumption behavior . S port s consumption i s a social function in which sport s consumers are highly influenced by their significant others when they make consumption decision s ( Cunningham & Kwon, 2003) . In their analysis , Mullin, Hardy, and Sutton (2007) found that less than 2% of collegiate and professional sports spectators in the United States attend an event alone. Because sports spectators highly rely on WOM communication, sport s organizations need to consider it as a strategic option in their business. Statement of Problem In the field of sport management, WOM has been examined as an outcome of other concepts such as game satisfaction ( Kuenzel & Yassim , 2007) , team identification ( Swanson, Gwinner, Larson , & Janda , 2003), relationship quality ( Kim & Trai l, 2011) , corporate social responsibility ( Walker & Kent , 2009) , and brand leadership (Chang & Ko, 2011 ) . These studies assume that sport s organizations can enhance consumers WOM recommendation behavior by improving customer s atisfaction, identification, and relationship quality , a mong other factors , and that such communication influence other consumer s . Nevertheless , none of the prior studies in the sport literature focused on the effectiveness of WOM from the information receivers standpoint. As discussed earlier, sport s organization s conduct WOM marketing to

PAGE 12

12 increase their communication effectiveness and ultimately profit s . To contribute to such efforts , researchers sh ould provide systematic information about WOM effectiveness . In the marketing literature, several scholars have point ed out the importance of examining WOM by focus ing on information receivers perspective s ( Bansal & Voyer, 2000; Brown & Reigen, 1987; Sweeney, Soutar , & Mazzarol, 2008 ). To fill the conceptual void, Sweeney and his colleagues (2008) proposed a conceptual model of WOM impact. The model includes four determinants of WOM influence ; ( a ) personal factors (e.g., source credibility), ( b ) interpersonal factors (e.g., tie strength), ( c ) situational factors (e.g., complexity of service), and ( d ) message characteristics (e.g., vividness of message). Those factors influence information receivers by reducing the risk of buying, improving the ir perception of the firm, and ultimately causing a greater likelihood o f buying ( Sweeney et al. , 2008) . However, although this is considered to be an appropriate framework of the present study , the model h as not been empirically tested. Purposes of the Study Accordingly , th e purpose of the present study is to examine the determinants of sport perceived WOM influence on their watching behavior by apply ing WOM impact model (Sweeney et al., 2008) . T he s pecific objectives are to investigate ( a ) the role of source (i.e., expertise and trustworthiness ) and message characteristics ( richness of message content and strength of message delivery ) in predicting viewer s perceived WOM influence and (b ) the moderating role s of in terpersonal factors (i.e., tie strength and homophily ) and receivers characteristics (i.e., involvement and susceptibility ) o n the relationship s between perceived influence a nd its predictors . Given the compelling need for an empirical examination of the effects of WOM on sport c onsumption behavior, t h e present research will make a significant contribution to the sport

PAGE 13

13 marketing literature . Furthermore, the present study will contribute to WOM literature in business marketing by empirically test ing conceptually developed model of WOM impact. S port organization s will also gain new insight s from th is study for developing effective promotional strategies .

PAGE 14

14 CHAPTER 2 LITERATURE REVIEW The Predicting Role of Source and Message Characteristics on Perceived Influence First, the author identif ied source and message characteristics that may influence based on extensive literature review . Specifically, message delivery are selected. Further discussion of each co nstruct is provided in the following section followed by research model and hypotheses. Source C haracteristics Since the early period of persuasion research, source credibility has been recognized as one of the most important conditions for successful persua r hetoric t heory, persuasiveness of arguments is determined by p athos, l ogos, and e thos (Aristotle, 1924). The se refer to the emotional appeal of a message, the logical appeal of a message, and the credibility of the sou rce, respectively. Later, t he Yale Communication Program , which dominated persuasion research in the 1950 s , also suggested that source credibility is a key factor of successful persuasion ( Hovland & Weiss, 1951; Kelman & Hovland, 1953). In more recent WOM studies, numerous researchers have argued that credible sources are persuasive and result in changes in inform s (B r own et al. , 2007 ; De Bruyn & Lilien, 2008; Pornpitakpan, 2004 ; Wangenheim & Bayon, 2004 ). Hovland, Janis, and Kelley ( 1953) suggested that s ource credibility is based on expertise and trustworthiness of the information sender and WOM impact model include s these two concepts . T he two concepts are discussed next . Expertise E xpertise is defined communicator is perceived to be a source of Hovland et al., 1953 , p. 23). The positive impact of a

PAGE 15

15 persuasive communication has been remarkably supported ( Clark, Wegener, Habashi, & Evans, 2012; De B ono & Harnish, 19 88; Moore, Hausknecht, & Thamodaran, 1986; Tobin & Raymundo, 2009 ) . It is commonly agreed that the opinion from experts is regarded as valid or trustable . In some cases, however, a highly credible source has less influence on information receivers compared to less credible source s ( Homer & Kahle , 1990). Specifically, source s with a high level of expertise ha ve less influence on a their opinion is weak (Bohner, Ruder, & Erb, 2002; Tormala, Briñol, & Petty, 2006). Also, an opinion given by an expert can be perceived as a biased message and has little influence on an information receiver when the receiver s carefully process the opinion (Chaiken & Maheswaran, 1994). In other words, a source s expertise positively inf luences persuasiveness only when the accuracy of information is ensured. O nce the accuracy is dimini shed by other factors (e.g., weak argument) , conversely , it w ill lead to even have a negative effect on persuasiveness. The h euristic systematic model ( Chaiken, 1980) and elaboration likelihood model (Petty & Cacioppo, 1986) suggest that a not possess a high level of motivation and ability to process opinions carefully. Therefore, there are sev such as issue involvement (Petty, Cacioppo, & Goldman 1981; Rhine & Severance 1970) , receiver s knowledge about the issue (Wood & Kallgren, 1988) and self monitoring (Snyde r, 1974). Therefore , against this background, it is hypothesized: H1 1. A s ource s expertise has a significant positive effect perceived influence . Trustworthiness an

PAGE 16

16 information s ender is trustworthy , the information receiver think s the information is valid and reliable. The d ifference between expertise and trustworthiness is that expertise stems from knowledge while trustworthiness is formed by personality ( McGinnies & Ward , 1980). Numerous studies supported that a trustworthy source has greater influence on receivers attitude compared to a less trustwo rthy one ( Eagly, Wood, & Chaiken, 1978; Levin & Cross , 2004 ; Mills & Jellison, 1967) . Moreover, trustworthiness is more important than expertise in some circumstances . For instance, McGinnies and War d ( 1980 ) conducted an experimental study using a persuasi ve message regarding international maritime boundaries. They manipulated the expertise and trustworthiness of the message source and collected data in four different countries. In two countries, trustworthiness showed a greater effect on persuasiveness of the message than expertise did. Accordingly, the author assumes that a source s trustworthiness has positive infl u ence on the receiver s decision making. H1 2. A s s trustworth iness has a significant positive effect on the information perceived influence. Message C haracteristics Sweeney and his colleagues (2008) believe that the vividness of a message is a key determinant of information receivers perceived WOM influence . In previous studies, researchers manipulated vividness in several different ways . For example, in their experimental study, Keller and Block ( 1997) manipulated vivid and nonvivid message s with anecdotal stories and statistical information , respectively . Several researchers assume that face to face communication gives vivid message and printed media give s nonvivid information ( Borgida & Nisbett, 1977; Herr, Kardes , & Kim , 1991). For the inconsistency of its operational definitions, the effects of message vividness on information rece ivers ha ve not been systematically examined to date (Taylor &

PAGE 17

17 Thompson, 1982). Alternatively, Sweeney, Soutar, and Mazzarol (2012) suggest analyzing the constructs of ric hness of message content and strength of message delivery. Richness of co ntent Richness of (Sweeney et al., 2012, p. 242) . T his concept was develop ed to explain the rational aspect of a WOM message. Since (1924) r hetoric t heory , the rational appeal of a message has been considered one of the most important determinants of message persuasiveness. It positively influences information receivers attitude s ( Petty & Cacioppo, 1986 ; Rosselli, Skelly, & Mackie, 1995). In particular , when they have the motivation and ability to carefully process the message, the quality and validity of the message content plays a critical role on message persuasiveness. Accordingly, it is hypothesized that: H1 3. Richness of content has a significant positive effect perceived influence. Strength of d elivery Strength of deliver y refers to (Sweeney et al., 2012, p. 242). T his concept was incorporated to explain an emotive aspect of WOM message. In addition to the rational appeal of a message, the authors considered the emotional appeal a n important determinant of message persuasiveness. In reality, c ompanies often depict their product s as fun, attractive, and exciting so that they can generate positive product related emotions and influence consumers feelings about the products ( Batra & Ray, 1985 ; Rosselli e t al. , 1995) . WOM research ers has also suggested that the emotional appeal of a WOM message plays a sign ificant role on the message attitude change ( Allsop, Bassett, & Hoskins, 2007; Davis & Mason , 2007 ). Therefore, it is assumed that WOM message

PAGE 18

18 has greater effects on an information receiver s decision making when it is delivered strongly. Therefore, it is hypothesized that: H1 4. Strength of delivery has a significant positive effect perceived influence. Figure 2 1. A proposed model for hypotheses H 1 The Moderating Roles of Tie Strength and Homophily People are not always influenced by WOM recommendation s , even if the source has expertise and the message is rich. S ince WOM is a social interaction , the relation ship between communicators is also a critical factor in determining the effectiveness of WOM ( Brown & Reigen , 1987) . In this study, the author incorporated tie strength and homophily as interpersonal factors to more clearly explain the relationship s between the aforementioned predictors (i.e., expertise, trustworthiness, richness of content, and strength of delivery) and perceived influence. Specific information about the modera tors is discussed next.

PAGE 19

19 Tie S trength Granovetter (1973) introduced the concept of tie strength to explain how people communicat e and spread information . He define d tie strength as combination of the amount of time, the emotional intensity, the intimacy (mutual confiding), and (p. 1361) . Tie strength can be categorized as either strong or weak. When co mmunicators are close friends with each other the tie between them is strong, while the tie between acquaintances is weak (Granovetter, 1983). R eceivers process the message differently depending on the strength of their tie s with the information senders. S trong tie sources are easily reachable because they have a greater motivation to help their communicators (Granovetter, 1983). If both strong and weak ties are available, people are more likely to contact the sources connected with strong ties than those w ith weak ties ( Brown & Reigen, 1987 ). Also, strong tie sources hav e a greater i mpact on making than weak tie counterparts (Brown & Reigen, 1987 ; Granovetter, 1983). Therefore, even if an informati on sender has expertise and his or her message is rich, information receiver s may perceive less influence when th ey are connected with a weak tie . Based on this theory, the author predict ed : H 2 . Tie strength play s a significant moderating role between perceived influence and its antecedents.

PAGE 20

20 Figure 2 2. A proposed model for hypotheses H 2 Homophily the degree to which individuals in a dyad are congruent on certain attributes (Rogers & Bhowmik , 1971, p. 526). This concept has received attention by researchers especially in WOM studies (Brown et al., 2007; Chu & Kim , 2011; McPherson, Smith Lovin , & CookSource , 2001) . Since people more frequently and intensely communicate with those who are demog raphically or behaviorally similar, WOM is more likely to occur among homogenous people (Brown & Reingen, 1987; De Bruyn & Lilien, 2008). De Bruyn and Lilien (2008) found that recommendations from similar people are more likely to generate interest in a product or service. Also , McPherson and his colleagues (2001) argue d tha t information has a senders and receivers is high . Thus, th e following hypothes is was developed.

PAGE 21

21 H 3 . Homophily plays a significant moderating role between perceived influence and its antecedents. Figure 2 3. A proposed model for hypotheses H 3 The Moderating Role of Involvement and Susceptibility Involvement According to the e laboration li kelihood m odel of persuasion (ELM; Petty & Cacioppo, 1986), people process received information in two distinct ways . First, they use the central route for careful consideration of arguments. When people change their attitude toward objects by processing arguments through the central route, the resulting attitude is relatively enduring. Second, the peripheral route requires people base their attitude off simple and peripheral cues such as attractiveness of the source. People d etermine whether they take a central or peripheral route based on their motivation and ability to process arguments. When people have both motivation and ability they take the central route, otherwise , they rely on peripheral cues.

PAGE 22

22 Although motivation and ability are proposed as determinants in the original model, numerous scholars have utilized involvement as a moderator variable. Involvement is defined as (Zaich kowsky, 1985, p. 342). Prior studies have supported that high involvement consumers evaluate information carefully and thoughtfully while low involvement consumers are affected by peripheral cues (Kai, Wang & Cheng Kiang, 2009; Park, Lee & Han, 2007; Petty , Cacioppo & Schumann, 1983). For example, Petty, Cacioppo, and Goldman (1981) showed that people conduct a careful evaluatio n of message quality under high involvement conditions, wh ereas they are influenc ed by peripheral cues under low involvement conditions. Thus, even if an information sender gives rich messages, an information receiver may not consider the message carefully if the receiver is not highly involved with a sporting event. Accordingly, the author developed the hypothes i s b elow. H 4 . Involvement plays a significant moderating role in the relationship between perceived influence and its antecedents.

PAGE 23

23 Figure 2 4. A proposed model for hypotheses H 4 Susceptibility Perceived WOM influence also depends on each information accept opinions from others (Sweeney et al., 2008) . C onsumer susceptibility to interpersonal influence is define s image with significant others through the acquisition and use of products and brands, the willingness to conform to the expectations of others regarding purchase decisions, and/ or the tendency to learn about products Bearden et al., 1989, p. 474). Consumer behavior researchers examined consumer susceptibility as a key characteristic , & Teel, 1989; Burnkrant & Cousineau, 1975; Deutsch & Gerard, 1955). Thus, a consumer with hig h susceptibility is more likely to be influenced by other consumers than is one with low

PAGE 24

24 susceptibility. This concept focuse ies, so it does not vary across situations (Bearden et al. , 1989). Therefore, the fifth hypothesis is developed: H 5 . Susceptibility play s a significan t moderating role in the relationship between perceived influence and its antecedents. Figure 2 5. A proposed model for hypotheses H 5

PAGE 25

25 Figure 2 6 . A proposed research model

PAGE 26

26 CHAPTER 3 METHODOLOGY Procedures The author utilize d Amazon Mechanical Turk (AMT) to recruit research participants. AMT is a n online marketplace to access and recruit potential participants for survey. AMT is an efficient and trustable method for data collection ( Buhrmester, Kw ang, & Gosling, 2011). In the present study, the data wa s collected from people who ha d received WOM about a sporting event (i.e., professional sport s , collegiate sport s , the Olympic Games, or s occer World Cup) in the past three months. T o collect reliable data , t he author recruited participants who were in the United States and had prior experience s taking AMT survey s (i.e., the number of approved is greater than or equal to 100 ; approval rate is over or equal to 95%). According to AMT, the average time to complete the survey was seven minutes and seven seconds. For their participation, $. 50 of compensation was given to each participant. A total of 452 people completed the questionnaire and 297 cases were identified as useful cases. Eighty cases were exclud ed because they ha d not actually watched the sporting event. Nineteen cases were excluded because they received WOM earlier than three months ago. Additional 56 cases were excluded because they ha d a missing value, complet ed the questionnaire in abnormally short time (i.e., within two minutes), tick ed the same number for almost all ques tions, or were identified as an outlier. Hair , Black, Babin, and Anderson ( 20 09 ) offer some suggested sample sizes based on the model complexity and the measurement model pro perty. They suggest that 150 is the minimum sample size for the model which has seven or fewer constructs, the modest item communalities (i.e., = .5), and no underidentified constructs. Because the proposed model in this study met these condition s, the a uthor concluded that the sample size of 297 was enough to conduct analyses.

PAGE 27

27 Instrumentation T he author took three steps t o select appropriate measurement scales for each variable. First, the author collected existing scales from previous literatures. Second, the author asked for evaluatio ns from a pane l of experts that included t wo sp ort management professors , one tourism professor and eight PhD students. Then, the author conducted a pilot study ( N = 40) and f inalized measurement scales. Specific details of the procedures are discussed next. Perceived I nfluence In persuasion studies, researchers commonly measure attitude change to assess the persuasiveness of communication ( O Keefe , 2002) . They evaluate s toward an object before and after communication and regard the difference as communication influence. On the other hand, several scholars measured a self report ed perceived persuasiveness. For example, Collins, Taylor, Wood, and Tho mpson ( 1988) measure d perceived persuasiveness by asking , In general, how persuasive do you think this message was? To what extent do you think your opinions on this issue were influenced by this message? Block and Keller (1995) ask ed th How likely are you to follow the recommendations? measure d intention to comply. In WOM research, Bansal and Voyer (2000) assess ed the influence of the s WOM using the In the present study, the author focused on WOM , which ha d already been sent and received. Hence, i t is appropriate to u s e perceived influence to measure WOM persuasiveness. Perceived influence was measured by three scale items ( Block & Keller , 1995 ; Collins et al., 1988 ; Wright , 1973) . The scale item format was se ven point Likert type 1 = not at all ; 7 = very much . The Cronbach s coefficient alpha of

PAGE 28

28 perceived influence was .94 and its AVE value was .85. These numbers met conditions suggested by Hair et al ( 20 09 ) . Expertise Expertise was measured using five seven point semantic differential scale items: Unknowledgeable Knowledgeable, Incompetent Competent, Not an expert Expert, Not experienced Experienced that were adopted from existing scales (Bansal & Voyer, 2000; Netemeyer & Bearden, 1992 ). The Cronbach s coefficient alpha of expertise .92 and its AVE value was .74 . Thus, both the Cronbach s coefficient alpha and AVE were greater than the standard suggested by Hair et al. ( 20 09 ) . Trustworthiness Trustworthiness was evaluated using five seven point semantic differential scale items: Dependable Undependable, Honest Dishonest, Reliable Unreliable, Sincere Insincere, Trustworthy Untrustworthy that were adopted from Ohanian s ( 1990) study . The Cronbach s coefficient alpha of trustworthiness was .91 and its AVE value was .73. Thus, both the Cronbach s coefficient alpha and AVE were greater than the standard suggested by Hair et al. ( 20 09 ) . Richness of C ontent Richness of c ontent was evaluated using four seven point Likert (1) strongly disagree to (7) strongly agree. Sweeney et al. (2012) conceptualized two aspects of message richness ; cognitive and emotive richness. In the present study, the author focused only on the cognitive aspect of message richness to avoid confounding effect of strength of message delivery. The Cronbach s coefficient alpha of richness of content was .76 and its AVE value

PAGE 29

29 was .56. Thus, both the Cronbach s coefficient alpha and AVE were greater than the standard suggested by Hair et al. ( 20 09 ) . Strength of D elivery Strength of delivery was evaluated using f our seven point Likert message was delivered i n an important manner in Sweeney et al. s ( 2012) study. The Cronbach s coefficient alpha of strength of delivery was .86 and its AVE value was .68. Thus, both the Cronbach s coefficient alpha and AVE were greater than the standard suggested by Hair et al. ( 20 09 ) . Tie St rength Tie strength is measured using four seven point Likert type scales om previous studies ( Bansal & Voyer, 2000; Frenzen & Davis, 1990 ). The Cronbach s coefficient alpha of tie strength was .89 which is greater than the standard suggested by Hair et al. ( 20 09 ) . Homophily Homophily was evaluated using four seven point Likert ranged from (1) not at all similar to (7) extremely similar (De Bruyn & Lilien, 2008; Gilly, Graham, Wolfinbarger, & Yale, 1998). The Cronbach s coefficient alpha of homophily was .85 which is greater than the standard suggested by Hair et al. ( 20 09 ) . Involvement Involvement was measured using Zaichkowsky s (1985 ) semantic differential scale items: Boring Exciting, Uninteresting Interesting, Worthless Valuable, Unappealing Appealing, Irrelevant Relevant, and Unimportant Important. The Cronbach s coefficient alpha of involvement was .93 which is greater than the standard suggested by Hair et al. ( 20 09 ) .

PAGE 30

30 Susceptibility Susceptibility was measured using seven seven point Likert previous studies (Bearden et al., 1989; Mangleburg, Doney , & Bristol, 2004). The Cronbach s coefficient alpha of susceptibility was .94 which is greater than the standard suggested by Hair et al. ( 20 09 ) . Data Analysis Descriptive statistics w ere performed to identify missing data and outliers, and also to check the assumption of normality. A series of confirmatory factor analyses (CFA) was performed to establish validity and reliability of the measurement scale. A s tructural equation model (SEM) and multiple group SEM analyses were then conducted to test research hypotheses . SPSS 20.0 was used for descriptive statistics and AMOS 20 .0 was used for CFA , SEM, and multiple group SEM . Descriptive Statistics Descriptive statistics were calculated for research constructs and the soci o demographic characteristics . Means , standard deviation, standard errors, skewness , and kurtosis were calculated. Measurement Model Test The author conducted confirmatory factor analysis (CFA) to examine reliability, valid ity, and theoretical relevance of scales for five constructs (i.e., perceived influence, expertise, trustworthiness, richness of content, and strength of delivery). To consider the appropriateness of the scales, a comparative fit index (CFI), standard root mean squared residual (SRMR), and root mean square error of approximation (RMSEA) were investigated. It is assumed that models with a good fit have a C FI value higher than .95 and an SRMR value of less than .08 (H u &

PAGE 31

31 Bentler, 1999). In addition , an RMSEA value s of less than .08 are considered an acceptable model fit (Brown e & Cudeck, 1992; Hu & Bentler, 1999), Cronbach s coefficient alpha greater than .70 indicates an acceptable consistency among subscale items for each constru ct; and an average variance extracted (AVE) value s greater than .50 indicate acceptable composite reliability of each construct ( Hair et al. , 20 09 ) . The author also examined item loading s , factor correlation s , and comparison s of squared correlation s of any two construct s with AVE value s to establish conv ergent and discriminant validities . The author concluded that convergent validity is established if item loading is equal to or greater than .40 (Nunnally & Bernstein, 1994) , a nd discriminant v alidity is established when correlation s among construct s are less than .85 (Kline , 2011 ) , and a squared correlation between two constructs is lower than the AVE for each construct (Fornell & Larcker, 1981) . Structural Equation Model ing Test A s tructural e quation m odeling (SEM) test was performed using AMOS 20.0 to examine the relationship between perceived influence and its antecedents (i.e., expertise, trustworthiness, richness of content, and strength of delivery) . Mul tiple G roup Structural Equation Model Test To examine the moderating effects of tie strength, homophily, involvement, and susceptibility, the author conducted the multiple group SEM test. Prior to the test, a measurement invariance test was performed to ensure that c onstructs were measured in a manner that was statistically the same across groups (Hair et al. , 20 0 9 ). After the measurement invarian ce test, the structural models we re examined to assess the moderating effects of four constructs. Following the exami nation of structural models, path analyses were conducted to investigate the moderating effect on paths between expertise and perceived influence, trustworthiness and perceived influence, richness of content and perceived influence, and

PAGE 32

32 strength of delivery and perceived influence. The fit indices that were utilized for the measurement model test were examin ed in SEM and multiple group SEM test s as well .

PAGE 33

33 CHAPTER 4 RESULTS Descriptive Statistics Description of the S ample Demographic characteristics of participants ( N = 297) are depicted in Table 4 1. Of the participants, 76.8% were male. The average age of the participants was 31 years old ( M = 30.96, SD = 8.6) , and 74.4 % of the participants were Caucasian. The major ity of participants received a recom mendation about a sporting event through face to face conversation (62.6% ), with 67.3% of participants receiving a recommen dation to watch a professional sporting event and 81.5% of participants receiving a recommendation to watch the sporting event on tel evision or the Internet. More than half of participants received a recommendation from their friends. Table 4 1. Demographic characteristics of participants Variable N % Gender Male 228 76.8 Female 69 23.2 Age 19 and under 4 1.3 20 29 157 52.9 30 39 98 33.0 40 49 23 8.1 50 and over 14 4.7 Ethnicity A frican American 21 7.1 Asian 35 11.8 Caucasian 221 74.4 Hispanic 1 8 6.1 Native American 1 0.3 Other 1 0. 3

PAGE 34

34 Table 4 1. Continued Variable N % Education No high school 1 0. 3 High school grad 27 9.1 Some college 77 25.9 Associate degree 22 7.4 147 49.5 Graduate degree 23 7.7 Income Under $15,000 33 11.1 $15,000 ~ $24,999 42 14.1 $25,000 ~ $34,999 29 9.8 $35,000 ~ $49,999 61 20.5 $50,000 ~ $74,999 69 23.2 $75,000 ~ $99,999 38 12.8 $100,000 ~ $149,999 20 6.7 $150,000 ~ $199,999 5 1.7 How recommendation was received Face to face 186 62.6 By e mail/text message 64 21.5 By telephone 25 8.4 Social Networking Service 22 7.4 Event type Professional 200 67.3 Collegiate 51 17.2 The Olympic Games 42 14.1 Soccer World Cup 4 1.3 How sporting event was watched On television/Internet 242 81.5 In person 55 18.5 Relationship with the recommender Friends 168 56.5 Family 78 26.3 Co worker 19 6.4

PAGE 35

35 Table 4 1. Continued Variable N % Relationship with the recommender Boy/Girl friend 17 5.7 Acquaintance 8 2.7 Fiancé 6 2.0 Other 1 0.3 Independent Variables and Dependent Variable Descriptive statistics for perceived influence of word of mouth and its antecedents are depicted in Table 4 2. The mean of perceived influence ranged from 5.14 to 5.29. Standard deviations ranged from 1.41 to 1.47. Skewness ranged from .88 to .67. Kurtosis ranged from .04 to .64. The means of antecedents ranged from 4.00 to 6.20. Standard deviations ranged from .915 to 1.59. Skewness ranged from 1.58 to .16. Kurtosis ranged from .59 to 3.41. The items for trustworthines s had the highest mean on the 7 point Likert type scale ( M = 6.10; SD = .875). The items for strength of delivery had the lowest mean ( M = 3.99; SD = 1.32). Table 4 2. Descriptive statistics for perceived influence and its antecedents Construct M S.D. Ku rt Perceived influence Did t he recommendation ha ve an influence on your decision to watch the sporting event ? 5.14 1.41 .85 .64 To what extent was your decision to watch the sporting event influenced by the recommendation ? 5.16 1.47 .67 .04 Did the recommendation help you to make a decision to watch the sporting event? 5.29 1.42 .88 .44 Expertise in the sporting event Not an expert Expert 4.69 1.59 .62 .26 Inexperienced Experienced 5.12 1.40 .82 .44 Unknowledgeable Knowledgeable 5.47 1.23 .95 .83 Unskilled Skilled 4.84 1.43 .61 .10

PAGE 36

36 Table 4 2. Continued Construct M S.D. Kurt Trustworthiness The person is Undependable Dependable 6.01 1.03 1.17 1.27 Dishonest Honest 6.17 .915 1.27 1.99 Unreliable Reliable 6.04 1.06 1.41 2.21 Untrustworthy Trustworthy 6.20 .946 1.58 3.41 Richness of content The recommendation about the sporting event was Informative 5.24 1.10 .14 .59 Clear 5.87 1.01 .68 .08 Specific 5.87 1.12 .85 .12 Strength of delivery The recommendation about the sporting event was delivered In a powerful tone 3.77 1.45 .16 .62 In a strong way 4.00 1.49 .19 .59 In an important manner 4.22 1.55 .12 1.27 Moderators Descriptive statistics for moderators are depicted in Table 4 3. The means of tie strength items ranged from 5.57 to 5.93. Standard deviations ranged from 1.06 to 1.31. Skewness ranged from 1.11 to .48. Kurtosis ranged from .46 to .64. The means of homophily items ranged from 5.03 to 5.38. Standard deviations ranged from 1.07 to 1.23. Skewness ranged from .44 to .31. Kurtosis ranged from . 4 to . 02 . The means of involvement items ranged from 4.90 to 5.66. Standard deviations ranged from 1.35 to 1.4 7. Skewness ranged from 1.32 to .43. Kurtosis ranged from .24 to 1.16. The means of susceptibility items ranged from 3.07 to 3.41. Standard deviations ranged from 1.53 to 1.66. Skewness ranged from .03 to .29. Kurtosis ranged from 1.06 to .7.

PAGE 37

37 Table 4 3. Descriptive statistics for moderators Construct M S.D. Kurt Tie strength How close are you with the person who recommended the sporting event ? 5.60 1.12 .48 .46 How likely are you to share personal confidence with the person? 5.57 1.31 .86 .23 How likely are you to extend everyday assistance to the person? 5.93 1.06 .74 .26 How likely are you to spend a free afternoon with the person? 5.77 1.30 1.11 .64 Homophily Considering your outlook on life, how similar are you and the person? 5.16 1.12 .44 .02 Considering your likes and dislikes, how similar are you and the person? 5.22 1.09 .42 .2 Considering your values how similar are you and the person? 5.38 1.07 .43 .4 Considering your experiences how similar are you and the person? 5.03 1.23 .31 .24 Involvement The sporting event is Boring Exciting 5.57 1.36 1.16 1.34 Uninteresting Interesting 5.66 1.35 1.32 1.66 Worthless Valuable 5.02 1.39 .45 .24 Unappealing Appealing 5.57 1.40 1.14 1.14 Irrelevant Relevant 5.25 1.36 .68 .15 Unimportant Important 4.90 1.47 .43 .2 Susceptibility It is important that others like the products and brands that I buy . 3.07 1.53 .4 .7 When buying products, I generally purchase those brands that I think others will approve of. 3.10 1.62 .29 .93 I like to know what brands and products make good impressions on others. 3.41 1.63 .03 1.06 I achieve a sense of belonging by purchasing the same products and brands that others purchase. 3.19 1.66 .28 .99

PAGE 38

38 Measurement Models A confirmatory factor analysis (CFA) was conducted to assess the psychometric properties and scale parsimony. Based on the results of the CFA, the author removed 4 items and left 17 observed variables for 5 latent variables. Table 4 4 shows the results of the measurement model. Cronbach s coefficient alpha ranged from .76 (richness of content) to .94 ( perceived influence). The average variance extracted (AVE) values ranged from .56 (richness of content) to .85 (perceived influence). Because all AVE value s were greater than .50, the composite reliability of the construct can be considered adequate ( Hair et al. , 2 0 09 ) . The model showed good fit ( 2 /df = 1.893, RMSEA = .055, CFI = .973, SRMR = .055). Convergent validity was established by high factor loading s in the present study. Each measurement scale item s loading was greater than the suggested value of . 50 (H air et al. , 20 09 ). To examine discriminant validity, correlation s among measured variables were analyzed. Correlations be tween variables ranged from .0 5 (trustworthiness and strength of delivery) to .5 4 (trustworthiness and richness of content) implying th at discriminant vali dity was established (Kline, 2011 ). In addition, each standardized correlation was found to be smalle r than AVE, which satisfied Fornell and Larcker s ( 1981) guideline . Table 4 4, 4 5, 4 6 illustrate the summarized results of the CFA. T able 4 4. Summary results for measurement model of perceived influence Construct AVE a Perceived influence Did t he recommendation ha ve an influence on your decision to watch the sporting event ? .92 .85 .94 To what extent was your decision to watch the sporting event influenced by the recommendation ? .91 Did the recommendation help you to make a decision to watch the sporting event? .93

PAGE 39

39 Table 4 4. Continued Construct AVE a Expertise in the sporting event Not an expert Expert .81 .74 .92 Inexperienced Experienced .92 Unknowledgeable Knowledgeable .87 Unskilled Skilled .84 Trustworthiness The person is Undependable Dependable .84 .73 .91 Dishonest Honest .85 Unreliable Reliable .87 Untrustworthy Trustworthy .85 Richness of content The recommendation about the sporting event was Informative .54 .56 .76 Clear .91 Specific .75 Strength of delivery The recommendation about the sporting event was delivered In a powerful tone .87 .68 .86 In a strong way .93 In an important manner .66

PAGE 40

40 Table 4 5. Correlation among items Items 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Exp 1 1.000 .546 .483 .511 .010 .013 .014 .013 .032 .008 .011 .066 .044 .078 .014 .017 .004 Exp 2 . 739 1.000 .645 .587 .016 .041 .035 .020 .024 .036 .035 .057 .035 .078 .018 .028 .012 Exp 3 .6 95 . 803 1.000 .503 .064 .066 .077 .067 .027 .075 .071 .030 .013 .066 .022 .027 .013 Exp4 . 715 .766 . 709 1.000 .014 .015 .024 .018 .019 .011 .009 .059 .026 .082 .011 .018 .007 Tru1 . 100 .1 27 . 252 . 118 1.000 .471 .576 .511 .127 .183 .130 .007 .000 .006 .097 .069 .083 Tru2 .1 17 . 202 . 257 . 121 .686 1.000 .417 .552 .097 .183 .096 .003 .000 .006 .071 .074 .066 Tru3 . 118 . 188 . 277 .156 .759 . 646 1.000 .534 .107 .176 .089 .014 .001 .013 .065 .045 .061 Tru4 .1 14 .1 4 0 . 258 . 135 .715 .743 .731 1.000 .144 .169 .106 .010 .001 .011 .067 .055 .078 Ric1 . 179 .1 54 . 163 .1 38 .356 .311 .327 .379 1.000 .238 .138 .110 .057 .130 .098 .074 .070 Ric2 . 092 . 191 . 274 . 105 .428 .446 .420 .411 .488 1.000 .480 .027 .024 .049 .125 .101 .099 Ric3 .1 04 .1 86 . 267 . 094 .360 .310 .299 .326 .371 . 693 1.000 .011 .013 .031 .076 .052 .066 S tr1 . 257 . 238 .1 72 . 242 .081 .050 .120 .102 .332 . 163 .106 1.000 .664 .325 .042 .069 .075 S tr2 .209 .187 .112 .162 .002 .002 .024 .027 .239 .155 .114 .815 1.000 .379 .030 .061 .058 S tr3 .280 .279 .256 .287 .075 .075 .115 .103 .361 .221 .176 .570 .616 1.000 .023 .043 .038 Inf1 .117 .135 .149 .107 .312 .267 .255 .258 .313 .353 .276 .205 .172 .152 1.000 .702 .728. Inf2 .132 .167 .163 .135 .262 .272 .213 .235 .272 .318 .229 .262 .246 .207 .838 1.000 .723 Inf3 .065 .111 .113 .086 .288 .256 .246 .279 .264 .315 .257 .273 .240 .196 .853 .850 1.000 Note: Correlations are under the diagonal, and squared correlations are above the diagonal.

PAGE 41

41 Table 4 6. Correlation among constructs Items 1 2 3 4 5 Expertise .739 .056 .055 .064 .024 Trustworthiness . 237 .730 . 296 .003 .108 Richness of content . 234 . 544 .564 . 048 .155 S trength of delivery . 253 . 054 . 218 .684 .075 Perceived influence .156 .328 .394 .274 .847 Note: The diagonal shows AVE scores of each construct. Correlations are under the diagonal, and squared correlations are above the diagonal. Structural Models After refining constructs and variables, the author analyzed the structural model. Overall, the model showed good fit of the data to the model ( 2 = 202.598, df = 107, 2 /df = 1.893, SRMR = .055, RMSEA = .055, CFI = .973). In addition, t he di rect path from trustworthiness 180 , and S.E . = . 114 ), richness of content (standardized = . 250 , S.E . = . 17 ) and strength of delivery ( 209 , S.E . = . 064 ) to perceived influence were si gnificant. However, t he direct path from expertise to perceived influence = . 002 , S.E . = . 06 ) was not significant. Table 4 7. Standardized total effect (Dependent variable: Perceived influence) * p < .05. ** p < .01. *** p < .001 Path standardized S.E . Critical Ratio p Expertise Perceived influence .002 .06 .036 .972 Trustworthiness Perceived influence .180 * .114 2.534 .011 Richness of content Perceived influence .250 ** .17 3.212 .001 Strength of delivery Perceived influence .20 9*** .064 3.442 .000

PAGE 42

42 * p < .05. * * p < .01. *** p < .001 . Figure 4 1. The direct effects of antecedents to perceived influence Moderating Effects To examine moderating effects of four constructs (i.e., tie strength, homophily, involvement, and susceptibility), the author had to create two comparable group s. When a moderator is a continuous variable, group s should be created based on the measurement score . If there are two clear peaks in the frequency distribution, the sample can be categorized into two gro ups around each mode. I n this study, however, all four possible moderators show unimodal distribution s . Two groups cannot be created around one mode because the groups should be very similar with each other . A n alternative grouping method is to delete some samples around the median so that the distribution of remaining sampl es becomes similar to bimodal one ( Hair et al. , 2009) . T he author applied this method and excluded around 60 samples close to the median of each construct . To keep an adequate sample for fit analysis, two groups were designed to have at least 100 samples ( Boomsma, 1985 ). The average group s mean score of tie strength items was between 2.75 and 5.5 0 and the strong tie strength group s scor e was between 6.0 and 7.0 in 7 point Likert type scale. The

PAGE 43

43 average group s mean score of homophily items was between 2.75 and 5.0, and the high homophily group s score was between 5.25 and 7.0. The average group s mean score of involvement items was between 1.0 and 5 .0 and the high involvement group s score was between 5.67 and 7.0. The low group s mean score of susceptibility items was between 1.0 and 2.5 and the high susceptibility group s score was between 3.5 and 7.0. Table 4 8 illustrates the sa mple size, mean score and standard error of each group. Table 4 8. Descriptive statistics for groups of moderators Construct Group Score n M S.E. Tie strength Average 2.75 5.5 123 4.7 0 .067 Strong 6.0 7.0 109 6.75 .027 Homophily Average 2.75 5.0 101 4.2 0 .055 High 5.25 7.0 129 6.03 .045 Involvement Average 1.0 5.0 107 4.06 .093 High 5.67 7.0 129 6.34 .04 Susceptibility Low 1.0 2.5 116 1.65 .048 High 3.5 7.0 121 4.71 .065 Tie S trength The result of m easurement invariance test showed that the difference in the chi square statistic for the measurement weight was in significant ( 2 = 19.416 , df = 12 , p = . 079 ), which implies that factor structure between the average tie and strong tie group s can be assumed to be invariant. T he difference in the chi square statistic for the structural weight was not significant either ( 2 = 20.706 , df = 16 , p = . 19 ) . Thus, the causal links in the structural model we re not significantly different between the two group s . The results le d the auth or to conclude that there is no significant moderating effect of tie strength on the relationship between perceived influence and its antecedents.

PAGE 44

44 Table 4 9 . Comparing the models and structural invariance between the average and strong tie strength groups Model Model fit indices 2 difference Moderating effect Unconstrained model 2 = 358.014 , df = 2 14 ; 2 / df = 1. 673 CFI = .944 ; RMSEA = .054 ; p < .000 Measurement weight 2 = 377.430 , df = 2 26 ; 2 / df = 1 .670 CFI = .9 41 ; RMSEA = .0 54 ; p < .000 1 9 . 416 (1 2 ), p = . 079 Structural weight 2 = 378.720 , df = 2 30 ; 2 / df = 1 .647 CFI = .9 42 ; RMSEA = .0 53 ; p < .000 20.706 (16), p = .19 Not supported Homophily The difference in the chi square statistic for the measurement weight was s ignificant ( 2 = 20.792 , df = 1 2 , p = . 054 ) suggesting that the factor structure between the average homophily and the high homophily group s can be assumed to be invariant. In addition, t he difference in the chi square statistic for the structural weight was significant ( 2 = 26.505, df = 16 , p = . 047 ) . Therefore, the causal links in the structural model wer e significantly different between the two group s . Following invariance tests, the author conducted path analyses. The results illustrate d significant moderating effects of homophily on the relationship between (1) expertise and perceived influence ( p = .003), (2) trustworthiness and perceived influence ( p = .001), (3) richness of content and perceived influence ( p = .001), and (4) strength of delivery and perceived influence ( p = .001). Table 4 10, Table 4 11, and Figure 4 2 summarize the results of the multiple group SEM analyse s. Table 4 10 . Comparing the models and structural invariance between the average and high homophily groups Model Model fit indices 2 diffe rence Moderating effect Unconstrained model 2 = 317.902 , df = 2 14 ; 2 / df = 1. 486 CFI = .9 62 ; RMSEA = .0 46 ; p < .000 Measurement weight 2 = 338.694 , df = 2 26 ; 2 / df = 1 .499 CFI = .9 59 ; RMSEA = .0 47 ; p < .000 20.792 (1 2 ), p = . 054 Structural weight 2 = 344.407 , df = 2 30 ; 2 / df = 1 .497 CFI = .9 58 ; RMSEA = .0 47 ; p < .000 26.505 (16), p = .047 Supported

PAGE 45

45 T able 4 10 . Continued Model Model fit indices 2 difference Moderating effect Expertise 2 = 369.535 , df = 2 41 ; 2 / df = 1 .533 CFI = .9 53 ; RMSEA = .0 48 ; p < .000 51.633 (30), p = .003 Supported Trustworthiness 2 = 374.945 , df = 2 41 ; 2 / df = 1 .556 CFI = .9 51 ; RMSEA = .0 49 ; p < .000 57.043 (30), p = .001 Supported Richness of content 2 = 374.684 , df = 2 41 ; 2 / df = 1 .555 CFI = .9 51 ; RMSEA = .0 49 ; p < .000 56.782 (30), p = .001 Supported Strength of delivery 2 = 374.811 , df = 2 41 ; 2 / df = 1 .555 CFI = .9 51 ; RMSEA = .0 49 ; p < .000 56.909 (30), p = .001 Supported Table 4 11. C ompar ing path coefficients and statistical significance between the average and high homophily groups Path S.E . Critical Ratio p Average and high homophily Average High Average High Average High Average High Expertise Perceived influence .195* .110 .094 .087 1.977 1.245 .048 .213 Trustworthiness Perceived influence .125 .166 .151 .175 1.347 1.552 .178 .121 Richness of content Perceived influence .239* .314** .286 .262 2.017 2.987 .044 .003 Strength of delivery Perceived influence .180 .241** .113 .090 1.646 2.831 .100 .005 * p < . 05 . ** p < . 01 . ***p < . 001 .

PAGE 46

46 * p < . 05 . **p < . 01 . ***p < . 001 . Figure 4 2 . The moderating effect o f homophily on the relationship between perceived influence and its antecedents Involvement The difference in the chi square statistic for the measurement weight was insignificant ( 2 = 17.478 , df = 12 , p = . 132 ) . Hence , factor structure between the average involvement group and the high involvement group can be assumed to be invariant. In addition, t he difference in the chi square statistic for the structural weight was significant ( 2 = 29.437 , df = 16 , p = . 021 ) , implying that the causal links in the structural model we re significantly variant between the two groups . The author then examined the difference in each path. The result showed si gnificant moderating effects of involvement on the relationship between (1) expertise and perceived influence ( p = . 046 ) , (2) trustworthiness and perceived influence ( p = . 027 ) , and ( 3 ) richness of message content and perceived influence ( p = . 031 ) . Table 4 10, Table 4 11, and Figure 4 3 summarize the results of the multiple group SEM analyses .

PAGE 47

47 Table 4 1 2 . Comparing the models and structural invariance between the average and high involvement groups Model Model fit indices 2 difference Moderating effect Unconstrained model 2 = 308.005 , df = 2 14 ; 2 / df = 1. 439 CFI = .9 67 ; RMSEA = .0 43 ; p < .000 Measurement weight 2 = 325.483 , df = 2 26 ; 2 / df = 1 .440 CFI = .9 65 ; RMSEA = .0 43 ; p < .000 17.478 (1 2 ), p = . 132 Structural weight 2 = 337.442 , df = 2 30 ; 2 / df = 1 .467 CFI = .9 62 ; RMSEA = .0 45 ; p < .000 29.437 (16), p = . 021 Supported Expertise 2 = 341.011 , df = 2 35; 2 / df = 1.451 CFI = .9 62 ; RMSEA = .0 44 ; p < .000 33.006 (21), p = . 046 Supported Trustworthiness 2 = 343.166 , df = 2 35; 2 / df = 1.460 CFI = .9 61 ; RMSEA = .0 44 ; p < .000 35.161 (21), p = . 027 Supported Richness of content 2 = 342.610 , df = 2 35; 2 / df = 1.458 CFI = .9 62 ; RMSEA = .0 44 ; p < .000 34.605 (21), p = . 031 Supported Strength of delivery 2 = 337.311 , df = 2 35; 2 / df = 1.435 CFI = .9 64 ; RMSEA = .0 43 ; p < .000 29.306 (21), p = . 11 Not supported Table 4 1 3 . C ompar ing path coefficients and statistical significance between the average and high involvement groups Path S.E . Critical Ratio p Average and high involvement Average High Average High Average High Average High Expertise Perceived influence .060 .128 .098 .105 .636 1.459 .525 .145 Trustworthiness Perceived influence .165 .200* .164 .175 1.586 2.035 .113 .042 Richness of content Perceived influence .245* .134 .303 .284 2.214 1.277 .027 .202 Strength of delivery Perceived influence .129 .406*** .109 .112 1.290 4.611 .197 .000 * p < .05 . ** p < .01 . ***p < .001 .

PAGE 48

48 * p < . 05 . **p < . 01 . ***p < . 001 . Figure 4 3 . The moderating effect of involvement on the relationship between perceived influence and its antecedents Susceptibility The difference in the chi square statistic for the measurement weight was insignificant ( 2 = 17.013 , df = 12 , p = . 149 ); therefore, structure between the low susceptibility group and the high susceptibility group s can be assumed to be invariant. On the other hand, t he difference in the chi square statistic for the structural weight was not significant ( 2 = 20.116 , df = 16 , p = . 215 ) , which suggests that the causal links in t he structural model we re not significantly different between the two groups . Accordingly, the author concluded that susceptibility has no significant moderating effect on the relationship between perceived influence and its antecedents.

PAGE 49

49 Table 4 14 . Comparing the models and structural invariance between the low and high susceptibility groups Model Model fit indices 2 difference Moderating effect Unconstrained model 2 = 357.773 , df = 2 14 ; 2 / df = 1. 672 CFI = .9 47 ; RMSEA = .0 53 ; p < .000 Measurement weight 2 = 374.786 , df = 2 26 ; 2 / df = 1 .658 CFI = .9 45 ; RMSEA = .0 53 ; p < .000 17.013 (1 2 ), p = . 149 Structural weights 2 = 377.889 , df = 2 30 ; 2 / df = 1. 643 CFI = .9 45 ; RMSEA = .0 52 ; p < .000 20.116 (1 6 ), p = . 215 Not supported

PAGE 50

50 CHAPTER 5 DISCUSSION General Discussion The Predicting Role of Source and Message Characteristics The purposes of this study were to investigate (1) the role of source and message characteristics in predi cting perceived WOM influence on sports viewers watching behavior and (2) the moderating role of int erpersonal factors and receiver characteristics on the relationship between perceived influence and its antecedents. For the first purpose, the author examined the main effects of source s expertise , source s trustworthiness, richness of message content, and strength of message delivery on receiver s percei ved influence on his or her watching behavior . T he data was collected from people who ha d received a WOM recommendation about sporting event (i.e., professional, col legiate, the Olympic Games, or s occer World Cup) in the past three months . The results showed significant main effects of trustworthiness, richness of content, and strength of delivery. However, there was no significant relationship betwe en expertise and perceived influence. These findings revealed unique characteristics of WOM in sport viewe rship. Usually , a source s expertise has positive effects on the receiver s decision making because people regard information from an expert as being valid or correct ( Clark et al., 2012 ). For instance, when people consider buying a new laptop, they seek information about laptops from someone who has a high level of expertise, because they think information from an expert is reliable and thus they can select a good laptop . In other words, a product s value can be guaranteed by an expert. A n expert in a sporting event , however, does not play the same rol e as a computer expert for several reasons. First, watching a sporting event does not usually have as much financial risk as high technology products. Thus, people do not need to conduct careful

PAGE 51

51 evaluation about the event before consumption. In particular, 81.5% of participants who watched sporting events on television or the Internet probably perceive d little financial risk and need ed no help from an expert . Second, an expert in a sporting event cannot guarantee the value of the event because the results of sporting ev ent s are almost always unpredictable. Also, sport viewers receive hedonic values (e.g., escape, drama) whereas consumers of high technology products receive utilitarian ones ( Hightower, Brady, & Baker , 2002; Hopkinson & P ujari , 1999). Accordingly , it is extremely difficult to guarantee the value of sporting events before consumption , even if the information source has plentiful experience and extensive knowledge. Thus, a source s expertise does not have a significant positive effect on sport watching behavior . On the other hand, the main effect of trustworthiness on perceived influence was found. As discussed in Chapter 2, trustworthiness is formed by personality whereas expertise is establishe d by knowledge. Thus, trustworthiness contributes to persuasion more than expertise does in some circumstances ( McGinnies & Ward , 1980 ). In addition, WOM in sport viewe rship tends to be an invitation rather than an information exchange because people enjoy watching sporting events with their family members, friends, and acquaintances. In fact, less than 2% of spectators of p rofessional or collegiate sport ing events attend the events by themselves ( Mullin et al. , 2007). Hence, it is rational that WOM has a greater influence on the receiver s decision making when trust exists between both communicators. In terms of message characteristics, both richness of content and strength of delivery showed significant positive effects on perceived influence . In past re search, rational and emotional message characteristics have been examined , particularly in experimental studies. In those studies, the richness of message content and strength of message delivery were controlled by researchers. On the other hand, the autho r focused on messages that had been actually sent

PAGE 52

52 and received by communicators in this study. Therefore, the findings of the current study can provide meaningful contributions to persuasion research. The Moderating Role of Interpersonal Factors and Recei ver C haracter istics Next, to accomplish the second purpose, the author conducted multiple group SEM analyses. As a result, significant moderating effects of homophily and involvement were found. For the high homophily group, richness of content and strengt h of delivery showed significant positive effects on perceived influence while richness of content also showed a significant positive effect for the average homophily group. According to the theory of social comparison, people inherently have a desire to evaluate their opinions and tend to compare their opinions with each other in order to get appropriate evaluations ( Festinger , 1954). When people are recommended to wa tch a sporting event, they attempt to get an appropriate evaluation of the event. In this situation, message content help s make an evaluation. If the recommen dation has rich content, an information receiver is more influenced by the information source no m atter how similar they are . In addition , Festinger (1954) suggests that the similarity between individuals play an important role during the process of evaluation. When people perceive the high level of similarity in value with the person to compare, the e valuation would be more stable. Besides, it has been suggested that when there is similarity between communicators, the information receiver is more likely to accept the information sender s opinion, because the similar person is an attractive source of in formation for the information receiver (Chu & Kim, 2011; Kandel , 1978; Touhey, 1975). When the attractive source strongly recommends a sporting event, the information receiver perceives more influence. This is why strength of delivery has a significant positive effect on the high homophily group. In contrast, when people perceive differences in value with the communicator, they tend to stop comparing and even communicating with the person ( Festinger , 1954). Interestingly, in this study, exper tise showed a significant negative

PAGE 53

53 effect on perceived influence for the average homophily group. When an information receiver is rec ommended a sporting event by an information sender , who has a relatively low level of similarity, the receiver assumes t hat he or she will not like the event as much as the sender does because they have different values. If the source is an expert in the event, the receiver m ay feel that the event is unlikely to be to his or her taste. In terms of involvement, the author foun d significant moderating effects on the relationship between (1) expertise and perceived influence, (2) trustworthiness and perceived influence, and (3) richness of content and perceived influence. For the average involvement group, richness of content sho wed significant positive effects , while trustworthiness showed a significant positive effect for the high involvement group . According to the disconfirmation model (Edwards & Smith, 1996), wh en people receive a message that is against their beliefs, they s earch their own memories and try to undermine evidence supporting the message. After the scrutiny, they decide on their final evaluation of the message. Following the model, the author explains that richness of content was the only factor hav ing a significant positive effect on perceived influence for people in the average involvement group because they were not highly involved with a sporting event and thus needed information to make proper evaluations regarding the event. On the other hand, richne ss of content did not have a significant effect on perceived influence for the high involvement group , but trustworthiness d id . One possible explanation for the insignificant relationship between richness of content and perceived influence is that highly i nvolved people already had enough information about the event. An a lternative explanation is that when highly involved people received a WOM message s about a sporting event, they gener ated positive thoughts and reduced the ability to process message conten ts. Mackie and Worth (1989) suggest that when people are in a positive mood , they are less able to process a message because the mood induce s positive thoughts in the memory and prevent s

PAGE 54

54 people from thinking critically. Similarly, some studies argue that happy feelings of an information receiver could reduce the capacity to process information thoroughly ( Schwarz, Bless, & Bohner, 1991) . To clarify the differences in process ing, further research should be conducted on WO M messages between high and low involvement viewer s . On the other hand, the author could not find significant moderating effects of tie strength and susceptibility. As shown in the previous chapter, the score s of tie strength items w ere high and the ir dis tribution s w ere negatively skewed. This implies that most of the participants received recommendation s from people connected with strong ties. Actually, over 80% of participants answered that their information sources were their family or friends whereas only 2.7% answered that their information sources were acquaintances. Thus, even though the author hypothesized that interpersonal factors work as moderators, tie strength might work as an ante cedent of perceived influence as some previous studies suggested (Bansal & Voyer, 2000; Sweeney et al. 2008). I n this research, however, the author focused only on people who had received recommendation s regarding a sporting event and actually watch ed the event. Thus , it may be possible to find a moderating effect if data can be collected from people who decided not to watch the event. In terms of susceptibility, the construct has been used in few sport management studies, so further research is needed to e valuate whether i t is useful in the sport s consumption context. Implication Theoretical Implication T he present study utilized WOM impact model (Sweeney et al., 2008) as its main framework . However, the author modified the model by creating moderators while t he original model considers all factors (i.e., personal, message, interpersonal , and situational factors) as antecedents of perceived influence s of WOM. In this study, the author supposed that absolute

PAGE 55

55 effects of WOM are determined by characteristic s of the message and message sender a nd that interpersonal factors and the receiver character istics moderate the main effects. The data supported the research model presented in this study . Therefore, the findings of the current study will be useful for fu ture research addressing WOM effects. Practical Implication T his study provided useful information for practitioners who try to utilize WOM communication. The results showed that the richness of content is the only factor that has influence on viewer s who do not have a high level of involvement. The factor also showed a significant positive effect on viewer s watching behavior even if there was not high homophily between the communicators. These results suggest that sports organization s should provide infor mation and experience s that make WOM message s richer to improve their influence. Moreover , the results also suggest ed that experts recommendations could have negative effects on viewer s watching behavior . Thus, sports organizations should not rely only on core fans who have a high level of expertise. Rather, they should provide information and stories that are understandable and useful for casual fans that have a low level of expertise when they engage in W OM communication. Limitation s and Future Research This study ha s some limitations. First , the results relied on participants retrospective evaluations of WOM communication . Thus, the ir evaluations could d iffer from the per ceptions they h eld right after the communication. Also, focal constructs were measured based on participants perceptions. T h e author did not acquire factual information about sourc es ( e. g. , demographics ) or WOM message (e.g., message volume, message content). T hus, it is unknown what kind of message was perceived as being rich or what kind of person was considered as an expert . For instance, one of the items measuring richness of content is the message was

PAGE 56

56 informative. Even if participants agreed with the statement , some of the m might get information about athletes who wo uld perform in the event wh ile other participants might hear detailed information about the event itself (e.g., date, time, place, and price). Therefore, further research is needed to investigate how people eval uate source and message characteristics . Furthermore , t he author focused on drawing a big picture of WOM influence in sport viewer ship, so some questions still remain. Why does expertise ha ve a negative effect on perceived influence when homophily between the communicator s is relatively low? Why were high involved viewer s not influenced by richness of content? F uture research should address these questions . The author suggests that an experimental method would work better than a survey t o examine such detailed theoretical relationship s . Also , future research should address how sport organizations can improve trustworthiness, richness of content, and strength of delivery. As discussed in Chapter 2, previous studies have often used WOM as a useful outcom e variable of other constructs. Researchers use WOM intention as a dependent variable because they assume that WOM brings more customers and increases revenues. Although a tremendous number of studies have examined antecedents of WOM intention based on thi s assumption, much less studies have examined whether the antecedents are truly effective for customer acquisition and revenue generation. Based on the findings of this research, researcher s can address how other constructs (e.g., satisfaction, service qua lity, and commitment ) lead to effective WOM rather than WOM intention. The refore, the author believes that future research can generate further useful information and knowledge .

PAGE 57

57 APP ENDIX SURVEY QUESTIONNAIRE Dear Participant; This survey aims to clarify antecedents of word of mouth impact on sport spectators . It would be greatly appreciated if you would simply complete the following questionnaire. It will take approximately 10 minutes to complete this survey. There are no known risks to you if you decide to participate in this survey . We guarantee that we wil l not obtain any identifying information from you. Your participation is voluntary and there is no penalty if you do not participate. Also, t here is no compensation or direct benefits to you for participating in the study. Regardless of whether you choose to participate, please let me know if you would like a summary of my findings. If you have any questions or concerns about completing the questionnaire or about being in this study, please contact us at the address below. You can withdraw your consent at a ny time with penalty. If you have any questions about your rights as a research participant, please contact the IRB at (352) 273 9600. Thank you again for your cooperation and the valuable information you are providing in this survey. Sincerely, A kira Asada Master s student Sport Management University of Florida a .asada @ufl.edu (352) 2 83 2667 Yong Jae Ko, PhD Associate Professor Sport Management Program University of Florida Rm.186A Florida Gym PO Box 118208 Gainesville, FL 32611 8208 yongko@hhp.ufl.edu (352) 392 4042x1277

PAGE 58

58 This section is about your sporting event attendance experience and your opinions of other Q1 . In the last three months, has anyone recommended you to watch a sporting event? a. Yes b. No Q2. What kind of sporting event was it? a. Professional b. Collegiate c. Olympic Games d. Soccer World Cup e. Other Q3. To the best of your recollection, when did you receive the recommendation? Day________ Month__________ Q4. Who did recommend you to watch the sporting event? (Please write down the name or initial of the person) __________________________ Q5. How did you receive the recommendation? a. Face to face b. By telephone c. By e mail/text message d. By social networking site (e.g., Facebook, Twitter, Google Plus) e. Other ______________ ___________________________ ________________________________________________ Q7. Please check the appropriate box that describes your feeling about watching the sporting event. Boring 1 2 3 4 5 6 7 Exciting Uninteresting 1 2 3 4 5 6 7 Interesting Worthless 1 2 3 4 5 6 7 Valuable Unappealing 1 2 3 4 5 6 7 Appealing Irrelevant 1 2 3 4 5 6 7 Relevant Unimportant 1 2 3 4 5 6 7 Important

PAGE 59

59 Q8. To the best of your recollection, what was the date of the event? Day________ Month__________ Year__________ Q9. Have you actually watched the sporting event? a. Yes (in person) b. Yes (on TV/Internet) c. No d. Not yet Q10. Please che ck the number that best describes your perception about the recommendation. Not at all Very much a. Did the recommendation have an influence on your decision to watch the sporting event? 1 2 3 4 5 6 7 b. To what extent was your decision to watch the sporting event influenced by the recommendation ? 1 2 3 4 5 6 7 c. Did the recommendation help you to make a decision to watch the sporting event? 1 2 3 4 5 6 7 Q10. Please che ck the number that best describes your perception about the recommendation. a. How did your attitude toward watching the sporting event change after the recommendation? Less favorable 1 2 3 4 5 6 7 More favorable Not at all Very much b. Did the recommendation have an influence on your attitude toward watching the sporting event ? 1 2 3 4 5 6 7 c. To what extent do you think your attitude toward watching the sporting event was influenced by the recommendation? 1 2 3 4 5 6 7

PAGE 60

60 This section is about the recommendation which you received. Q11. Please check the number that best describes the recommendation you received. The recommendation about the sporting event Strongly disagree Strongly agree a. Informative 1 2 3 4 5 6 7 b. Clear 1 2 3 4 5 6 7 c. Specific 1 2 3 4 5 6 7 d. Elaborate 1 2 3 4 5 6 7 e. Explicit 1 2 3 4 5 6 7 The recommendation about the sporting event Strongly disagree Strongly agree a. In a powerful tone 1 2 3 4 5 6 7 b. In a strong way 1 2 3 4 5 6 7 c. In an important manner 1 2 3 4 5 6 7 d. Enthusiastically 1 2 3 4 5 6 7

PAGE 61

61 This section is about the person who recommended you to watch the sporting event. Q12. Please che ck the appropriate box that best describes the person who recommended the sporting event. Und ependable 1 2 3 4 5 6 7 D ependable Dish onest 1 2 3 4 5 6 7 H onest Unr eliable 1 2 3 4 5 6 7 R eliable Ins incere 1 2 3 4 5 6 7 S incere Unt rustworthy 1 2 3 4 5 6 7 T rustworthy Not an expert 1 2 3 4 5 6 7 E xpert Ine xperienced 1 2 3 4 5 6 7 E xperienced Unk nowledgeable 1 2 3 4 5 6 7 K nowledgeable Uns killed 1 2 3 4 5 6 7 S killed

PAGE 62

62 This section is about the relationship between you a nd the person who recommended the sporting event. Q 14 . Which of the items below best describe the relationship between you and the person who recommended the sporting event ? 1. Family 2. Friend 3. Close friend 4. Boy/Girl friend 5. Fiancé 6. Acquaintance 7. Co worker 8. Othe r ( ) Q15. On average, how often do you talk with the person who recommended the sporting event ? 1. Daily 2. Weekly 3. Monthly 4. Yearly 5. Less than once a year Q16. How many years have you known the person who recommended the sporting event ? Y ear s ___________ M onth s __________ Q17. Please check the number that best describes your perception about the relationship between you and the person who recommended the sporting event. Not close at all Ext remely close a. How close are you with the person who recommended the sporting event ? 1 2 3 4 5 6 7 Very unlikely Very likely b. How likely are you to share personal confidence with the person? 1 2 3 4 5 6 7 c. How likely are you to extend everyday assistance to the person? 1 2 3 4 5 6 7 d. How likely are you to spend a free afternoon with the person? 1 2 3 4 5 6 7 Q18. Please check the number that best describes your perception about the relationship between you and the person who recommended the sporting event. Not at all similar Extremely similar a. Considering your outlook on life, how similar are you and the person? 1 2 3 4 5 6 7 b. Considering your likes and dislikes, how similar are you and the person? 1 2 3 4 5 6 7 c. Considering your values how similar are you and the person? 1 2 3 4 5 6 7 d. Considering your experiences how similar are you and the person? 1 2 3 4 5 6 7

PAGE 63

63 This section is about your consumption behavior and demographics. Q19. Please check the number that describes your consumption behavior. Strongly disagree Strongly agree a. It is important that others like the products and brands that I buy. 1 2 3 4 5 6 7 b. When buying products, I generally purchase those brands that I think others will approve of. 1 2 3 4 5 6 7 c. I like to know what brands and products make good impressions on others. 1 2 3 4 5 6 7 d. I achieve a sense of belonging by purchasing the same products and brands that others purchase. 1 2 3 4 5 6 7 Q20. I am 1. Male 2. Female Q21. How old are you? ______ ____ Q22. What is your ethnicity? 1. African American 2. Asian 3. Caucasian 4. Hispanic 5. Native American 6. Other ( ) Q23. What is the h ighest level of education you have completed? 1. No high school 2. High school grad 3. Some college 4. Associate degree 5. 6. Graduate degree 1. Under $15,000 2. $15,000 ~ $24,999 3. $25,000 ~ $34,999 4. $35,000 ~ $49,999 5. $50,000 ~ $74,999 6. $75,000 ~ $99,999 7. $100,000 ~ $149,999 8. $150,000 ~ $199,999 9. $200,000 and above

PAGE 64

64 L IST OF REFERENCES Allsop, D. T., & Bassett, B. R. (2007). Word of mouth research: Principles and applications. Journal of Advertising Research, 47 (4), 398 411. Anderson, E. W. (1998). Customer satisfaction and word of mouth. Journal of service research , 1 , 5 17. Aristotle, R. (1924). trans. W. Rhys Roberts. The Works of Aristotle (Chicago: Encyclopedia Britannica, 1952), 1378a20 30 . Bansal, H. S., & Voyer, P. A. (2000). Word of mouth processes within a services purchase decision context. Journal of S ervice R esearch , 3 , 166 177. Batra, R., & Ray, M. L. (1986). Affective responses mediating acceptance of advertising. Journal of Consumer Research, 1 3 (2), 234 249. Bearden, W. O., Netemeyer, R. G., & Teel, J. E. (1989). Measurement of consumer susceptibility to interpersonal influence. Journal of Consumer Research, 15 , 473 481. Block, L. G., & Keller, P. A. (1995). When to accentuate the negative: The effects of per ceived efficacy and message framing on intentions to perform a health related behavior. Journal of Marketing Research , 32 , 192 203. Bohner, G., Ruder, M., & Erb, H. (2002). When expertise backfires: Contrast and assimilation effects in persuasion. The Briti sh Journal of Social Psychology / the British Psychological Society, 41 , 495 519. Boomsma, A. (1985). Nonconvergence, improper solutions, and starting values in LISREL maximum likelihood estimation. Psychometrika , 50 , 229 242. Borgida, E., & Nisbett, R. E. (1977). The Differential Impact of Abstract vs. Concrete Information on Decisions1. Journal of Applied Social Psychology , 7 , 258 271. Brown, J. J., & Reingen, P. H. (1987). Social ties and word of mouth referral behavior. Journal of Consume r Research , 14 , 350 362. Brown, J., Broderick, A. J., & Lee, N. (2007). Word of mouth communication within online communities: Conceptualizing the online social network. Journal of interactive marketing , 21 (3), 2 20. Browne, M. W., & Cudeck, R. (1992). Alternative ways of assessing model fit. Sociological Methods & Research , 21 , 230 258. Buhrmester, M., Kwang, T., & Gosling, S. D. (2011). Amazon's Mechanical Turk a new source of inexpensive, yet high quality, data? Perspectives on Psychological Science , 6 , 3 5. Burnkrant, R. E., & Cousineau, A. (1975). Informational and normative social influence in buyer behavior. Journal of Consumer Research, 2 , 206 215.

PAGE 65

65 Chaiken, S. (1980). Heuristic versus systematic information processing and the use of source versu s message cues in persuasion. Journal of Personality and Social Psychology , 39 , 752 766. Chaiken, S., & Maheswaran, D. (1994). Heuristic processing can bias systematic processing: Effects of source credibility, argument ambiguity, and task importance on at titude judgment. Journal of Personality and Social Psychology, 66 , 460 473 . Chang, Y., & Ko, Y. J. consumption. 2011 Sport Marketing Association Conference, Houston, TX. Chu, S., & Kim, Y. (2011). Determinants of consumer engagement in electronic word of mouth (eWOM) in social networking sites. International Journal of Advertising, 30 (1), 47 75. Clark, J. K., Wegener, D. T., Habashi, M. M., & Evans, A. T. (2012). Source Expertise and Persuasion The Effects of Perceived Opposition or Support on Message Scrutiny. Personality and Social Psychology Bulletin , 38 (1), 90 100. Collins, R. L., Taylor, S. E., Wood, J. V., & Thompson, S. C. (1988). The vividness effect: Elusive or illusory? Jour nal of Experimental Social Psychology, 24 , 1 18. Cunningham, G. B., & Kwon, H. (2003). The theory of planned behaviour and intentions to attend a sport event. Sport Management Review , 6 , 127 145. Davis, B. H., & Mason, P. R. (2007). More than the words: Using stance shift analysis to identify crucial opinions and attitudes in online focus groups. Journal of Advertising Research, 47 (4), 496 506. DeBono, K. G., & Harnish, R. J. (1988). Source expertise, source attractiveness, and the processing of persuasive information: A functional approach. Journal of Personality and Social Psychology, 55 , 541 546. De Bruyn, A., & Lilien, G. L. (2008). A multi stage model of word of mouth influence through viral marketing. International Journal of Research in Marketing , 25 , 151 163. Deutsch, M., & Gerard, H. B. (1955). A study of normative and informational social influences upon individual judgment. Journal of Abnormal and Social Psychology, 51 , 629 636. Eagly, A. H., Wood, W., & Chaiken, S. (1978). Causal inferences about communicators and their effect on opinion change. Journal of Personality and Social Psychology, 36 , 424 435. Edwards, K., & Smith, E. E. (1996). A disconfirmation bias in the evaluation of arguments. Journal of Personality a nd Social Psychology, 71 , 5 24. Festinger, L. (1954). A theory of social comparison processes. Human relations, 7(2), 117 140.

PAGE 66

66 Fisher, E. (2013, April 1). Arizona Diamondbacks: Having some fun. SportsBusiness Journal. Retrieved March 15 , 201 4 , from http://www.sportsbusinessdaily.com/Journal/Issues/2013/04/01/In Depth/Diamondbacks. aspx?hl=viral%20&sc=0 . Fornell, C., & Larcker, D. F. ( 1981). Structural equation models with unobservable variables and measurement error: Algebra and statistics. Journal of Marketing Research, 18 , 382 388 . Frenzen, J. K., & Davis, H. L. (1990). Purchasing behavior in embedded markets. Journal of Consumer Research , 17 , 1 12. Gilly, M. C., Graham, J. L., Wolfinbarger, M. F., & Yale, L. J. (1998). A dyadic study of interpersonal information search. Journal of the Academy of Marketing Science , 26 , 83 100. Granovetter, M. S. (1973). The strength of weak ties. A merican journal of sociology , 78, 1360 1380. Granovetter, M. S. (1983). The strength of weak ties: A network theory revisited. Sociological theory , 1 (1), 201 233. Hair , J. F. , Black, W. C., Babin, B. J., & Anderson, R. E., (20 0 9 ). Multivariate data analysis . Upper Saddle River, NJ: Prentice Hall. Harrison Walker, L. J. (2001). The measurement of word of mouth communication and an investigation of service quality and customer commitment as potential antecedents. Journal of Service Research , 4 , 60 75. Herr, P. M., Kardes, F. R., & Kim, J. (1991). Effects of word of mouth and product attribute information on persuasion: An accessibility diagnosticity perspective. Journal of Consumer Research , 17, 454 462. Hightower Jr, R., Brady, M. K., & Baker, T. L. (2 002). Investigating the role of the physical environment in hedonic service consumption: an exploratory study of sporting events. Journal of Business Research, 55 , 697 707. Homer, P. M., & Kahle, L. R. (1990). Source expertise, time of source identificatio n, and involvement in persuasion: An elaborative processing perspective. Journal of Advertising , 19 (1) , 30 39. Hopkinson, G. C., & Pujari, D. (1999). A factor analytic study of the sources of meaning in hedonic consumption. European Journal of Marketing, 33 , 273 294. Hovland, C. I., Janis, I. L., & Kelley, H. H. (1953). Communication and persuasion; psychological studies of opinion change . Westport, Conn: Greenwood. Hovland, C. I., & Weiss, W. (1951). The influence of source credibility on communication effectiveness. Public opinion quarterly , 15 , 635 650.

PAGE 67

67 Hu, L., & Bentler, P. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling a Multidisciplinary Journal, 6 , 1 55. Kai, W., Wang, E. G., & Cheng Kiang, F. (2009). Influence of Web Advertising Strategies, Consumer Goal Directedness, and Consumer Involvement on Web Advertising Effectiveness. International Journal of Electronic Commerce , 13 (4), 67 95. Kandel, D. B. (1978). Similarity in real life adolescent friendship pairs. Journal of personality and social psychology, 36 , 306 312. Keller, P. A., & Block, L. G. (1997). Vividness effects: A resource matching perspective. Journal of Consumer Research , 24 , 295 304. Kelman, H. C., & Hovland, C. I. (1953). " Reinstatement" of the communicator in delayed measurement of opinion change. The Journal of Abnormal and Social Psychology , 48 , 327 335 . Kim, Y. K., & Trail, G. (2011). A conceptual framework for understanding relationships between sport consumers and sport organizations: A relationship quality approach. Journal of Sport Management, 25 , 57 69. Kline, R. B. (2011). Principles and practice of structural equation modeling . New York: Guilford. Kozinets, R., de Valck , K., Wojnicki, A., & Wilner, S. (2010). Networked narratives: Understanding word of mouth marketing in online communities. Journal of Marketing, 74 , 71 89. Kuenzel, S., & Yassim, M. (2007). The effect of joy on the behaviour of cricket spectators: the med iating role of satisfaction. Managing Leisure , 12 , 43 57. Lazarsfeld, P. F., & Katz, E. (1955). Personal influence. New York , 174. Levin, D. Z., & Cross, R. (2004). The strength of weak ties you can trust: The mediating role of trust in effective knowledge transfer. Management Science, 50 (11), 1477 1490. Liu, Y. (2006). Word of mouth for movies: Its dynamics and impact on box office revenue. Journal of M arketing , 70 (3) , 74 89. Mackie, D. M., & Worth, L. T. (1989). Processing deficits and the mediation of positive affect in persuasion. Journal of P ersonality and S ocial P sychology , 57 , 27 40. susceptibility to peer influence. Journal of Retailing, 80 , 101 116. Mangold, W. G., & Faulds, D. J. (2009). Social media: The new hybrid element of the promotion mix. Business horizons , 52 , 357 365.

PAGE 68

68 McGinnies, E., & Ward, C. D. (1980). Better liked than right: Trustworthiness and expertise as factors in credibility. Personality and Social Psychology Bulletin, 6 (3), 467 472. McPherson, M., Smith Lovin, L., & Cook, J. M. (2001). Birds of a feather: Homophily in social networks. Annual Review of Sociology, 27 (1), 415 444. Mills, J., & Jellison, J. M. (1967). Effect on opinion change of how desirable the communication is to the audience the communicator addressed. Journal of Personality and Social Psychology, 6 (1), 98 101. Moore, D. L., Hausknecht, D., & Thamodaran, K. (1986). Time compression, response opport unity, and persuasion. Journal of Consumer Research, 13 , 85 99. Mullin, B. J., Hardy, S., & Sutton, W. A. (2007). Sport marketing . Champaign, IL: Human Kinetics Murray, K. B. (1991). A test of services marketing theory: consumer information acquisition activities. Journal of Marketing , 55 , 10 25. Netemeyer, R. G., & Bearden, W. O. (1992). A comparative analysis of two models of behavioral intention. Academy of Marketing Science Journal, 20 , 49 59. Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory. New York: McGraw Hill. Ohanian, R. (1990). Construction and validation of a scale to measure celebrity endorsers' perceived expertise, trustworthiness, and attractiveness. Journal of Advertising, 19 (3), 39 52. O'Keefe, D. J. (2002). Persuasion: Theory & research . Thousand Oaks, CA: Sage Publications. Park, D. H., Lee, J., & Han, I. (2007). The effect of on line consumer reviews on consumer purchasing intention: The moderating role of involvement. International Journal of Electronic Commerce , 11 (4 ), 125 148. Petty, R. E., & Cacioppo, J. T. (1986). Communication and persuasion: Central and peripheral routes to attitude change . New York: Springer Verlag . Petty, R. E., Cacioppo, J. T., & Goldman, R. (1981). Personal involvement as a determinant of argument based persuasion. Journal of Personality and Social Psychology, 41 , 847 855. Petty, R. E., Cacioppo, J. T., & Schumann, D. (1983). Central and peripheral routes to advertising effectiveness: The moderating role of involvement. Journal of Consumer Research, 10 (2), 135 146 . Pornpitakpan, C. (2004). The persuasiveness of source credibility: A critical review of five decades' evidence. Journal of Applied Social Psychology , 34 , 243 281.

PAGE 69

69 Rhine, R. J., & Severance, L. J. (1970). Ego involvement, discrepancy, source credibility, and attitude change. Journal of Personality and Social Psychology, 16 , 175 190. Rogers, E. M., & Bhowmik, D. K. (1970). Homophily heterophily: Relational concepts for communication research. The Public Opinion Quarterly, 34 , 523 538. Rosselli, F., Skelly, J. J., & Mackie, D. M. (1995). Processing rational and emotional messages: The cognitive and affective mediation of persuasion. Journal of Experimental Social Psychology , 31 , 163 190. Schwarz, N., Bless, H., & Bohner, G. (1991). Mood and persuasion affective states influence the processing of persuasive commu nications. Advances in Experimental Social Psychology, 24 , 161 199. Snyder, M. (1974). Self monitoring of expressive behavior. Journal of Personality and Social Psychology, 30 , 526 537. Swanson, S. R., Gwinner, K., Larson, B. V., & Janda, S. (2003). Motivations of college student game attendance and word of mouth behavior: the impact of gender differences. Sport Marketing Quarterly , 12 , 151 162. Sweeney, J. C., Soutar, G. N., & Mazzarol, T. (2008). Factors influencing word of mouth effectiveness: receiver perspectives. European Journal of Marketing , 42 , 344 364. Sweeney, J. C., Soutar, G. N., & Mazzarol, T. (2012). Word of mouth: measuring the power of individual messages. European Journal of Marketing , 46 , 237 257. Taylor, S. E., & Thompson, S. C. (1982). Stalking the elusive "vividness" effect. Psychological Review, 89 , 155 181 . Tobin, S., & Raymundo, M. (2009). P ersuasion by causal arguments: The motivating role of perceived causal expertise. Social Cognition, 27 (1), 105 127. Tormala, Z. L., Briñol, P., & Petty, R. E. (2006). When credibility attacks: The reverse impact of source credibility on persuasion. Journal of Experimental Social Psychology , 42 , 684 691. Touhey, J. C. (1975). Interpersonal congruency, attitude similarity, and interpersonal attraction. Journal of Research in Personality, 9, 66 73. Trail, G., & James, J. (2001). The motivation scale for sport consumption: Assessment of the scale's psychometric properties. Journal of Sport Behavior, 24 (1), 108 127. Trusov, M., Bucklin, R. E., & Pauwels, K. H. (2008). Effects of word of mouth versus traditional marketing: findings from an internet social networking site. Robert H. Smith School Research Paper No. RHS , 06 065.

PAGE 70

70 Villanueva, J., Yoo, S., & Hanssens, D. M. (2008). The impact of marketing induced versus word of mouth customer acquisition on customer equity growth. Journal of Marketing Research , 45 , 48 59. Walker, M., & Kent, A. (2009). Do Fans Care? Assessing the Influence of Corporate Social Responsibility on Consumer Attitudes in the Sport Industry. Journal of Sport Management , 23 . 743 769 Wangenheim, F. v., & Bayon, T. (2004). The effect of word of mouth on services switching: Measurement and moderating variables. European Journal of Marketing, 38 , 1173 1185. Wood, W., & Kallgren, C. A. (1988). Communicator Attributes and Persuasion Recipients' Access to Attitude Relevant Information in Memory. Personality and Social Psychology Bulletin , 14 (1), 172 182. Wright, P. L. (1973). The cognitive processes mediating acceptance of advertising. Journal of Marketing Res earch, 10 , 53 62. Ye, Q., Law, R., Gu, B., & Chen, W. (2011). The influence of user generated content on traveler behavior: An empirical investigation on the effects of e word of mouth to hotel online bookings. Computers in Human Behavior , 27 , 634 639. Zai chkowsky , J. L. (1985). Measuring the involvement construct. Journal of Consumer Research, 12 , 341 352.

PAGE 71

71 BIOGRAPHICAL SKETCH Akira Asada earned his Master of Science degree (sport management) in the C ollege of H ealth and Human Performance in University of Florida. He received his Bachelor of Science degree in sport sciences from Waseda University in Japan. His primary research interest is in consumer behavior , specifically communicative behavior , in sport spectator settings . He is currently working on a research project focusing on word of mouth behavior in sport viewership with Dr. Yong Jae Ko . The ultimate goal of his research is to provide better understandings of sports consumer behavior and develop the strategy to manag e sports teams fan bases . He is a member of the N orth A merican S ociety for S port M anagement and S port M arketing A ssociation . He will continuously study sport management in the doctoral program in the C ollege of Health and Human Performance in University of Florida .