<%BANNER%>

Adoption of Twitter and its effectiveness in e-WOM

Permanent Link: http://ufdc.ufl.edu/UFE0043775/00001

Material Information

Title: Adoption of Twitter and its effectiveness in e-WOM
Physical Description: 1 online resource (271 p.)
Language: english
Creator: Son, Hyunsang
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: adoption -- ewom -- hedonic -- newmedia -- perceivedfit -- productcategory -- socialinfluence -- socialmedia -- sourcecredibility -- sourcesimilarity -- tam -- tpb -- tra -- twitter -- utilitarian
Journalism and Communications -- Dissertations, Academic -- UF
Genre: Mass Communication thesis, M.A.M.C.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: As a one of the fastest growing areas for online marketing, social media, specifically Twitter, a social networking site, could provide prominent opportunities for marketers. However, little scholarship has empirically investigated why consumers use Twitter and how to utilize Twitter as a brand information exchanging tool. This study mainly contributes to three different areas. First, the study investigates the motivation associated with Twitter by comparing it to other product-information sources such as advertising, thus providing a relative eWOM effectiveness perspective. Second, factors that influence eWOM intentions from sender and receiver perspectives are analyzed thus offering both marketer and consumer contexts. Third, the study tries to identify appropriate types of products about which people share information, often by making comparisons with other product categories. Ultimately, this study attempts to answer questions about what and how people decide to adopt or exchange product-related information. This study employed 4 independent variables-perceived usefulness, perceived ease of use, conformity to subjective norm, and demographical factors (age, education level and gender)-for Twitter adoption. In addition, 5 independent variables-perceived similarity, perceived credibility, product category (utilitarian and hedonic), perceived fit (utilitarian and hedonic), and demographic factors (age, education level and gender)-which might affect consumer attitude toward marketing messages in Twitter, were examined in this study. A pretest was conducted before the main test. For the main test, this study employed an online survey by utilizing an online consumer panel for U.S. consumers based on the results of the pretest. Statistical analyses found that perceived usefulness affected individuals' attitudes toward Twitter and their Twitter usage. Also, the perceived ease of use influenced individuals' attitudes toward Twitter directly and Twitter usage indirectly. However, individual conformity to subjective norms and demographical factors did not affect individuals' attitude toward Twitter or their Twitter usage. In terms of eWOM-related independent variables, perceived similarity, perceived credibility, product category of hedonic dimension, and perceived fit of hedonic product dimensions affected consumers' evaluation of brand attitudes. In addition, perceived credibility, perceived fit of utilitarian dimension, and perceived fit of hedonic dimension affected consumers' eWOM-spreading intention.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Hyunsang Son.
Thesis: Thesis (M.A.M.C.)--University of Florida, 2011.
Local: Adviser: Chan-Olmsted, Sylvia M.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2013-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0043775:00001

Permanent Link: http://ufdc.ufl.edu/UFE0043775/00001

Material Information

Title: Adoption of Twitter and its effectiveness in e-WOM
Physical Description: 1 online resource (271 p.)
Language: english
Creator: Son, Hyunsang
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: adoption -- ewom -- hedonic -- newmedia -- perceivedfit -- productcategory -- socialinfluence -- socialmedia -- sourcecredibility -- sourcesimilarity -- tam -- tpb -- tra -- twitter -- utilitarian
Journalism and Communications -- Dissertations, Academic -- UF
Genre: Mass Communication thesis, M.A.M.C.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: As a one of the fastest growing areas for online marketing, social media, specifically Twitter, a social networking site, could provide prominent opportunities for marketers. However, little scholarship has empirically investigated why consumers use Twitter and how to utilize Twitter as a brand information exchanging tool. This study mainly contributes to three different areas. First, the study investigates the motivation associated with Twitter by comparing it to other product-information sources such as advertising, thus providing a relative eWOM effectiveness perspective. Second, factors that influence eWOM intentions from sender and receiver perspectives are analyzed thus offering both marketer and consumer contexts. Third, the study tries to identify appropriate types of products about which people share information, often by making comparisons with other product categories. Ultimately, this study attempts to answer questions about what and how people decide to adopt or exchange product-related information. This study employed 4 independent variables-perceived usefulness, perceived ease of use, conformity to subjective norm, and demographical factors (age, education level and gender)-for Twitter adoption. In addition, 5 independent variables-perceived similarity, perceived credibility, product category (utilitarian and hedonic), perceived fit (utilitarian and hedonic), and demographic factors (age, education level and gender)-which might affect consumer attitude toward marketing messages in Twitter, were examined in this study. A pretest was conducted before the main test. For the main test, this study employed an online survey by utilizing an online consumer panel for U.S. consumers based on the results of the pretest. Statistical analyses found that perceived usefulness affected individuals' attitudes toward Twitter and their Twitter usage. Also, the perceived ease of use influenced individuals' attitudes toward Twitter directly and Twitter usage indirectly. However, individual conformity to subjective norms and demographical factors did not affect individuals' attitude toward Twitter or their Twitter usage. In terms of eWOM-related independent variables, perceived similarity, perceived credibility, product category of hedonic dimension, and perceived fit of hedonic product dimensions affected consumers' evaluation of brand attitudes. In addition, perceived credibility, perceived fit of utilitarian dimension, and perceived fit of hedonic dimension affected consumers' eWOM-spreading intention.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Hyunsang Son.
Thesis: Thesis (M.A.M.C.)--University of Florida, 2011.
Local: Adviser: Chan-Olmsted, Sylvia M.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2013-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0043775:00001


This item has the following downloads:


Full Text

PAGE 1

1 ADOPTION OF TWITTER AND ITS EFFECTIVENESS IN e WOM By HYUNSANG SON A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS IN MAS S COMMUNICATION UNIVERSITY OF FLORIDA 2011

PAGE 2

2 2011 Hyunsang Son

PAGE 3

3 To my beloved grandmother Young Sook Hwang my family, friends, and Young Eun

PAGE 4

4 ACKNOWLEDGMENTS commitmen t, and time. I would like to thank all of the people who encouraged me to First and foremost, I would like to express my deep gratitude to my advisor and thesis chair, Dr. Sylvia Chan Olmsted, Professor of Telecommunication, Associated Dean for Research, and senior research associate of the Public Utility Research Center at the University of Florida (PURC) for her enduring support. I learned everything from her, including how to capture a research idea, how to organize research topics, and how to elaborate upon my idea and write a paper. I am very fortunate to have her as my advisor, thesis chair, and mentor. From lectures to research meetings, her insights were invaluable in the development and completion of this work. It was great honor for me to complete my thesis under her guidance and to coordinate several conference papers with her. I would like to express my gratitude to all my thesis committee members. I am thankful to Dr. David Ostroff, Professor and Chair of the Department of Telecommunication. Despite his busy schedule as a department chair, he provided me in building and my theoretical framework of social infl uence perspective. In particular, he offered me great guidance with my data analysis. I also want to express my special and sincere gratitude to other faculty members who provided me with insightful lectures, comments, and individual meetings during my mas throughout my graduate studies. Through her lectures, regular research meetings, and

PAGE 5

5 from working as her research assistant, I was able to expand my research area throughout various contexts, including public relations and health communications. The knowledge and insight of Dr. Spiro Kiousis, the Chair of the Department of Public Relations, were particularly helpful in building and improving my perspective toward various persu asion theories. I also want to thank Dr. Johanna Cleary for her insightful lectures and discussions, which helped me to build mass communication theories. I would like to thank all of my Korean friends in the College of Journalism and Communications. I es pecially appreciate Jinsoo Kim, Chunsik Lee, Dae Hee Kim, Sun young Park, Hyejoon Rim, Doori Song, Moon Hee Cho, and Hanna Park for their academic advice, research opportunities and helping me recruit samples for my research; the Korean Communigators: Jing hong Ha, Dae wook Kim, Sooyeon Kim, Junga Kim, Jin sook Im, Eun Hwa Jung, Jaejin Lee, Eun Go, Kyung Gook Park, Yoo Jin Chung, Wansup Jung, Eun Soo Rhee, Jung Min Park, Jiyoung Kim, Hyunji Lim, Kang Hoon Sung, Jihye Kim, and Seul Lee; and the Korean Baptist Church community for their advice and support. Also, I would like to express my sincere gratitude to my undergraduate mentors. I thank Dr. Jae shin Lee, Dr. Minkyu Lee, and Dr. Dong Kyu Sung, whose encouragement made it possible for me to pursue graduate studies. I will always keep Also, I would like to express my appreciation to my friends in Korea w ho shared my concerns: BBang, Ang, Sir. Jung saeng Lee, Kwang, Jungman, JJiwan, and Yong woon.

PAGE 6

6 My greatest appreciation goes to my parents and family members. My father, Changwook Son, my mother, Dr. Myunghee Lee, and my sister, Hyun Ji Son, receive my de epest gratitude and love for their dedication and many years of support. They always loved and believed in me. Also, I appreciate my grandfather, Yoonmo Son, grandmother, Do young Yoon, and other family members for their support and love. I must acknowled ge the most heartfelt thanks to God for generosity. Whenever I face difficulty, God saves me. Last but not least, I would like to thank my Muse, Young Eun, who shared every single joy and sorrow with me since we met in high school. She is my inspiration, especially during the time we have shared while traveling together throughout our Journalism and Communications at the University of Florida and are proud Gators.

PAGE 7

7 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ .......... 12 LIST OF FIGURES ................................ ................................ ................................ ........ 14 LIST OF ABBREVIATIONS ................................ ................................ ........................... 15 ABSTRACT ................................ ................................ ................................ ................... 16 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 18 WOM and eWOM ................................ ................................ ................................ ... 18 Social Media as a new eWOM Marketing Tool ................................ ....................... 19 Purpose of the Study ................................ ................................ .............................. 24 Limited Academic Research ................................ ................................ ............. 24 Research Directions ................................ ................................ ......................... 25 2 THEORETICA L FRAMEWORK ................................ ................................ .............. 28 Social Networking Site Adoption Studies ................................ ................................ 28 Current Research Trend ................................ ................................ ................... 28 Media Use and adoption Research ................................ ................................ .. 29 Consumer Characteristics in Media Adoption ................................ ................... 32 Social Influence ................................ ................................ ................................ 32 Social influence in new media adoption ................................ ........................... 33 Technology Acceptance Model ................................ ................................ ........ 35 Revision of TAM ................................ ................................ ............................... 37 Integration of Social Influence and TAM ................................ ........................... 38 WOM and eWOM ................................ ................................ ................................ ... 39 Word of Mouth ................................ ................................ ................................ 39 Electronic Word of Mouth ................................ ................................ ................. 41 3 LITERATURE REVIEW AND RESEARCH QUESTIONS ................................ ....... 57 Twitter Usage Motivation ................................ ................................ ........................ 57 Testing Social Influence in Twitter Adoption ................................ ..................... 57 Testing T echnology Acceptance Model in Twitter adoption .............................. 59 Perceived usefulness ................................ ................................ ................. 59 Perceived ease of use ................................ ................................ ............... 60 Consumer Related Characteristics ................................ ................................ ... 61 Age ................................ ................................ ................................ ............ 61

PAGE 8

8 Education level ................................ ................................ ........................... 62 Gender ................................ ................................ ................................ ....... 62 Attitude Toward Twitter and Actual Usage of Twitter ................................ ........ 63 WOM and eWOM Related Factors ................................ ................................ ......... 64 Communicator Characteristics ................................ ................................ ......... 64 Perceived similarity ................................ ................................ .................... 64 Source credibility ................................ ................................ ........................ 65 Product Related Characteristics ................................ ................................ ....... 67 Hedonic and utilitarian product category ................................ .................... 67 Perceived fit ................................ ................................ ............................... 71 Attitude Toward the Brand, eWOM Intention and Purchase Intention .............. 72 Consumer Rel ated Characteristics ................................ ................................ ... 73 4 METHOD ................................ ................................ ................................ ................ 81 Measurement ................................ ................................ ................................ .......... 81 Pretest ................................ ................................ ................................ .................... 81 Participants and Procedure ................................ ................................ .............. 81 Reliability Tests ................................ ................................ ................................ 82 Regression Anal ysis for Twitter Adoption ................................ ......................... 82 Descriptive Statistics of the eWOM Product Category and Perceived Fit ......... 84 Principal Component Analys is of Product Category and Perceived Fit ............ 85 Main Test ................................ ................................ ................................ ................ 86 Instrument Development ................................ ................................ ......................... 87 Measures ................................ ................................ ................................ .......... 87 Prior Experience of Obtaining Information from Twitter ................................ .... 88 Frequency of Obtaining Brand Informatio n from Twitter ................................ ... 88 Actual Twitter Usage ................................ ................................ ........................ 88 Conformity to Subjective Norm ................................ ................................ ......... 89 TAM Related Measures ................................ ................................ .................... 89 Perceived usefulness ................................ ................................ ................. 90 Perceived ease of use ................................ ................................ ............... 91 Attitude Toward Twitter ................................ ................................ ..................... 91 Perceived Source Similarity ................................ ................................ .............. 91 Perceived Source Credibility ................................ ................................ ............ 92 Product Category ................................ ................................ ............................. 93 Perceived Fit ................................ ................................ ................................ .... 93 Attitude Toward the Brand ................................ ................................ ................ 94 eWOM Spreading Intention ................................ ................................ .............. 94 Purchase Intention ................................ ................................ ........................... 95 Consumer Characteristic Factor s ................................ ................................ ..... 95 Data Collection, and Procedure ................................ ................................ .............. 96 Participants ................................ ................................ ................................ ............. 97 Statistic al Analysis ................................ ................................ ................................ .. 99 Overview ................................ ................................ ................................ .......... 99 Validity and Reliability Test ................................ ................................ ............. 100 Val idity test ................................ ................................ ............................... 100

PAGE 9

9 Reliability test ................................ ................................ ........................... 102 Structural Equation Modeling ................................ ................................ ......... 102 Advantages of structural equation modeling (SEM) ................................ 102 Testing model fit ................................ ................................ ....................... 103 Additional Analysis ................................ ................................ ......................... 104 5 RESULTS ................................ ................................ ................................ ............. 120 Twitter Adoption Results ................................ ................................ ....................... 120 Descriptive Statistics ................................ ................................ ...................... 120 Correlations Analysis ................................ ................................ ...................... 120 Factor Analysis ................................ ................................ ............................... 121 Exploratory factor analysis ................................ ................................ ....... 121 Confirmatory factor analysis ................................ ................................ ..... 122 Validity and reliability ................................ ................................ ............... 123 Struct ural Equation Modeling ................................ ................................ ......... 124 Additional A nalysis ................................ ................................ ......................... 127 eWOM Related Results ................................ ................................ ........................ 128 Descriptive Statistics ................................ ................................ ...................... 128 Correlations Analysis ................................ ................................ ...................... 129 Factor Analysis ................................ ................................ ............................... 129 Principal component analysis ................................ ................................ ... 130 Exploratory factor analysis ................................ ................................ ....... 130 Confirmatory factor analysis ................................ ................................ ..... 131 Validity and reliability ................................ ................................ ............... 132 Structural Equation Modeling ................................ ................................ ......... 132 Additional Analysis ................................ ................................ ......................... 136 6 DISCUSSION ................................ ................................ ................................ ....... 165 Summary of Findings for Twitter Adoption ................................ ............................ 165 Effects of Conformity to Subjective Norm on Attitude toward Twitter and Usage ................................ ................................ ................................ .......... 166 Effects of Perceived Usefulness on Attitude toward Twitter and Usage ......... 169 Effects of Perceived Ease of Use on Attitude toward Twitter and Usage ....... 171 Effects of Consumer Demographic Variables on Attitude toward Twitter and Usage ................................ ................................ ................................ .......... 173 Effects of Attitude toward Twitter on Usage ................................ .................... 175 Summary of Findings for eWOM Related Variables ................................ ............. 175 Effects of Perceived Similarity on Attitude toward the Brand and eWOM Intention ................................ ................................ ................................ ...... 176 Effects of Perceived Credibility on Attitude toward the Brand and eWOM Intention ................................ ................................ ................................ ...... 178 Effects of Product Category on Attitude toward the Brand and eWOM Intention ................................ ................................ ................................ ...... 179 Effects of Perceived Fit on Attit ude toward the Brand and eWOM Intention ... 181

PAGE 10

10 Effects of Consumer Demographic Variables on Attitude toward the Brand and eWOM Intention ................................ ................................ ................... 183 Effects of Attitude toward the Brand on eWOM intention and Purchase Intention ................................ ................................ ................................ ...... 185 7 CONCLUSION ................................ ................................ ................................ ...... 190 Twitter Adoption Re lated Implication ................................ ................................ .... 191 Theoretical Implications ................................ ................................ .................. 191 Integrating adoption theories in Twitter context with actual usage using gene ral sample ................................ ................................ ..................... 191 Conformity to subjective norm in attitude toward Twitter and usage ........ 192 The importance of perceived usefulness in attitude toward Twitter and usage ................................ ................................ ................................ .... 193 The importance of perceived ease of use in attitude toward Twitter and usage ................................ ................................ ................................ .... 194 Consumer characteristics in attitude toward Twitter and usage ............... 196 Importance of perceived usefulness in new media adoption .................... 197 Impo rtance of perceived ease of use in new media adoption .................. 198 Industrial Implications ................................ ................................ ..................... 198 Importance of perceived usefulness i n new media adoption .................... 199 Importance of perceived ease of use in new media adoption .................. 199 eWOM Related Implication ................................ ................................ ................... 200 Theoretical Implication ................................ ................................ ................... 200 Testing consumer eWOM behavior in Twitter context with actual usage using general sample ................................ ................................ ............ 20 2 Importance of perceived similarity in attitude toward the brand and eWOM intention ................................ ................................ .................... 204 Importance of perceived credibility in att itude toward the brand and eWOM intention ................................ ................................ .................... 205 Importance of product category in attitude toward the brand and eWOM intention ................................ ................................ ................................ 206 Importance of perceived fit in attitude toward the brand and eWOM intention ................................ ................................ ................................ 207 Consumer demographic in attitude toward the brand, eWOM intention and purchase intention ................................ ................................ ......... 209 Industrial Implication ................................ ................................ ....................... 209 Strategic use of Twitter as an alternative marketing tool .......................... 209 Strategic use of perceived similarity ................................ ......................... 210 Strategic use of perceived credibility ................................ ........................ 211 Strategic us e of utilitarian vs. hedonic product category .......................... 211 Strategic use of perceived fit ................................ ................................ .... 212 Limitations ................................ ................................ ................................ ............. 212 Limitations for Twitter Adoption Study ................................ ............................ 212 Limitations for eWOM Related Study ................................ ............................. 214 Suggestions for Future Research ................................ ................................ ......... 214 Future Research Direction for New Media Adoption ................................ ...... 214

PAGE 11

11 Future Research Direction f or eWOM Research ................................ ............ 215 APPENDIX A PRETEST QUESTIONNAIRE ................................ ................................ ............... 217 B MAINTEST QUESTIONAIRE ................................ ................................ ................ 225 C UNIVERSITY OF FLORIDA INSTITUTIONAL REVIEW BOARD INFORMED CONSENT APPROVAL ................................ ................................ ........................ 235 Protocol Submission Form ................................ ................................ .................... 235 Informed Consent ................................ ................................ ................................ 238 LIST OF REFERENCES ................................ ................................ ............................. 240 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 271

PAGE 12

12 LIST OF TABLES Table page 2 1 Technology acceptance model, social influence related literatures in adoption study ................................ ................................ ................................ ................... 45 2 2 eWOM related literature ................................ ................................ ..................... 47 3 1 Summary of Twitter adoption related hypotheses and research questions ........ 74 3 2 Summary of eWOM perspective related hypotheses and research questions ... 75 4 1 Descriptive summary of pretest questionnaire ................................ .................. 105 4 2 R eliability t est of pretest questionnaire ................................ ............................. 105 4 3 Correlation matrix of pretest items ................................ ................................ .... 106 4 4 Regression analysis of pretest ................................ ................................ .......... 107 4 5 Descriptive statistics of product category ................................ ......................... 108 4 6 Descriptive statistics of perceived fit ................................ ................................ 108 4 7 Correlation matrix (Product category) for pretest ................................ .............. 109 4 8 Correlation matrix (Perceived fit) for pretest ................................ ..................... 109 4 9 Principal component analysis with product category ................................ ........ 110 4 10 Principal component analysis with perceived fit ................................ ............... 110 4 11 Operational definition of main constructs and measurement ............................ 111 4 12 Summarizes the original constructs included and their operational definition ... 113 4 13 The comparison of the sample profile with industrial report .............................. 116 4 14 Comparing model fit index ................................ ................................ ................ 118 5 1 Twitter adoption related measurement, descriptive statistics, skewness and kurtosis ................................ ................................ ................................ ............. 138 5 2 Initial correlation matrix among Twitter adoption variables ............................... 139 5 3 Model fit of exploratory factor analysis of Twitter adoption related variables .... 140 5 4 Model fit of confirmatory factor analysis of Twitter adoption variable ................ 140

PAGE 13

13 5 5 Results of final factor loading and reliability test (Twitter adoption) .................. 141 5 6 Correlation matrix for final validity of constructs (Twitter adoption). .................. 142 5 7 Significant parameter estimates of model ................................ ......................... 143 5 8 Summary of hypothesis testing for the Tw itter adoption study ......................... 144 5 9 Direct, indirect and total effects of major variables ................................ ........... 146 5 10 Mediation effect of attitude toward Twitter ................................ ........................ 147 5 11 eWOM related measurement, descriptive statistics, skewness and kurtosis, ... 148 5 12 Initial Correlation matrix am ong eWOM variables ................................ ............. 150 5 13 The results of principal component analysis of product category ..................... 151 5 14 The results of principal co mponent analysis of perceived fit ............................. 151 5 15 Model fit of exploratory factor analysis of eWOM related variables .................. 152 5 16 Model fit of confirmatory factor analysis of eWOM related variables ................ 152 5 17 Results of final factor loading and reliability test (eWOM) ................................ 153 5 18 Correlation matrix for final validity of constructs (eWOM) ................................ 156 5 19 Significant parameter estimates of model for eWOM ................................ ....... 157 5 20 Direct, indirect and total effects of major variables ................................ ........... 158 5 21 Summary of hypothesis testing for the eWOM in Twitter study ........................ 159 6 1 Result summary for hypotheses ................................ ................................ ....... 187 6 2 Result summary for research question ................................ ............................. 189

PAGE 14

14 LIST OF FIGURES Figure page 1 1 Classification of social media ................................ ................................ .............. 27 2 1 The theory of reasoned action ................................ ................................ ............ 54 2 2 The theory o f planned behavior ................................ ................................ .......... 54 2 3 Technology acceptance model ................................ ................................ ........... 55 2 4 United t heory of a cceptance and u se of t echnology (UTAUT) ............................ 55 2 5 Decomposed t heory of p lanned b ehavior ................................ ........................... 56 3 1 Proposed conceptual research model for Twitter adoption in this study ............. 77 3 2 Proposed model for Twitter adoption ................................ ................................ .. 78 3 3 Proposed conceptual model for the Twitter in eWOM p erspective ..................... 79 3 4 Proposed model for the Twitter in eWOM perspective ................................ ....... 80 4 1 Flow of statistical analysis ................................ ................................ ................ 119 5 1 Scree plot of Twitter adoption variables ................................ ............................ 161 5 2 Structural model including observed pathway ................................ .................. 162 5 3 Scree plo t of eWOM related variables ................................ .............................. 163 5 4 Proposed model for the Twitter in eWOM perspective ................................ ..... 164

PAGE 15

15 LIST OF ABBREVIATION S eWOM Electronic Word of Mouth UTAUT United Theory of Acceptance and U se of Technology TAM Technology Acceptance Model TP B Theory of Planned Behavior TRA Theory of Reasoned Action WOM Word of Mouth

PAGE 16

16 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 Arts in Mass Communication ADOPTION OF TWITTER AND ITS EFFECTIVENESS IN e WOM By Hyunsang Son December 2011 Chair: Sylvia M. Chan Olmsted Major: Mass Communication As a one of the fastest growing areas for online marketing, social media, specifically Twitter, a social networking site, could provide prominent opportunities for marketers. However, little scholarship has empirically investigated why consumers use Twitter and how to utilize Twitter as a brand in formation exchanging tool. This study mainly contributes to three different areas. First, the study investigates the motivation associated with Twitter by comparing it to other product information sources such as advertising, thus providing a relative eWOM effectiveness perspective. Second, factors that influence eWOM intentions from sender and receiver perspectives are analyzed thus offering both marketer and consumer contexts. Third, the study tries to identify appropriate types of products about which pe ople share information, often by making comparisons with other product categories. U ltimately, this study attempts to answer questions about what and how people decide to adopt or exchange product related information. T his study employed 4 independent vari ables perceived usefulness, perceived ease of use, conformity to subjective norm, and demographical factors (age, education level and gender) for Twitter adoption. In addition, 5 independent variables perceived similarity, perceived credibility, product ca tegory (utilitarian and

PAGE 17

17 hedonic), perceived fit (utilitarian and hedonic), and demographic factors (age, education level and gender) which might affect consumer attitude toward marketing messages in Twitter, were examined in this study. A pretest was cond ucted before the main test. For the main test, this study employed an online survey by utilizing an online consumer panel for U.S. consumers based on the results of the pretest. Statistical analyses found that perceived usefulness itudes toward Twitter and their Twitter usage. Also, the perceived indirectly. However, individual conformity to subjective norms and demographical factors did not affe In terms of eWOM related independent variables, perceived similarity, perceived credibility, product category of hedonic dimension, and perceived fit of hedonic product dimensions affected con credibility, perceived fit of utilitarian dimension, and perceived fit of hedonic dimension spreading intention.

PAGE 18

18 CHAPTER 1 INTRODUCTION Due to the rapid development in new communication technology (i.e., Web 2.0, social media), traditional marketing strategies, using seller created product information (Park, Lee, & Han, 2007) such as advertising are no longer as effective in influencing consumer decision making as the y were several decades ago. This trend may be explained by two reasons. First two way communication platforms such as the Internet and even more progressive forms of social media have been developed (Sohn, 2009b; Trusov, Bucklin, & Pauwels, 2009 ). Today co nsumers do not wait for product information from television advertisements, but actively seek information through the Internet (Pew Research Center, 2011; Prendergast, Ko, & Yeun, 2010). Second, skepticism among consumers is widespread when compared with t he past decade. Fo r example, according to Trusov et al. (2009), 40% fewer people agree that advertisements are an appropriate way to obtain new product information (Nail, 2005). WOM and eWOM As one of the most effective marketing strategies to overcome con sumer seeking efforts, word of mouth (WOM) marketing has drawn enormous attentio n from both practitioners (e.g., Jaffe, 2007; Kelly, 2007 ; Misner, 1999; Rosen 2000 2009 ; Sernovitz, 2009 ) and researchers (e.g., Tr usov et al. 2009 ). Two major findings from empirical research help explain why people react more positively or actively toward WOM than traditional marketing strategies such as advertisements: vividness of information (Anderson, 1998; Feldman & Lynch, 19 88; Herr, Kardes, & Kim, 1991; Kisielius & Sternthal, 1984; Lau & Ng, 2001; McGill & Anand,

PAGE 19

19 1989 ) and perceived credibili ty of information source (Allsop Bassett, & Hoskins, 2007). Anderson (1998) indicated that WOM could convey vivid, novel, and pleasant experiences when people share positive feelings or negative complaints. Also, consumers may the challenges and changes in the marketing environment, it is necessary to test empirically alternative strategies designed or adapted for the era of the Internet and to overcome consumer skepticism toward advertising. Acknowledging the importance of WOM many companies have applie d the concept of WOM to include the online environment Both companies and customers now use online conversations as a tool to spread product, brand, and service information. This newly emerged concept is called electronic word of mouth (eWOM) which often take forms in on line forums (e.g., Prendergast et al. 2010), online reviews (e.g., Zhu & Zhang, 2010 ), e mail (e.g., Phelps, Lewis Mobilio, Perry, & Raman, 2004), and SMS messages (Okazaki, 2008, 2009 ); eWOM has also become a tool useful in spreading pro duct, brand, and service information in the world of Web 2.0 Social Media as a new eWOM Marketing Tool One growing area of new media studies in new media in the context of the Web 2.0 environment focused on the emergence of social media, particularity amo ng U.S. consumers (Radwanick, 2011). Nine of every 10 U.S. Internet users use social media, and 25.3% of mobile subscribers use social media via their mobile phone (Flosi, 2011). In particular, in 2010, social networking sites (SNSs) ranked as the second m ost

PAGE 20

20 (20.2%); SNSs captured even more time than e mail usage (11.0%) and experienced a 15% increase in unique visitors compared with 2009 (Dennen, 2011). Specifically, o ne of the most prominent SNSs, Facebook, reached 153.9 million 2011). Twitter, a microblog ging service, has also grown rapidly and drawn more than 20 million unique visitors per month in 2010, a 26% increase from 2009 (Radwanick, 2011). Considering the burgeoning of online communication tools, WOM marketing is particularly useful in the online environment because the Internet provides abundant opportunities for consumers to share their opinion s preference s for, or previous experience s related to specific products or brands ( Trusov et al., 2009 ). For example, 59% of customers said that they freq uently forward information found on the Internet to colleagues, pee rs, family, or friends (Allsop et al., 2007). Market research data also support this assumption; for example, the Hotmail service provided by Microsoft expanded its user base to 1 million s ubscribers in the first 6 months, 2 million only 2 months after its launch, and eventually 11 million within 18 months by spreading its brand name at the bottom of all the outgoing e mail messages sent through the Hotmail service (Dobele, Toleman, & Beverl and, 2005). More recently, several researchers have argued that social media, one of the fastest growing areas for online marketing, could provide prominent opportunities for marketers (e.g., Chu & Kim, 2011; Hennig Thurau et al., 2010; Interactive Advert ising Bureau, 2009; Kaplan & Haenlein, 2010; Lenhart, Purcell, Smith, & Zickuhr, 2010; Libai et al., 2010; Mangold & Faulds, 2009; Nielsen, 2011; Vollmer & Precourt, 2008; Wuyts

PAGE 21

21 Dekimpe, Gijsbrechts, & Pieters 2010 ). F Vollmer and Precourt (2008) indicated that SNSs are ideal tools for eWOM and spreading brand information, because consumers willingly create and deliver it to their friends, classm ates, and other acquaintances. Mangold and Faulds (2009) also argued that th e use of SNSs helps companies to communicate with consumers. Furthermore, scholars have begun to test empirically the effectiveness of social media eWOM in influencing an Due to their nature as eme rging media platforms, SNSs do not have a universal definition across various disciplines. One popular source has referred to social media as openness, conversation, comm 5). Mayfield (2008) also classified social media into seven basic forms: social networks (e.g., Cyworld, MySpace, Facebook), blogs, wikis (e.g., Wikipedia), podcasts (e.g., Apple iTunes), forums, co ntent communities (e.g., Flickr.del.icio.us), and microblogging (Twitter), whereas Nielsen (2011) categorized social media as social networks (e.g., gaming, media sha ring (e.g., YouTube, Flickr), and microblogs (e.g.,Twitter). Also, based applications that build on the ideological and technological foundations of Web 2.0, and that allow the cre ation and exchange of User Generated Content" (p. 61), and differentiated social media types based on the degree of self presentation/self disclosure and social presence/media richness (Table 1 1). More specifically, Kaplan and Haenlein (2010) indicated th at social media are classified into a total of six different

PAGE 22

22 types: blogs, collaborative projects (e.g., Wikipedia), social networking sites (e.g., Facebook, Twitter), content communities (e.g., YouTube), virtual social worlds (e.g., Second Life), and virt ual game worlds (e.g., World of Warcraft) Among various types of social media, this study focused on microblogging. sized blogging, where and through the mobile phone time updates (Nielsen, 2011). As the definitions illustrate, the most prominent features of microblogging involv e connecting with others. Users of mi small group of founders who send out invitations to join the site to the members of their 2009 p. 90), and likewise new members send invitations to their personal net work and newer members continue to repeat this process. Therefore, considering the similarity between invitations in social media and eWOM referrals, it can be assumed that eWOM marketing would be effective in microblogging. In particular, although much o f the literature tend to examine SNSs and microblogging in the same context (Radwanick, 2011; Dennen, 2011; Flosi, 2011), this study focused on Twitter (microblogging) rather than Facebook (SNS) services because and the efficacy of eWOM in the context of microblogging. Twitter has attracted a great deal of attention as a new communication tool. The growing use of Twitter is extraordinary; with more than 40 million users and a 1,400% increase in use between April 2008 and April 2009 ( Milstein, Chowdhury, Hockmuth,

PAGE 23

23 Lorica, & Magoulas, 2008; Xifra & Grau, 2010) More recently, Twitter traffic now reaches 200 million tweets (messages) a day and 20.6 million U.S adults accessing Twitter at least once a month in 2011 ( S eiple, 2011; Twitter, 2011 ) More importantly, Twitter is an especially effective tool for marketers for three reasons: users frequently mentioned brand s on Twitter and topics frequently discussed on Twitter are of consumption nature ; it is easy to make co nnections with consumers ; there is a well defined mechanism for WOM marketing (Jansen, Zhang, Sobel, & Chowdury, 2009; Kwak, Lee, Park, & Moon, 2010 ; Seiple, 2011; Mulcahy, 2011 ). Spe cifically a total of 26% of online discussion s contains certain brand na mes (Nielsen, 2011) Also, unlike SNSs such as Facebook and MySpace, Twitter does not rely on reciprocation to create relationships. A user can follow (create a relationship) any other user, but the user being followed does not necessarily have to follow b ack. Therefore, it is easy for marketers to following or being followed. Being a on Twitter means that the user agrees to receive all the messages (called ) from thos e the individual follows (Kwak et al. 2010). Consequently once a company creates a relationship with a consumer, it is easy to spread its message to that consumer. Finally the most prominent feature of Twitter, Retweet, is a mechanism for followers with one click of the mouse or by using the well defined symbol R T ; this helps (Kwak et al., 2010). Inde ed, according to industry report, 53% of Twitter users use Twitter for information from others ( Seiple, 2011). Despite this phenomenal Web 2.0 empirical

PAGE 24

24 research into the business us e of Twitter has rarely been conducted. Accordingly, this study is designed to investigate the factors that might play a role in the eWOM process. The results of the study can contribute to our understanding of the marketing utilizing of an emerging media and its implication. Purpose of the Study Limited Academic Research This study attempts to fill the theoretical gap in two areas: consumer motivation for using microblogging, Twitter, and factors affecting eWOM intention in the context of microblogging First, due to the nature of SNSs as emerging media platforms, the motivations for using them as one category of social media have not been thoroughly scrutinized empirically. Although previous researchers have examined the driving forces of new media adopt ion among consumers such a s the Internet (e.g., Cha, 2009b ; J. Lee, 2003; S. Lee, 2007 ; Papacharissi & Rubin, 2000), terrestrial digital television (Chan Olmsted & Chang, 2006), and even broadband service adoption (Chan Olmsted, Li, & Jung, 2005), limited studies have examined the motivation for using microblogging from the eWOM or other marketing perspective even if some scholars have suggested the possibility of using social media as effective eWOM tools ( Kozinets, D e Valck, Woinicki, & Wilner, 2010; Tru sov et al., 2009). In addition, while previous studies have addressed various online consumer behaviors, such as digital auctions (Dholakia, Basuroy, & Soltysinski, 2002), software downloading (Hanson & Putler, 1996), and even music piracy on the Internet (Ki, Chang, & Khang, 2006), online recommendation, eWOM, has not been investigated thoroughly. From the perspective of WOM, eWOM is one of the most revolutionary developments in its history However, it has been suggested that eWOM

PAGE 25

25 behaviors di ffer from traditional WOM processes when the context is expanded to the online environment (Cheema & Kaikati, 2010). For example, Cheema and Kaikati (2010) argued that eWOM is less powerful for a product category that has uniqueness as a characteristic, wh ereas initial WOM researchers consistently reconfirm that people rely more on WOM when they consider purchasing unique or innovative products ( Rogers, 1995; Ryan & Gross, 1943; Whyte, 1954 ). Also, although previous researchers investigated the effectivenes s and motivations of eWOM in various online environments, including online forums (e.g., Prendergast et al., 2010), online product reviews ( Lee & Youn, 2009; Zhu & Zhang, 2010 ), and mob ile SMS messages (Okazaki, 2008, 2009), SNSs have not been examined as an eWOM tool from both theoretical and empirical perspectives. Research Directions This study is expected to yield three main contributions. First, the study investigates the motivation associated with Twitter by comparing it to other product information s ources such as advertising thus providing a relative eWOM effectiveness perspective. Second, factors influencing eWOM intention from both sender and receiver perspectives is analyzed thus offering both marketer and consumer contexts. Third, the study trie s to identify appropriate types of products about which people share information, often by making comparisons with other product categories. Given an in depth understanding of Twitter as an eWOM tool, ultimately, this study attempts to answer questions abo ut what and how people decide to adopt or exchange product related information (Frenzen & Nakamoto, 1993 ; Sohn, 2009a, 2009b). Therefore, theoretically, this study investigates adoption studies and eWOM literature to provide a better understanding of the s ignificant factors associated with online consumer behavior

PAGE 26

26 and, practically, this study helps companies construct appropriate marketing strategy for using SNSs.

PAGE 27

27 Social presence/ Media richness Low Medium High Self presentation/ Self disclosure High Blogs Social networking sites (e.g., Facebook) Virtual social world (e.g., Second Life) Low Collaborative projects (e.g., Wikipedia) Content communities (e.g., YouTube) Virtual game world (e.g., World of Warcraft Figure 1 1 Classification of s ocial m edia by social presence/media richness and self presentation/self disclosure ( Source: Kaplan & Haenlein, 2010)

PAGE 28

28 CHAPTER 2 THEORETICAL FRAMEWORK This study is composed of two main frameworks T he first framework is used to investigate the adoption factor s of Twitter. The second framework is used to examine Twitter as an eWOM tool and the effectiveness of eWOM as a marketing tool. Specifically t he first purpose of this study is to examine how consumer attitudes toward the new media platform Twitter are f ormed and linked to adoption intention accordingly. I n this section, the integration of Theory of Reasoned Action (TRA), Theory of Planned Behavior (TPB) and Technology Acceptance Model (TAM) are introduced and explained in detail in the context of social networking site (Twitter) In addition, in terms of eWOM perspective, perceived similarity, perceived credibility, product category, perceived fit are introduced for the antecedents of consumer eWOM behavior in Twitter. Social Networking Site Adoption St udies Current Research Trend Reflecting the importance and growth of social media the current research trend in this area focuse s mainly on three dimensions: the characteristics of SNSs (e.g., Ellison, Steinfield, & Lampe, 2007), the ro le s of SNSs in various contexts (e.g., Ellison et al 2007; Tong, Van Der Heide Langwell, & Walther, 2008) and the motivations for using social media (e.g., Raacke & Bonds Raacke, 2008). According to Ellison et al. (2007) social network sites (SNSs) can be differentiated based on their initial purpose s at the early stage of diffusion of SNSs: work related context (e.g., LinkedIn.com), romantic relationship initiation (Friendster.com), sharing interest (Myspace.com, Cyworld.com) or communication among col lege students (Facebook.com).

PAGE 29

29 Several researchers also examined the role s of SNSs in terms of communication perspective such as social capital (Ellison et al., 2007) and interpersonal impression formation (Tong et al., 2008). Ellison et al. (2007) found a strong positive relationship between Facebook usage and social capital. Tong et al. (2008) focused more on the personal level and the role of Facebook in terms of interpersonal impression formation. Another important research trend of SNSs is on how the specific user demographics of social media affect individual motivation for adoption. While traditional media spread messages across all ages (Cha, 2007 2009 a ), social network users were highly concentrated among teenagers and people in their 20s to 30s. For example, 40.3% of Facebook users and 46.6% of Twitter users were in the age group of 18 to 34 (Radwanick, 2011). Those age groups are particularly important for marketers since they are attractive consumer groups (Cha, 2007 2009a ), and they exhibit hi gher intention to purchase products and services online than older generations (Akhter, 2003). However, due to the nature of emerging media, the driving force of SNSs (e.g., Twitter) adoption was not fully tested empirically. Media Use and adoption Resear ch It now is widely believed that technology and new media adoption is not a single function of technology factors. Rather, it is a media adoption function that includes social context, media accessibility, availability of communication partners, previous experience with media usage, individual life style, consumer innovativeness, characteristics of media, or even the fit between task and technology (Daft & Lengel, 1984; Dishaw & Strong, 1999; Fulk, Schmitz, & Steinfield, 1990; Lee, 2003; Markus, 1987; Sala ncik & Pfeffer, 1978). Before examining the framework of the Theory of Planned Behavior (TPB) and Technology Acceptance Model (TAM), it is worth

PAGE 30

30 reviewing the assumptions of traditional media adoption and usage theories to compare and illustrate the differ ent aspects of traditional media adoption studies. Note that this research mainly subscribes to the foundations presented in TRA, TPB, and TAM. According to Lee (2003), one of the most prominent traditional media usage studies is Rational Choice Model s of media (Daft & Lengel, 1984, 1986 ; Lengel & Daft, 1989 ). This model assume s explained by rational choices that best suit their needs. Therefore, two main factors tion richness and social presence. Social presence is the level of perceived presence that people view as sociable, warm, sensitive, personal, or intimate during the media usage process and interact with others by using specific media. Therefore, by using media that represent their need for appropriate levels of social presence, media usage patterns can be explained (Short, Willimans, & Christie, 1976). Media richness theory also indicated the different doption and usage of media. According to Daft and Lengel (1984) people rationally select media to fulfill their need for different information. Therefore, regarding the speed of feedback, types of channel employed, and personal ness of source, people selec t the most desirable media for their communication purpose among face to face communication, telephone, and p rint media (Daft & Lengel, 1984, 198 6; Lengel & Deft, 1989 ). However, the basic assumption of the rational choice model that individuals make ratio nal and objective evaluation of their own needs and tasks for media use has been questioned. For example, Fulk et al. (1990) insisted that decision making related media usage was neither always rational nor always efficiency motivated.

PAGE 31

31 Uses and Gratificat ion (U&G) theory also suggests that multiple motivations, including sociability, contribute to new media adoption such as the mobile phone usage (Wei, 2008). Similar to rational choice models but including somewhat different motivations such as social infl uence, U&G examine s the various motivations for adopting media based on the assumption that individuals use the same mass medium for different purposes (Severin & Tankanrd, 2001 ), M any scholars pointed out that U&G is also one of the most ideal theoretica l reasoning methods to reveal the role of psychological and behavioral tendencies in media usage ( Ko, Cho, & Roberts, 2005; Korgaonkar & Wolin, 1999; Lin & Cho, 2010). Drawing on the uses and gratification theoretical framework, Ko et al. (2005) indicated that U&G has been considered as axiomatic theory that can apply to many adoption behaviors from both traditional media such as radio (Mendelsohn, 1964), newspapers and magazines (Elliott & Rosenberg, 1987; Licheterstein & Rosenfeld, 1984), and television ( Babrow, 1987; Conway & Rubin, 1991; Rubin, 1981, 1983 1984) and nontraditional media (the so Greenberg, 1985; LaRose & Atkin, 1988), VCR (Cohen, Levy, and Golden, 1988; Levy, 1987), telephones (Dimmic 1995), pagers (Leung & Wei, 1999), e mail (Dimmick, Kline, & Stafford, 2000), the Internet ( Ko et al., 2005; Lin, 1999; Pap acharissi & Rubin, 2000), satellite radio (Lin, 2010), and online radio (Lin, 2 009). U & G theory has identified a variety of motivational factors through decades of research (e.g., surveillance, sociability, diversion, escape, arousal, instrumentality, reassurance, and companionship) (Wei, 2008).

PAGE 32

32 Consumer Characteristics in Media Ad option Another media adoption research trend has focused on the importance of consumer demographic characteristics ( Atkin & LaRose, 1994; Atkin Neuendorf Jeffres, & Skalski 2003 ; Dutton Rogers, & Jun 1987; Krugman, 1985; LaRose & Atkin, 1992; Lin, 199 8; Steinfield, Dutton, & Kovaric, 1989) and psychological traits such as innovativeness (e.g., Chan Olmsted & Chang, 2006; Chan Olmsted, Li, & Jung, 2005 ; Chang, Lee, & Kim, 2006). For example, in terms of personal computer adoption, those with higher soc ioeconomic status are more likely to use a personal computer at home (Atkin & LaRose, 1994; Dutton et al., 1987; Steinfield et al., 1989; Chan Olmsted & Chang, 2006). Males are more likely to use online games (Chang et al., 2006) and the Internet (Ernst & Young, 1999), whereas no gender difference was found in SNS usage and shopping behavior on SNSs (Cha, 2009 a, 2010). Social Influence The term social influence in media adoption studies (Fulk et al., 1990) has varying meanings and there is hard to measure in an academic setting. The concept of social influence initially was explained in two different dimensions: informative (informational) and normative (Cialdini & Trost, 1999). Informative social influence is referred to (Deutsch & Gerard, 1955, p. 629), whereas normative social influence is considered as the tendency of be liked or accepted by others and conforms to what individuals believe to be the norms of the grou (Cialdini, 1984, p. 13). Drawing from these basic concepts of social influence (not necessarily in terms of technology and media usage), several researchers have

PAGE 33

33 suggested that it has different effects in various contexts such as deindividuation effect (Festinger, Pepitone & Newcomb, 1952), social identity ( Elias, Appiah, & Gong, 2008; Tajfel & Turner, 1 986; Turner, 1982; Mastro, 2003 ), social categorization ( Brewer & Gaertner, 2004; Hogg, 2004; Hogg & Reid, 2006; Hogg & Terry, 1995 ), and social comparison ( Buunk & Gibbons, 2007; Buunk & Oldersma, 2001; Festinger, 1954 ; Gulas & McKeage, 2000 ). For example, Festinger et al. (1952) examined social influence by proposing the concept of deindividuati on that individuals tend to lose their personal identity and merge into a group or crowd. They reported that despite the various moderators and predictors of social influence, social influence on media usage and technology acceptance has not been tested fully. More recently, Cialdini and Goldstein (2004) differentiated three motivations that make individuals highly sensitive to social influence, such as accuracy seeking, affiliation, and maintenance of a positive self concept. Although these concepts oft en occur simultaneously in real world situations, they have different operational definitions. Accuracy seeking is an individual tendency to pursue and react toward uncertain social situations. Maintaining a positive self ts to increase their self highly related to a motivation for engaging in norms and behaviors in order to obtain desirable social relationships. Among them, current researc h is particularly related to the affiliation factor of social influence, which is also highly related to Fishbein and that discussed later. Social influence in new media adoption To provide an alternative theoretical reasoni ng from the findings of rational choice

PAGE 34

34 adoption that individual perceptions toward certain media can be formulated not only for objective features of media but also decid ed by the attitude, statements, and behaviors of other people by using the term social influence (Fulk et al., 1990). Fulk et al. (1990) proposed the social influence model of technology use (SIMoTU) by suggesting that the individual behavior of media and technology adoption is not always based on the rational evaluation or objective need for media usage, but also the social influence from coworkers, group behavior norms and other social roles in the society. As previous mentioned, social influence on indiv particular on normative social influence, has been examined using two theoretical frameworks: The Theory of Reasoned Action (TRA) and the Theory of Planned Behavior (TPB). Although these theories were not designed for technolo gy and media adoption, both provided useful explanations for the role of normative social influence in various consumer behaviors. The Theory of Reasoned Action (TRA), initially proposed by Fishbein and Ajzen or and attitudes toward the behavior is determined by two factors: attitudes and subjective norms. Attitude toward a behavior is that most people who are important to him think one should or should not perform the influence. Therefore, the behavior al intention decided by the conjoint influence of attitude and subjective norm affect actual behavior ( Ajzen & Fishbein, 1980; Fishbein & Ajzen, 1975 ) (Figure 2 1 for a conceptual model of TRA). Adopting the TRA constructs

PAGE 35

35 and theoretical premise, one of the co developers of TRA, Ajzen (1991) developed the Theory of Planned Behavior (TPB), adding a third construct of behavioral intention, perceived behavioral control, which demonstrated the assumption that a variety of human behaviors are controlled by vol ition (Ajzen, 2002). As illustrated earlier while the TRA and TPB were not originally designed to reveal an technology or media adoption; each theory was applied in various technology and media usage motivation studies (e.g., Davis 1989; Kwon & Chon, 2009; Lin, Chuan, & Rivera, 2009 ; Venkatesh & Davis, 2000; Venkatesh & Morris, 2000 ). Figure 2 2 illustrated the basic conceptual model of TPB. In particular, focusing on the social influence of new media and technology adoption, sever al studies have indicated the importance of the social influence dimension (e.g., Carr, 2008 ; Homburg, Wieseke, & Kuehnl, 2010; Kwon & Chon, 2009; Lin et al 2009; Park, Kwan & Cheong, 2007 ). For example, Homburg et al. (2010) revealed that pressure from adopt a new sales technology packet. Park et al. (20 07 ) also found that the effects of compliance with school policies include a positive relationship between Internet based course management sy stems. Carr (2008) and Lin et al. (2009) also found that both messaging usage pattern. Technology Acceptance Model While TRA and TPB revealed the mechanism of various individual behaviors within a variety of disciplines, neither focused solely on technology or media adoption but general behavior or behavioral intention. Also, individual behavior and behavioral intention are not entirely affected by social influence, as well as formed by their own

PAGE 36

36 beliefs of technology. Therefore, the concept of the Technology Acceptance Model (TAM) emerged based on TRA and TPB and focused exclusively on technology related studies (Davis, 1989). After an in depth examination of TRA and TP B, Davis (1989) proposed a model to predict audience usage of information technologies. Davis Bagozzi, & Warshaw (1989) demonstrated the concepts of perceived usefulness and perceived ease of use (Figure 2 ee to which a person believes that using a particular system would enhance his or her job degree to which a person believes that using a particular system would be free of ef (p. 320). In sum, when individuals consider the technology useful, they will start to think of the positive use performance relationship; thus, it is linked to the adoption intention (Lee, 2003) and the higher perceptions of perceived usefulness and perceived ease of use lead to a more favorable attitude toward the technology and intention to adopt (Davis, 1989; Davis et al., 1989). According to Park (2010), a number of studies have investigated the role of perceived ease of use and perceived usefulne ss within the information and communication technology contexts, such as word processing programs, spreadsheet software, and operating systems (e.g., Chau, 1996; Davis, 1993; Davis et al., 1989; Doll, Hendrickson, & Deng, 1998; Mathieson, 1991 ). In adoptin g the basic concepts of TAM, several researchers indicated the effectiveness of TAM to predict consumer adoption of the Internet and Internet based technologies, such as distance learning programs (Lee, 2003; Lee et al., 2003), e commerce (Jiang, Hsu, & Kl ein, 2000), telemedicine (Chau & Hu, 2002; Karahanna, Straub, & Chervany, 1999), e learning tools (Park et al., 2007),

PAGE 37

37 and digital library systems (Hong, Thong, Wong, & Tam, 2002; Thong, Hong, & Tam, 2002). More specifically, in terms of SNSs, shopping be havior in SNSs (Cha, 2009 a ) and adoption behavior of SNSs (Cha, 2010) have also been scrutinized using the TAM constructs. Revision of TAM Although TAM has already been applied in various contexts of adoption study (e.g., Hong et al., 2003 ; Jiang et al., 2000; Lee, 2003; Lee et al., 2003 ), there have been revision s in recent years (Park, 2010). First, Venkatesh, Morris, Davis, and Davis (2003) developed the unified theory of acceptance and use of technology (UTAUT), which integrates eight preexisting techn ology acceptance models, including TAM, TRA, TPB, and diffusion theory (Rogers, 1995). Specifically, Venkatesh et al. (2003) included the constructs of perceived usefulness, perceived ease of use (both from the TAM), and relative advantage (from diffusion theory) and inserted slightly different dimensions of performance expectancy. More importantly, subjective norm (from the TRA and the TPB) and other social and situational factors (e.g., age, gender, experience, and voluntariness of use) were considered. A dopting this revised framework many studies revealed the adoption factors, including traditional TAM factors (e.g., perceived usefulness, ease of use), social influence, and consumer demographics, within various media contexts ( Homburg et al. 2010; Lin e t al., 2009; Putzke, Schoder, & Fishbach, 2010; Park, 20 10 ; Park et al., 2007 ). For example, Putzke et al. (2010) found that both perceived usefulness and willingness to invest effort for a mass customized newspaper were conjointly related to consumer inte ntion to adopt a mass customized newspaper and also indicated the moderating role of gender. Kwon and Chon (2009) indicated that social influence in particular, maintaining a positive self image dimension was a

PAGE 38

38 rrestrial and satellite mobile TV usage, with consideration given to uses and gratification variables. Integration of Social Influence and TAM literature (e.g., Carr, 2008; Hombu rg et al., 2010; Kwon & Chon, 2009; Lin et al., 2009; Park et al., 2007 ), this study also emphasized the constructs from TAM (perceived ease of use and perceived usefulness) and social influence (conformity to subjective norm) from TPB and TRA. Although bo th social influence theory (TRA, TPB) and TAM have been tested and retested in various contexts of new media adoption and successfully validated by several different disciplines, each theory offers limited scope to fully understand new media adoption. TAM could not fully explain why people feel specific technology (e.g., Internet) is useful or easy, but simply asked whether respondents considered a technology or service to be useful or easy (Baaren, Wijngaert & Huizer, 2011). In other words, situational f actors (e.g., social influence) and personality traits (e.g., user demographics) have not been discussed in TAM (Baaren et al., 2011). Indeed, Mathieson (1991) revealed that each model has unique strengths and weaknesses depending upon the situation. TAM had a slight advantage for empirical technology, whereas TPB provided more specific information about the users' behaviors. Taylor and Todd (1995) also compared TAM's and TPB information technology adoption. For example, TAM might be more useful for predicting usage alone, whereas TPB could provide the theoretical explanation that considers situational factors (e.g., social influence). Adopting th e important role of conformity toward the norm, Lucas and Spitler (1999) found that social norms were a significant

PAGE 39

39 factor affecting ease of use and usefulness of technology that finally influenced consumer intention to use technology. Figure 2 4 and Figur e 2 5 illustrate the decomposed model of TPB and TAM. Therefore, to overcome the weakness of the preexisting conceptual framework, this study proposed an integration of TAM and TPB (but was simpler than Taylor and new media adoption within the Twitter context ( F igure 2 6). Variables were taken from TAM (perceived usefulness, perceived ease of use), TPB (conformity to social norm), and consumer factors (gender, age, educational level). Table 2 1 displays important media adoption literatures including TAM, TPB and social influence dimensions. WOM and eWOM Word of Mouth WOM communication has been exa mined in various studies (Herr et al. 1991), it is defined by the following words (Haywood, 1989): informal, noncommer cial, post purchase behavior and exchange, flow of information, communication, and conversation, among others (Goyette, Richard, Bergeron, & Marticotte, 2010), theory of the strength of weak ties (SWT) provides the theoretical reasonin g of the effectiveness of WOM. Granovetter (1973) differentiated the strength of interpersonal ticular, innovation among consumers was spread most effectively through weak ties rather than strong ties. Drawing upon the basic concept of ties, several researchers emphasized the effectiveness of weak ties as a WOM cue (Duhan, Johnson, Wilcox, & Harrell 1997).

PAGE 40

40 C ommunication scholars have recognized the importance of WOM particularly in terms of diffusion in innovation studies ( Rogers, 1995; Ryan & Gross, 1943). Westbrook about the ownership, usage, or characteristics of particular goods and services or their to person communication between a perceived noncommercial communicator and receiver regardi Whyte (1954) indicated that the term WOM means a communication between individuals about a certain topic. Arndt (1967) also defined it as a situation in which personally connected people commun icate about a product, brand, or service. It is worth noting the original idea of the WOM concept (Whyte, 1954; Katz & Lazarsfeld, 1955). Whyte (1954) revealed that WOM played an important role in consumer purchasing behavior for a newly invented product in the 1950s: the air conditioner. The purchasing of air conditioners was not random, and purchases revealed an interesting pattern. Also, Katz and Lazarsfeld (1955) indicated that the influence of WOM was two times more powerful than that of radio adverti sing, four times more powerful than direct marketing, and seven times more powerful than newspaper or magazine advertisements. Using this basic definition, many studies have revealed the powerful influence of WOM in the evaluation of alterative product cho ices and consumer behavior, especially compared to commercial information or neutral third party information such as Consumer Reports. In particular, a great deal of marketing research has focused on the various effects of WOM communication strategies, suc h as consumer attitudes (Brucks, 1985),

PAGE 41

41 consumer risk taking (Woodside & Delozier, 1976), short term and long term product evaluation (Bone, 1995), and consumer choice of product (Lau & Ng, 2001). In the communication field, Hong and Yang (2009) revealed t hat in terms of the public relations perspective, corporate reputation and relational satisfaction are critical Note that a l though many researchers have investigated the effect of WOM, its antecedents and moderators have been less tested (De Matos & Rossi, 2008). Electronic Word of Mouth Drawing on the importance and effectiveness of WOM in the consumer decision making process, the Internet has been widely used as an important sales channel (Cheema & Papatla, 2010 ). According to a recent survey from the U.S. Census Bureau, total U.S. online retail commerce exceeded $41.525 million in 2010, a 13.6% increase over the previous year (Winters, Detlefsen, & Davie, 2010) and online purchasing behavior is high ly affected by online formation seeking activities such as electronic word of mouth (eWOM) information. Not only online consumers, but also general consumers rely on online information. Recognizing the importance of online information concerning consumer b ehavior, previous researchers have examined a new dimension of WOM that uses online platforms such as online product reviews and forums (Bick art & Schindler, 2001; Dellarocas, 2004), which corresponds to the increasing interest in the Internet as an inform ation source ( Pew Research Center, 2005; Prendergast et al., 2010). Moreover, a recent survey conducted by the Pew Research Center (2011) revealed that the Internet is the second most widely used news source, following television news, and among 18 to 29 y ear olds, the Internet has surpassed television as

PAGE 42

42 the main news source. Therefore, it is reasonable to assume that people will try to obtain product related information through the Internet. Thus, word of mouth marketing is a particularly important featu re of the Internet (Trusov et al., 2009). According to Sohn (2009a), previous literature demonstrated that eWOM is an outcome of psychological motives such as innovativeness (e.g., Henning Thurau, Gwinner, Walsh, & Gremler, 2004; Phelps et al 2004; Sun, Youn, Wu, & Kuntaraporn, 2006) and the social tie among individuals on the Internet (e.g., Steyer, Garcia Bardidia, & Quester, 2006; Vilpponen, Winter, & Sundquist, 2006). The effectiveness of WOM in the online setting has received attention in different c ontexts, such as movie reviews (Basuroy, Chatterjee, & Ravid, 2003; Dellarocas, Zhang, & Awad, 2007). The accessibility diagnosticity model (Feldman & Lynch, 1988; Herr et al. 1991; Lau & Ng, 2001) provided a theoretical explanation of the effectiveness o f WOM compared to mass communication messages. With face to face (vivid) presentation, and decision making process (Feldman & Lynch, 1988; Herr et al., 1991; Kisielius & Sternthal, 1984; Lau & Ng, 2001; McGill & Anand, 1989). Electronic word of by potential, actual, or former customers about a product or company, which is made available to a multitude of people Thurau et al., 2004, p. 39). In the online environment, the concept of e WOM has emerged with the possibilities of unlimited network ing through the Internet (Smith, Coyle Lightfoot, &

PAGE 43

43 Scott 2007) and enabling re al time activity, conversation, or correspondence about a product (Kliatchko, 2008). Several studies have confirmed that eWOM can overcome the drawbacks of WOM communication based on two factors: measurability and transparency (Godes & Mayzlin, 2004; Park & Kim, 2008; Rezabakhsh, Bornemann, Hansen, & Schrader, 2006 ). For example, Godes and Mayzlin (2004) indicated that online WOM behavior could be a product review or c omment in online forums, whereas it is difficult to measure the underlying dynamics of WOM since offline WOM information is usually exchanged in personal communication channels (Park & Kim, 2008). Also, using online channels, consumers obtain more power to overcome information asymmetries and achieve market transparency (Rezabakhsh et al., 2006). Although about 90% of WOM communication has accrued in the offline context (Keller & Berry, 2006), eWOM in the Internet context could change the way WOM influence consumer behavior. Many studies have reconfirmed the effectiveness of WOM in consumer behavior, unlike in mass media, information transfer through interpersonal channels such as informal conversation or discussion is invisible (Rosen, 2000). However, due t o pervasive use of the Internet, researchers can now measure eWOM through a blog or social networking site such as Facebook or Twitter that allows individuals to write a comment or express their opinion to various stakeholders (Sohn, 2009b). The eWOM examp les include blogs, online reviews (Dellarocas, 2003 ; Dellarocas et al. 2007; Duan, Gu, & Whinston, 2008), online discussion forums (Bickart &

PAGE 44

44 Schindler, 2001; Chevalier & Mayzlin, 2006; Chiou & Cheng, 2003; Dellarocas, 2004), and news groups. Okazaki (200 8) indicated that WOM marketing has expanded in the computer mediated environment, which includes e mail, blogs, community sites, online review (Dellarocas, 2003; Dellarocas et al. 2007; Duan et al. 2008), discussion forums (Bickart & Schindler, 2001 ; Ch evalier & Mayzlin, 2006; Chiou & Cheng, 2003; Dellarocas, 2004), news groups, and even mobile SMS message based WOM mark eting strategies (Okazaki, 2008, 2009). Table 2 2 summarizes important eWOM related literatures.

PAGE 45

45 Table 2 1. Technology acceptance mo del, social influence related literatures in adoption study Context Author (year) Method Significant moderators, Predictors Main Findings Mass Customized Newspaper adoption (TAM) Putzke et al (2010 ) Survey Willingness to invest effort, gender, perceived u sefulness Willingness to invest effort for mass customized products, perceived usefulness were significant predictor. For women, base category satisfaction was significantly affects to PU. Instant messaging (UTAUT) Lin et al (2009) Survey Effort expecta ncy, social influence, and peer influence Effort expectancy, social influence and peer influence were significantly affected to messaging usage Internet based course management system (TAM) Park et al (2007) Survey Perceived ease of use, perceived usefulness, compliance with school policy Perceived ease of use, perceived usefulness and compliance with school policy played a significant role to courseware adoption VoIP adoption (TAM + U&G) Park (2010) Survey Pe rceived ease of use, perceived usefulness, motivation for communicatio n, instrumental use Perceived ease of use affects perceived usefulness and perceived usefulness has positive association with actual VoIP use. Also, instrumental and communication motiva tion has positive relationship with perceived ease of use, perceived usefulness and usage of VoIP Mobile Internet service Jiang (2009) Survey Beliefs about mobile Internet, quality perception Belief about mobile internet and quality perception has linear relationship with adoption intention

PAGE 46

46 Table 2 1. Continued Context Author (year) Method Significant moderators, Predictors Main Findings Instant Messaging Carr (2008) Interview Organization members were introduced to instant messaging primarily by peers and utilized for peer interaction Sales technology adoption Homburg et al (2010) Survey Perceived Usefulness, perceived ease of use, perceived adoption pressure Perceived usefulness, perceived ease of use and perceived adoption pressure of manager paly an important role to new sales technology adoption

PAGE 47

47 Table 2 2 eWOM related literature Context Author (year) Method Significant moderators, Predictors Main Findings Importance of eWOM Rezabakhs h et al (2006) Survey Willingness to invest effort, gender, perceived usefulness e WOM empowered consumers to solve (1) asymmetries information about the product (2) boost market transparency. Online review of Electronic product (PMP) Park & Kim (2008) Experi ment Consumer expertise, cognitive fit (types of reviews), number of reviews The volume of reviews are positively related to show preference Online review of movie Dellarocas et al (2007) Second ary analy sis Volume of reviews The volume of reviews are positively related to movie sales Online newsgroup and TV shows Godes & Mayzlin (2004) Second ary analy sis Volume of reviews The volume of reviews are positively related to show preference Factors influencing e WOM in Online forum Prenderga st et al (201 0) Survey Source similarity, persuasivenes s, attitude toward the forum topic, initial attitude toward the forum are strong predictors Information sources in product evaluation Cheema & Papatla (2008) Survey Produ ct type (hedonic vs. utilitarian), Internet experience Consumers consider online information source is relatively important than offline sources in utilitarian product evaluation than hedonic product evaluation For the consumers with high level of experien ce, offline information is more important

PAGE 48

48 Table 2 2. Continued Context Author (year) Method Significant moderators, Predictors Main Findings Effectiveness of e WOM strategy comparing with traditional marketing strategy Trusov et al (2009 ) Second ary A nal ysis eWOM WOM referrals have more effects than traditional marketing. Factors of mobile WOM Okazaki (2008) Survey M obile cam paig n Brand commitment, relationship with mobile device, group person connectivity, perceived value of campaign Commitment to th e promoted brand, relationship with mobile device, group person connectivity are all important antecedent of referral. Entertainment value is more strong predictor than purposive value Actual usage of mobile base WOM Okazaki (2009) Survey Mobile ca m paig n Interpersonal connectivity, self identification with the mobile device, Affective brand commitment For brand commitment, WOM is more affective than mobile WOM, however in terms o willingness to make referrals, mobile WOM lead more persuasive power Online consumer behavior in book sales Huang & Chen (2006) Experi ment Sales volume, customer reviews, evaluations of others, (expert review vs. consumer review) The recommendations of peer influence more than expert. eWOM in China Xue & Zhou (2011) Experi ment Positive/negative WOM, involvement Positive/negative WOM, involvement, prior experience are significant moderator

PAGE 49

49 Table 2 2. Continued Context Author (year) Method Significant moderators, Predictors Main Findings eWOM in brand attitude Lee et al (2009) Experi ment Positive/negative (extremely positive, moderate, extremely negative) Extremely positive reviews increased brand attitude. Extremely negative reviews had strong influence than positive WOM. eWOM in apartment review Lee & Youn (2009) Exper i ment eWOM platform (independent product review website, website, personal blog), eWOM valence (positive vs.negative Participants exposed to the review on personal blog were less likely to recommend the product to friends than those exposed to the review on other platform (contrary to previous assumption) In the negative eWOM, there was no significant difference eWOM in brand choice facilitation Hung & Li (2007) Second ary analy sis eWOM in virtual community In the virtual community setting, eWOM hig hly affect toward consumer brand selection Online consumer behavior in book sales Huang & Chen (2006) Experi ment Sales volume, customer reviews, evaluations of others, (expert review vs. consumer review) The recommendations of peer influence more than exp ert. eWOM in brand trust of mobile phone Xingyuan et al (2010) Survey WOM WOM increase brand trust

PAGE 50

50 Table 2 2. Continued Context Author (year) Method Significant moderators, Predictors Main Findings eWOM motivation to leave a online review Hennig thrau et al (2004) Survey Platform assistance, venting negative feelings, concern for other consumers, extraversion/posi tive self enhancement, social benefits, helping the company, advice seeking Platform assistance, venting negative feelings, concern for other consumers, extraversion/positive self enhancement, social benefits, helping the company, advice seeking are significant moderators eWOM motivations to read consumer review Hennig Thurau & Walsh (2003) Survey Obtaining buying related information, s ocial orientation through information, community membership, remunerate, to learn to consume a product Obtaining buying related information, social orientation through information, community membership, remunerate, to learn to consume a product are all the significant predictor of eWOM reading motivation eWOM motivations to write review Hennig Thurau et al (2004) Survey Platform assistance, venting negative feelings, concern for other consumers, extraversion/posi tive self enhancement, social benefits, eco nomic incentives, helping the company, advice seeking Platform assistance, venting negative feelings, concern for other consumers, extraversion/positive self enhancement, social benefits, economic incentives, helping the company, advice seeking are all the important predictor of motivation for writing review

PAGE 51

51 Table 2 2. Continued Context Author (year) Method Significant moderators, Predictors Main Findings Predictor s of online information searching in automobile purchase Ratchford et al (2003) Survey Se cond ary data analy sis Age, education level Young and more educated people are more frequently use the Internet as a information source Motivation for online opinion seeking Goldsmith & Horowitz (2006) Survey Perceived risk, influence of others, price con sciousness, ease of use, cool, to get information, saw on TV Perceived risk, influence of others, price consciousness, ease of use, to get information, saw on TV are all significant predictor of motivation for on line opinion seeking Product involvement as a moderator of eWOM Xue & Zhou (2011) Experi ment Product involvement, previous WOM experience, negative/positive WOM For higher product involvement participants, negative comments are more powerful than positive comment Online forum discussion Bickart & Schindler (2001) Experi ment Online forum Online forum influenced more than corporate webpage. Online forum discussion Chiou & Cheng (2003) Experi ment Brand image For low image brands, negative WOM hurt more than high image brand their brand image Online review of movie Basuroy et al (2003) Second ary analy sis eWOM, star, budget Negative reviews hurt performance more than positive reviews help performance Online review Duan et al (2008) Second ary analy sis The n umber of ratings, rating quality (high vs. low) Unlike previous literature, higher ratings do not lead to higher sales whereas the number of ratings highly related the movie sales

PAGE 52

52 Table 2 2. Continued Context Author (year) Method Significant moderators, Predictors Main Findings Online feedback Dellarocas (2003) Concep tual Age, education level Online feedback is a good opportunity for operations research, management science research eWOM Online book reviews Chevalier & Matzlin (2006) Second ary analy sis V olumes of review, review quality (high/ low) The volume of reviews are highly related to firms profit, negative review affects more than positive Online consumer review Chen & Xie (2008) Second ary analy sis eWOM eWOM, in particular online consumer review i s influential to consumer reaction Online review in movie market Liu (2006) Second ary analy sis The review valence (positive/ negative) The number of review Volume of review affects movie sales whereas their valence affect limitedly eWOM in online communi tie s (blog) Kozinets et al (2010) Qualitat ive Evaluation, explanation, embracing, endorsement There are four different strategies for the communal reference expression Online professor evaluation Edwards et al (2009) Experi ment Expectancy Expectancy eff ects play on mediate role Online review Park et al (2007) Experi ment Involvement, quantity of review, quality of review For low involvement consumers, quantity of review affects more than quality eWOM Sohn (2009a) Experi ment Social network density, info rmation valence In the dense social network, positive and negative information valence did not differ whereas in the non dense network situation, positive product information is more valuable than negative product information

PAGE 53

53 Table 2 2. Continued Contex t Author (year) Method Significant moderators, Predictors Main Findings eWOM in online forum Sohn (2009b) Survey Social norm, perceived value of information Social norm affects eWOM intention eWOM in SNSs Chu & Kim (2011) Survey Trust, normative influenc e, informational influence Trust, normative influence, informational influence are positively associated eWOM behavior

PAGE 54

54 Figure 2 1. The t heory of r easoned a ction Figure 2 2. The t heory of p lanned b ehavior Belief of Consequences Evaluation of Beliefs Normative Beliefs Motivation to Comply Stimulus Conditions Attitude Subjective Norm Behavioral Intention Behavior Belief of Consequences Evaluation of Beliefs Normative Beliefs Motivation to Comply Control Beliefs Perceived Impo rtance of Control Attitude Subjective Norm Perceived Behavioral Control Behavioral Intention Behavior

PAGE 55

55 Figure 2 3. Technology a cceptance m odel Figure 2 4. United t heory of a cceptance and u se of t echnology (UTAUT) External Variables Perceived Usefulnes s Attitude P erceive e ase of Use Behavioral Intention to Use Actual Use Performance Expectancy Social Influence Facilitating Conditions Behavioral Intention Effort Expectancy Use Behavior Experience Voluntariness of use Age Gender

PAGE 56

56 Figure 2 5. Decomposed t heory o f p lanned b ehavior

PAGE 57

57 CHAPTER 3 LITERATURE REVIEW AND RESEARCH QUESTIONS This chapter provides the proposed measurements and the specific theoretical and send electro nic word of mouth (eWOM) messages through Twitter Twitter Usage Motivation Testing Social Influence in Twitter Adoption Although the initial premise of TAM highly rel ies on the social influence dimensions predicting an beha vior (Davis, 1989), the social influence dimensions were rarely tested in relevant empirical studies T his research uses an integrated approach of TRA, TPB, and TAM, and the original assumption of perceived subjective norm (i.e., normative social influence ) that influenced intention to adopt new media (Fishbein & Ajzen, 1975). Indeed, the original theoretical framework of social influence revealed that normative (subjective) social influence affects a situation that is new and ambiguous (Asch, 1951). This t rend of prominent effects of social influence on unfamiliarly situation, also particularly reconfirmed in the context of early stages of experience of technology ( Lin et al., 2009; Venkatesh et al., 2003). havioral intention to act in a certain detail, Fishbein and Ajzen (1975) indicated that attitude and subjective norm both influence behavior intention. Also, conformity to the social norm influences new media adoption in various contexts, such as new distance learning technology adoption (Lee,

PAGE 58

58 in using technology such as Twitter coul d be influenced by the level of their conformity tendency. Applying the logic of TRA and TPB (Fishbein & Ajzen, 1975) from the early stage of adoption study (e.g., Mathieson, 1991; Taylor & Todd, 1995 ) to Internet technology based media (e.g., Carr, 2008; Chu & Kim, 2011; Lee et al., 2003 ; Lin et al., 2009; Park et al., 2007 ), many studies reconfirmed the role of the conformity tendency toward an adoption intention and usage. For example, Lin and colleagues (2009) demonstrated the effects of p messaging choice: when the majority of others use an instant messaging program which differs from their current program, they might change their choice of instant messaging program (Chung & Nam, 2007 ). Carr (2008) also indicated that organization instant messaging usage is for peer interaction. Park et al. (2007 ) also indicated that compliance with school policy played a s courseware adoption. Based upon these research findings, it is plausible that if participants have higher tendency of compliance toward the subjective norm, then they might adopt Twitter more easily. In addition considering the fact that SNSs, especially Twitter, can still be considered emerging media with a short history that are highly influenced by social norm (Asch, 1951; Venkatesh et al., 2003), thus it is expected that the attitude toward the SNSs and intent ion to adopt the SNSs can vary by the degree of conformity toward social norms (Lee, 2003). Thus, the following hypothesis is postulated:

PAGE 59

59 H1 attitude toward Twitter and (b) Twitte r usage. Testing Technology Acceptance Model in Twitter adoption perceived utility (Davis, 1989; Schiffman & Kunuk, 1978). According to the Technology Acceptance Model (TAM), t wo main dimensions affect attitudes or intentions to adopt products with newly developed technology: perceived usefulness and perceived ease of use (Davis, 1989 ; Davis et al., 1989). Perceived usefulness ual believes that using a particular system would Perceived u sefulness This refers to the degree of personal belief that the usage of technology can ). Gefen, Karahanna, & Straub (2003) conceptual defini ) c oncept of relative advantages. Considering perceived usefulness, the significant prediction role of perceived usefulness is reconfirmed in shopping on social networking web sites (Cha, 2009a) and online video platforms (Cha, 2009 b ) and. Also, in the recent context of mobile phone adoption, the dimensions of perceived usefulness are measured by three differe nt aspects of perceived utility: perceived utility of new technology, perceived utility of a new service, and perceived utility of a handset (Teng, Lu, & Y u, 2009). Putzke et al. (2010) tested perceived usefulness in consumer adoption of a mass customized newspaper and found that it played a significant prediction role. In addition, as

PAGE 60

60 supported by previous literature, the positive relationship between pe rceived usefulness and technology adoption intention was reconfirmed in the context of Internet technology based media, such as course manag ement systems (Park et al., 2007 ) and voice over Internet protocol (VoIP) (Park, 2010). Therefore, it can be assumed that it can be assumed that individuals who initially perceive Twitter to be greatly useful exhibit a favorable attitude toward Twitter and use it more frequently. Thus the following hypothesis is developed for testing the perceived usefulness: H2 : Percei attitude toward Twitter and (b) Twitter usage. Perceived ease of use that using a particular system would 323). Davis (1989) indicated the similarity of perceived ease of use and self efficacy, cost, effort, and complexity ( Lee, 2003; Lee et al., 2003). Adopting this conceptual definition, much previous lit erature revealed a positive relationship between perceived ease of use and media adoption intention, such as shopping in social networking (Cha, 2009a ) and e commerce service ( Gefen & Straub, 2000; Lee, Park, & Ahn, 2001). Indeed, according to Lee (2003), recent studies in new media and technology adoption indicate the importance of a simple interface design (Raskin, 1997), represented by the slogan as KISS (Keep It Simple, Stupid) (Preec e Rogers, & Sharp, 2002). In particular, in the Web browsing context, simple and easy to use have been the primary factors in determin ing which Web browser to use (Gambon, 1998). As Twitter is primarily Web based and requires an Internet connection, it is expected that the factor of perceived ease of use play an important role in Twitter

PAGE 61

61 adoption affect ing both attitudes toward Twitter and actual usage. Also, recent research in Internet based technology such as voice over Internet protocol (VoIP) (Park, 2010), Internet based course mana gement system (Park et al., 2007 ), al so revealed a positive association of perceived ease of use and adoption intention of new media platform. Thus, the following hypothesis was proposed: H3. toward Twitter and (b) T witter usage. Consumer Related Characteristics Age Consumer demographic factors in the online environment also have been investigated. Traditionally, in terms of communication technology adoption, age was negatively associated to consumer adoption behavior (e.g., Chan Olmsted et al., 2005; Lin, 1998). Drawing upon this role of age in an online environment, Ratchford, Lee, and intention to employ information search behavior o n the Internet. They concluded that younger and more educated people tend to use Internet searching in automobile purchasing decisions. This result is consistent with other social networking research which generally found that age has a negative relations hip with online shopping on soc ial networking sites (Cha, 2009a ) and e commerce (Akhter, 2003). Also consistent with other findings on online technology adoption using the Internet (Madden & Savage, 2000), online chat rooms and webcasts use have a negative relationship with the age group (Lin, 2004; Peter, Valkenburg, & Schouten, 2005). Previous research has also indicated that older people are reluctant to adopt new technology and exhibit negative perceptions toward new technology ( Cha, 2009b; Gilly & Zeithaml, 1985;

PAGE 62

62 Pommer, Berko witz, & Walton, 1980 ). Drawing on the basic concept of adoption in this inverse relationship of age and technology usage, He and Mykytyn (2007) expanded its context to adoption of online payment methods and shopping on soc ia l networking sites (Cha, 2009a ). Accordingly, the following hypothesis is postulated : H4 Twitter usage Education l evel Ratchford et al. (2003) also revealed the linear relationshi p between education level and online information searching. Therefore, it is expected that education level and attitude toward Twitter adoption have positive relationships with education levels. H5 attitude toward Twitter and (b) Twitter usage. Gender Gender differences in new media adoption have been studied for over a decade. In the early era of new media diffusion, males consistently used new media more frequently than females ( Dutton et al., 19 87; Jeffres & Atkin, 1996 ; LaRose & Atkin, 1988 ). Consistent with the results of previous studies, Internet usage was also prominent among males (Ernst & Young, 1999 ). However, in terms of current usage status of the Internet, there was no significant diff erence between males and females ( Dennen, 2011; Radwanick, 2011) Furthermore, regarding Internet based, advanced technology usage, such as instant messaging, females use instant messaging more frequently for obtaining socialization gratification (Leung, 2 001 ). Considering social media usage patterns, recent industry reports revealed that the usage gap between genders did not differ significantly, and females even spend more time on SNSs than

PAGE 63

63 males ( Dennen, 2011; Radwanick, 2011 ). Corresponding with industr y reports ( Dennen, 2011; Radwanick, 2011 ), Cha (2009 a ) revealed that, in terms of frequency, males and females ar e not statistically different. It seem that there was conflicting results regarding the traditional new media adoption study and SNSs specific usage studies Therefore, the following research question is proposed: RQ1: Twitter usage? Attitude Toward Twitter and Actual U sage of Twitter Although established relationships be tween attitude toward and purchase intention or have been discussed in previous literatures particularly from marketing advertising ( e.g., Brown & Stayman 1992 ; McKenzie & Lutz 1989 ) and new media adoption study research context (e.g., Vishwanath & Gold haber, 2003 ) Ajzen and Fishbein (2005) argued that there are possible differences between the attitude and actual behavior of individuals. Though most empirical studies have shown a positive relationship between attitude and behavior (Ajzen & Fishbein, 2 005), it is also expected that consumer evaluation of Twitter and their actual usage of Twitter can vary ( Ajzen & Fishbein, 1977; Fishbein & Ajzen, 1975; Perloff, 2010). This study measured multiple dimensions of fically, attitude toward the Twitter and actual Twitter usage ( daily usage, hourly usage, and frequency of obtaining brand information ) were evaluated. T hus, it can be postulated that favorable Twitter attitude can be transferred to Twitter usage. In othe r words, attitude toward Twitter and Twitter usage would be positively associated, the following hypotheses are proposed:

PAGE 64

64 H6 WOM and eWOM R elated F actors Communicator Characteristi cs Perceived similarity One of the most important factors affecting WOM is perceived communicator characteristics such as similarity (Brown & Reingen, 1987; Gilly, Craham, Wolfnbarger, & Yale, 1998 ; Price, Feick, & Higie, 1989; Wangenheim & Bayon, 2004). A ccording to Wangenheim and Bayon (2004), several studies have provided theoretical explanations for why people prefer messages from those who share similar characteristics. First, to compare themselves with others increases when they encounter people they consider to have similar characteristics. This is because consumers assume that similar people will share similar needs and preferences. Second, the source attractiveness model (Ke lman, 1961) explained that receivers more closely identify themselves with similar sources. up hypothesis suggested that the effect of information could differ from the congruity of the image of the communicator and perceived i mage of the receiver. Empirical studies in the WOM context reconfirmed that the perceived source similarity induces more influential power toward the consumer product and brand selection (Brown & Reingen, 1987; Gilly et al., 1998; Price et al., 1989; Wange nheim & Bayon, 2004). Similarity influences behavior intention, and this logic is applicable, too, in the context of the online environment. Particularly, in the eWOM environment of online forums, Prendergast and colleagues (2010) found that perceived simi larity of the communicators were positively correlated with the attitude toward the eWOM platform and purchasing intention.

PAGE 65

65 Likewise, in the Twitter environment, the similarity among users (both the following and followers), particularly in terms of shar ing similar interests, might affect the persuasion power of the product information, as previous literatures from both traditional WOM and eWOM indicated that higher perceived source similarity of group members was positively associated with the individual attitude and behavior intention ( Prendergast et al., 2010; Wangenheim & Bayon, 2004, 2007 ). Therefore, the following hypothesis was postulated: H7 The perceived similarity between a Twitter user and Twitter friends who tweet about a branded product will be positively related to (a) attitude toward the brand and (b) eWOM spreading intention Source credibility expert in a certain topic can also affect the evaluation of WOM (Wangenheim & Bayon, 2004). Various scho lars in particular communication area have found that higher credibility of a spokesperson lead s to more persuasive power (e.g., Hovland, Janis, & Kell e y, 1953; Jones, Sinclair, & Courneya, 2003; Kiousis, 2001 ). Based on the research trends in source credi bility and expertise, this study proposes the concept of source credibility based on the model of Petty and ration likelihood model (ELM). The ELM has two basic notions First, when people are exposed to a certain kind of persuasive message, they might c and reasoning (central cue). Or the audience may be influenced by a peripheral cue, such as mood or fee ling (Petty & Cacioppo, 1984). Adopting this logic, many factors have been scrutinized, incl uding personal involvement, argument repetition, and characteristics of source (Jones, et al., 2003). Hovland and colleagues conducted a systemic study ( Hovland et al., 1953) which

PAGE 66

66 determined that source credibility represents the effects of the different characteristics of the communicator on the processing of a message (Kiousis, 2001) D esirable source credibility has been investigated in various contexts such as political communication (Iyengar & Valentino, 2000) and effective advertisement (Gotlieb, & S arel, 1991; Yoon, Kim, & Kim, 1998 ). For example, Iyengar and Valentino (2000) revealed that it is political party in making an effe ctive political advertisement. Go tlieb and Sarel (1991) considered the source credibility concept in terms of advertisement design and also confirmed that overall source credibility leads to more favorable attit udes toward the advertisement. In addition, this relatively high impact of sou rce credibility on the persuasion message is even confirmed by cross cultural comparison between samples from the United States and those from Korea. S ource credibility consisted primarily of two key features: source expertise and attractiveness. The conce pt of message source credibility mainly from the source expertise was reconfirmed in the WOM context (Bansal & Voyer, 2000; Wangenheim & Bayon, 2004). For example, Bansal and Voyer purchasing decisions. Wangenheim and Bayon (2004) also confirmed the effects of higher credible sources on service switching. Likewise, it is expected that perceived source credibility as people consider highly credible sources on Twitter to be opinion leaders (Jacoby & Hoyer, 1981), and they perceive superior source credibility as an indication of product quality (Gilly et al., 1998). As a result, we expected that:

PAGE 67

67 H8 The perceived sou rce credibility of Twitter friends who tweet about a brand spreading intention. Product Related Characteristics Hedonic and u tilitarian product c ategory Traditional WOM researchers mainly focused on tangible (Arndt, 1967; Sheth, 1971) and relatively new products (Whyte, 1954). Overtime, r esearchers have expanded their scope of investigating a WOM to intangible products such as services WOM is particularly important in t he service industry since intangibility leads to difficulty in prepurchase trials (Berry, 1980; Zeithaml, 1981; Zeithaml, Berry, & Parasuraman, 1993; Zeith aml, Parasuraman, & Berry, 1985 ). Also, due to high levels of complexity and perceived risk (Berry, 1 980; Zeithaml et al., 19 81; Zeithaml et al., 1985), WOM information is considered credible among consumers since it is independent from organizations (Silverman, 2001; Sweeney, Soutar, & Mazzarol, 2008). A recent survey from Nielsen (2011) indicated that electronics and computer equipment are the products most highly influenced by eWOM (29%), followed by beauty care and clothes (27%), financial products services (27%), telecommunication services (25%), and travel and entertainment (22%). Also, WOM is part icularly powerful in experience goods ( Godes & Mayzlin, 2004 ; Granovetter, 1973 ) such as telecommunication services. E mpirical research has shown that the WOM effect is prominent in the service context because of its intangible and experiential nature ( De Matos & Rossi, 2008;

PAGE 68

68 Murray, 1991 ; Zeithaml et al. 1993 ). In summary, product categories can vary in their level of influence in WOM and eWOM, including intangible services like movies (Basuroy et al., 2003) and tangible products like computers (e.g., He rr et al., 1991; Laczniak, DeCarlo, & Ramaswami, 2001; Lee, Rogers, & Kim, 2009 ), automobiles (Mizerski, 1982; Ratchford, Lee, & Talukdar, 2003), and mobile phones (Xingyuan, Li, & Wei, 2010). Research has suggested that consumer purchasing motivations hav e mainly focused on two different motivatio ns: utilitarian and hedonic ( Babin & Darden, 1995; Ba bin, Darden, & Griffin, 1994; Cha, 2009 a ). The definitions of utilitarian and hedonic moti ves differ between researchers. For example, Batra and Ahtola (1990) d efined utilitarian aspect as non emotional outcome, task oriented, and cognitive, whereas the hedonic has aspects sensory attribute fantasy, and emotive feeling. Empirically, utilitarian motivation s are typically ial the s are ). Psychologically, these two consumer motivations (utilitarian and hedonic) are neither mutually exclus ive nor equally salient. Some brands or product categories generally consider utilitarian motivation to be more salient than hedonic motivation (e.g., selecting toothpaste to prevent going to the dentist versus choosing a luxury fash ion brand for its appe arance). In addition, utilitarian and hedonic considerations affect simultaneously when consumers are purchasing products. For example, when consumers consider purchasing an automobile, they might think about the both gas mileage (utilitarian) and sporty d esign (hedonic) concurrently (Dhar & Wertenbroch, 2000). This differentiation

PAGE 69

69 between utilitarian and hedonic motivation is frequently mentioned in several examples of marketing literature focusing on different motivations for purchasing behavior, i.e., th e two different dimensions of utilitarian (product oriented) and hedonic (experience oriented) aspects ( Babin et al. 1994; Hirschman & Holbrook, 1982; Holbrook & Hirschman, 1982). Drawing upon the concepts of utilitarian and hedonic motivations, differe nt effects of fostering attitudes toward products have been investigated by various disciplines such as marketing, sociology, psychology, economics (Barta, & Ahtola, 1990; Chitturi, Raghunathan, & Mahajan, 2007; Dhar & Wertenbroch, 2000; Okada, 2005), and communication research (Cha, 2009 a ). and hedonic motivation is utilized when marketers design the marketing strategy. For example, Cha (2009 a ) revealed that shopping for r eal goods on social networking networks is more dependent on the utilitarian value, whereas shopping for virtual items is more relevant for hedonic motivations. Similarity, in terms of brand evaluation and the spread of eWOM on Twitter, product category mi ght also play a role Generally, online consumers tend to seek utilitarian values rather than hedonic values, as the online shopping environment lacks multisensory attributes for hedonic dimensions (Reibstein, 2002). However, online information seeking beh avior or the spread of eWOM would be different from the actual purchasing behavior of consumers, as the recent survey by Nielson (2011) indicates that consumer eWOM activities are associated with both utilitarian (e.g., electronics and computer equipment) and hedonic (e.g., entertainment, travel service) dimensions. Considering the mixed results from online purchasing

PAGE 70

70 literatures and eWOM research, this study investigates the differences and similarities among factors that affect the evaluation of brands an d eWOM spreading intention in the Twitter context. Previous literature from both the industry and academic perspectives have dealt with different product categories. For example, Cha (2009 a ) used the product categories of books, tickets, DVDs, clothing and accessories, computers and accessories, a nd video games. Allsop et al. (2007) selected restaurants, computers, movies, vehicles, nutrition and healthy eating, health care providers, financial products/services, and vacations; Graham and Havlena (2007) stu died the product categories of auto, retail, soft d rinks, technology, and travel. From the industry viewpoint, Neilson (2011) revealed a ranking of the most frequently discussed product categories in the online environment: electronics and computer equipme nt (29%), beauty care and clothing (27%), finance products/services (27%), telecommunication services (25%), and t ravel and entertainment (22%). ComScore (2007) reported the percentage of review viewers who subsequently made a service purchase: restaurant (41%), hotels (40%), travel (27%), automotive (24%), home (19%), medical (14%), and legal (8%). The influence of an online review on the purchase decision also varied by product type: restaurant (79%), hotels, (87%), travel (84%), automotive (78%), home (7 3%), medical (76%), and legal (79%). Therefore, it is unclear which product category would be most frequently discussed on Twitter and the varying effects of product category may have on consumer brand obtaining behavior. RQ2 How does the product categor attitude toward the brand and (b) eWOM spreading intention?

PAGE 71

71 Perceived fit In most marketing literature, consumer evaluation of brands is moderated by perceived fit or similarity between extension and parent brand (e.g., Aaker & Keller, 1990; Bhat & Reddy, 2001; Park, Jaworsk, & Macinnis; 1986 ). Previous literature consistently shows that the better the fit between the extended product and the parent brand, the more positive evaluations will be toward the brand exte nsions. The categorization and schema theory (Aaker & Keller, 1990; Bhat & Reddy, 2001; Boush & Loken, 1991 ) justifies this logic: When consumers sense a close relationship between the original brand and the extended brand, they might associate both with t heir preexisting perceptions, thoughts, and categories. Therefore, the greater the perceived fit between the original brand and extended brand, the more favorable the evaluation toward the brand extension will be. Similarly, when the company launched the n ew products or services, individual evaluation of perceived fit was similar to that of the parent brand (Boush et al., 1987; Papadimitriou, Apostolopoulou, & Loukas, 2004). Drawing on this important role of perceived fit in the context of SNSs, Cha (2007, 2009 a ) indicated that developing shopping services in the SNSs could be considered a category extension. Therefore, Cha (2007, 2009 a perception of the fit between SNSs and individual items (virtual items vs. real items) on th e shopping services of SNSs influenced their attitude toward the shopping services in the SNSs. Likewise, as Twitter is a microblogging service, when individuals are exposed to the product or service information on Twitter, if the fit between Twitter and t he items are high, they might have a more favorable attitude toward the product or service. This process can be explained by the reasoning that introducing product information on Twitter can be considered a category extension as described in most marketing

PAGE 72

72 literature (Cha, 2009 a the individual items introduced would affect the attitude toward the brands and their eWOM spreading intention. Applying this logic, the following hypothesis is propo sed for both the utilitarian and hedonic product category: H9 Perceived fit between Twitter and utilitarian product category information is intention on Twitter H10 Pe rceived fit between Twitter and hedonic product category information is intention on Twitter Attitude T oward the B rand eWOM I ntention and Purchase I ntention Adopting the argument that there are possible difference s between the attitude and actual behavior of individuals (Ajzen & Fishbein, 2005), this study measured multiple dimension s of attitude and intention. Specifically attitude toward the brand, eWOM i ntention and purchase intention were evaluated. Although e stablished relationships between attitudes toward advertising and attitudes toward the brand and purchase intention have been well documented in prior literature including marketing, advertising co ntexts ( e.g., Brown & Stayman 1992; MacKenzie & Lutz 1989) and Internet advertising ( e.g., Cho & Leckenby 1999 ; Cho & Cheon, 2004; Ko et al. 2005 ) Previous studies have consistently found positive relationships between attitudes and behavioral intent ion in particular for purchase intention This study also expects that these relationships occur within attitude toward brand, eWOM intention and purchase intention

PAGE 73

73 Thus, the following hypotheses are proposed: H11 Attitude toward the brand is positively spreading intention H12 intention H13 Consumer Related Characteristics The relations hip between consumer characteristics and usage of Twitter as a tool for eWOM was also investigated. Specifically three sets of research questions were proposed in terms of age, education level and gender: RQ3 ward the brand, (b) eWOM spreading intention and (c) purchase intention? RQ4 eWOM spreading intention and (c) purchase intention? RQ5 de toward the brand, (b) eWOM spreading intention and (c) purchase intention? Figure 3 1, Figure 3 2, Figure 3 3 and Figure 3 4 describe the proposed model to predict the intention to use Twitter and eWOM motivations and Table 3 1 summarized the hypothese s and research questions in this study.

PAGE 74

74 Table 3 1. Summary of Twitter adoption related hypotheses and research questions Dimension Independent Variable Hypothesis Twitter Adoption Subjective Norm H1a. Conformity to subjective norm will be positively as toward Twitter. H1b. Conformity to subjective norm will be usage. Perceived Usefulness H2a. Perceived usefulness of Twitter will attitude toward Tw itter. H2b. Perceived usefulness of Twitter will Twitter usage Perceived Ease of Use H3a. Perceived ease of use will be toward Twitter. H3b. Perceived ease of use will be posi usage. Age H4a. Age is negatively associated with H4b. Age is negatively associated with Education Level H5a. Education level is positively itude toward Twitter H5b. Education level is positively associated with his/her Twitter usage Gender RQ1a. attitude toward Twitter? RQ1b. Twitter usage? Attitude toward Twitter H6. Attitud e toward Twitter is positively

PAGE 75

75 Table 3 2. Summary of eWOM perspective related hypotheses and research questions Dimension Independent Variable Hypothesis Twitter as a marketing Tool Perceived Similarity H7 a. The pe rceived similarity between a who tweet about a branded product will toward the brand H7 b. The perceived similarity between a who tweet ab out a branded product will spreading intention Source Expertise H8a. The perceived source credibility of Twitter friends who tweet about a branded product will be positively related H8b. The perceived source credibility of Twitter friends who tweet about a branded product will positively related to Product Category RQ2a How does the product category of a towa rd the brand? RQ2b How does the product category of a spreading intention? Perceived fit H9a. Perceived fit between Twitter and hedonic product category information is toward t he brand. H9b. Perceived fit between Twitter and hedonic product category information is spreading intention. H10a: Perceived fit between Twitter and utilitarian product category information is positively associated wi toward the brand H10b: Perceived fit between Twitter and hedonic product category information is spreading intention

PAGE 76

76 Table 3 2. Continue d Dimension Independent Variable Hypothesis Attitude tow ard the brand H11 Attitude toward the brand is positively intention H12. Attitude toward the brand is positively intention. eWOM intention H13. eWOM intention is positively associated Age RQ3 a. attitude toward the brand? RQ3 b. eWOM spreading intention? RQ3c. purchasing intention? Educational level RQ4a How does edu cational level brand? RQ4 b. How does educational level intention? RQ4c. How does educational level ? Gender RQ5a atti tude toward the brand? RQ5 b. eWOM spreading intention? RQ5c. purchasing intention ?

PAGE 77

77 Personal Trait dimension TAM dimension Soci al Influence dimension Figure 3 1 Proposed conceptual research model for Twitter adoption in this study Perceived Ease of Use Conformity to Norm Perceived Usefulness Attitude toward Twitter Twitter usage D emographical factors (Age, education level gender)

PAGE 78

78 Figure 3 2. Proposed model for Twitter adoption Perceived Ease of Use Conformity to Norm Perceived Usefulness Attitude toward the Twitter T witter U sage D emographic al factors (Age education level gender) H1a H1b H2a H3a H4a H5a RQ1a 6a H4b H5b RQ1b 6a H2b H3b H6

PAGE 79

79 Figure 3 3 P roposed conceptual m odel for the Twitter in eWOM p erspective eWOM sender characteristics (p erceived similarity, source expertise) Product related Characteristics (Product category, Perceived fit of utilitarian and hedonic product ) Consumer related characteristics (Age, gender, educational level) eWOM spreading intention Attitude Toward the Brand Purchase Intention Demographic factors (Age educational level gender)

PAGE 80

80 Figure 3 4 Proposed model for the Twitter in eWOM perspective Perceived Similarity Perceived Credibility Product Category : Utilitarian Perceived fit: Utilitarian Attitude toward the Brand eWOM Spreading Intention Perceived fit: Hedonic H7a H9b H8a H8b H7b RQ2a RQ2b H9a H10a H10b Purchase Intention H12 H11 H13 Demographic factors (RQ3, RQ4, RQ5) (Age, educational level gender) Product Catego ry : Hedonic

PAGE 81

81 CHAPTER 4 METHOD Measurement This study ado pted several types of variables from traditional media adoption studies based on the social influence perspective and eWOM literatures: For the first model, the attitude toward Twitter and actual Twitter usage (dependent variable), conformity to norm, perc eived usefulness, and perceived ease of use were measured. For the second eWOM research model, the attitude toward the brand and eWOM spreading intention (dependent variable), perceived similarity, perceived source expertise, product category, and perceive d fit were measured. Also, consumer related variables such as age, gender, income, and educational level are included in the study Pretest A pretest of 30 undergraduate students was carried out before the main test. The purpose of the pretest was to con firm the validity and reliability of the measurements Another purpose of the pretest was to ensure the wording and flow of the survey questionnaire to prevent potential confusion and misunderstanding before the main test. The questionnaire was created ba sed primarily on three well developed theoretical frameworks: TRA, TPB, and TAM. Participants and Procedure A total of 30 undergraduate students from a large southeastern university participated in the in class survey for extra course credit. The ages of the participants for the pretest were 19 to 36 years. The mean age was 21.42 ( SD = 2.99). The majority of participants were females (79.3%, n =24), and 20.7% of subjects were males

PAGE 82

82 (n = 6); in addition, 79.3% of participants were Caucasian (n = 23), 1 0.3 % were Hispanic (n = 3), and 10.3% were Asian (n = 3). All of the constructs in the questionnaire except demographic questions were assessed using multiple items with a 7 point Likert scale. The constructs included multiple indicators, such as (1) confor mity to norm, (2) perceived usefulness, (3) perceived ease of use, (4) attitude toward Twitter, (5) perceived similarity, (6) perceived credibility, (7) product category, (8) perceived fit, and (9) Retweet intention. Table 4 1 provides a descriptive summar y of questionnaire items. Reliability Tests used to estimate reliability in particular, for internal consistency that represents inter correlations among items. This study adopt ed the following criteria that are most Cha, 2009b; Chang 2 005; Hair, Anderson, Tatham, & Black, 1 998 ; Kline, 2005 ). The criterion was .60 ( Cha, 2009 b ; Hair et al., 1998 ). The results of the pretest indicated that all measurements used in the pretest among undergrad uate students were above the 2005 ). Table 4 2 displays the results of the reliability check for the pretest. Regression Analysis for Twitter Adoption Based on the a bove reliability test results, this study averaged the items for each of constructs, such as (1) conformity to norm, (2) perceived usefulness, (3) perceived ease of use, (4) attitude toward Twitter, (5) perceived similarity, (6) perceived credibility,

PAGE 83

83 (7) product category, and (8) perceived fit. Before conducting a regression analysis, a Pearson co rrelation matrix was examined. Table 4 3 shows the correlation matrix among variables. A regression analysis was then employed for the pretesting of 30 particip ants. More specifically there was a high correlation between perceived usefulness and perceived ease of use (r = .75, p < .01) and attitude toward Twitter (r = .78, p < .01). In addition, perceived ease of use was found to correlate with perceived useful ness (r = .78, p < .01) and attitude toward Twitter (r = .88, p < .01). In addition, considering other variables, there were no extremely high correlations among variables that would have made it possible to conduct a regression analysis (Chang, 2005). An other important role for the regression analysis of the pretest was to determine t he variable for Twitter usage. ways: days per week and hours per day. Although the two items were correlated (r = .75 p < .01) and measured the related theoretical concept of Twitter usage, this study used for mainly two reasons. First, most industrial reports used both constructs (Nie lson, 2011; Webster, 2010, 2011). Second, the mean scores and standard deviations for daily measurement ( M = 3.90, SD = 2.94) and hours per day ( M = 2.59, SD = 1.48) were different. Therefore, three regression analyses were conducted, which considered att itude toward Twitter and each Twitter usage variable (daily, hourly) using the dependent variables. First, the regression model for Twitter attitude was significant, with F (3, 25) = 42.93, p < .01, and explained 84% of the variance (R 2 = .84). All three p redictor

PAGE 84

84 variables, conformity to subjective norm ( = .20, p p p < .001) -were significant predictors of attitude toward Twitter. By using weekly usage as a dependent variable, the regression model was also significant, with F (3, 25) = 10.90, p < .01, and explained 57% of the variance (R 2 p < .05). However, in terms of hourly based Twitter usage, none of the predictors were significant. Table 4 4 show s the results of two regression analyses, with attitude toward Twitter and Twitter usage (both day based and hour based) as dependent variables. Therefore, all of the measurements that were used in the pretest were also used in the main test since all me asurements were reliable. Descriptive Statistics of the eWOM Product Category and Perceived Fit Another important purpose of the pretest was to differentiate dimen sions of the product category. In order to describe the discrete theoretical dimensions of t he product category and perceived fit, it is necessary to report descriptive statistics regarding the product categories most frequently discussed on Twitter in terms of brand information and their perceived fit between Twitter and brand i nformation obtain ing activity. In terms of product category, participants obtained movie information most frequently ( M = 4.56, SD = 1.80), followed by clothing information ( M = 4.45, SD = 1.97), travel information ( M = 4.41, SD = 1.72), and restaurant information ( M = 4.4 1, SD = 1.80). However, the respondents to the pretest rarely obtained health care service information ( M = 2.48, SD = 1.24) or computer information ( M = 2.76, SD = 1.62). perceived fit between Twitter and information obtaining too ls, they felt Twitter was most appropriate for obtaining movie information ( M = 4.79, SD = 1.52), followed by

PAGE 85

85 restaurant information ( M = 4.41, SD = 1.66). For obtaining computer equipment information ( M = 3.07; SD = 1.39) and health care provider informat ion, Twitter was seen as less useful ( M = 3.00, SD = 1.24). Detailed descriptive statistics can be found in Table 4 5 and Table 4 6. Principal Component Analysis of Product Category and Perceived Fit Traditionally, the product category could be separated i nto several criteria within marketing and communication research, such as tangible vs. intangible product differentiation (e.g., Arndt, 1967; Sheth, 1971 ; Zeithaml et al. 1993) and utilitaria n vs. hedonic product (e.g., Babin & Darden, 1995; Ba bin et al. 1994; Cha, 2009 a ). Drawing upon this theoretical differentiation, a principal component analysis with Direct Oblimin ld differ by product category. Principal component analysis was emp loyed, since this method is appropriate for capturing as much information as possible using only a few components (Park, Dailey, & Lemus, 2002 ). Direct Oblimin rotation was selected, as there was significant correlation among many of the items for each pro duct category and perceived fit constructs (Fabrigar, Wenger, MacCallum, & Strahan, 1999) (Table 4 7 and Table 4 8). However, the results of the principal component analysis did not correspond with previous pro duct category differentiation. As Table 4 9 a nd Table 4 10 summarizes, the principal component analysis yielded two factors in terms of product category (factor 1: travel service, movie, electronics, telecommunication service, restaurant, and clothing; factor 2: computer, automobile, financial servic e/product, and health care providers). Regarding perceived fit, two factors were obtained; however, the constructs were different (factor 1: financial service/product, health care providers, automobile,

PAGE 86

86 restaurant, clothing, and computer; factor 2: movie, electronics, travel, and telecommunication service). This means that, in terms of item composition, questionnaire respondents did not differentiate between tangible vs. intangible or utilitarian vs. hedonic product dimensions. In particular, factor loadin g of health care providers in terms of product category (.49 v s. .52) was not severe. The financial service/product and health care provider items were loaded together with automobile and computer in both product category and perceived fit dimensions. Thes e results might have been caused by the wording of "financial service/product," which may have industrial reports (Nielson, 2011; Webster, 2010, 2011 ). Therefore, for the m ain test, the word product was eliminated. Following the same logic, the wording of "health care providers" was also changed to "health care service" and used in the main test, which employed real consumers as participants Main Test The objectives of this research are to identify consumer motivation for the adoption of Twitter and to investigate the possible direction for Twitter as an effective marketing tool. To achieve these goals, a survey method was adopted for data collection, as survey s were appropr iate for investigating research questions in the context of realistic settings and allows for generalizability of the research findings (Wimmer & Dominick, 2006). Traditionally, survey research was conducted via mail, telephone, and the Internet. Among th ese three survey methods, the telephone survey was excluded, as this study employs many questions. The telephone survey is considered appropriate only for short questionnaires ( Babbie, 2001 ; Wimmer & Dominick, 2006 ). The mail survey was

PAGE 87

87 excluded as a data collecting method, as it generally has a low response rate (Wimmer & Dominick, 2006). Therefore, this research will be conducted via the Internet for two reasons: (1) efficiency in data collecting procedures and (2) online connection requirement for Twitte r usage. The online survey is more effective than mail or telephone surveys because it requires less time to develop, implement, and process (Wimmer & Dominick, 2006); therefore, the cost of conducting this research is relatively inexpensive. Also, the rec ruiting process is simple and convenient, because participants receive questionnaires through email or a Web site ( Spizziri, 2000; Wimmer & Dom inick, 2006 ). In addition, it is logical to use an online survey to investigate Twitter related issues, as an Int ernet connection is necessary for both online surveys and Twitter usage. Therefore, this study was conducted by online survey using an email invitation that includes the link to a Web based survey screen. Note that t he order of sets of questions were rand omly mixed into the set of question through the Qualtrics survey program to minimize potential bias. Instrument Development Measures Table 4 11 and Table 4 12 summarize the constructs included and their operational definitions. All the measurements used in this study were conceptualized based on established theoretical framework s both from new media adoption and eWOM literature Particularly, this study used multiple items for each of the constructs except qualifying question to minimize measurement erro r. Also, the wording of this question was tested through the pretest

PAGE 88

88 Prior Experience of O btaining Information from Twitter T o recruit actual and active user s that have obtained brand information via Twitter, one qualifying question was used to assess whe ther the respondents have obtained individual Frequency of Obtaining Brand Information from Twitter To measure the frequency of brand information obtaining behavior via Twitter the respondent s were asked how often they get brand information using a seven point The scale was used previously in social media related stud ies (Cha, 2009 a 2010). Actual Twitter Usage Twitter usage status was measured by three different constructs. Once the qualifying question isolated the participants who have experience obtaining brand information in Twitter, t hey were asked to answer how frequently they use Twitter each weak (1 = one day to 7 = seven days). A dditional question s were also used for measuring participants consumption of Twitter in hours per day. Previous literature measured new media usage such as the Internet (LaRose et al., 2006 ) Facebook (Ellison et al., 2007; Tong et al., 2008) and SNSs (Cha, 2010) have used two different method s in message usage m easuring One is asking directly how many hours consumer usuall y used the media wit h open ended question (LaRose, Lai, Lang, & Wu 2005 ; Tong et al., 2008). Another method is using a seven point Likert scale (Cha, 2010; Ellison et al., 2007). This study used the measure of 3

PAGE 89

89 s since it is particularly designed to measure an the extend to which SNSs was integrated into his/he r daily activities (Ellison et al., 2007). Also this type of measure has been adopted by several social media usage literatures both from industry reports (Webster, 2010; 2011; Radwanick, 2011; Nielson, 2011) and academic papers ( Ellison et al, 2007; Cha, 20 10 ). Finally, to assess the usage pattern in the context of brand information a statement was created using a seven point Likert scale. Conformity to Subjective Norm Conformity to subjective norm is defined as the level of overall tendency to the should not perform the behavior in question. ( Lee, 2003; Lee et al., 2003; Mathieson, 1991 ). The variab le of conformity to subjective norm was measured here with three seven point Likert scale s (1 = strongly disagree 7 = strongly agree) adopted from Lee (2003) and Lee et al. (2003) with modification for the Twitter contex t. The wordings of questionnaire were shoul = .95). TAM Related Measures TAM related measures were consisted of two dimensions: perceived usefulness and perceived ease of use. This construct has been tested in various context s from to traditional TAM studies (e.g., Davis, 1989; Davis et al., 1989) and internet based

PAGE 90

90 technology adoption such as internet based course mana gement system (Park et al., 2007 ), distance learning program (Lee, 2003; Lee et al., 2003), instant messaging (Lin et al., 2009) to VoIP adoption s (Park, 2010). The scales validatin g and reliability have been reconfirmed in a variety of different languages such as German (Baaren et al., 2011) or geographic region including South Korea (Kwon & Chon, 2009). This study adopted the original version of TAM (Davis, 1989) rather than Venka Technology (UTAUT) for two reasons: higher reliability of scales and more precise operational definition of constructs followed the suggestion of Putzke, Schoder, & Fishbach (20 10). Perceived usefulness dimensions in the which a person believes that using a particular system would enhance hi s or her job items were adopted from previous studies to measure perceived usefulness of Twitter (Davis, 1989; Cha, 2009 a, 2009b ). Participants were asked to express their feeling of agreement for each statement in seven point Likert scale s (1 = strongly disagree to 7 = strongly agree) with revision to Twitter environment. The statements were obtain ther, these six it .94.

PAGE 91

91 Perceived ease of use P believes that using a particular system would be free of effort (Davis, 1989, p. 320) ons toward Twitter in terms of ease of use were assessed via a six seven point Likert scale s ranging from 1 (strongly disagree) to 7 (strongly agree) with changing some wording. The wording of questions were Retweet on Twitter is (C .94). Attitude T oward Twitter were measured by Chen and Wells (1999), items, seven point Likert sclae (1 = strongly agree, 7 = strongly disagree). The statements wer e attitude measures of Twitter were highly reliable (Cronbac .90). Perceived Source Similarity Perceived source similarity refers to the degree of similarity with others that each individual has i n terms of specific shared characteristics (Rogers, 1995 ), or homophily of certain attributes (Brown & Reingen, 1 987). To measure the general perception of item, seven

PAGE 92

92 point Likert scale s ranging from 1 = strongly disagree, to 7 = strongly agree borrowed from previous literature ( Gilly e t al., 1998; Prendergast et al., 2010; Wangenheim & Byon, 2004, 2007) was used with slight modification s The statements were dislikes, my Twitter friends = .93). Perceived Source Credibility Source credibility was consist ed of two dimensions : trustworthiness and expertise (Kiousis, 2001) They are defined as trustworthiness of source (Petty & Cacioppo, 1986) rel To measure perceived source credibility, six item, seven point Likert scale s anchored from previous relevant studies (Prendergast et al., 2000; Wangenh eim & Bayon, 2004) were used. Specifically the following questions were included in the information given by my Twitter friends information given by my Twitter friends is powerful In addition, four items were created to measure the knowledge and exp ertise specifically for each product categories (hedonic: restaurant service, utilitarian: computer equipment). The statements of items were

PAGE 93

93 and four created measurements these eight items formed a highly reliable scale, Product Category To verify the appropriate product categories for spreading brand information from the marketer perspective, a total of 10 product categories were tested (computer equipment, clothes, finance services, restaurants, telec ommunication services, movies, healthcare services, electronics) based on relevant industry report s from industry (Nielson, 2011; Webster, 2010, 2011). Specifically we assessed self reported frequency for obtaining brand related information i n each product category. The statement used was product category using a seven point Likert scale with 1 being strongly disagree and 7 being T wer e also tested through the Perceived Fit Perceived fit was measured through various scales used within marketing literature particular ly in brand extension evaluation (e.g., Aaker & Keller, 1990; Martinez & Pina, 2010; Voss Spangengerg, & Grohmann, 2003 ). Most of traditional perceived fit measurements were bipolar scales (e.g., Voss et al., 2003). For example, Voss and point bipolar items with being 1 being not at all logical to 7 being logical T he original work from Aaker and Keller (1990) also employed bipolar scale anchored

PAGE 94

94 R ecent stud ies in examining SNSs for marketing purpose s on the other hand used a seven point Likert scale that directly assessed participants perception of fit between SNSs and shopping behavior of virtual vs. real item (Cha, 2009 a ). Using this approach the current study also directly assessed d fit between Twitter and the usefulness of Twitter for obtaining brand information. S pecific ally the participants were asked to indicate their agreement of the perceived fit between Twitter and the individual items that they obtained brand information for point Likert scale anchored by strongly disagree (1) to strongly agree (7 Attitude Toward the Brand General attitude toward the obtained brand information in Twitter were accessed by four item, seven a ttitude toward the brand (Kim & Chan Olmsted, 2005) and brand attitude in the online environment (Lee, Kim, & Chan Olmsted, 2010). The four item brand attitude eWOM Spreading Intention This study used eWOM spreading intention (forwarding brand related information) as a one of th e most important dependent variable s Three items constituted the measure eWOM spreading intention Twitter, I want to Retweet it to my friends after reading the tweeted brand information from Twitter frie information, I will encourage him or her to Tweet after reading the tweeted brand

PAGE 95

95 Tweet or Retweet the interesting product related information after reading the tweeted brand information from Twitter friends Three items were adopted by previous lit eratures (Okazaki, 2008, 2009 ). Participants indicated their level of agreement with three statements in sev en Purchase Intention In addition to Twitter adoption and the eWOM related variables, intention to purchase for the product that the respondents obtained brand information for assessed using four item s : Twitter, I will purchase the product next ti called me last night to get the advice in his/her search for a product I would recommend him/her to buy the produ ct The measure was based on (2010) item with a seven Consumer Characteristic Factors This study suggests three consumer related var iables such as age, educational level and gender ( Chan Olmsted & Chang, 2006; Chan Olmsted & Jung, 2005 ; Chang & Chan Olmsted, 2010). In the current study, age is assumed to be negatively associated with Twitter usage (e.g., Chan Olmsted & Juang, 2005) sin ce older people are reluctant to adopt new technology and more likely to exhibit negative perceptions

PAGE 96

96 toward new technology (Gilly & Zeithaml, 1985; Pommer, et al, 1980; Cha, 2009b). For other re also measured. Data Collection, and Procedure To investigate the consumer motivation for using Twitter and its effectiveness as a marketing tool, this study adopted an online survey utilizing a national consumer panel. Consumer panel was selected becaus e the current study focuses on the perception of Twitter adoption and its implication for marketing strategy of real consumer rather than perceptions among students (Washburn, Till, & Priluck, 2004). Recognizing that the advantages (such as ease in recruit ing participants and efficiency) and the disadvantages (such as the sampling issue) of conducting research online might be overcome via the popularity of using the Internet, many studies on both marketing context (e.g., Bachmann Elfrink, & Vazzana 1996; I lieva, Baron, & Healey, 2002; Taylor, 2000) and communication research (Andrews, Nonnecke, & Preece, 2003; Yun & Trumbo, 2000) have employed the method of online survey. Also, as access to the Internet is a prerequisite for using microblogging, it is appr opriate to conduct online survey through the email. A total of 458 consumers randomly selected from a national consumer panel maintained a leading research firm in the United States received the invitation e mail for the survey. The e mail contained the s urvey link so subjects could participate in the survey by clicking on the link during the period from June 6 to June 10, 2011. When participants clicked on the survey link, an informed consent form was displayed to obtain their agreement to participate in the research. To ensure that this study employed only the consumers that have experienced obtaining brand information form Twitter, a qualifying question that assessed whether the contacted panelists have used

PAGE 97

97 Twitter for obtaining brand information was us ed. Specifically, subjects were asked to y. Among the 453 panel members, 136 subjects have not use Twitter for brand information and were excluded from this study. Nine surveys were also excluded since participants did not completed survey. As a result, a total of 308 questionnaires were used for statistical analysis (69.5%). Participants As stated previously, a total of 308 responses were used for statistical analysis. Before testing the hypotheses and research questions, the user profile of the participants in this study was briefly discussed an d compared with recent industry surveys in terms of SNSs and Twitter usage pattern (Cha, 2009b). The majority of the samples were females (75.6%, n = 233) whereas 24.0% of participants were males (n = 74); 28.9% of the participants were in ages between 18 to 24, 45.1% were 25 to 34, 10.1% were 35 to 44, 8.9% were 45 to 55, and 4.2% were 55 or above. Seventy seven participants answered they were Caucasians (n = 239), 8.8% were Asian (n = 27), 6.8% were Hispanic (n = 21), 4.9% were African American and 1.6% (n = 5) were others. Out of 308 respondents, 16.5% (n = 51) were high school graduate or less, 36.1% (n = 111) and 35.1(n = 108) hold 1 and 4 years college diploma respectively, and 12.0% (n = 37) had completed graduate or professi onal degree. In terms of annual household income, 17.2% (n = 53) of subjects answered that they earn less than $25,000, 33.2% were between $25,000 to $50,000 (n = 102), 23.7% were ranged from $50,000 to $75,000 (n = 73), 13.3% (n = 41) were $75,000 to $100 ,000 and 11.4% (n = 35) said they earn more than $100,000.

PAGE 98

98 As for Twitter usage, the respondents stated that they use Twitter on average 5.3 days per we e k ( SD = 1.90). When it comes to usage in hours / minute, 9.4% respondents used Twitter for less than 1 0 minutes per day (n = 29), 27.3% used between ten 10 from 30 minutes (n = 84), 32.1% used 30 minutes to 1 hour (n = 99), 15.9% used 1 hour to 2 hours (n = 49), 7.1% used 2 hours to 3 hours (n = 22), 3.2% used 3 hours to 4 hours (n = 10) and 4.9% used mor e than 4 hours per day (n = 15). This study also compared the demographic characteristics of the respondents with the user profile from recent industry reports. Table 4 13 illustrates the user profile of this study against the information gathered from in dustry reports published by leading marketing firms (Webster, 2010; 2011). The biggest difference between the current information. The sample for this study seemed to obtain more brand information comparing with the results from the industry reports that measured the percentage of users obtaining brand information in Twitter and among general social media user (69.5% vs. 51% vs. 16%). Next, the sample of current study has a h igher percentage of female participants than industry report (75.6% vs. 54%). Also, participants who were between the ages of 25 34 accounted for 45.1% of the sample in this study whereas the age group represented only 28% of the subjects of the industry r eport. In terms of ethnicity, this study employed more Caucasian (77%) than market research data (55%). Using the criteria for U.S. Census occupation categories, the largest percentage occupation of participants were management (16.4%) and followed by educ ation, training or library (12.7%). All other demographic characteristics were fairly similar between this study and the industry report.

PAGE 99

99 Statistical Analysis Overview This study employed several statistical analyses, including exploratory factor analysis, test. In addition, to test the proposed hypotheses and research questions, a correlations analysis, structural equation modeling (SEM), analysis of variance (ANOVA), and regression analysis were conducted for testing consumer related variables (gender, age, and educational level). Figure 4 1 shows the specific flow of statistical analyses for our study. Three separate statistical analyses were conducted. First, all measu rements of this study were put in together, and a reliability test and correlations analysis were conducted to verify the relationship between variables. Variables related to Twitter adoption and eWOM spreading intention variables were then analyzed separa tely. For the Twitter adoption variables, perceived usefulness, perceived ease of use, and conformity to subjective norm were analyzed with three main dependent variables: attitude toward Twitter, Twitter usage, and frequency of obtaining brand information Next, to confirm the extent to which measurements reflect the appropriate meaning for each theoretical concept, exploratory and confirmatory factor analyses were conducted. For the main hypotheses and research questions testing, SEM was adopted, and, fin ally, analysis of variance was conducted to test the categorical variable (gender). In terms of eWOM related research questions and hypotheses, the main process was similar to those for Twitter adoption research; however, one factor analysis (both explorat ory and confirmatory) process was added to verify the product category and perceived fit measures.

PAGE 100

100 Following SEM, ANOVA and regression analysis were employed for testing consumer demographic variables (gender, age, educational level). Validity and Reliabi lity Test To verify the extent to which the measurements reflect the real meaning of the abstract concepts of definition and to confirm that the same object generates the same result each time, validity and reliability tests were conducted prior to the mai n statistical analysis for test hypotheses and research questions. For validity testing, an exploratory factor analysis and confirmatory factor analysis were employed. To test reliability, stical programs. Validity test Validity refers to the extent to which the measurement represents the exact meaning of the concept. Among several validity tests, including predictive validity, content validity, and convergence validity, this study particul arly focused on the associated with one another without significant associations with other items measuring a different concept" (Chang, 2005, p. 83). Although each co nstruct for both for Twitter adoption and eWOM related variables had a solid theoretical background (e.g., Arndt, 1967; Davis, 1993; Lee, 2003; Lee et al., 2003; Dellarocas, 2003), measurements were based on the different contexts of adoption studies and e WOM studies; therefore, principal component analysis, exploratory factor analysis, and confirmatory factor analysis were performed to check the validity of measurements. Principal component analysis. The purpose of principal component analysis is variable reduction, for data reduction and fewer numbers (Snook & Gorsuch, 1989). In

PAGE 101

101 detail, all questionnaire items for, in particular, product category related variables were put into the factor analysis using principal component analysis with Direct Oblimin rota tion. This was done because the dimensions were not based on solid theoretical measurements, but, rather, on the industrial report (Webster, 2010, 2011). Direct Oblimin rotation was selected, as it is recommended that Oblimin rotation be used when the fact ors correlate to each other (Fabrigar et al., 1999; Park et al., 2002). Exploratory factor analysis. We employed exploratory factor analysis (EFA), using maximum likelihood estimation with Oblique rotation. Oblique rotation was selected because variables w ere correlated to each other (Hong, Milik, & Lee, 2003). Although the majority of measures used in this study has been verified and were based on solid theoretical background, EFA was conducted to determine the appropriate scale dimensions. The model fit was tested via the statistic program SPSS 19. The sample size is important when conducting exploratory factor analysis: less than 50 is very poor; 100 is poor; 200 is fair; and more than 300 is good (Comery & Lee, 1992; J. Kim, M. Kim, and Hong, 2009; Tab achnick & Fidell, 1996;). In the current study, a total of 308 participants were used, which is considered a sufficient sample size. To determine the appropriate factor numbers, three criteria were used: (1) less than .08 of the root mean square error of approximation (RMSEA), (2) the changes in RMSEA values were less than .01, and (3) it was recommended to use a fewer factor model if the changes in RMSEA were marginal. In interpreting the rotated factors, following the logic of Jun, Cho, and Kwon (2008), this study excluded item loads higher than 0.4 with other factors and concentrated on the items that loaded at least 0.5 on only one factor.

PAGE 102

102 Confirmatory factor analysis After conducting EFA to confirm the validity of measurements, a confirmatory fact or analysis (CFA) using maximum likelihood was conducted. The interpretation of the confirmatory factor analysis was the same as for the exploratory factor analysis. Reliability test Reliability refers to the item to s for the represents the extent of correlations among scales) was used, The general criteria for interpretation of reliability are as follows: a reliability coefficient around 90 is excellent; around .80 is very good; and values around .70 are adequate (Hair et al., 1998; Kline, 1998). reliability coefficients of at least .70 were used in the statistica l analysis (Cha, 2009a, 2009b; Chang, 2005; Hair et al., 1998; Kline, 2005). The constructs for the reliability test were conformity to subjective norm, perceived usefulness, perceived ease of use, attitude toward Twitter, actual Twitter usage, perceived s imilarity, perceived credibility, product category (utilitarian vs. hedonic), perceived fit (utilitarian vs. hedonic), attitude toward the brand, eWOM intention, and purchase intention. Structural Equation Modeling Advantages of structural equation modeli ng (SEM) To test the hypotheses and research questions, this study adopted structural equation modeling (SEM) for data analysis for three primary reasons. First, even if multiple regression could directly answer the causal relationship between the independ ent variables and dependent variables, SEM can consider the dependence

PAGE 103

103 relationship among independent variables (Chang, 2005). Second, SEM can control the measurement errors of multiple items. Three, SEM contains not only observed variables but also unobse rved (latent) variables, allowing the researcher to conduct a more accurate and sophisticated statistical analysis. Therefore, to test the proposed hypotheses, a structural equation modeling (SEM) method was used. It has been suggested that SEM can control measurement error with greater sophistication, resulting in more accurate results (Browne & Cudeck, 1993; Bentler, 1990; Tucker & Lewis, 1973; West, Finch, & Curran, 1995). Testing model fit Two representative methods to evaluate model fit: 2 using fit indices, were used to test the hypotheses in this study, since various model fit indices, such as the root mean square error of approximation (RMSEA), Tucker Lewis index ( TLI), and Comparative Fit Index (CFI), can overcome the weakness of 2 t hat makes it sensitive to the sample size and 2 rejects the null hypothesis conservatively. Therefore, this study used RMESA, TLI, and CFI, as well as 2 (2009) recommendation. These methods are highly reliable in terms of model fit indices criteria, and the model is relatively simple (Browne & Cudeck, 1993). Previous studies have reported that RMESA indices of less than .05 indicate a strong fit, whereas those ranging from .05 to .10 represent a moderately model fit, and TLC and CFI of more than .90 represent a good model fit (Bentler, 1990; Browne & Cudeck, 1993; Tucker & Lewis, 1973; West et al., 1995). Table 4 14 shows the representative methods for model fit testing.

PAGE 104

104 Additional Analysis In a ddition to SEM, analysis of variance (ANOVA) and regression analysis were applied in terms of gender difference, since gender was measured via categorical variable (Chan Olmsted & Chang, 2006; Chan Olmsted et al., 2005; Chang & Chan Olmsted, 2010; Chang et al., 2006). A one way ANOVA was conducted for gender difference, and multiple regression for age and educational level was employed.

PAGE 105

105 Table 4 1. Descriptive summary of pretest questionnaire Variable No. of Item No. of case Minimum Maximum Mean SD Confo rmity to Norm 3 29 1.00 6 3.45 1.33 Perceived Usefulness 6 29 1.00 7 3.55 1.51 Perceived Ease of Use 6 29 1.00 7 4.31 1.63 Attitude Toward Twitter 5 29 1.00 6.40 4.17 1.55 Perceived Similarity 6 29 1.00 5.67 3.69 1.40 Perceived Credibility 8 29 1.00 6.25 3.85 1.12 Product Category 10 29 1.00 5.60 3.61 1.17 Perceived Fit 10 29 1.00 6.50 3.77 1.18 Table 4 2 R eliability test of pretest questionnaire Variable No. of item Conformity to Norm 3 .79 Perceived Usefulness 6 .94 Perceiv ed Ease of Use 6 .96 Attitude Toward Twitter 5 .91 Perceived Similarity 6 .99 Perceived Credibility 8 .92 Product Category 10 .89 Perceived Fit 10 .93 Retweet Intention 3 .75

PAGE 106

106 Table 4 3. Correlation matrix of pretest items Mean SD 1 2 3 4 5 6 7 8 1 3.45 1.33 .18 .31 .44* .21 .40* .28 .24 2 3.55 1.51 .18 .75** .78** .71** .72** .65** .55** 3 4.32 1.63 .30 .75** .88** .68** .71** .74** .53** 4 4.17 1.55 .44* .78** .88** .72** .79** .81** .69** 5 3.69 1.40 .21 .71** .68* .72** .74** .52** .56** 6 3.85 1.12 .40* .72** .71** .79** .74** .48** .44* 7 3.90 2.94 .28 .65* .74** .81** .52** .48** .75** 8 2.59 1.48 .24 .55** .53** .69** .56** .44* .76** p < .05; ** p < .01 1 = Conformity to subjective norm, 2 = perce ived usefulness, 3 = perceived ease of use, 4 = Attitude toward Twitter, 5 = perceived similarity, 6 = perceived source expertise, 7 = Twitter usage (days per weak), 8 = Twitter usage (hours per day )

PAGE 107

107 Table 4 4. Regression analysis of pretest Dependent V ariable Independent Variable SE t p value Twitter Attitude Conformity to Norm .20 .09 2.39 .025 Perceived Usefulness .30 .13 2.45 .022 Perceived Ease of Use .59 .12 4.61 .000 F (3, 25) 42.93*** R 2 .84 Adjusted R 2 .82 Twitter U sage (days per week) Conformity to Norm .07 .31 .52 .606 Perceived Usefulness .22 .39 1.11 .279 Perceived Ease of Use .55 .38 2.63 .014 F (3, 25) 10.90*** .07 .31 .52 .606 R 2 .57 Adjusted R 2 .52 Twitter usage (hours per day) Conformi ty to Norm .11 .19 .63 .531 Perceived usefulness .35 .24 1.43 .165 Perceived ease of use .23 .23 .91 .371 F (3, 25) 4.38*** R 2 .35 Adjusted R 2 .27 *p< .05, **p< .01, ***p< .001

PAGE 108

108 Table 4 5. Descriptive statistics of product category Pro duct Category Minimum Maximum Mean SD Movie 1 7 4.56 1.80 Clothes 1 7 4.45 1.97 Restaurant 1 7 4.41 1.80 Travel information 1 7 4.41 1.72 Electronics 1 6 3.72 1.53 Telecommunication Service 1 7 3.28 1.39 Financial product/ service 1 6 3.07 1.44 Aut omobile 1 7 3.00 1.75 Computer equipment 1 7 2.76 1.62 Healthcare providers 1 5 2.48 1.24 Table 4 6. Descriptive statistics of perceived fit Product Category Minimum Maximum Mean SD Movie 1 7 4.79 1.52 Clothes 1 7 4.41 1.66 Restaurant 1 7 4.24 1.60 Travel information 1 7 4.17 1.39 Electronics 1 7 3.83 1.51 Telecommunication Service 1 7 3.55 1.43 Financial product/ service 1 7 3.48 1.60 Automobile 1 7 3.17 1.47 Computer equipment 1 6 3.07 1.39 Healthcare providers 1 7 3.00 1.24

PAGE 109

109 Table 4 7. Correlation matrix (Product category) for pretest ` Mean SD 1 2 3 4 5 6 7 8 9 10 1 2.76 1.62 .24 .70* .29 .42* .16 .69** .10 .34 .36 2 4.45 1.97 .24 .34 .54** .72** .49* .13 .55** .42* .44* 3 3.07 1.44 .70** .34 .39* .51** .30 .57** .14 .28 .50** 4 4.41 1.72 .29 .54** .39* .76** .67** .52** .76** .40* .74** 5 4.41 1.80 .42* .72** .51** .76** .51** .57** .49** .51** .66** 6 3.28 1.39 .16 .40* .30 .67** .51** .13 .69** .29 .81** 7 3.00 1.75 .69** .13 .57** .52** .57** 13 .24 .48** .50** 8 4.55 1.80 .10 .55** .14 .76** .49** .69** .24 .36 .70** 9 2.48 1.24 .34 .42* .28 .40* .51** .29 .48** .36 .35 10 3.72 1.53 .36 .44* .50** .74** .66** .81** .51** .70** .35 p < .05; ** p < .01 1 = Computer equipment, 2 = clo thes, 3 = finance product/services, 4 = travel information, 5 = restaurants, 6 = telecommunication service, 7 = automobiles, 8 = movies, 9 = health care providers, 10 = electronics Table 4 8. Correlation matrix (Perceived fit) for pretest Mean SD 1 2 3 4 5 6 7 8 9 10 1 3.07 1.39 .55** .43* .29 .48* .54** .44* .28 .34 .33 2 3.83 1.51 .55** .63** .56** .73** .60* .54** .53** .54* .49** 3 3.17 1.47 .43* .63** .58** .63** .68** .77** .42* .90** .42* 4 4.17 1.39 .29 .56** .58* .68** .78** .57** .83** .52* .85** 5 4.41 1.66 .48** .73** .63** .68** .59** .75** .66** .54** .61** 6 3.55 1.43 .54** .60** .69** .78** .59** .63** .63** .66** .75** 7 3.48 1.60 .44* .54** .77** .57** .75** .63** .43 .81** .58** 8 4.79 1.52 .2 8 .53** .42* .83** .66** .63** .43* .29 .83** 9 3.00 1.44 .34 .54** .90** .52* .54** .66** .81** .29 .39* 10 4.24 1.60 .33 .49** .42* .85** .60** .76** .59** .83** .39** p < .05; ** p < .01 1 = Computer equipment, 2 = clothes, 3 = finance product/ services, 4 = travel information, 5 = restaurants, 6 = telecommunication service, 7 = automobiles, 8 = movies, 9 = health care providers, 10 = electronics

PAGE 110

110 Table 4 9. Principal component analysis with product category Variable Factor1 Factor2 Travel ser vice .89 .47 Movie .88 .15 Electronics .86 .50 Telecommunication Service .84 .19 Restaurant .78 .64 Clothes .71 .33 Computer equipment .23 .89 Automobile .37 .88 Finance service / product .37 .82 Healthcare providers .49 .52 Table 4 10. Princip al component analysis with perceived fit Variable Factor1 Factor2 Financial service/product .93 .47 Healthcare providers .91 .38 Automobile .88 .54 Restaurant .76 .73 Clothes .75 .60 Computer equipment .62 .36 Movie .44 .94 Electronics .52 .94 Travel service .61 .93 Telecommunication service .77 .77

PAGE 111

111 Table 4 11. Operational definition of main constructs and measurement Variable Conceptual D efinition Measurement Conformity to Subjective Norm The level of overall tendency to the perso people who are important to him think one should or should not perform the behavior in question. 3 items seven point Likert scale (Lee 2003; Lee et al., 2003; Mathieson, 1991). Perceived Usefulness The level of a person believes that using a particular system would enhance his or her job performance. 6 items seven point Likert scale (Davis, 1989; Lee, 2003; Cha, 2009 b ) Perceived Ease of Use The level of an individual believes that using a particular system would be free of real a nd mental effort. 6 items seven point Likert scale (Davis, 1989; Lee, 2003; Cha, 2009b) Attitude toward Twitter The level of individual attitude toward Twitter. 5 items seven point Likert scale (Chen & Wells, 1999; Prendergast et al., 2010). Actual usage time of Twitter The level of individual usage pattern of the Twitter 2 items seven point Likert scale Perceived Source Similarity The degree to which individuals are similar in terms of certain shared characteristics specifically for eWOM se nder (Twitter friends). 6 items seven point Likert scale (Wolfinbarger & Gilly, 1993; Prendergast et al., 2010; Wangenheim & Byon, 2004; 2007). Perceived Source Credibility The level of perceived source trustworthiness and expertise of eWOM sender (Twitte r friends). 6 items seven point Likert scale (Kiousis, 2001; Gurhan Canli & Maheswaren, 2000; Prendergast et al., 2010; Wangenheim & Bayon, 2004). Product Category The product category of consumer mainly obtained information from Twitter 10 items, seven point Likert scale (Created for this study) Perceived Fit Twitter and the individual items that they obtained brand information 10 items, seven point Likert scale (Created for this study). Attitude Toward the Brand The level of individual perception of favorable attitude toward the brand 4 items seven point bipolar scale (Kim & Chan Olmsted, 2005; Lee et al., 2010).

PAGE 112

112 Table 4 11. Continued Variable Conceptual Definition Measurement eWOM Spreading Intention The lev el of individual intention for eWOM spreading 3 items seven po int Likert scale (Okazaki, 2008, 2009). Purchase Intention Individual intention for purchase of product that obtained brand information 4 items seven point Likert scale (Prendergast et al., 2 010) Age Age of a respondent 1 item Educational Level Education level of respondent. 1 item Gender Gender of participants. 1 item Income income. 1 item

PAGE 113

113 Table 4 12 Summarizes the original constructs included and the ir operational definition Construct Item Qualifying Question Have you ever get brand information from your Twitter friends? (Both from individual and company account ) Frequency of obtaining brand information How often do you get information about a brand from your Twitter friends? Twitter Usage How many days during a typical week do you use Twitter? How many hours during a typical day do you use Twitter? Conformity to subjective Norm Generally speaking, I would do what my group members think I should do Generally speaking, I would do what my Twitter friends think I should do in the Twitter environment Generally speaking, I would do what others think I should do in the online environment Perceived Usefulness I find Twitter useful in my life Use of Tw itter improves my performance. Use of Twitter makes it easier to obtain product information Use of Twitter to obtain product information increases my productivity Use of Twitter enables me to accomplish tasks more quickly Use of Twitter enhances the effe ctiveness in product information search Perceived Ease of Use Tweet, Mention and Retweet on Twitter is easy Learning to use Twitter is easy for me It is easy to get information on Twitter. I find Twitter to be flexible to interact with. It is easy for me to become skillful at using Twitter. I find Twitter easy to use. Attitude Toward Twitter Twitter makes it easy for me to build a relationship with the online community I would like to communicate with my Twitter friends again in the future ied with the services provided by Twitter I feel comfortable in using Twitter I feel surfing on Twitter is a good way for me to spend my time. Perceived Source Similarity In terms of outlook on life, my Twitter friends are similar to me. In terms of lik es and dislikes, my Twitter friends are similar to me. In terms of values and experiences, my Twitter friends are similar to me. In terms of tastes for products, my Twitter friends are similar to me. In terms of preferences and value, my Twitter friends ar e similar to me. Overall, my Twitter friends are similar to me

PAGE 114

114 Table 4 12. Continued Construct Item Perceived Source Credibility I feel the tweeted product information given by my Twitter friends is strong I feel the tweeted brand information given by my Twitter friends is convincing I feel the tweet brand information given by my Twitter friends is persuasive I feel the tweet brand information given by my Twitter friends is powerful. My Twitter friends have knowledge about computer equipment in genera l My Twitter friend is an expert in the area of computer equipment. My Twitter friends have knowledge about restaurants in general. My Twitter friend is an expert in the area of restaurants. Product Category I often try to obtain product information abou t computer equipment I often try to obtain product information about clothes. I often try to obtain product information about finance services. I often try to obtain product information about travel information. I often try to obtain product information a bout restaurants. I often try to obtain product information about telecommunication services. I often try to obtain product information about automobiles. I often try to obtain product information about movies. I often try to obtain product information abo ut healthcare services. I often try to obtain product information about electronics. Perceived Fit Twitter is a good medium to learn about computer equipment. Twitter is a good medium to learn about clothes. Twitter is a good medium to learn about finance services Twitter is a good medium to learn about travel information. Twitter is a good medium to learn about restaurants. Twitter is a good medium to learn about telecommunication services. Twitter is a good medium to learn about automobiles. Twitter is a good medium to learn about movies. Twitter is a good medium to learn about healthcare services. Twitter is a good medium to learn about electronics. Attitude Toward the Brand Unfavorable/favorable Bad/good Dislike/like Negative/positive

PAGE 115

115 Table 4 12 Continued Construct Item eWOM spreading intention: After reading the tweet brand information from If I find interesting product information on the Twitter, I want to Retweet it to my friends after reading the tweeted brand information fr om Twitter friends. If somebody asks me for advice about interesting product information, I will encourage him or her to Tweet after reading the tweeted brand information from Twitter friends. I would recommend my friends and family to Tweet or Retweet in interesting product related information after reading the tweeted brand information from Twitter friends. Purchase Intention After considering the product information on my Twitter, it is very likely that I will buy the product After considering the prod uct information on my Twitter, I will purchase the product next time I need a product After considering the product information on my Twitter, I will definitely try the product. If my friend called me last night to get the advice in his/her search for a product. I would recommend him/her to buy the product

PAGE 116

116 Table 4 13. The comparison of the sample profile with industrial report Variable Category Current Study Year of 2011 Industry Report (2011 Twitter Users) a Industry Report (General Social Media Use rs) b Obtaining brand information in Twitter Yes No 69.5% (n = 317) 30.5% (n = 136) 51% 1) 49% 1 16% 1) 84% 1) African American Gender Male 24.1% (n = 74) 46% 1) Female 75.6% (n = 233) 54% 1) Age 12 to 17 13% 1) 15% 2) 18 to 24 28.9% (n = 89) 21% 1) 9% 2) 25 to 34 45.1% (n = 139) 28% 1) 18% 2) 35 to 44 10.1% (n = 31) 20% 1) 25% 2) 45 to 55 8.9% (n = 27) 11% 1) 19% 2) 55 and over 4.2% (n = 13) 7% 1) 13% 2) Ethnicity Caucasian 77% (n = 239) 55% 1) African Americ an 4.9% (n = 15) 22% 1) Hisp anic 6.8% (n = 21) 15% 1) Asian 8.8% (n = 27) 3% 1) Others 1.6% (n = 5) 5% 1) Education High School/less 16.5% (n = 51) 12% 1) 1 3 years college 36.1% (n = 111) 23% 1) 4 year college 35.1% (n = 108) 30% 1) Some Graduate 12. 0% (n = 37) 17% 1) Annual Income Under $25K 17.2% (n = 53) 17% $25K to $50K 33.2% (n = 102) 14% $50K to $75K 23.7% (n = 73) 23% $75K to $100K 13.3% (n = 41) 11% Over $100K 11.4% (n = 35) 13% Twitter Usage: Days per week 5.34 (SD = 1. 90) 3.17 1.47 Twitter Usage: Hours per day Less than 10 m inutes 10 30 min 9.4% (n = 29) 27.3% (n = 84) 30 min 1 hour 32.1% (n = 99) 1 2 hours 15.9% (n = 49) 2 3 hours 7.1% (n = 22) 3 4 hours 3.2% (n = 10)

PAGE 117

117 Table 4 13. Continued Variable Category Current Study Year of 2011 Industry Report (2011 Twitter Users) a Industry Report (General Social Media Users) b More than 4 hours 4.9% (n = 15) Source: The Edison Research / Arbitron Internet and Multimedia St udy, ( Webster, 2010 2011) Note: a: Twitter users are defined as people who use Twitter at least once per month (Webster, 2011) b: Daily unique visitors from 19 different social network sites (Facebook, LinkedIn Myspace, Twitter and etc.) were gather ed an d calculated 1) Source: Twitter usage in America: 2010 (Webster, 2010). 2) Source: Study: Ages of Social Network Users (Pingdom, 2010).

PAGE 118

118 Table 4 14. Comparing model fit index Model F it Indices Sensitive for the N umber of S ample Considering Simplicity of M odel Criteria for I nterpretation GFI O X above .90 AGFI O O X PGFI O O X NFI O X above .90 RFI O X above .90 IFI O X above .90 NNFI(TLI) X O above .90 CFI X X above .90 PNFI O O X PCFI X O X RMSEA X O less than .08 = fair less than .05 = good Source: Kim, Kim, &Hong (2009).

PAGE 119

119 Figure 4 1. Flow of statistical a nalysis ( FB: f requency of obtaining brand information, TU: Twitter usage, CN: c onformity to subjective norm, PU: p erceived usefulness, PE: p erceived ease of use, AT: a ttitude toward Tw itter, SS: p erceived source similarity, SC: p erceived source credibility, PC: p roduct category, PF: p erceived fit, AB: a ttitude toward the brand, eI: eWOM speeding intention, PI: p urchase intention ).

PAGE 120

120 CHAPTER 5 RESULTS This chapter presents the results of the hypothesis testing and research question. It consists of mainly two main parts. First, hypotheses and research questions addressing the adoption of Twitter are discussed. Second, the results of the hypotheses testing and research question s on eWOM rela ted variables are provided. Each part contain s factor analysis and structural equation modeling for the main variables. A dditional analyses were also conducted such as ANOVA and regression for consumer characteristic variables such as age, gender and educa tional level. Twitter Adoption Results Descriptive Statistics Before conducting the main statistical analysis, two types of descriptive statistics of mean and the standard deviation (SD) were performed. The results are displayed in T able 5 1. Among varia bles, the mean value of perceived ease of use was relatively high er (M = 5.80, SD = 1.17) than perceived usefulness (M = 4.94, SD = 1.24). Also SD = 1.47) whereas their frequen cy of obtaining brand related information was slightly low (M = 3.13, SD = 1.47). However, their daily Twitter usage per week (M = 3.88, SD = 1.37) and hourly usage per day (M = 5.34, SD = 1.91) was high. Correlation s Analysis Before the main statistical testing for the Twitter adoption variables, a correlation matrix was constructed to check the relationship between variables (Table 5 2). As expected, all measurements were significantly correlated each other. In particular, main

PAGE 121

121 dependent variables attit ude toward Twitter was significantly correlated with conformity to norm (r = .31, p < .01), perceived usefulness (r = .65, p < .01), and perceived ease of use (r = .73, p < .01). Also, participants daily Twitter usage per week was associated with conformit y to norm (r = .14, p < .05), perceived usefulness (r = .36, p < .01), and perceived ease of use (r = .43, p < .01). Also, hourly Twitter usage per day was related to conformity to norm (r = .24, p < .01), perceived usefulness (r = .40, p < .01), perceived ease of use (r = .28, p < .01). Therefore all measurements were included into structural equation modeling both for factor analysis and hypotheses and research questions testing. Factor Analysis Exploratory factor analysis E xploratory and confirmatory fa ctor analyses were conducted to verify the measurement dimensions. In terms of the factor number decision the results of exploratory factor analysis reconfirmed the five dimensions: conformity to subjective norm, perceived usefulness, perceived ease of us e, attitude toward Twitter and actual Twitter usage. To determine the factor number a few statistics were reviewed. F irst, the scree plot revealed that four factors or five factors are appropriate for the model since the graph for eigenvalue seemed to shar p descent (Figure 5 1). Also, between four satisfied the following criteria : (1) The value is less than .08 RMSEA, (2) the changes in RMSEA values were less than .01 ; and (3) it was recommended to use a fewer factor model if the changes in RMSEA were marginal (Hu & Bentler, 1999). Therefore it was concluded that the five factors model was most appropriate for the Twitter adoption study. In interpreting the rotated facto rs, foll owing the logic of Jun et al., (2008), this

PAGE 122

122 study excluded the items load higher than 0.4 with and other factors and concentrated on the items that loaded at least 0.5 on only one factor. Based on the results of the EFA, each factor was labeled in conf ormity to subjective norm, perceived usefulness, perceived ease of use, attitude toward Twitter and Twitter usage. (Table 5 3) Confirmatory factor analysis To confirm the four different dimensions of Twitter adoption related variables, a confirmatory fact or analysis (CFA) using maximum likelihood method was conducted. hierarchical behavior rather than a one dimension activity that was mainly influenced by conformity to subject ive norm, perceived usefulness and perceived ease of use. Therefore, the tested. The results indicated a good model fit with five dimensions (conformity to subjective norm, perceived u sefulness, perceived ease of use, attitude toward Twitter 2 = 731,383, df = 199, p < .001) Comparative Fit Index (CFI) was .91, Tucker Lewis index (TLI) was .89 and RMSEA was .093 whereas one factor model was not appropriate for the data ( 2 = 2804.106, df = 209, p < .001, CFI = .556, TLI = .509, RMSEA = .201). T he initial factor loadings for Twitter adoption variables ranged from .69 to .94. However, even if the results for both exploratory factor analysis and confirmatory factor analysis produced the five dimensions of measurements, the model fit was relatively method for model revision, both standardized residuals and modification indices were analyzed for model revision based on the standardized residuals and theoretical

PAGE 123

123 background in previous stud ies The results for standardized residuals and modification indices suggested that six items have high correlated errors with other variables. A total of six items were removed one by one to observe the model fit change The deleted items wer e for 2 = 192.015, df = 71, CFI = .968, TLI = .960. RMSEA = .071) and sat isfied that criteria for good model fit evaluation (CFI, TLI > .90, RMSEA < .80). Therefore, the results of the CFA presented a model with five constructs in seven different dimensions that were the same as for EFA: conformity to subjective norm, perceiv ed usefulness, perceived ease of use, attitude toward Twitter and Twitter usage. Table 5 4 shows the results of model fit of confirmatory factor analysis and T able 5 5 shows the final factor loading. The results of confirmatory factor analysis indicated t hat all the items statistically evaluated the each constructs as intended and factor loading of each item is high (.59 to .94). Validity and reliability variables were assesse d. Table 5 6 shows the correlations between the measurement items and latent variables that are highly correlated with each others switching the recommended criteria proposed by Gefen and Straub (2005).

PAGE 124

124 The r eliability of all measurements was also assess ed via two representative ranged from .88 to .93 satisf ying the criteria for good reliability (above .70) recommended by previous research (Nun n aly, 1978). Also, all the measu reliability was tested by check ing the factor loadings (> 0.5 ) ( Cha, 2009b; Rivard, 1988 ). The confirmatory factor analysis yielded factor loadings ranged from .59 to .95. Both Twitter adoption study were reliable. Structural Equation Modeling After confirming that the five different factors have the adequate fit in measurement model, a structural equation modeling was conducted for testing the hypotheses and research questions using the five variables: conformity to subjective norm, perceived usefulness, perceived ease of use, attitude toward Twitter and actual Twitter usage. A critical assumption for conducting SEM is multivariate normality which means that the scale items a nd their combined measurements should be statistically normally and kurtosis were first assessed. It was found that the skewness have ranged from 1.68 to .906, and kurto sis from .93 to 3.81; this satisfied the criteria for normality for the Twitter adoption model (ske wness < 2, kurtosis < 4). (Hong et al. 2003). ( T able 5 1) After confirming multivariate normality of the structural model, correlation s analysis with exog enous variables was performed. Traditionally the correlation value above .90 may the problem of multicollinearity (Hair et al., 1998). The current model c orrelation

PAGE 125

125 value s are ranged from .24 to .72 ( p < .01) Therefore there is no multicollinearity prob lem in the structural model. (Table 5 6) After the correlation matrix was displayed, a structural equation model was performed to test hypotheses and research questions. The overall model fit indices suggested that the model adequately fit the data ( 2 = 2 76.322, d.f. = 105, p < .001; CFI = .958, TLI = .946 RMSEA = .073), having above .90 of CFI and TLI (Bentler 1990) and lower than .08 of RMSEA (Browne & Cudeck, 1993). The results of the structural model are presented in Table 5 7, Table 5 8 and Table 5 9, with the s tandardized path coefficients. H1a and H1b predicted that the conformity toward the subjective norm is positively related to the attitude toward Twitter Hypothesis 1b were n ot supported according to the modeling results ( T able 5 8). In terms of perceived usefulness and attitude toward Twitter as well as Twitter usage, the results of the structural equation model indicated that the perceived usefulness is a significant predi ctor of the attitude toward Twitter ( = .30, p < .05). Also, perceived usefulness is positively associated with the per day) ( = .22, p < .001) and frequency of obtaining brand information ( = .38, p < .001). However, this relationship was not exhibited in daily based Twitter usage. Thus, H ypothesis 2a was supported whereas H ypothesis 2b was partially supported. In terms of hypotheses 3a and 3b, which predicted a positive relationship between perceived ease of use and attitude toward Twitter ( H3a) and the (H3b), only H ypothesis 3a was confirmed. Perceived ease of use was a significant predictor of the Twitter ( = .69, p < .001).

PAGE 126

126 The fourth set of hypotheses suggested a significant effect of the attitude toward Twitter and Twitter usage. The finding indicated that one attitude toward Twitter was positively associated with his/her daily Twitter usage ( = .61, p < .001), hourly Twitter usage per day ( = .33, p < .01) and their behavioral frequency to obtain brand related information in Twitter ( = .24, p < .05). Therefore, H ypothesis 4 was supported. As illustrated in Table 5 9, perceived ease of use had a direct effect on attitude toward Twitter, but not on Twitter usage. However attitude toward Twitter was found to affect Twitter Usage. Therefore the mediation effect of attitude toward Twitter was evaluated in this context Specifically, the model indica ted that through perceived ease of use did not direct ly predict the the construct did influence the through attitude toward Twitter as a mediator. T he path coefficient of perc daily usage per week, with attitude toward Twitter as a mediator was .42(.69*.61), hourly usage per day was .22(.69*.33), and frequency of obtaining brand information was .16(.69*.24) (Table 5 10) T his study reveal ed the role of conformity to subjective norm, perceived Conformity to subjective norm did not affect directly or indirectly attitude toward Twitter and Twitter usage ( H ypoth esis 1a and H ypothesis 1b were rejected). In terms of perceived usefulness, it affects the usage ( H ypothesis 2a and H ypothesis 2b were supported). In terms of perceived ease of use, it was positively relate d attitude toward Twitter but not his / her Twitter usage pattern. However, perceived ease of use was positively associated with Twitter

PAGE 127

127 usage while Twitter attitude is acting as a mediating variable ( H ypothesis 3a was supported, H ypothesis 3b wa Twitter is a significant predictor of their actual Twitter usage ( H ypothesis 4 was supported). In sum when people consider Twitter to be useful, they tend to exhibit more favorable attitude towar d Twitter and use Twitter more frequently. Also, individuals who felt it is easy to use Twitter tend to have more positive feeling of Twitter but this feeling did not directly translate to Twitter usage. Nevertheless the perceived ease of u se regarding Twitter does indirectly influence their Twitter usage when favorable attitude toward Twitter is present. Figure 5 2 shows the final structural model, including the significant pathways' observed values Additional A nalysis This study employed two additional statistics to analyze consumer related variables such as age, education level and gender. Twitter usage (daily usage, hourly usage, and frequency of obtain ing brand information) was performed using consumer demographics as independent variables (age, education level). However, age and education level were not found to be statistically significant in terms of attitude toward Twitter (adjusted R squared value of .01, p = .07), daily usage of Twitter (adjusted R squared value of .005, p = .17), hourly usage of Twitter (adjusted R squared value of .003, p = .59) and brand related behavior in obtaining information (adjusted R squared value of .005, p = .17). Also a one way analysis of variance (ANOVA) was performed by using gender as an independent variable and attitude toward Twitter, Twitter usage (frequency of

PAGE 128

128 obtaining brand information, daily usage, and hourly usage) for dependent variables. Although mean di fference existed in attitude toward Twitter ( M male : 5.57, SD = 1.13 vs. M female : 5.53, SD = 1.07), days per week usage ( M male : 5.61, SD = 1.85 vs. M female : 5.26, SD = 1.92), hours per day usage ( M male : 3.28, SD = 1.56 vs. M female : 3.09, SD = 1.44) and freq uency of brand information obtaining behavior ( M male : 4.08, SD = 1.38 vs. M female : 3.81, SD = 1.37), the results for gender difference showed that there were no significant differences in terms of attitude toward Twitter ( F = .06, p = .80), daily usage ( F = 1.86, p = .17), hourly usage ( F = 1.02, p = .31), and frequency of behavior in obtaining brand information ( F = 2.25, p = .13). ( Hypothesis 4, Hypothesis 5 and R esearch Q uestion 1 were not supported). eWOM R elated R esults Descriptive Statistics Before co nducting the main statistical analysis, two types of descriptive statistics of mean and standard deviation (SD) were performed and displayed in T able 5 11. Among all variables, the mean value s for perceived similarity (M = 5.19, SD = 1.10) and perceived c redibility were relatively high (M = 5.00, SD = 1.09). Also participants were more frequency obtaining brand information of hedonic product (M = 5.15, SD = 1.22) than utilitarian product (M = 4.37, SD = 1.44). The dominant role of the hedonic product is al so exhibited in consumer perception of perceived fit between Twitter and each product category the mean value for perceived fit of hedonic product is 5.25 ( SD = 1.16) and utilitarian product is 4.55 ( SD = 1.31). Finally the mean refers for th e three ma in dependent variables, attitude toward the brand (M = 5.39, SD = 1.19), eWOM intention (M = 5.13, SD = 1.37) and purchasing intention (M = 4.93, SD = 1.12) are relatively high as well.

PAGE 129

129 Correlation s Analysis Similar to the Twitter adoption analysis, a corr elation matrix was created to verify the relationship among variables (Table 5 12). As expected, the results of correlation s analysis yield statistically significant results for each variable. Specifically the main dependent variables for eWOM related re search, attitude toward the brand, eWOM intention and purchase intention were significantly correlated with independent variables (perceived similarity, perceived credibility, product category and perceived fit). The correlation s analysis yielded that atti tude toward the brand was significantly correlated with perceived similarity (r = .53, p < .01), perceived credibility (r = .58, p < .01), product category (r = .49, p < .01), and perceived fit (r = .59, p < .01). eWOM intention was also highly correlated with perceived similarity (r = .57, p < .01), perceived credibility (r = .72, p < .01), product category (r = .60, p < .01) and perceived fit (r = .72, p < .01). Finally, correlation s analysis confirmed the statistically significant correlation between pur chase intention and perceived similarity (r = .50, p < .01), perceived expertise (r = .68, p < .01), product category (r = .62, p < .01), and perceived fit (r = .69, p < .01). Therefore all measurements were included into structural equation model for both factor analysis and hypotheses and research question testing (Table 5 12). Factor Analysis The purpose of the factor analysis here is to confirm the measurement dimensions first this study employed principal component analysis for product category and perceived fit to verify the virtual/real, utilitarian/hedonic product differentiation. Then the results of principal component analysis were considered to conduct exploratory factor analysis and confirmatory analysis for final factor loadings.

PAGE 130

130 Principal c omponent a nalysis T his study expected certain level of product category differentiation such as tangible vs. intangible product differentiation (e.g., Arndt, 1967 ; Sheth, 1971 ; Zeithaml et al. 1993) or utilitarian vs. hedonic product (e.g., Babin & Darden, 1995; Babin et al., 1994 ; Cha, 2009 a ). To verify the theoretical differentiation of product category, this study employed a principal component analysis with Direct Oblimin eWOM behavior would be different based on the factor of product category. Principle component analysis was employed since this method is appropriate for capturing as much information as possible b y using a few comp onents (Park et al., 2002) and Direct Oblimin rotation was selected since there was significant correlation among many of items for each product category and perc eived fit constructs (Fabrigar et al., 1999). The results of principal component yield utilitar ian vs. hedonic differentiation rather than tangible vs. intangible product category differentiation. Table 5 13 and Table 5 14 show the results of principal component analysis for product category and perceived fit yield two factors successfully. The firs t factor (utilitarian dimension) refers to five items: healthcare services, financial services, automobiles, telecommunication services and computer equipment whereas the second factor (hedonic dimension) consists with other five items: movie, restaurant, clothes, electronics and travel information. This differentiation of utilitarian and hedonic dimensions was included for further exploratory and confirmatory factor analysis. Exploratory factor analysis Including utilitarian and hedonic dimension different iation, an exploratory factor analysis with maximum likelihood extraction was conducted to determine factor

PAGE 131

131 numbers. Same criteria used in Twitter adoption research for verifying factor differentiation was applied: less than .08 RMSEA, changes less than .0 1 in RMSEA values and using fewer factor if the RMSEA changes were marginal (Hu & Bentler, 1999). First, the number of factor s was evaluated via scree plot graph in terms of its descent (Figure 5 3 ) It yielded nine factor. Next, based on the RMSEA value and its change rate, the nine factor model was confirmed following excluded the item load higher than 0.4 with other factors and concentrated on the items that loaded at least 0.5 on only one factor. The following nine f actors were confirmed in this study: perceived similarity, perceived credibility, product category (utilitarian), product category (hedonic), perceived fit (utilitarian), perceived fit (hedonic), attitude toward the brand, eWOM intention and purchase inten tion ( T able 5 15) Confirmatory factor analysis To confirm the nine different perceptions of eWOM related variables, a confirmatory factor analysis (CFA) using maximum likelihood method was applied. This itter is a multi dimensional and hierarchical behavior rather than one dimension activity that could be predicted by perceptions of appropriation in spreading brand in formation in Twitter according to product category (utilitarian vs. hedonic). Therefore, the related behavioral pattern based on the results of the pretest and exploratory factor analysis was performed. The results indicated a good model fit with utilitarian vs. hedonic differentiation for product category and perceived fit. The specific factors are perceived similarity,

PAGE 132

132 perceived credibility, product category (utilitarian) product category (hedonic), perceived fit (utilitarian ), perceived fit (hedonic), attitude toward the brand, eWOM spreading intention and actual purchase intention ( 2 = 2538.326, d.f = 909, CFI = .87, TLI = .86, RMSEA = .076). Table 5 16 shows the results of model fit of the confirmatory factor analysis and T able 5 17 shows the final factor loading. The results of the confirmatory factor analysis indicated that all the items statistically evaluated the each constructs as intended and the factor loading of each item is high (.61 to .92). Validity and reliab ility After confirming the factors for the discriminant validity through the correlation s analysis of the latent variables was presented The result of correlation s analysis was displayed in Table 5 18 It indicated th at all the latent variables were highly correlated. Reliability of final measurements was satisfying the baseline for good reliabilit y (above .70) suggested by previous literature (Nun n aly, 1978). In addition, all s were above 0.5 (from .67 to .91) satisfying the criteria for reliability test (Rivard, 1988). Therefore, it was concluded that the variables for eW OM (T able 5 17 and Table 5 18). Structural Equation Modeling After confirming the nine different factors with the adequate fit of measurement model, a structural equation modeling was conducted for testing hypotheses and research questions using the nine variables: perceived similarity, perceived credibility, product category (utilitarian), product category (hedonic), perceived fit (utilitarian),

PAGE 133

133 perceived fit (hedonic) attitude toward the brand, eW OM intention and actual purchase intention A critical assumption for conducting SEM is multivariate normality which states that scale items and their combined measurements should be statistically normally distributed (Joreskog, 1973). Therefore, the vari assessed. It was found that the data here ranged from 1.41 to .15, in skewness and .93 to 2.56 in kurtosis satisfying the criteria for normality for the structural equation model (skewness < 2, kurtosis < 4). (Ho ng et al., 2003). ( T able 5 11) After verifying the multivariate normality of structural model, correlation s analysis with exogenous variables was performed. Previous studies have argued that the correlation value above .90 may cause a problem of multicoll inearity (Hair et al., 1998). For this model of eWOM related variables all correlation value s ranged between 39 and .73 ( p < .01) I t was concluded that there is no multicollinearity problem in the structural model. (Table 5 18) Once the correlation mat rix was displayed, a structural equation model was performed to reveal the relationship between variables and to test the hypotheses and research questions. The overall model fit indices suggested that the model strongly fit the data for explaining the rel ationship of variables ( 2 = 169.055, d.f. = 89, p < .001; CFI = .983, TLI = .971, RMSEA = .054), with above .90 CFI and TLI (Bentler et al., 1990) and below .08 RMSEA (Browne & Cudeck, 1993). The results of the structural model are presented in Table 5 19 and T able 5 20, with th e s tandardized path coefficients. Figure 5 4 shows the final structural model, including the significant pathways' observed values H7a and H7b predicted that the

PAGE 134

134 perceived similarity between respondents and their Twitter friends who Tweet brand related in formation would be positively associated with their attitude toward the brand and eWOM intention. The results of SEM indicated that perceived similarity is a but it did not influence the individual i ntention for spreading eWOM message ( = .21, p < .001; = .047 p = .31, respectively). Therefore, H ypothesis 7a was supported and H ypothesis 7b wa s not supported. ( See T able 5 21 for the hypothesis testing summary). In terms of perceived credibility and its effects on attitude toward the brand and eWOM intention, perceived credibility was highly associated with both attitude toward the brand ( = .19, p < .05) and eWOM spreading intention ( = .28, p < .001). Thus, H ypothesis 8a and Hypothesis 8b were bo th supported. RQ2 investigated whether there is a difference between utilitarian and hedonic product attitude toward a tweeted brand and eWOM intention. The results of SEM suggested that there was a difference in the attitude toward the brand but dependent on the product category. Specifically obtaining brand information for a hedonic product (e.g., movie and restaurant ) played a role in the toward the brand ( = .16, p < .001 ). But the utilitarian product perception toward the brand. Also, obtaining brand related information on Twitter for either utilitarian or hedonic product intention. This study also expected that consumer perception of fit between product dimensions (utilitarian vs. hedonic) and the usage of Twitter for brand information would be positively associated attitude toward the brand and his/ her eWOM

PAGE 135

135 spreading inte ntion ( H ypothesis 9a, Hypothesis 9b and H ypothesis 10a, Hypothesis 10b). In terms of perceived fit between utilitarian products and Twitter, the variable successfully predicted eWOM intention ( = .16 p < .05). However, perceived fit was not related attitude toward the brand in the context of utilitarian products In addition, the results of SEM indicated that perceived fit of hedonic products and Twitter positively influence d both ind = .29, p < .001) and eWOM intention ( = .18, p < .01). Hence, H ypothesis 9a was not supported while H ypothesis 9b, H ypothesis 10a and 10b were supported. Next, the pathways from attitude toward the brand to eWOM int ention and purchase intention were examined Regarding the relationships predicted in H ypothesis 11, Hypothesis 12, and Hypothesis 13, the suggested that there were significant by positive relationships between attitude toward the brand and eWOM intention ( = .33 p < .001), attitude toward the brand and purchase intention ( = .24 p < .001) and eWOM intention and purchase intention ( = .46 p < .001). In summary, this study found the role of perceived similarity, perceived credibility, utilitarian produc t category, hedonic product category, perceived fit (utilitarian), perceived fit (hedonic) to be significant in brand, eWOM intention and purchasing intention. Perceived similarity only influence attitude tow ard the brand ( H ypothesis 7a was supported whereas H ypothesis 7b was rejected). Regarding perceived credibility, both attitude toward the brand and eWOM intention were positively associated with perceived credibility ( H ypothesis 8a and Hypothesis 8b were s upported). In terms of product category, only the hedonic product dimension was related attitude toward the brand. The study also

PAGE 136

136 revealed that p erceived fit was only positively related to eWOM intention in the context of utilitarian product s, while perceived fit was significant in affect ed both attitude toward the brand and eWOM intention for hedonic products ( H ypothesis 9b, Hypothesis 10a, Hypothesis 10b were supported whereas H ypothesis 9a was not supported). Also, as predicted attitude t oward the brand did positively influence eWOM intention and purchase intention ( H ypothesis 11, H ypothesis 12) and eWOM intention was a significant predictor of purchase intention ( H ypothesis 13). Additional A nalysis This study employed two additional sta tistics to investigate consumer related variables such as age, education level, and gender. the brand, eWOM intention, and purchase intention by using age and education level for independent variables. Age and education were not found to be statistically significant in terms of attitude toward the brand (adjusted R squared value of .001, p = .45), eWOM intention (adjusted R squared value of .006, p = .91), and their purchase intention (adjusted R squared value of .001, p = .43). (Research Q uestion 3 and Research Q uestion 4 was not supported). Also, a one way analysis of variance (ANOVA) was performed by using gender as e toward the brand, eWOM intention and purchase intention. While mean difference existed in terms of attitude toward the brand ( M male : 5.35, SD = 1.17 vs. M female : 5.41, SD = 1.20), eWOM intention ( M male : 5.21, SD = 1.36 vs. M female : 5.10 SD = 1.38) and p urchase intention ( M male : 5,12 SD = 1.10 vs. M female : 4.86, SD = 1.12), the results for gender difference showed no significant differences in terms of attitude toward the brand ( F = .17 p = .68), eWOM

PAGE 137

137 intention ( F = .37 p = .55), and purchase intention ( F = 3.09, p = .08). Thus, R esearch Q uestion 5 was not supported.

PAGE 138

138 Table 5 1 Twitter adoption related measurement, descriptive statistics, skewness and kurtosis Variable Items Min Max Mean SD Skewness Kurtosis Conformity to Norm NORM1 1 7 4.03 1.57 .176 .277 NORM2 1 7 3.88 1.59 .036 .876 NORM3 1 7 4.07 1.57 .109 .762 Perceived Usefulness PU1 1 7 4.64 1.50 .425 .394 PU2 1 7 4.66 1.45 .429 .357 PU3 1 7 4.86 1.33 .473 .018 PU4 1 7 5.03 1.38 .752 .494 PU5 1 7 5.23 1.32 .891 724 PU6 1 7 5.33 1.37 1.03 1.05 Perceived Ease of Use PE1 1 7 5.72 1.28 1.21 1.40 PE2 1 7 5.86 1.24 1.47 2.26 PE3 1 7 5.69 1.24 1.04 .946 PE4 1 7 5.71 1.19 1.22 2.05 PE5 1 7 5.75 1.25 1.27 1.73 Attitude toward Twitter AT 1 1 7 5.57 1.28 1.23 1.52 AT2 1 7 5.83 1.21 1.68 3.82 AT3 1 7 5.59 1.23 1.16 1.59 AT4 1 7 5.84 1.20 1.40 2.30 AT5 1 7 4.88 1.55 .630 .234 Twitter usage for days per week TU1 1 7 5.34 1.91 .884 .431 Twitter usage for hour s per day TU2 1 7 3.13 1.47 .906 .592 Frequency of obtaining brand information FB 1 7 3.88 1.37 .491 .223

PAGE 139

139 Table 5 2. Initial correlation matrix among Twitter adoption variables Mean SD 1 2 3 4 5 6 7 1 3.88 1.37 .24** .47** .29** .41* .32** .33** 2 3.99 1.47 .24** .36** .17** .31** .14* .24** 3 4.96 1.21 .47** .36** .46** .65** .36* .40** 4 5.76 1.10 .29** .17** .46** .73** .43** .28** 5 5.54 1.08 .41** .31** .65** .73** .49** .40** 6 7 5.34 3.13 1.91 1.47 .32** .33** .14 .24** .36** .40** .43** .28** .49** .40** .43** .43** 1: Frequency of obtaining brand information, 2: Conformity to social norm. 3: Perceived usefulness, 4: Perceived ease of use, 5: Attitude toward Twitter, 6: Twitter usage (days per week), 7: Twit ter usage (hours per day). p < .05, ** p < .01

PAGE 140

140 Table 5 3. Model fit of exploratory factor analysis of Twitter adoption related variables Model 2 d.f p RMSEA Model 1 (1 factor model) 2649.209 170 .000 217 Model 2 (2 factors model) 1260.445 151 .000 .154 Model 3 (3 factors model) 588.146 133 .000 .106 Model 4 (4 factors model) 283.597 116 .000 .068 Model 5 (5 factors model) 219.302 100 .000 .062 Table 5 4. Model fit of confirmatory factor analysis of Twitter adoption variable Model 2 d.f CFI TLI RMSEA Model 1 (1 factor model) 2804.106 209 .556 .509 .201 Model 2 (5 factors model) 731.383 199 .909 .894 .093 Model 3 (5 revised factors mod el) 275.575 109 .960 .949 .071

PAGE 141

141 Table 5 5 Results of final factor loading and reliability test (Twitter adoption) Variable Item Standardized Factor Loading Conformity to subjective Norm Generally speaking, I would do what my group members think I should do. .89*** .93 Generally speaking, I would do what my Twitter friends think I should do in the Twitter environment. .86*** Generally speaking, I would do what others think I should do in the online environment. .94*** Perceived Usefulness I find Twitter useful in my life. .82*** .92 Use of Twitter makes it easier to obtain product information .89*** Use of Twitter to obtain product information increases my productivity .90*** Use of Twitter enables me to accomplish tasks more quickly .85*** Perceived Ease of Use Tweet, Mention and Retweet on Twitter is easy. .84*** .95 Learning to use Twitter is easy for me. .95*** It is easy for me to become skillful at using Twitter. .89*** I find Twitter easy to use. .94* ** Attitude Toward Twitter I would like to communicate with my Twitter friends again in the future. .81*** .88 provided by Twitter. .82*** I feel comfortable in using Twitter. .89*** Actual Twitter Usage H ow many days during a typical week do you use Twitter? .63*** How many hours during a typical day do you use Twitter? .60*** How often do you get information about a brand from your Twitter friends? .59*** Note: *** p < .001

PAGE 142

142 Table 5 6. Correlation matrix for final validity of constructs (Twitter adoption). Variable Items CN PU PE AT FB TUD TUH Conformity to Norm (CN) 1 .93** .32** .16** .22** .21** .15** .21** 2 .92** .30** .13* .18** .21** .08 .21** 3 .95** .33** .13* .22** .22** .16** .24** Perceived Usefulness (PU) 1 .30** .88** .25** .43** .40** .29** .38** 2 .33** .91** .33** .47** .46** .27** .35** 3 .33** .92** .32** .48** .43** .30** .36** 4 .24** .89** .37** .51** .42** .33** .28** Perceived Ease of Use (PE) 1 .13** .32** .89** .67** .28** .41** .26** 2 .16** .33** .95** .66* .23** .35** .22** 3 .15** .30** .92** .67** .19** .34** .17** 4 .10 .34** .95** .68** .19** .39** .24** Attitude toward Twitter (AT) 1 .22** .52** .57** .91** .34** .45* .31** 2 .19** .43** .77** .89** .29** .45** .31** 3 .20** .46** .62** .89** .30** .45** .29** Frequency of obtaining brand information (FB) 1 .24** .47** .29** .41** .32** .33** Twitter Usage: Days per week (TUD) 1 .14* .36** .43** .49** .32** .43** Twitter Usage: Hours per day (TUH) 1 .24** .40** .28** .40** .33** .43** Note: p < .05, ** p < .01

PAGE 143

143 Table 5 7. Significant parameter estimates of model Parameter Parameter Estimates Perceived usefulness Attitude toward Twitter .296 (.297)*** Perceived ease of use Attitude toward Twitter .657 (.687)*** Attitude toward Twitter Frequency of obtaining brand Information .303 (.236)* Attitude toward Twitter Twitte r usage (Days per week) 1.09 (.607)*** Attitude toward Twitter Twitter usage (Hours per day) .460 (.333)** Perceived usefulness Frequency of obtaining brand information .471 (.383)*** Perceived usefulness Twitte r usage (Hours per day) .291 (.222)** Note: Number represents non standardized parameter estimates, Standardized parameter estimates displayed in parenthesis. p < .05; ** p < .01; *** p < .001

PAGE 144

144 Table 5 8. Summary of hypothesis testing for the Twitter ad option study Independent Variable Dependent Variable Path coefficient (Standardi zed) SE P Dependent Variable H1a Conformity to subjective norm Attitude toward Twitter .037 .032 .39 Not supported H1b Conformity to subjective norm Twitter usage (days per week) .007 .076 .904 Not supported Twitter usage (hours per day) .101 .061 .081 Not supported Frequency of obtaining brand information .061 .055 .276 Not supported H2a Perceived usefulness Attitude toward Twitter .297 .045 .001*** Supported H2b Per ceived usefulness Twitter usage (days per week) .038 .121 .590 Not supported Twitter usage (hours per day) .222 .097 .003** Supported Frequency of obtaining brand information .383 .088 .001*** Supported H3a Perceived ease of use Attitude toward Twit ter .687 .047 .001*** Supported H3b Perceived ease of use Twitter usage (days per week) .104 .175 .306 Not supported Twitter usage (hours per day) .124 .138 .235 Not supported Frequency of obtaining brand information .110 .124 .271 Not supported Note: p < .05, ** p < .01, *** p < .001

PAGE 145

145 Table 5 8. Continued Independent Variable Dependent Variable Path coefficient (Standardi zed) SE P Result H4 Age Twitter usage (days per week) N.A N.A N.A Not supported Twitter usage (hours per day) N.A N.A N.A Not supported Frequency of obtaining brand information N.A N.A N.A Not supported H5 Education Level Twitter usage (days per week) N.A N.A N.A Not supported Twitter usage (hours per day) N.A N.A N.A Not supported Frequency of obtaining brand info rmation N.A N.A N.A Not supported RQ1 Gender Twitter usage (days per week) N.A N.A N.A No Twitter usage (hours per day) N.A N.A N.A No Frequency of obtaining brand information N.A N.A N.A No H6 Attitude toward Twitter Twitter usage (days per week) .607 .222 .001*** Supported Twitter usage (hours per day) .333 .172 .007** Supported Frequency of obtaining brand information .236 .154 .048* Supported Note: p < .05, ** p < .01, *** p < .001

PAGE 146

146 Table 5 9. Direct, indirect and total effects of major v ariables Independent Variable Dependent Variable Direct Effect Indirect Effect Total Effect Perceived Usefulness Attitude toward Twitter .297 .000 .297 Twitter usage (days per week) .038 .180 .218 Twitter usage (hours per day) .222 .099 .320 Frequen cy of obtaining brand information .383 .070 .453 Perceived Ease of Use Attitude toward Twitter .687 .000 .687 Twitter usage (days per week) .104 .417 .312 Twitter usage (hours per day) .124 .228 .105 Frequency of obtaining brand information .110 .162 .052 Conformity to Subjective Norm Attitude toward Twitter .037 .000 .037 Twitter usage (days per week) .007 .022 .029 Twitter usage (hours per day) .101 .012 .113 Frequency of obtaining brand information .061 .009 .069 Attitude Toward Twitter Twitter usage (days per week) .607 .000 .607 Twitter usage (hours per day) .333 .000 .333 Frequency of obtaining brand information .236 .000 .236

PAGE 147

147 Table 5 10. Mediation effect of attitude toward Twitter Independent Variable Mediation Variable Depen dent Variable Standardized Path Coefficient Perceived ease of use Attitude toward Twitter Twitter usage (days per week) .417 Twitter usage (hours per day) .228 Frequency of obtaining brand information .162

PAGE 148

148 Table 5 11. eWOM re lated me asurement, descriptive statistics, skewness and kurtosis, Variable Items Min Max Mean SD Skewness Kurtosis Perceived Similarity SIM1 1 7 5.12 1.32 .97 .94 SIM2 1 7 5.22 1.26 1.00 1.31 SIM3 1 7 5.07 1.33 .99 .87 SIM4 1 7 5.28 1.19 1.09 1 .66 SIM5 1 7 5.18 1.22 .95 1.35 SIM6 1 7 5.24 1.31 .95 .90 Perceived Credibility CR1 1 7 5.07 1.22 .78 1.15 CR2 1 7 5.12 1.26 1.02 1.65 CR3 1 7 5.05 1.32 .97 1.05 CR4 1 7 4.94 1.31 .85 1.13 CR5 1 7 4.96 1.35 .83 .78 CR6 1 7 4 .63 1.48 .46 .12 CR7 1 7 5.18 1.36 .87 .82 CR8 1 7 4.69 1.49 .59 .05 Product Category PC1 1 7 4.69 1.63 .51 .53 PC2 1 7 5.08 1.53 .86 .23 PC3 1 7 4.24 1.73 .21 .93 PC4 1 7 4.93 1.59 .69 .18 PC5 1 7 5.21 1.49 .95 .55 PC6 1 7 4.43 1.63 .39 .62 PC7 1 7 4.32 1.71 .28 .86 PC8 1 7 5.51 1.35 1.14 1.33 PC9 1 7 4.15 1.73 .15 .97 PC10 1 7 5.02 1.58 .86 .02 Perceived Fit PF1 1 7 4.90 1.45 .66 .11 PF2 1 7 5.12 1.41 .94 .99 PF3 1 7 4.24 1.64 .24 .65 PF4 1 7 4.98 1.48 .78 .40 PF5 1 7 5.38 1.33 1.13 1.64 PF6 1 7 4.83 1.34 .66 .55 PF7 1 7 4.61 1.45 .42 .15 PF8 1 7 5.65 1.32 1.41 2.56 PF9 1 7 4.17 1.67 .19 .70 PF10 1 7 5.13 1.36 .86 .86 Attitude toward the Brand AB1 1 7 5. 33 1.35 .87 .81 AB2 1 7 5.42 1.28 .85 .80 AB3 1 7 5.42 1.30 .90 .97 AB4 1 7 5.41 1.28 .85 .81

PAGE 149

149 Table 5 11. Continued Variable Items Min Max Mean SD Skewness Kurtosis eWOM spreading intention eI1 1 7 5.30 1.38 1.10 1.02 eI2 1 7 4.93 1.58 .7 6 .01 eI3 1 7 5.16 1.47 .91 .35 Purchase Intention PI1 1 7 4.82 1.30 .38 .10 PI2 1 7 4.92 1.25 .63 .48 PI3 1 7 4.85 1.28 .40 .28 PI4 1 7 5.11 1.18 .48 .35

PAGE 150

150 Table 5 12. Initial Correlation matrix among eWOM variables Mean SD 1 2 3 4 5 6 7 1 5.19 1.10 .64** .48** .53** .53** .57** .50** 2 4.96 1.09 .64** .69** .72** .58** .72** .68** 3 4.76 1.24 .48** .69** .70** .49** .60** .62** 4 4.90 1.15 .53** .72** .70** .59** .72** .69** 5 5.39 1.19 .53** .58** .49** .59** .68** .67** 6 7 5.13 4.93 1.37 1.12 .57** .50** .72** .68** .60** .62** .72** .69** .68** .67** .75** .75** 1: Perceived similarity, 2: Perceived credibility. 3: Product category, 4: Perceived fit, 5: Attitude toward the brand, 6: eWOM intention, 7 : purchase intention. p < .05, ** p < .01

PAGE 151

151 Table 5 13. The results of principal component analysis of product category Variable Factor1 Factor2 Healthcare services .92 .10 Financial services .85 .03 Automobiles .8 5 .03 Computer equipment .69 .06 Telecommunication services .61 .20 Movie .15 .98 Restaurant .02 82 Clothes 15 69 Electronics .32 58 Travel information .40 47 Table 5 14. The results of principal component analysis of perceived fit Variable Factor1 Factor2 Healthcare servi ces .9 7 .1 2 Financial services .96 .0 9 Automobiles 80 .12 Computer equipment .62 .32 Telecommunication services .6 1 32 Movie .1 6 .9 9 Restaurant .02 90 Clothes 21 .6 6 Electronics .3 2 64 Travel information .4 2 52

PAGE 152

152 Table 5 1 5 Model f it of exploratory factor analysis of eWOM related variables Model 2 d.f p RMSEA Model 1 (1 factor model) 5601.808 945 .000 .127 Model 2 (2 factors model) 4574.207 901 .000 .115 Model 3 (3 factors model) 3720.697 858 .000 .104 Model 4 (4 factors model) 3173.512 816 .000 .097 Model 5 (5 factors model) 2685.281 775 .000 .090 Model 6 (6 factors model) 2197.248 735 .000 .081 Model 7 (7 factors model) 1839.763 696 .000 .073 Model 8 (8 factors model) 1497.378 658 .000 .065 Model 9 (9 factors model) 1278. 589 621 .000 .058 Table 5 16 Model fit of confirmatory factor analysis of eWOM related variables Model 2 d.f CFI TLI RMSEA Model 1 (1 factor model) 6060.427 945 .599 .561 .133 Model 2 (7 factors model) 3017.230 924 .836 .816 .086 Model 3 (9 factors model with utilitarian, hedonic differentiation) 2538.326 909 .872 .855 .076

PAGE 153

153 Table 5 17. Results of final f actor loading and reliability test (eWOM) Variable Item Standardized Factor Loading Perceived Source Similarity In terms of outlook on life, my Twitter friends are similar to me. .87** .93 In terms of likes and disli kes, my Twitter friends are similar to me. .82** In terms of values and experiences, my Twitter friends are similar to me .81 ** In terms of tastes for products, my Twitter friends are similar to me .79** In terms of preferences and value, my Twi tter friends are similar to me. .86** Overall, my Twitter friends are similar to me. .88** Perceived Source Credibility I feel the tweeted product information given by my Twitter friends is strong .85 ** .92 I feel the tweeted brand information given by my Twitter friends is convincing .87 ** I feel the tweet brand information given by my Twitter friends is persuasive. .84 ** I feel the tweet brand information given by my Twitter friends is powerful. .84** My Twitter friends have knowled ge about computer equipment in general .61 ** My Twitter friend is an expert in the area of computer equipment .67** My Twitter friends have knowledge about restaurants in general .76** My Twitter friend is an expert in the area of restaurants .77* Product Category (U tilitarian) I often try to obtain product information about healthcare services .79 ** .9 1 I often try to obtain product information about finance services .78 ***

PAGE 154

154 Table 5 17. Continued Variable Item Standardized Factor Load ing I often try to obtain product information about automobiles .83 ** I often try to obtain product information about telecommunication services .86 ** I often try to obtain product information about computer equipment. .80 ** Produ ct Category (H edonic) I often try to obtain product information about movies. .78 ** .87 I often try to obtain product information about restaurants. .76 ** I often try to obtain product information about clothes .71 ** I often try to obtain product information about electronics .77** I often try to obtain product information about travel information. .75** Perceived fit (Utilitarian) Twitter is a good medium to learn about healthcare services .82** .91 Twitter is a good medium to learn about finance services. .81 ** Twitter is a good medium to learn about automobiles. .85** Twitter is a good medium to learn about computer equipment. .83** Twitter is a good medium to learn about telecommunication services. .81** Perceived f it (Hedonic) Twitter is a good medium to learn about movies. .80 ** .9 0 Twitter is a good medium to learn about restaurants .83 ** Twitter is a good medium to learn about clothes. .76 ** Twitter is a good medium to learn about electronics .81**

PAGE 155

155 Ta ble 5 17. Continued Variable Item Standardized Factor Loading Twitter is a good medium to learn about travel information .80 ** Attitude toward the brand Unfavorable/favorable .90** .94 Bad/good .91** Dislike/like .91 ** Negat ive/positive .86 ** eWOM spreading intention If I find interesting product information on the Twitter, I want to Retweet it to my friends after reading the tweeted brand information from Twitter friends. .92 ** .92 If somebody asks me for advice abo ut interesting product information, I will encourage him or her to Tweet after reading the tweeted brand information from Twitter friends. .86** I would recommend my friends and family to Tweet or Retweet in interesting product related information after reading the tweeted brand information from Twitter friends .90 Purchase Intention After considering the product information on my Twitter, it is very likely that I will buy the product .90 ** .91 After considering the product information on my Tw itter, I will purchase the product next time I need a product. .84 ** After considering the product information on my Twitter, I will definitely try the product. .84 ** If my friend called me last night to get the advice in his/her search for a product I would recommend him/her to buy the product .84** Note: ** p < .01

PAGE 156

156 Table 5 18. Correlation matrix for final validity of constructs (eWOM) 1: Perceived Similarity, 2: Perceived Credibility, 3: Product Category (UT), 4: Product Category (HD), 5: Perceived Fit (UT), 6: Perceived Fit (HD), 7: Attitude Toward Brand, 8: eWOM Intention, 9: Purchase Intention. ** p < .01 Variable 1 2 3 4 5 6 7 8 9 Perceived Similarity .64** .39** .51** .45** .54** .53** .57** .50** Perceived Credibility .64** .62** .68** .66** .67** .58** .72** .68** Product Category (UT) .39** .62** .73** .70** .50** .39** .51** .57** Product Category (HD) .51** .68** .73** .60** .69** .54** .61** .59** Perceived Fi t (UT) .45** .66** .70** .56** .73** .49** .64** .63** Perceived Fit (HD) .54** .67** .50** .69** .73** .62** .70** .65** Attitude Toward Brand .53** .58** .39** .54** .49** .62** .68** .67** eWOM Intention .57** .72** .51** .61** .64** .70** .67* .73** Purchase Intention .50** .68** .57** .59** .63** .65** .67** .75**

PAGE 157

157 Table 5 19. Significant parameter estimates of model for eWOM Parameter Parameter Estimates Perceived Similarity Attitude toward the Brand .225 (.205)*** Perceived Credibility Attitude toward the Brand .211 (.190) Product Category (Hedonic) Attitude Toward the Brand .163 (.163)* Perceived Fit (Hedonic) Attitude Toward the Brand Perceived Credibility eWOM intention .303 (.287)*** .318 (.275)*** Perceived fit (Utilitarian) eWOM intention Perceived fit (Hedonic) eWOM intention .157 (.162)* .192 (.177)** Attitude toward the brand eWOM intention .347 (.334)*** Attitude toward the brand Purchase intention .228 (.238)*** eWOM intention Purchase intention .421 (.456)*** Note: Number represents non standardized parameter estimates, Standardized parameter estimates displayed in parenthesis. *< p 05; ** p < .01; *** p < .001

PAGE 158

158 Table 5 20. Direct, indirect and total effects of major variables Independent Variable Dependent Variable Direct Effect Indirect Effect Total Effect Perceived Similarity Attitude toward the brand .205 .000 .205 eWOM intention .047 .068 .116 Purchase intention .033 .101 .068 Perceived Credibility Attitude toward the brand .190 .000 .190 eWOM intention .275 .063 .338 Purchase intention .112 .199 .311 Product Category (Utilitarian) Attitude toward the brand .082 .000 .082 eWOM intent ion .066 .027 .093 Purchase intention .237 .062 .175 Product Category (Hedonic) Attitude toward the brand .163 .000 .163 eWOM intention .077 .054 .131 Purchase intention .106 .099 .007 Perceived Fit (Utilitarian) Attitude toward the brand .0 42 .000 .042 eWOM intention .162 .014 .176 Purchase intention .001 .090 .089 Perceived Fit (Hedonic) Attitude toward the brand .289 .000 .289 eWOM intention .177 .096 .273 Purchase intention .095 .193 .288 Attitude toward the brand eWOM intentio n .334 .000 .334 Purchase intention .238 .152 .390 eWOM intention Purchase intention .456 .000 .456

PAGE 159

159 Table 5 21. Summary of hypothesis testing for the eWOM in Twitter study Independent Variable Dependent Variable Path coefficient (Standardi zed) SE P Result H7 a Perceived Similarity Attitude toward the brand 205 .064 .001*** S upported H7 b Perceived Similarity eWOM intention .047 .054 .312 Not supported H8 a Perceived Credibility Attitude toward the brand .190 .0 82 .001 ** Supported H 8b Perceived Cr edibility eWOM intention 275 .069 011 ** S upported RQ2a Product Category ( Hedonic ) Attitude toward the brand 163 .0 85 .05* Yes (hedonic) RQ2b Product Category eWOM intention N.A N.A N.A No H9a Perceived fit (Utilitarian) Attitude toward the brand .042 .078 .620 Not supported H9b Perceived fit (Utilitarian) eWOM intention 162 .064 .015 Supported H10a Perceived fit (Hedonic) Attitude toward the brand .289 .087 .001*** Supported H10b Perceived fit (Hedonic) eWOM intention .177 .073 .008** Supported H11 Attitude toward the brand eWOM intention .334 .054 .001*** Supported H12 Attitude toward the brand Purchase intention .238 .057 .001*** Supported H13 eWOM intention Purchase intention .456 .072 .001*** Supported RQ3a Age Attitude toward the brand N .A N.A N.A No RQ3b Age eWOM intention N.A N.A N.A No RQ3c Age Purchase intention N.A N.A N.A No RQ4a Education level Attitude toward the brand N.A N.A N.A No RQ4b Education level eWOM intention N.A N.A N.A No RQ4c Education level Purchase intention N. A N.A N.A No RQ5a Gender Attitude toward the brand N.A N.A N.A No p < .05, ** p < .01, *** p < .001

PAGE 160

160 Table 5 21. Continued Independent Variable Dependent Variable Path coefficient (Standardi zed) SE P Result RQ5b Gender eWOM intention N.A N.A N.A No RQ5 c Gender Purchase intention N.A N.A N.A No p < .05, ** p < .01, *** p < .001

PAGE 161

161 Figure 5 1 Scree plot of Twitter adoption variables

PAGE 162

162 Demographical factors (Age, education level gender) 2 = 276.322, d.f. = 105, p < .001; CFI = .958, TLI = .946, RMSEA = .073 Figure 5 2. Structural model including observed pathway UD: Twitter usage (days per week), UH: Twitter usage (hours per day), FB: Twitter usage (frequency of brand information obtain ing behavior) Perceived Ease of Use Conformity to Norm Percei ved Usefulness Attitude toward the Twitter T witter Usage .297*** .687*** UH: .222**, FB: .383 *** UD: .607*** UH: .333** FB: .154*

PAGE 163

163 Figure 5 3. Scree plot of eWOM related variables

PAGE 164

164 Figure 5 4 Proposed model for the Twitter in eWOM per spective Perceived Similarity Perceived Credibility Product Category : Utilitarian Perceived fit: Utilitarian Attitud e toward the Brand eWOM Spreading Intention Perceived fit: Hedonic .205*** .162* .190*** .275*** .163* .289*** .177** Purchase Intention .238*** .334*** .456*** Demographic factors (RQ3, RQ4, RQ5) ( Age educational level gender) Product Category : Hedonic

PAGE 165

165 CHAPTER 6 DISCUSSION This chapter has two purposes. First, the results of the hypotheses and research questions are summarized Second, individual contributions of each independent variable are discussed and possible explanations are suggested if the independent variables were not statically signi ficant. Although Twitter usage and its business implications have gained increasing attention in the marketplace, little empirical research has been conducted specifically to address the adoption and marketing utility of Twitter. This study conducted mult iple sets their actual usage of Twitter, and its effectiveness for marketing purposes. Table 6 1 and Table 6 2 show the summary of findings for both the Twitter adoption investigation and eWOM study. To test the hypotheses and research questions, this study adopted SEM, regression, and ANOVA as statistical methods for Twitter adoption and eWOM related variables. Summary of Findings for Twitter Adoption Hypotheses 1 to 6 a nd R esearch Q uestion 1 addressed the factors that affect norms, perceived usefulness, perceived ease of use, and consumer characteristics (age, education level, and gender ) were used as independent variables; daily Twitter usage per week, hourly Twitter usage per day, and the frequency of obtaining brand related information were used as dependent variables. Also, one path relationship between attitude toward Twitter and usa ge was assessed. To test the relationship between

PAGE 166

166 independent and dependent variables, this study employed SEM (conformity to subjective norm, perceived usefulness, perceived ease of use, attitude toward Twitter), multiple regression (age and education lev el), and one way ANOVA (gender). In terms of SEM, H ypothesis 2a (perceived usefulness and attitude toward Twitter), H ypothesis 3a (perceived ease of use and attitude toward Twitter), and H ypothesis 6 (attitude toward Twitter and Twitter usage) were strong ly supported. Hypothesis 2b (perceived usefulness and Twitter usage) was partially supported since perceived usefulness was positively associated with hourly usage and the frequency of behavior in obtaining brand information was not associated with daily T witter usage. However, H ypothesis 1a and H ypothesis1b (conformity to subjective norm) were not supported. In the regression analysis of demographic variables (age, education level), neither er ( H ypothesis 4a, H ypothesis 5a) and their Twitter usage ( H ypothesis 4b, H ypothesis 5b). In addition, RQ1 investigated gender differences among attitude toward Twitter ( R esearch Q uestion 1a) and actual Twitter usage ( R esearch Q uestion 1b). The result of ANOVA found that there was no difference between male and female subjects both in terms of attitude toward Twitter and Twitter usage. Table 6 1 and Table 6 2 summarize the empirical tests of variables related to Twitter adoption. Effects of C onformity to Subjective N orm on Attitude t oward Twitter and U sage The current study proposed two hypotheses in terms of conformity to subjective oward Twitter. Second, the relationship between daily, hourly usage and the frequency of behavior regarding obtaining brand information

PAGE 167

167 is assumed to have a positive rela tionship. While conformity to subjective norm p < .01), daily usage of Twitter (r = .14, p < .05), hourly usage (r = .24, p < .01) and frequency of behavior regarding obtaining brand in formation (r = .24, p < .01), the results of SEM rejected both hypotheses in the current study. Specifically, regarding the results of SEM, H ypothesis 1 (conformity to subjective norm) was not supported, meaning that conformity to subjective norm did not influence behavior (e.g., Baaren et al., 2011; Lin et al., 2009; Lu, Yao, & Yu, 2005) this study did not find significant conformity to the subjective norm in predicting Twitter adoption and usage. It is plausible that conformity to the subjective norm might contribute to the finding For example, the majority of previous literatures from social science, including communication and marketing areas, dealt with the concept of social influence via two representative methods: directly asking individual perception through a survey q uestionnaire (e.g., Davis, 1989; Davis et al., 1989) or manipulating by experimental condition (e.g., Vishwanath, 2009). A lternative methods such as network analysis (Katona, Zubcsek, & Sarvary, 2011; Lee et al., 2003) might be more appropriate for evaluat ing social influence in this context Indeed, Katona et al. (2011) recently revealed that the diffusion of SNSs was highly related to personal influence by utilizing a network analysis approach.

PAGE 168

168 Another possible explanation for this result might be that th ere are different dimensions of social influence. In this study, social influence was referred to as the level that one should or should not perform the behavior. Socia l influence in this study was measured using an original scale that dealt with conformity to subjective norm from TRA and TPB (Fishbein & Ajzen, 1975; Lee, 2003; Lee et al., 2003; Venkatesh & Davis, 2000). This is similar to the concept of social norm (Cia ldini, 1984, 1994) and many of social influence constructs in new media adoption studies such as Personal Digital Assistant (PDA) adoption (Nasco, Kulviwat, Kumar and Bruner, 2008), text messaging services (Pedersen & Nysveen, 2003) and mobile commerce (Pe dersen, 2005) focused on the social influence for conformity to perceived dominant norm. The results of subjective norm and attitude toward the media / intention to use a media platform varied (e.g., Nasco et al., 2009, Venkatesh & Davis, 2000; Hsu, Yen, C hiu, & Chang, 2006). For example, Nasco et al. (2009) and Venkatesh and Davis (2000) found positive relationships between conformity to subjective norm and attitude toward media platform, whereas Hsu et al. (2006) did not find effects of conformity to subj ective norm. In addition, social influence might vary within specific contexts such as perceived popularity in society (Zhou, 2008) or peer influence versus social influence (Lin et al., 2009). Indeed, Kwon and Chon (2009) differentiated social influence into three different dimensions: affiliation, positive self image, and perceived popularity. Therefore, it is plausible that Twitter, with its unique media characteristics, might be more relevant only to certain dimensions of social norm.

PAGE 169

169 The r esearch cont ext of this study also should be addressed. While various new media platforms and information technology have been examined for over a decade, it was found that for non Internet based products (i.e., PDA), conformity to subjective norm construct influenced consumer attitude and usage pattern (Nasco et al., 2008), whereas attitude and behavioral intention of Internet based service (i.e., blog) was not associated with social influence (Hsu & Lin, 2008). Another possible explanation for the results is that the current adoption stage of Price, 1992). For example, Fisher and Price (1992) found that superordinate group influence significantly affected the intention for consum through both personal and normative outcomes from early adoption. However, although this study considered Twitter as an emerging media, only actual users of Twitter were recruited as participants. Finally, regarding the adopti on stage, the sample of this study might be different from previous literatures. In detail, a variety of new media adoption studies (e.g., Eckhardt, Laumer, & Weitzel, 2009; Katona et al., 2011) differentiated the adopter and non adopter in terms of partic ipants Eckhardt et al. (2009) found that peer influence significantly affect non case for adopters. Therefore, considering the sample of this study included only actual users of Twitter, this might explain why the level of conformity did not predict H ypothesis 1a) and their usage ( H ypothesis 1b). Effects of Perceived Usefulness on Attitude t oward Twitter and Usage This study investigated one of the most im portant variables that the expectation of a system would enhance his or her job

PAGE 170

170 performance. First, in order to increase validity, this study adopted the original version of TAM measurements for perceived usefulness (Davis, 1989) rather than Venkatesh et (UTAUT) recommended by Putzke, Schoder, & Fishbach (2010) to successfully predict of Twitter. Second, although perceived usefulness in previous literatures focused on the original definition of perceived usefulness to increase their job performance (e.g., Davis et al., 1989; Lee, 2003; Venkatesh et al., 2003), this study modified the it ems specifically for behavior to obtain brand information as well as the general perception of itional TAM rmation related perceptions. As expected, our study confirmed that high reliability and validity factor analysis (standardized factor loading ranged .82 to .90), and participant s of this study exhibited relatively high levels of perceived usefulness ( M = 4.94, SD = 1.24). That is, the measurement of perceived usefulness for a brand specific concept was Twitter was useful to obtain product/brand related information. In terms of hypothesis testing, as expected, all four items of perceived usefulness p < .001) and actual

PAGE 171

171 p = .383, p < .001), c orresponding with previous literature form TAM (Davis et al., 1989) and UTAUT (Venkatesh et al., 2003). Indeed, perceived usefulness is one of the most technologies and the ir intention or usage in various Internet technology based new media platforms such as distance learning adoption (Lee et al., 2003), PDA (Nasco et al., 2008), blog (Hsu et al., 2008), online shopping service (Vijayasarathy, 2004), and mobile commerce (Wu & Wang, 2005). Thus, H ypothesis 2a (attitude toward Twitter) and H ypothesis 2b (Twitter usage) were supported. Effects of Perceived Ease of Use on Attitude t oward Twitter and Usage This study also tested the perceived ease of use in predicting individuals toward Twitter and their Twitter usage. For an average user or consumer of new communication technology, new information technology generally is considered to be costly or unaffordable (Kang, 2003). Therefore, if the medium is easy to use, users might exhibit higher intention to use or higher actual consumption of new media. Adopting this logic, a variety of studies reconfirmed that the effects of perceived ease of ology based services such as blog (Hsu & Lin, 2008). This study reconfirmed that perceived ease of use influenced attitude toward p < .001). In fact, r p < .001 vs p < .001). The descriptive statistics also showed that individuals were more likely to foster favorable attitude toward Twitter for ease of use than usefulness ( M ease = 5.80, SD = 1.17 vs. M usefulness = 4.93, SD = 1.24 vs.).

PAGE 172

172 Note that perceived actual usage. It is not surprising that there were mixed results in terms of the relationship between perceived ease of use and attitude toward new media or Intention to use (Vijayasarathy, 20 04). Particularly for the Internet based media platforms or services such as on line shopping usage (Vijayasarathy, 2004), mobile commerce (Wu & Wang, 2005), only attitude toward online shopping was significantly influenced by perceived ease of use where u sage intention was not directly influenced by perceived ease of use. Note that even though perceived ease of use did not relate directly to Twitter usage, two alternative statistical method s correlations analysis and a mediation analysis provided possib le clues on its indirect role in affecting Twitter usage Specifically, i correlations with daily usage of Twitter (r = .40, p < .01), hourly usage of Twitter (r = .24, p < .01), and frequenc y of obtaining brand information (r = .24, p < .01), in addition to attitude toward Twitter (r = .72, p < .01). B ased on these findings, it might be argued influence d Twitter usage through the attitude toward Twitter. More accurately, this mediation effect of attitude toward Twitter was evaluated by SEM. The path coefficient Twitter as a mediator, was .42 (.69*.61), hourly usage per day was .22 (.69*.33), and frequency of obtaining brand information was .16 (.69*.24). (Hypothesis 3a was supported, H ypothesis 3b was not supported).

PAGE 173

173 This result might be explained by the adoptions stage of Twitter (e.g., Eckhardt, Laumer, & Weitzel, 2009; Katona et al., 2011). As previously mentioned, this study initially recruited those users of Twitter who specifically obtained brand related information and might already be familiar with Twitter usage. Th us, the perceived ease Effects of Consumer Demographic V ariable s on Attitude t oward Twitter and Usage In terms of consumer characteristics, this study adopted three variables: age, educatio n level, and gender. Age and education levels were analyzed with regression Contrary to previous studies such as cable modem broadband adoption (Chan Olmsted et al., 2005), terrestrial dig ital television (Chan Olmsted & Chang, 2006), and e book adoption (Jung, Chan Olmsted, Park, & Kim, 2011), age and education levels examination of correlations analysis a lso found that there was no significant correlation between age, education level and Twitter attitude, and Twitter usage, respectively, except weak correlation between education level and the frequency of behavior in obtaining brand information (r = .11, p < .05). (Hypothesis 4a, H ypothesis 4b, H ypothesis 5a, and H ypothesis 5b were not supported). Gender difference based on the ANOVA result indicated that there was no difference regarding attitude toward Twitter and Twitter usage, contrary to previous TAM related studies in the context of mass customized newspaper adoption (Putzke et al., 2010) or recent mobile TV adoption (Kwon and Chon, 2009). Contrary to previous M male :

PAGE 174

174 5.73, SD = 1.18 vs. M f emale : 5.76, SD = 1.07) as well as their daily Twitter usage ( M male : 5.61, SD = 1.85 vs. M f emale : 5.26, SD = 1.92), hourly usage ( M male : 3.28, SD = 1.56 vs. M f emale : 3.09, SD = 1.44), and frequency of behavior in obtaining brand infor mation ( M male : 4.08, SD = 1.38 vs. M f emale : 3.81, SD = 1.36). Thus, regarding R esearch Q uestion 1, there was no difference between males and females in terms of attitude toward Twitter or Twitter usage. One possible explanation why gender based differences did not predict the dominance of sample in the study. Specifically males were 24% of participants (n = 74) whereas females were 75.6% (n = 233). Therefore, a different proportion of par might affect the result. In addition, in the context of Twitter usage it might also be plausible that there was usage in certain types of computer related technology (Jung et al., 2011). Indeed, Jung et al (2011) found that there was no gender difference in e book adoption. Twitter usage among the general population has drastically increased in recent years, reaching 200 million accounts in 2010 (Twitter, 2011). Simultaneously, Facebook, another represent ative of SNSs usage, reached 160 million visitors each month about three out of every four Internet users (Lipsman, Mudd, Rich, & Bruich, 2011). Also, this study employed only current user of Twitter, it is possible that demographic characteristics such as age and education level no longer play important roles to predict Yang, Morris, Teevan, Adamic, and Ackerman (2011) recently revealed that the cultural

PAGE 175

175 differenc es among participants (China, India, United States, and United Kingdom) than demographic variables, including age and gender. (Hypothesis 4a, H ypothesis 4b, H ypothesis 5a, H ypothesis 5b were not supported). Effects of Attitude toward Twitter on Usage Although many studies regarding new media adoption measured attitude toward media and intention for adoption, the current study measured actual usage for three different dimens ions (days per week, hours per day, frequency of obtaining brand information). This study particularly measured actual usage of Twitter rather than intention, employing only current users of Twitter through the qualifying question. The results indicate tha t attitude toward Twitter was a significant factor affecting daily Twitter p p < .001), and frequency of p < .05). Thus, H ypothesis 6 was supported. Summary of Fi ndings for eWOM Related Variables Twitter. Hypotheses 7 to 13 and R esearch Q uestions 2 to 5 were empirically tested via SEM, multiple regression and ANOVA. Perceived similarity, p erceived credibility, product category, perceived fit, and consumer characteristics (age, education level, gender) were used as the independent variables, and attitude toward the brand, eWOM intention, and purchase intention were used as dependent variable s. In terms of SEM, results for perceived similarity, perceived credibility, product category and perceived fit were tested. This study found that H7a (perceived similarity and attitude toward the brand), H8a (perceived credibility and attitude toward the brand),

PAGE 176

176 H8b (perceived credibility on eWOM intention), H9b (perceived fit of utilitarian product on eWOM intention), H10a (perceived fit of hedonic product on attitude toward the brand), H10b (perceived fit of hedonic product on eWOM intention), H11 (atti tude toward the brand on eWOM intention), H12 (attitude toward the brand on purchase intention), and H13 (eWOM intention on purchase intention) were strongly supported. Considering the results of multiple regression and ANOVA, RQ2 (gender and product categ However, H7b (perceived similarity on eWOM intention), H9a (perceived fit of utilitarian product category on attitude toward the brand), RQ2b (product category and eWOM int ention), RQ3 (age), RQ4 (education level), RQ5 (gender) were not supported. Table 6 1 and Table 6 2 summarize the empirical tests of hypot heses and research questions. Effect s of Perceived Similarity on Attitude t oward the Brand and eWOM Intention This s tudy adopted perceived similarity, one of the most important variables in the context of WOM and eWOM study (e.g., Brown & Reingen, 1987; Gilly, Craham, Wolfnbarger, & Yale, 1998; Price, Feick, & Higie, 1989; Wangenheim & Bayon, 2004) regarding communicato r characteristics. In terms of perceived similarity ( H ypothesis 7), only attitude toward the brand was p < .001); this yielded different results from previous eWOM studies in the context of blogs (e.g., Prender gast et al., 2010) or forums (e.g., Dellarocas, 2004). Most previous studies within the context of the online environment revealed a positive relationship between perceived similarity and consumer willingness to spread eWOM (Dellarocas, 2004; Prendergast e t al., 2010; Wangenheim & Bayon, 2004). However, Chu and Kim (2011) recently revealed that homophily, a

PAGE 177

177 eWOM behavior. The discrepancy might be rationalized from the per spective of eWOM cost. Although most WOM and eWOM literature has focused on the benefits and motivations of WOM and eWOM communication in consumer behavior, several studies have discussed the inevitable cost of WOM (Cheema & Kaikati, 2010; Frenzen & Nakam oto, 1993). In particular, Cheema and Kaikati (2010) suggested that the uniqueness is a special psychological trait (Lynn & Harris, 1997; Snyder & Fromkin, 1977; Tian, Bear den, & Hunter, 2001) that creates a preference for distinct and unique product comparisons with common products (Bloch, 1995; Simonson & Nowlis, 2000). While eWOM via Twitter helps foster a favorable attitude toward the brand and to inspire greater purchas e intention, it also may contribute to a sense of loss of reactions to eWOM messages could be me effort to eWOM in Twitter. Also, regardi ng the relationship between perceived source similarity and attitude toward the brand, eWOM intention, and purchase intention, correlations analysis provides some possible explanation. Although source similarity did not significantly WOM intention in SEM, source similarity had significant correlations with eWOM intention (r = .57, p < .01) and purchase intention (r = .50, p < .01). Further,

PAGE 178

178 perce ived credibility was put into the multiple regression. The model yielded an adjusted R squared value of 61.3% and perceived similarity w as significant predictor of eWOM p < .05). However, additional regression result indicated that Therefore, while perceived similarity did not significantly predict = .134, p < .05). Effects of Perceived Credibility on Attitude t oward the Brand and eWOM Intention This study adopted source credibility as an impor tant independent variable to in Twitter. Regarding measurement items, the current study modified the items specifically for Twitter context. In detail, four items from com municator credibility measurements in eWOM were adopted and modified for message senders in Twitter tweeted brand information given by my Twitter friends is powerf were created to evaluate the knowledge and expertise specifically to select from each product category (hedonic: restaurant, utilitarian: computer equipment). The statements computer equipment in

PAGE 179

179 Consistent wi th our prediction, the results of the reliability test, exploratory factor analysis, and confirmatory factor analysis found and satisfied value of standardized factor loadings (.61 to .87). That is, the measurement of source credibility for brand specific concept in Twitter was successfully evaluated by our measurement. Regarding perceived credibility, this study reconfirmed its important role in 2004). To be specific, this study revealed a positive association between perceived credibility, attitude toward p < .001), and intention to spread eWOM p < .001). In other words, brand attitude and individuals' willingness to spread eWOM were highly influenced by the credibility of the information sender ( H ypothesis 8). This finding re new media environment. Effects of Product Catego ry on Attitude t oward the Brand and eWOM Intention Our study used the product differentiation of utilitarian and hedonic differentia tion specifically based on products function (Hirschman & Holbrook, 1982; Holbrook & Hirschman, 1982; Verhagen, Boter, & Adelaar, 2010) rather than consumer perceived shopping value (Babin et al., 1994). A majority of previous studies regarding eWOM in b log or customer reviews only focused on specific product with experiment design (e.g., Lee et al., 2009; Mizersk, 1982; Ratchford et al., 2003). For example, Lee and colleagues (2009) found the importance of negative eWOM in attitude toward the laptop comp uter and purchase

PAGE 180

180 intention. Mizerski (1982) and Ratchford et al. (2003) empirically tested the importance of eWOM within automobile context. However, this study adopted multiple product categories obtained from industry report that most frequently discuss ed on online environment (healthcare services, finance services, automobiles, telecommunication services, computer equipment, movies, restaurants, clothes, electronics and travel information). Next, questionnaire s vel of brand component analysis, exploratory factor analysis and confirmator y factor analysis were performed and confirmed that consumers perceived product nature based on the utilitarian vs. hedonic inherit. The results of reliability and validity tests indicated that measurement successfully evaluated product category differenti ation (utilitarian product between .71 to .78). In terms of the SEM of product cate gory, an interesting finding should be addressed. In line with prior differentiation between utilitarian and hedonic products (Hirschman & Holbrook, 1982; Holbrook & Hirschman, 1982; Reibstein, 2002; Verhagen et al., 2010), this study found that Twitter is more important in mediation of brand p < .001) ( R esearch Q uestion 2). Indeed, descriptive statistics indicated that consumers more often obtained hedonic product related brand information ( M = 5.15, SD = 1.22) th an brand information of utilitarian product ( M = 4.36, SD = 1.44).

PAGE 181

181 One possible explanation for only the significant influence of the hedonic product category on (Hagtvedt & Pat rick, 2009). Hagtvedt and Patrick (2009) found that hedonic association for brands was a key factor of brand extendibility; therefore, a hedonic brand leads to a more favorable attitude toward brand extension evaluation than a utilitarian brand. Likewise, this study is based on the concept of category extension: that with Twitter providing brand information, consumers might be more favorable to brand information of a hedonic product. ble explanation that hedonic products easily lead to a higher likelihood of being brand lover. Following this notion, hedonic product category may foster a more favorable attitude toward the brand in Twitter environment comparing with the utilitarian produ ct category. While only the hedonic product category predicted attitude toward the brand, additional correlations analysis found that hedonic product category predicted eWOM p < .001), and p < .001) and util itarian p < .01), p p < .01). Thus, R esearch Q uestion 2a confirms the difference between utilitarian and hedonic pr oduct category differentiation. Effects of Perceived Fit on Attitude t oward the Brand and eWOM Intention C onsidering the findings in hypotheses 9 and10, we also adopted perceived fit construct, one of the most frequently discussed in previous marketing an d brand related study (e.g., Aaker & Keller, 1990; Cha, 2009a; Chang & Chan Olmsted, 2010; Bhat & Reddy, 2001; Park et al., 1986).

PAGE 182

182 T fit of Twitter and each product category ( utilitarian vs. hedonic) with a focus on category extension rather than product extension. In detail, questionnaires were created to directly ask each category that obtained from factor analysis and confirmatory f actor analysis. Reliability and validity tests performed in this study confirmed that these items actually measured the concept of perceived fit of product (both utilitarian and hedonic product) and Twitter. Specifically, perceived fit of utilitarian dimen loading ranged from .81 to .85, and hedonic dimension also enjoyed high value of Regarding hypotheses 9 and 10 perceived fit between Twitter and the product spreading intention. This is consistent with most previous studies in the area of perceived fit and brand extension. Specifically, perceived fit of utilitarian product was not positively related to attitude toward the brand, whereas the hedonic dimension o f the p < .001). p < .05) and p n. One possible theoretical reasoning that perceived fit of hedonic product

PAGE 183

183 brand exte ndibility (Hagtvedt & Patrick, 2009), similar to the result from R esearch Q uestion 2 (product category). Hagtvedt and Patrick (2009) indicated that if the brand has more hedonic related associations, then it lead s to more favorable attitude toward the bran d extension evaluation. A lthough this study adopted the logic of category extension rather than product extension, consumers might feel more positive when they obtained hedonic product related brand information in Twitter than utilitarian product. Indeed, the descriptive statistics revealed that consumers generally considered Twitter a good medium for obtaining brand information on a hedonic product more than a utilitarian product ( M hedonic = 5.25, SD = 1.16 vs. M utilitarian = 4.55, SD = 1.31). In additio n to SEM, correlations analysis provided that utilitarian perceived fit had p < .01), eWOM intention (r = .64, p < .01) and purchase intention (r = .63, p < .01). Therefore, altho ugh consumers indicated their perception of perceived fit of hedonic product and Twitter was more important than utilitarian fit, perceived fit between utilitarian product and Twitter was somewhat important. Effects of Consumer Demographic Variables on At titude t oward the Brand and eWOM Intention This study additionally conducted two statistical analyses to investigate the consumer related variables such as age, education level and gender. However, none of attitude toward the brand, eWOM intention and purchase intention by multiple regression and ANOVA. Additional correlations analysis did not find significant correlations between age and attitude toward brand, eWOM intention, and purchase intention, as well as between

PAGE 184

184 education level and attitude toward the brand, eWOM intention, and purchase intention. Also chi square test indicated that there was no significant difference among gender in terms of the three dependent variables. Based on these findings, it c an be argued that consumer demographic characteristics such as age, education level and gender did not eWOM context within Twitter with caution. However, regarding inconsist ent results of consumer characteristics in eWOM context (e.g., Cheung, & Law, 2009; Heung, 2003; Lp, Lee, & Law, 2011; Ratchford et al., 2003), it could be argued that demographics did not serve as a consistent predictor r. For example, some previous literature found that young and highly educated people used more Internet (Ratchford et al., 2003), specifically travel information sharing behavior (Lp et al., 2011) whereas some studies did not find significant difference am ong age and education level for travel web site usage (Cheung & Law, 2009; Heung, 2003). Gender differences were not found in SNSs usage (Cha, 2009b) and traveling information sharing web site (Lp et al., 2011), and that was reconfirmed in this study. Sim ilar to the demographic characteristic in the pattern of Twitter adoption study, one possible explanation of failing to predicting individual eWOM behavior in Twitter by demographic characteristic is a female dominant sample of this study. Specifically, ma les respondents only comp osed 24% of the participants (n = 74) Therefore, the in the adoption of computer related technology (Jung et al., 2011). I t is plausible that eWOM behavior in

PAGE 185

185 Internet technology based service, Twitter, also did not affect tech anxiety for female consumers. Also, current adoption stage of SNSs and Twitter might affect the l ack difference s in age, education level and gender in consumer eWOM behavior in Twitter. As previously mentioned, Twitter usage among general population has drastically increased recent year, reached 200 million accounts in 2010 (Twitter, 2011) many consum ers are familiar with Twitter and its role in obtaining brand related information or as an eWOM spreading tool. Therefore, their attitude toward brand, eWOM behavior or purchase intention might be similar regardless of characteri stics. Effects of Attitude t ow ard the Brand on eWOM intention and Purchase Intention Regarding the prediction role of consumer attitude and actual behavioral intention, this study investigated the relationship among attitude toward the brand, eWOM intenti on and purchase intention. Although many studies in eWOM context used specific fictitious brand to test the effect of eWOM in attitude toward brand and actual purchase intention in experiment setting (e.g., Lee & Youn, 2009; Lee et al., 2009), this study a sked consumer s about their brand attitude when consumers encountered brand information forwarded by their Twitter friends both form personal and company account. T he descriptive statistics here indicated that the respondents exhibited relatively high scor e of bran attitude ( M = 5.39, SD = 1.19), eWOM intention ( M = 5.13, SD = 1.37), and purchase intention ( M = 4.83, SD = 1.11) That is consumers exhibited favorable attitude toward the brand, eWOM intention, and purchase intention when they received brand r elated information in Twitter.

PAGE 186

186 p < .001), and purchase p g intention was p < .001). The results Fishbein, 1977, 2005; Fishbein & Ajzen, 1975; Perloff, 2010). Thus H ypot hesis 11, 12 and 13 were supported.

PAGE 187

187 Table 6 1. Result summary for hypotheses Hypothesis Result H1a. Conformity to subjective norm will be positively associated with Not supported H1b. Conformity to subjective norm will b e positively associated with Not Supported H2a. Perceived usefulness of Twitter will be positively associated with Supported H2b. Perceived usefulness of Twitter will be positively associated with Twitter usage. Supported attitude toward Twitter. Supported Twitter usage. Not supported H4a. Age is negatively asso Not supported Not supported toward Twitter. H5b. Education level is positively associ Not supported Not supported usage. Supported Twitter friends who tweet about a branded p roduct will be positively Supported Twitter friends who tweet about a branded product will be positively N ot supported H8a. The perceived source credibility of Twitter friends who tweet toward the brand. Supported H8b. The perceived source credibility of Twitter friends who tweet about a br spreading intention. Supported H9a. Perceived fit between Twitter and utilitarian product category brand. Not supported H9b. Perceived fit between Twitter and utilitarian product category intention. Supported H10a. Perceived fit between Twitter and hedonic product category tude toward the brand. Supported H10b. Perceived fit between Twitter and hedonic product category intention. Supported

PAGE 188

188 Table 6 1. Continued Hypothesis Result H11. Attitude toward the brand eWOM spreading intention. Supported purchase intention. Supported intention Supported

PAGE 189

189 Table 6 2. Result summary for research question Research Question Result toward Twitter? No difference usage? No difference RQ2a How does the product cat egory of a the brand? Hedonic played a role in affecting whereas utilitarian product did not. RQ2b How does the product category of a spreadi ng intention? No difference toward the brand? No difference spreading intention? No difference intention? No difference RQ 4a. How does educational level influences No difference RQ4b. How does educational level influences No difference purchasing intention? No d ifference toward the brand? No difference spreading intention? No difference purchasing intention? No difference

PAGE 190

190 CHAPTER 7 CONCLUSION This chapter discusses the thesis findings in the context of academic contributions and practical applications. It also addresses limitations and suggestions for future research. Our study consisted of two main parts: factors affecting Twitter adoption and its utilization of eWOM as a marketing tool. This study was conducted in consideration of the recent burgeoning of SNSs, specifically, Twitter, an ideal tool for alternative marketing. Little empirical research has been conducted to investigat e consumer motivations for using Twitter and their eWOM behavior. For the initial stage of Twitter related study, this study investigated several as a brand information obtaining tool and its poten tial for marketing purposes To obtain a more in depth as a product/brand information communication tool, current study integrated several different theories that expl ain consumer behavior specific to online environment: theory of reasoned action, theory of planned behavior, technology acceptance model, and eWOM. Based on the multiple sets of empirical test, including SEM, multiple regression, ANOVA, and correlations a nalysis, our study found a list of variables affecting discusses how the findings can be applied from the perspective of academic contribution and iew. Also, limitations and suggestions for future research are provided.

PAGE 191

191 Twitter Adoption Related Implication Theoretical Implication s In terms of academic contribution of Twitter adoption, this study identified six different aspects First, it empiricall y tested SNSs adoption based on integrating theoretical background of social influence, TRA, TPB, and TAM by measuring various dimensions of usage including the frequency of behavior in obtaining brand information rather than adoption intention employing a general sample. Second, somewhat different from previous studies, it found the con cept conformity to subjective norm was no longer an important predictor of new media adoption. Third, this study validated again the importance of perceived usefulness. Fourth, this study reconfirmed the impact of the perceived ease of use. Lastly, the study shows a different result from previous studies characteristics such as age, education level, and gender did not influence Twitter att itude and usage. Integrating adoption theories in Twitter context with actual usage using general sample behavior of a new media platform (e.g., Homburg et al., 2010; Lee, 2003 ; Lee et al., 2003; Park, 2010), using TPB, TRA, and TAM, the adoption of social media, particularly SNSs, has rarely been tested empirically. This study proposed integrating models of TPB, TRA, and TAM, and including a demographic characteristic construct to predict different constructs of actual Twitter usage, including daily usage per week, hourly usage per day, and the most important item of the frequency of behavior in obtaining

PAGE 192

192 brand information, rather than adoption intention by recruiting current Twitter users. Although previous studies revealed the consistent attitude of new media and usage intention (e.g., Davis et al., 1989) some studies indicated the possibilit y of a discrepancy & Ajzen, 1975; Perloff, 2010). Therefore, this study analyzed various factors influencing attitude toward the new media, Twitter, and their actual usage to provide more in depth understanding of consumer behavior on Twitter, particularly behavior in obtaining brand related information. Also, this study employed general U.S. consumers by utilizing a national consumer panel. Previous studies in the con text of media adoption mostly used student samples that might lack external validity due to sampling issues (e.g., Lin et al., 2009). While the student sample was used for the pre test, the general population sample was used for the main test. The purpose of the pre test was to test validity and reliability; the results of exploratory factor analysis and confirmatory factor analysis yielded almost identical questionnaire items. A brief comparison of the results between student sample for pre test and genera l sample for main test were somewhat different. Therefore, it should be noted that although student sample was used frequently in new media especially social media adoption studies t heir interpretation and generalization need caution. C onformity to subje ctive norm in attitude toward Twitter and usage not significantly relate to their adoption intention or usage of Twitter. Our study contradicts previous findings that the d egree of conformity to subjective norm was positively associated with new media adoption such as distance learning (Lee, 2003;

PAGE 193

193 degree of conformity to subjective norm such as experiment or network analysis would SNSs attitude and usage. Also different dimensions of social influence such as peer influence should be analyzed. The im portance of perceived usefulness in attitude toward Twitter and usage As expected, perceived usefulness was found to be a significant factor affecting reconfirmed t toward Twitter and their usage of Twitter. Specifically, this study verified the perceived usefulness construct in the social media context According to TAM (Davis, 1989; Davis e t al., 1989) one of the most important constructs to influence individual perceived utility is perceived usefulness M any studies have confirmed the important predicting role of perceived usefulness in media adoption, particularly among Internet based new technology. However, the majority of previous studies measured perceived usefulness expanded the conceptual dimension of perceived usefulness in brand related variables such ed the original TAM constructs of perceived usefulness utilizing results from the exploratory factor analysis, confirmatory factor analysis and SEM It validated the role of perceived usefulness regarding brand information obtaining behavior attitude toward the Twitter and usage. Thus, as one of the theoretically important perspectives of adoption study, perceived usefulness was the strongest predictor among Twitter adoption related

PAGE 194

194 variables, including perceived ease of use, age, education level, and gender. Specifically, the result of SEM reconfirmed that perceived usefulness induced the usage of Twitter. The importance of perceived ease of use in attitude toward Twitter and usage Consistent with previous adoption literatures, the effect of the p erceived ease of use was strongly related to though it did not link to Twitter usage. T heoretically, two contributions might be addressed in this study: the indirect effects of perceived ease of use to ind and adoption stage of Internet based new media platform. First, although the results of empirical test of perceived ease of use and attitude toward the new media and usage intention, the results of correlations analysis an d mediation analysis of SEM provided that perceived ease of use indirectly influenced Twitter usage rather via attitude toward Twitter rather than directly impacted to actual usage. In other words, although individuals were not directly influenced by their perception of ease of use when they tried new media platforms, they might still be indirectly influenced. Second, the adoption phase of an Internet based media platform should be addressed. Specifically, regarding the more recent Internet based media plat forms or services, including on line shopping usage (Vijayasarathy, 2004), mobile commerce (Wu & Wang, 2005), only attitude toward the online shopping was significantly influenced by perceived ease of use where using intention was not directly influenced b y perceived ease of use, while in the initial stage of the Internet burgeoning, literatures in terms of Internet based new media platforms such as distance learning program (Lee, use for

PAGE 195

195 each media platforms. Therefore, the prominent usage of the Internet, social media and SNSs, perceived ease of use no longer influenced individual intention for using Internet service based media since audiences have experienced the Internet to a c ertain degree. This study identifies perceived usefulness and perceived ease of use as the two this study empirically tests the relationship among variables. The re sults of this study generally support the adoption studies, namely traditional media adoption and Internet based media. However, in relation to attitude and behavioral measures, the effects of perceived usefulness and ease of use are different. Perceived usefulness positively influences both consumer attitude toward Twitter and their actual usage whereas perceived ease not adopt certain types of high tech, Internet b ased new media platforms available in markets. System characteristics, including usefulness and ease of use, play a role in the consumer decision making process for adoption. When consumers foster attitude toward certain new media services, they are primar ily driven by ease of use, and perception of usefulness directly influences their usage. Considering the services that are useful but complex, or easy to use but less helpful to consumers, it is necessary to understand different expectations for ease of use and usefulness through the attitudinal stage and actual behavioral stage. Furthermore, going beyond the TAM, researchers should consider the

PAGE 196

196 media adoption. Traditionally, many media adoption studies focused on the similarities of new media that could rep lace or substitute old ones (e.g., Lee & Leung, 2008). functional uniqueness and their likelihood of using the new video platform instead of television (Cha, 2009b). Likewise, in the Twitter context, at the attitudinal stage, consumers exhibit more favorable attitudes when consumers consider Twitter as similar to other easy to use Internet based services. However, in the actual behavioral stage, fundamental functional similarity w usage. Moreover, Jung et al. (2011) proposed that the concept of perceived hybrid functionality refers to the com bining of relatively new reading platforms and old print media by explaining e book reader adoption. Similarly, in the SNSs context, perceived usefulness and ease of use are not simple concepts regarding how to use Twitter. Twitter is a relatively new plat form for communication and shares similarities with other web (Jung et al., 2011) functions compared with other new media platforms, particularly for sending brand related information to consumers. Consumer characteristics in attitude toward Twitter and usage Our study investigated age, education level and gender as audience demographic related independent variables in addition to traditional adoption variables form TPB and TAM. Age and education level were measured by Likert scale whereas gender was asked by categorical variable. C ontrary to previous adoption study contexts such as cable modem broadband adoption (Chan Olmsted et al., 2005), terrestrial digital

PAGE 197

197 television (Ch an Olmsted & Chang, 2006) or e book adoption (Jung, et al., 2011), were not associated with either attitude toward Twitter or usage. The results of regression and ANOVA including demographical variab les may imply that Internet based services or new media platforms (Jung et al., 2011) As previously mentioned, previous technology related adoption study revealed that females reluctant to adopt new media ser vices or platform. However, in the Internet service based platforms, it has been pointed out that there were no gender differences. penetration of Internet usage among peopl e already familiar to the Internet and its applied services such as Twitter. Specifically, Twitter usage among the general population has drastically increased in recent years, reaching 200 million accounts in 2010 (Twitter, 2011). Also, the sample of thi s study only recruited the current user of Twitter. Therefore, if we employed non adopter simultaneously and compare two groups, demographical characteristics might Importance of perceived usefulness in n ew media adoption In terms of perceived usefulness in Twitter adoption and usage, our study confirmed the role of perceived usefulness in predict ing Twitter and their usage. Two strategic implications could be concluded, especi ally for marketer s When media companies launched new media platform, they need to job related performance but also to help their purchasing behavior including obtainin g brand related information. Also, perceived usefulness strongly influenced attitude

PAGE 198

198 toward Twitter and further actual usage including frequency of brand information alternat ive brand information spreading tool, marketers should recognize the importance of perceived usefulness construct to predict individual attitude toward Twitter, frequency of behavior in obtaining brand related information Importance of perceived ease of u se in new media adoption Findings of this study also suggested the importance of perceived ease construct r particularly for behavior in obtaining brand related information. This study found that perceived ease of use from can be concluded for media marketer s When media companies launched new media platform, t hey should consider design various context. For example, the interface of new media platform should be simple and intuitive for individuals. Also, regarding brand information spreading tool, marketers should recognize the importance of perceived ease of use construct to predict individual attitude toward Twitter and frequency of behavior in obtaining brand related information and design marketing campaigns acc ordingly Industrial Implications In this section, in terms of Twitter adoption results addressed the importance of Twitter and their usage. Specifically, the author illu strated how theoretical findings linked

PAGE 199

199 For practical purpose, it is note that too methodology oriented terms are not discussed here. Importance of perceived usefulness in new media adoption Considering perceived usefulness in Twitter adoption and usage, our study confirmed previous literatures that emphasizing perceived usefulness to predict designer and marketer. When media companies launched new media platform, they boost thei r job related performance but also to help their purchasing behavior including obtaining brand related information. Also, perceived usefulness strongly influenced attitude toward Twitter and further actual usage including frequency of brand information obt alternative brand information spreading tool, marketers should recognize the importance of perceived usefulness construct to predict individual attitude toward Twitter, frequenc y of behavior in obtaining brand related information as well. Importance of perceived ease of use in new media adoption Findings of this study also suggested that the importance of perceived ease usage of Twitter particularly for behavior in obtaining brand related information. This study found that perceived ease of ves, two strategic implications obtained for media platform designer and marketer. When media companies launched new media platform,

PAGE 200

200 example, the interface of new media platform should be simple and intuitive for brand information spreading tool, marketers should recognize the importance of perceived ease of use construct to predi ct individual attitude toward Twitter and frequency of behavior in obtaining brand related information. Also the findings in this study will be of interest to Internet technology based new media platform designers as well as managers. A better understandi ng of the help designers create attractive new media services. This will lead to favorable consumer attitudes, the success of these services, and, consequently, increased co nsumer usage. eWOM Related I mplication Theoretical Implication Regarding the academic contribution of eWOM from the perspective of Twitter, the results of this study addressed six theoretical implications. First, the study empirically for the first time test the theoretical background of social influence to obtain brand related information using general sample. Second, the results show that perceived similarity was an important predictor of attitude toward a brand wherea intention. Third, this study validated the importance of perceived credibility both for attitude toward the brand and eWOM spreading intention. Fourth, this study investigated the product category of utilitari behavior in Twitter for the first time Fifth, our study employed perceived fit of utilitarian providing a

PAGE 201

201 different result from previous stud ies or eWOM spreading intention. More specifically, this study investigated the significant prediction role of p erceived credibility, perceived similarity, product category (hedonic dimension), and perceived fit purchase intention using a real consumer sample. This finding illustrates consumer information obtaining and spreading behavior Traditionally, attribution theory indicates that consumers form preferences about a product after considering information a bout stimulus (e.g., brand, advertisement), (perceived fit in this case), or a combination of these three factors (Kelley, 1973; Laczniak et al., 2001). This study examined h ow different influences of online brand information go beyond traditional eWOM perspectives and attribution theory that solely investigated each of stimulus, person and circumstance that affect online consumer behavior. Our findings based on perspective o f eWOM in the consumer behavior by exhibiting the different impact of perceived similarity, perceived credibility, product category and perceived fit. Specifically, previous researchers found that higher source similarity and source credibility independent brand and his or her purchase intention. This study empirically tested the simultaneous effects of perceived source similarity and perceived source credibility and added

PAGE 202

202 perceived product and fit dimensions to o btain a model with more depth to explain consumer behavior in the context of social media (i.e., Twitter in this case). Interestingly, perceived similarity and perceived credibility had different influences in terms of a d eWOM spreading intention. Specifically, perceived similarity only influenced a favorable attitude toward the brand whereas perceived credibility had a positive impact on both the brand attitude and eWOM spreading intention. This provides a clue to help u s understand when and why a consumer responds differently to a marketing message provided via Twitter. Consumers prefer a communicator who is similar to them and this improves their attitude toward a brand whereas they exhibit a greater intention to spread product information when they receive information from an expert. For example, when a company launches a new brand or product and wants to foster a favorable attitude toward this brand, it is more effective to use a similar source in the stage when it is building brand awareness. On the contrary, if a marketer wants to spread brand information, it is better to use an expert representative or a medium such as an official page for its brand or use a third party endorsement. Testing consumer eWOM behavior in Twitter context with actual usage using general sample Although several studies have investigated the eWOM behavior and reconfirmed the importance of perceived similarity (Prendergast et al., 2010; Wangenheim & Bayon, 2004, 20 07), perceived credibility (Bansal & Voyer, social media context has rarely been tested empirically. In addition, this study adopted the constructs of utilitarian vs. hedonic prod uct differentiation, perceived fit of utilitarian /

PAGE 203

203 hedonic dimensions, and consumer demographic characteristics in the testing of brand related behavior in the context of Twitter. D ifferentiation from previous studies, this study measured three construct s of dependent variables such as attitude toward the brand, eWOM intention and purchase intention. Although previous study revealed a consisten t effect of attitude toward the brand, eWOM intention and purchase intention, some studies have indicated the pos & Fishbein, 1977; Fishbein & Ajzen, 1975; Perloff, 2010). Therefore, this study analyzed various factors affecting attitude toward the new media, Twitter and their actua l usage to prov ide more in depth understanding of consumer behavior on Twitter particularly in behavior in obtaining brand related information. T o enhance external the validity of this study, the survey rather than experiment research method were em ployed using general consumer sample. Note that most previous literatures in the eWOM context (e.g., Lee & Youn, 2008; Lee et al., 2009) adopted the experiment method. Experiment has advantage for testing causal relationship but has limitation in in exter nal validity. Thus this study used the survey method to obtain in depth and more real ized behavior within Twitter context. In detail, as a dependent variable, attitude toward the eneral attitude of brand that obtained information in Twitter rather than specific product or brand that commonly used in an experiment setting. Similarly, other major dependent variables such as eWOM intention and purchase intention also were measured by perception when they encountered brand related information in Twitter.

PAGE 204

204 Also, this study recruited general consumer s in the United States Previous studies in the context of eWOM have used student sample s that might lack extern al validity due to sampling issues in using student sample s (e.g., Chu & Kim, 2011). T his study used general population sample in the main test although student sample was employed in pre test for testing validity and reliability. The results of principal component analysis, exploratory factor analysis and confirmatory factor analysis yielded almost identical questionnaire items. However there is a slight difference between pretest results that using student sample and main test that using general consumer. Thus, it is necessary to recognize the difference between student sample and general sample in the interpretation of results in eWOM and brand literature. Importance of perceived similarity in attitude toward the brand and eWOM intention Based on previou s literature, our study provided the in depth understanding of online consumer behavior particularly brand related information processing in Twitter. T his study contributed to the traditional source similarity literature one of the most important factors in various marketing, advertising and communication area s Although favorable attitude toward the brand and help consumers lead purchase intention, the degree or direction o f the effects of source similarity could diverse depend on the context. The results of this study indicated that the effect of source similarity could be perceived simi larity, one of the most important variables in the context of WOM and eWOM study (e.g., Brown & Reingen, 1987; Gilly, Craham, Wolfnbarger, & Yale, 1998;

PAGE 205

205 Price, Feick, & Higie, 1989; Wangenheim & Bayon, 2004) regarding communicator characteristics. The anal yses of SEM indicated that perceived source similarity of ; however this favorable attitude toward the brand was not associated with eWOM intention. Therefore, the result here c hallenges the notion of a relationship between and eWOM intention that has been frequently discussed both in WOM and eWOM literatures. This result might be explained by the concept of the need for uniqueness that was pointed out as a cost of WOM behavior which creates a preference for distinct and unique product comparisons with common products (Bloch, 1995; Lynn & Harris, 1997; Simonson & Nowlis, 2000 ; Snyder & Fromkin, 1977; Tian et al., 2001). W hen consumers obtain brand related information through Twitter, this might hel p foster a positive feeling toward the brand. However, it might also cause the dilution of ati, 2010). Therefore, although consumer exhibited favorable attitude toward Twitter, they might be reluctant to spread eWOM since it could dilute their possession of uniqueness Importance of perceived credibility in attitude toward the brand and eWOM in tention Consistent with previous marketing and communication literature, the positive effect of perceived credibility between attitude toward the brand and eWOM intention was strongly supported by this study. As o ne of the most important findings, perceive d credibility was a strong predictor among the tested variables that produced high path coefficient and t value through the SEM.

PAGE 206

206 As described earlier, this study modified the traditional perceived credibility related activ ity within the Twitter context to create appropriate measurement items and modif y pre existing items. Our study contributed to the traditional concept of source credibility which has been consistently pointed out as one of the most important factors infl uencing s such as advertising, marketing and communication. This study also found perceived source credibility to be highly associated with attitude toward the brand and thei r purchasing intention. That is, when consumers perceived the eWOM brand information sender are reliable, they exhibited more favorable attitude toward the brand and more p ositive intention to purchase. Importance of product category in attitude toward th e brand and eWOM intention Our study suggested product category differentiation of utilitarian and hedonic brand and purchase intention within the Twitter environment. A s an exploratory investigation of product category differentiation from an eWOM perspective in the Twitter context, this study contributed to existing literature in two areas : reconfirming the traditional utilitarian vs. hedonic product differentiation and the effect of product category toward brand attitude and purchase intention. First, previous literatures have reviewed the utilitarian vs. hedonic differentiation from two aspects: shopping motivation (Barta & Ahtola, 1990) and product differentiation ( Babin & Darden, 1995; Babin et al., 1994; Cha, 2009a). This study adopted the product differentiation of utilitarian vs. hedonic perspective rather than the shopping motivation perspective. The results of principal component analysis

PAGE 207

207 successfully different iated the utilitarian product category (healthcare services, finance services, automobiles, computer equipment, and telecommunication services) from the hedonic product category (movies, restaurants, clothes, electronics and travel information) Further ex ploratory factor analysis and confirmatory factor analysis also reconfirmed that differentiation. Thus, our finding supports the notion of product differentiation from marketing literature in that varied depend ent on th e product category. Second, based on the product category of utilitarian vs. hedonic differentiation, the effects of each product category were tested through SEM. The results of SEM indicated that only hedonic product category was positively associated w eWOM spreading intention. That is, as an exploratory investigating of eWOM behavior within Twitter context, hedonic products carried more weight in affecting eWOM behavior specifically when the consumers received brand related inf ormation. Thus, our study contributes to the eWOM research within the Twitter context, specifically through the adoption of brand related variables by demonstrating appropriate product category for spreading brand related eWOM m essage in Twitter. Importan ce of perceived fit in attitude toward the brand and eWOM intention Consistent with previous marketing literature, the effects of perceived category fit between Twitter and each product category (utilitarian vs. hedonic) was strongly supported by this stud y. As mentioned earlier, this study adopted the logic of category fit (Cha, 2009a) rather than traditional parent brand extended brand fit since the category fit perspective was more appropriate for analyzing Internet based service extension such as shop ping behavior on SNSs (Cha, 2009a).

PAGE 208

208 Theoretically, this study reconfirmed the importance of perceived fit in the category extension context for the utilitarian product category H igher perceived fit lead s to higher intention for eWOM spreading intention a nd hedonic dimensions of perceived fit fostered both favorable attitude toward the brand and eWOM intention. As described earlier, this different effect of utilitarian vs. hedonic dimension might be explained by the concept of brand extendibility (Hagtvedt & Patrick, 2009) which posited that the more hedonic associations of product lead to more favorable attitude toward brand extension. That is, when consumers are exposed to hedonic product information, they tend to exhibit more favorable attitude toward t he brand and more willingness to purchase than utilitarian product information. Indeed, the results of SEM indicated that perceived fit of hedonic product category successfully predicted both attitude toward the brand and eWOM spreading intention whereas p erceived fit of utilitarian product category was only associated with eWOM spreading intention. Adopting the importance of perceived fit, this study also noted the concept of Specifically when com panies spread brand related information in Twitter, there might be a need for a series of statements about the possible connections or links between Twitter and each product category. Indeed, Bridges et al., (2000) indicated that when participants were exp osed to the exploratory link condition, they perceived a higher level of perceived fit comparing with no exploratory link conditions. Therefore, if a company insert s exploratory link s to their brand related information in Twitter, it would increase consume perceived fit between Twitter and the product category The correlation may ultimately foster favorable attitude toward the brand and increas e purchase intention.

PAGE 209

209 Consumer demographic in attitude toward the brand, eWOM intention and pur chase intention This study also conducted regression analysis and ANOVA to investigate the role of the demographic variables such as age, education level and gender. Although some previous studies identified the importan ce role of certain demographic varia bles in our regression analysis and ANOVA failed to show any significance in this study. That is, demographic characteristics no longer influenced both Unlike of the correlate with attitude toward the brand, eWOM intention or purchase intention. Industrial Implication In this section, based on the empirical finding s the industri al application is elaborate d five strategic implications are derived from the results. Specifically the potential strategic usage of Twitter as a marketing tool strategic use of perceived similarity, strategic use of perceived credibility, strategic use for product category and strategic use of perceived fit were elaborated. Strategic use of Twitter as an alternative marketing tool Corresponding with industrial reports (Neilson, 2011; Webster, 2010, 2011), Twitter may be an effective tool for communicat ing brand information with consumers The number of Twitter accounts may reached 200 million in recent years (Bennett, 2011) more than 200 million tweets were sent per day, including brand/product information; this is an enormous increase from only 65 mill ion tweets per day in 2010 (Twitter, 2011). Twitter is not only a medium for communicating with peers and acquaintances but also a tool for obtaining information. For instance, recent articles

PAGE 210

210 from The New York Times and Business Insider indicated that people actively seek information via Twitter rather than a news web site (Hamburger, 2011), and the number of people who obtain information from Twitter rather than Google has increased (Miller, 2011). More importantly, when compared with other social me dia such as Facebook or LinkedIn Twitter has a relative advantage for marketing purposes. For example, it is noteworthy that 42% of Twitter users utilize Twitter to learn about products/services, 41% of Twitter users provide opinions about products/serv ices, 31% of users ask for opinions about products/services, and 21% users even purchase certain products and services on Twitter (Bennett, 2011). Indeed, this study also found that 69.5 % (n = 317) of respondents obtained brand related information on Twit ter, whereas 30.5% (n = 136) did not use Twitter as a brand information tool. Therefore, when marketing practitioners formulate marketing strategy, Twitter should be seemed as an effective platform for conv eying brand information. Strategic use of perceiv ed similarity When marketer designed brand related message in Twitter, they should consider message. The results of this study revealed that perceived source similarity did However, no relationship between was found This result implies that marketers should consider either the effectiveness of eWOM and the cost of eWOM simulta neously. Consumers were generally more favorable to brand information from a communicator who is similar to them However, in certain circumstance or certain product category, similarity between consumers and communicator s might cause

PAGE 211

211 ss by the marketer s when they develop marketing strategy in the Twitter environment. For example, it might be reasonable to avoid spreading brand information of luxury product in Twitter since luxury or hig h image brand s need a great amount of uniqueness that differentiates themselves in consumers mind On the contrary, though it has somewhat hedonic dimensions, everyday product such as restaurant information or travel information would be appropriate in th e Tweeting environment through a brand information spreading mechanism Strategic use of perceived credibility Regarding the effect of perceived credibility on brand attitude, eWOM intention, ived credibility was an important factor affecting consumers attitude toward the brand and eWOM intention. Therefore, in terms of designing a brand information delivery strategy, marketers should consider devising credible messages rather than "like me" messages. For example, if a marketer wishes to spread brand information in a tweet, the representative of the message should be considered credible rather than similar to the consumer. Possible directions for boosting credibility for brand related informat ion include using third party endorsements (Dean & Biswas, 2001) or another credible representative, such as an expert (Gotlieb & Sarel, 1991; Homer & Kahle, 1990; Yoon et al, 1998). Strategic use of utilitarian vs. hedonic product category From a marketin g perspective, the finding here shows the utility of Twitter as a branding platform for products with dominant hedonic nature. Again, Twitter seems to provide the best environment of hedonic product marketing. Also regarding product category, marketers sh ould note that hedonic product information was more effective

PAGE 212

212 than utilitarian product information when discussed in the Twitter environment. Therefore, when devising a marketing strategy or advertising media strategy for Twitter, practitioners should cons ider designing the appropriate messages considering the type of product category for which to spread brand related information Strategic use of perceived fit Industry practitioners should also recognize the importance of perceived fit in lead ing to more fa vorable attitude toward the brand, eWOM intention and purchase intention particularly considering the different effects of perceived fit of utilitarian vs. hedonic product. Our study showed that perceived fit of utilitarian product category and hedonic pro duct category were important predictor of consumers eWOM behavior in the Twitter context. Based on this finding, marketer should consider ways to improve exploratory link s. Limitations Limitations for Twitter Adoption Study Several limitations of this study should be acknowledged in terms of Twitter adoption related study. This study employed the specific SNS context of Twitter. Although Twitter was selected based on ce rtain criteria, including the effectiveness of using Twitter for marketing purposes, Twitter has a relatively small number of users compared with Facebook the most prominent SNS. This study did not cover a broader scope of social media, such as Wikipedia or YouTube Therefore, the results of our study are limited in terms of generalizability in the context of SNS s Also, in terms of sample of this study, several limitations should be addressed. First, although this study employed general sample rather t han student sample, the

PAGE 213

213 sample size was limited in comparing differences between groups through SEM. Bandalos (1997) suggested that a minimum sample to compare group differences is 150 whereas our study included a total of 307 sample. Thus, though ANOVA wa s conducted to compare gender difference, considering females were dominant (75.6%, n = 233), whereas 24.0% of participants were males (n = 74) in the analysis, SEM could not be executed. Second, this study employed consumer panel to recruit participants b y online survey. Therefore, due to the nature of online survey and the use of online consumer panelists who were somewhat compensated for completing the survey, it is possible that the subjects might not be representative of the general users of Twitter us age. Third, this study employed only current user of Twitter through a qualifying question. Therefore, this study did not include non adopter of Twitter that could measure adoption intention rather than actual usage. In addition to the sampling issue, the adoption stage of Twitter should be addressed. Although Twitter has emerged as an accepted med ium the notion of adoption stages adoption behavior highly influenced their adoption stage of new media. However, this adoption stage. Thus, consumer characteristics such as innovativeness did not included in this study. Also, different dimensions of s ocial influence were not considered in this study. Previous literatures indicated that there might be a different social influence including peer influence. However, this study solely focused on the social influence of conformity to subjective norms.

PAGE 214

214 Last ly, individual difference s including cultural difference, involvement and issue relevance were not consider ed in our study. Limitations for eWOM Related Study Among limitations in terms of the eWOM study sampling issues were similar to that of the Twitte r adoption study Specifically, this study only employed one SNSs context of Twitter also this study recruited a limited sample to compare groups through SEM. Finally our study used a consumer panel through an online survey. For the eWOM specific study r elated limitations, this study inquired about general perceptions of brand and purchase intentions for products featuring brand information. This might result in potential bias that threatens the external validity. Also, this study mainly focused the eWOM related information on Twitter. Therefore, this study can answer the question about how to veness but not who is most influenced by eWOM message. Suggestions for Future Research Future Research Direction for New Media Adoption As far as future research is concerned, to increase external validity, several replication studies should be conducted For example, the total sample size of this study was somewhat limited when considering group differences using structural equation modeling. Although this study conducted a one way ANOVA test for investigating gender differences in both Twitter adoption and eWOM related research, Bandalos (1997) suggested that the minimum sample to compare group differences is 150 for

PAGE 215

215 SEM. However, as previously mentioned, in this study, females were dominant (75.6%, n = 233) in the analysis. Next, regarding the variable of social influence specifically related to Twitter adoption, future research could include another dimension of social influence, such as peer influence. Indeed, Lin et al. (2009) revealed the effectiveness of both social influence and peer influence in Although this study did not find a predicting role of conformity to a subjective norm, future research should revisit a variety of social influences, including peer influence. Also considering the fact tha t this study only employed survey method to investigate the social influence, alternative method including network analysis for measuring accurate social influence should be revisited (e.g., Kantona et al., 2011). Cultural differences also provide a promis ing future research direction (Straub et al., 1997; Yang et a l. 2011). Indeed, in terms of application for TAM, Straub et al. (1997) indicated that TAM could be varied depend on geographical region (Japan, Switzerland and United States) in terms of predic ting technology use. Also, focusing SNSs, Yang et al. (2011) revealed that there was a regional difference among and China) were relatively more active than Western coun tries (United States and In addition, different adoption stages that comparing adopter vs. non adopter should be also revisited (Jung et al., 2011). Future Research Dir ection for eWOM Research For the eWOM related study, this study inquired about the general perceptions of brand and purchase intentions for products featuring brand information. This might

PAGE 216

216 result in potential bias that threatens internal validity. Alternat ive methods, such as an experimental design with a specific product category or brand name, or field experiments in a natural setting, would be useful for future research and improving internal validity. Another promising direction for future study is to i nvestigate the role of individual differences --in determining the consumer profile of those considered most influential and important for eWOM marketing. For example, Price, Feick and Guskey (1995) revealed a relationship between marketplace involvement, altruistic values, and traditional WOM behavior. The results indicated that people who held more altruistic values and involvement exhibited more intention for WOM behavior. To apply these results to eWOM beha vior in a social media context, it would be useful to investigate individual differences for understanding online consumer behavior in Twitter. Also, another marketing utilization for eWOM should be considered in future research. This study employed SNSs specifically Twitter as an eWOM marketing tool, however, there might be various media or a combination of them in eWOM to build effective marketing strategies. For example, Hong and Rim (2010) found that using corporate web sites positively influenced a c onsumer WOM intention. SNSs Google+ or corporate blog should be investigated in near future.

PAGE 217

217 APPENDIX A PRETEST QUESTIONNAIRE T hank you for agreeing to participate in t his brief survey. I am Hyunsang Son, a graduate student in the College of Journalism and Communications at the University of Florida. Your input for this study about consumers' attitude toward Twitter is much appreciated. A ll responses from this survey w ill be anonymous and used for academic research only. Please respond to the following questions. (Please remember all of your answers are completely anonymous.) 1. Have you ever get brand information from your Twitter friends? (Both from person and comp any account) Yes No Very All the Rarely Time 2. How often do you get information about a brand from your Twitter friends? _____:____ _:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 Please indicate how much you agree or disagree with each of the following statements. Strongly Stro ngly Disagree Agree 1. Generally speaking, I would do what my group members think I should do. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 2. Generally speaking, I would do what others think I Should do in the online environment. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 3. Generally speaking, I wou ld do what my Twitter friends think I should do in the Twitter environment.. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 Please indicate how much you agree or disagree with each of the following statements.

PAGE 218

218 Strongly Strongly Disagree Agree 1. Use of Twitter enables me to accomplish tasks more quickly. _____:___ __:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 2. Use of Twitter improves my performance. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 3. Use of Twitter to obtain product information increases my productivity. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 4. Use of Twitter enhances the effectiveness in product inf ormation search. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 5. Use of Twitter makes it easier to obtain product information _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 6. I find Twitter useful in my life _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 Please indicate how much you agree or disagree with each of the following statements. Strongly Strongly Disagree Agree 1. Tweet, Mention and Retweet on Twitter is easy _____:_____:_____:_____ :_____:_____:_____ 1 2 3 4 5 6 7 2. Learning to use Twitter is easy for me _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 3. It is ea sy to get information on Twitter _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 4. I find Twitter to be flexible to interact with _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 5. It is easy for me to become skillful at using Twitter _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 6. I find Twitter eas y to use _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 Please indicate how much you agree or disagree with each of the following statements.

PAGE 219

219 Strongly Strongly Disagree Agree 1. Twitter makes it easy for me to build a relationship with the online community _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 2. I would like to communicate with my Twitter friends again in the future. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 the services provided by Twitter _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 4. I feel comfortable using Twitter _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 5. I feel surfing on Twitter is a good way for me to spend my time _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 How many days during a t ypical week do you use Twitter? Never One day Two days Three days Four days Five days Six days Seven days How many hours during a typical day do you use Twitter? Never Less than 10 minutes 10 30 minutes 30 mi nutes 1 hour 1 2 hours 2 3 hours 3 4 hours More than 4 hours a day The following questions concern only your Twitter friends who tweet "brand related information."

PAGE 220

220 Please indicate how much you agree or disagree with ea ch of the following statements. Strongly Strongly Disagree Agree 1. In terms of outlook on life, my Twitter friends are similar to me _____ :_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 2. In terms of likes and dislikes, my Twitter friends are similar to me _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 3. In terms of values and experiences, my Twitter friends are similar to me _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 4. In terms of taste s for products, my Twitter friends are similar to me _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 5. In terms of preferences and value, my Twitter friends are similar to me _____ :_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 6. Overall, my Twitter friends are similar to me _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 Please indicate how much you agree or disagree with each of the following statements. Strongly Strongly Disagree Agree 1. I feel th e tweeted product information given by my Twitter friends is strong _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 2. I feel the tweeted brand information given by my Twitter fri ends is convincing _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 3. I feel the tweeted brand information given by my Twitter friends is persuasive _____:_____:_____:_____:_____:_____:_ ____ 1 2 3 4 5 6 7 4. I feel the tweeted brand information given by my Twitter friends is powerful. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 5. My Twitter friends have knowledge about compute r equipment in general. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 6. My Twitter friends is an expert in the area of computer equipment. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 7. My Twitter friends have knowledge about _____:_____:_____:_____:_____:_____:_____

PAGE 221

221 restaurants in general. 1 2 3 4 5 6 7 8. My Twitter friends is an expert in the area of restaurants _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 Please indicate how much you agree or disagree with each of the following statements. Strongly Strongly Disagree Agree 1. I often try to obtain product information abou t computer equipment _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 2. I often try to obtain product information about clothes _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 3. I often try to obtain product information about finance product / services _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 4. I often try to obtain product information about travel information. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 5. I often try to obtain product information about restaura nts. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 6. I often try to obtain product information about telecommunication services. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 7. I often try to obtain product information about automobiles _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 8. I often try to obtain product information about movies _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 9. I often try to obtain product information about health care providers _____:_____:_____: _____:_____:_____:_____ 1 2 3 4 5 6 7 10. I often try to ob tain product information about electronics. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 Please indicate how much you agree or disagree with each of the following statements Strongly Strongly Disagree Agree 1. T witter is a good medium to learn about computer equipment _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 2. Twitter is a good medium to learn about clothes _____:_____:_____:_____:_____ :_____:_____ 1 2 3 4 5 6 7

PAGE 222

222 3. Twitter is a good medium to learn about finance product / services _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 4. Twitter is a good medium to learn about travel information. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 5. Twitter is a good medium to learn about restaurants. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 6. Twitter is a good medium to learn about telecommunication services. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 7. Twitter is a good medium to learn about automobiles _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 8. Twitter is a good medium to lea rn about movies _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 9. Twitter is a good medium to learn about health care providers _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 10. Twitter is a good medium to learn about electronics. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 Please indicate how much you agree or disagree with each of the following statements. Strongly Strongly Disagree Agree 1. If I find interesting product information on the Twitter, I want to Retweet it to my friends after regarding the tweeted brand information from T witter friends _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 2. If somebody asks me for advice about an interesting product information, I will encourage him or her to Tweet after reading the tweeted brand information from twitter friends _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 3. I would recommend my friends and family to Tweet or Retweet in interesting product related information after reading the tweeted brand i nformation from twitter friends _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7

PAGE 223

223 In order to effectively evaluate the survey responses, please answer the following questions about yourself. Please answer the following questions by filling in the blank or checking one option. Gender: Male Female Age: ________________ Year in School: Freshmen Junior Senior Graduate Other (Please specify __________________ ) Major: Advertising Journalism Public Relations Telecommunication Other (Please specify __________________ ) Ethnic ity: Arabic Asian Black/African American Caucasian Hispanic/Latino Other Thank you for your interest and participation in this important study. You will not be identified individually within the survey, and any information you provide w ill remain strictly anonymous. If your instructor offers extra credit for participating in this study, however, we ask that you provide the following information. We will forward it to your instructor for only extra credit.

PAGE 224

224 First, Please indicate y our UFID ______ ______ ______ _______ _____ ______ ______ ______ Please indicate your UFL e mail address (@ufl.edu). ______________________________________ Please enter prefix and number of your class (e.g., RTV 6508). ____________________________ ___________ Thank you for participation. Your opinions are extremely important to us, Again, we appreciate your valuable time.

PAGE 225

225 APPENDIX B MAINTEST QUESTIONAIRE T hank you for agreeing to participate in this brief survey. I am Henry Son, a gr aduate student in the College of Journalism and Communications at the University of Florida. Your input for this study about consumers' attitude toward Twitter is much appreciated. A ll responses from this survey will be anonymous and used for academic research only. No email or IP addresses or other identifying information will be connected to your responses at any time. 1. Have you ever get brand information from your Twitter friends? (Both from person and company account) Yes No Very All the Rarely Time 2. How often do you get information about a brand from your Twitter friends? _____:____ _:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 Please indicate how much you agree or disagree with each of the following statements. Strongly Stro ngly Disagree Agree 1. Generally speaking, I would do what my group members think I should do. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 2. Generally speaking, I would do what others think I Should do in the online environment. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 3. Generally speaking, I wo uld do what my Twitter friends think I should do in the Twitter environment.. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7

PAGE 226

226 Please indicate how much you agree or disagree with e ach of the following statements. Strongly Strongly Disagree Agree 1. Use of Twitter enables me to accomplish tasks more quickly. _____:____ _:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 2. Use of Twitter improves my performance. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 3. Use of Twitter to obtain p roduct information increases my productivity. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 4. Use of Twitter enhances the effectiveness in product info rmation search. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 5. Use of Twitter makes it easier to obtain product information _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 6. I find Twitter useful in my life _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 Please indicate how much you agree or disagree with each of the following statements. Strongly Strongly Disagree Agree 1. Tweet, Mention and Retweet on Twitter is easy _____:_____:_____:____ _:_____:_____:_____ 1 2 3 4 5 6 7 2. Learning to use Twitter is easy for me _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 3. It is e asy to get information on Twitter _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 4. I find Twitte r to be flexible to interact with _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 5. It is easy for me to become skillful at using Twitter _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 6. I find Twitter ea sy to use _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7

PAGE 227

227 Please indicate how much you agree or disagree with each of the following statements. Strongly Strongly Disagree Agree 1. Twitter makes it easy for me to build a relationship with the online community _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 2. I would like to communicate with my Twitter friends again in the future. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 h the services provided by Twitter _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 4. I feel comfortable using Twitter _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 5. I feel surfing on Twitter is a good way for me to spend my time _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 How many days during a t ypical week do you use Twitter? Never One day Two days Three days Four days Five days Six days Seven days How many hours during a typical day do you use Twitter? Never Less than 10 minutes 10 30 minutes 30 mi nutes 1 hour 1 2 hours 2 3 hours 3 4 hours More than 4 hours a day

PAGE 228

228 The following questions concern only your Twitter friends who tweet "brand related information." Please indicate how much you agree or disagree with each of the following statements (remember the statements only apply to those Twitter friends who tweet brand related info). Strongly Strongly Disagree Agree 1. In terms of outlook on life, my Twitter friends are similar to me _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 2. In terms of likes and dislikes, my Twitter fri ends are similar to me _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 3. In terms of values and experiences, my Twitter friends are similar to me _____:_____:_____:_____:_____:_____:___ __ 1 2 3 4 5 6 7 4. In terms of tastes for products, my Twitter friends are similar to me _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 5. In terms of preferences and value, my Twitter friends are similar to me _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 6. Overall, my Twitter friends are similar to me _____ :_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 Please indicate how much you agree or disagree with each of the following statements. Strongly Strongly Disagree Agree 1. I feel the tweeted product information given by my Twitter friends is strong _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 2. I feel the tweeted brand information given by my Twitter friends is convincing _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 3. I feel the tweeted brand infor mation given by my Twitter friends is persuasive _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 4. I feel the tweeted brand information given by my Twitter friends is powerful. _____:_ ____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7

PAGE 229

229 5. My Twitter friends have knowledge about computer equipment in general. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 6. My Twit ter friends is an expert in the area of computer equipment. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 7. My Twitter friends hav e knowledge about restaurants in general. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 8. My Twitter friends is an expert in the area of restaurants _____:_____:_____:_____:_____ :_____:_____ 1 2 3 4 5 6 7 Please indicate how much you agree or disagree with each of the following statements. Strongly Strongly Disagree Agree 1. I often try to obtain product information about computer equipment _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 2. I often try to ob tain product information about clothes _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 3. I often try to ob tain product information about finance services _____:_____:____ _:_____:_____:_____:_____ 1 2 3 4 5 6 7 4. I often try to ob tain product information about travel information. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 5. I often try to o btain product information about restaurants. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 6. I often try to ob tain product information about telecommunication services. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 7. I often try to ob tain product information about automobiles _____:_____:_____:_____:_____:_____: _____ 1 2 3 4 5 6 7 8. I often try to ob tain product information about movies _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 9. I often try to ob tain product information about health care services _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 10. I often try to ob tain product information about electronics. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7

PAGE 230

230 Please indicate how much you agree or disagree with each of the following statements Strongly Strongly Disagree Agree 1. Twitter is a good medium to learn about computer equipment _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 2. Twitter is a good medium to learn about clothes _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 3. Twitter is a good medium to learn about finance services _____:___ __:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 4. Twitter is a good medium to learn about travel information. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 5. Twitter is a good medium to learn about restaurants. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 6. Twitter is a good medium to learn about tele communication services. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 7. Twitter is a good medium to learn about automobiles _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 8. Twitter is a good medium to learn about movies _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 9. Twitter is a good medium to learn about health care services _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 10. Twitter is a good medium to learn about electronics. _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 Please indicate how you feel about brand information in general on my Twitter timeline Unfavorable _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 Favorable Bad _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 Good Dislike _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 Like Negative _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 Positive

PAGE 231

231 Please indicate how much you agree or disagree with each of the following statements. Strongly Strongly Disagree Agree 1. If I find interesting product information on the T witter, I want to Retweet it to my friends after regarding the tweeted brand information from T witter friends _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 2. If somebody asks me for advice about an interesting product information, I will encourage him or her to Tweet after reading the tweeted brand information from twitter friends _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 3. I would recommend my friends and family to Tweet or Retweet in interesting product related information after reading the tweeted brand information from twitter friends _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 After considering the product information on my Strongly Strongly Disagree Agree 1. It is very likely that I will buy the product _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 2. I will definitely try the product _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 3. I will purchase the product next time I need a product _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7 4 Suppose that a friend called you last night to get you advice in his/her search for a product. Would you have recommended him/her to buy the product? _____:_____:_____:_____:_____:_____:_____ 1 2 3 4 5 6 7

PAGE 232

232 In order to effectively evaluate the survey responses, please answer the following questions about yourself. Ple ase answer the following questions by filling in the blank or checking one option. 1. What is your g ender ? Male Female 2. What is your current age? 18 to 19 20 to 24 25 to 34 35 to 44 45 to 54 55 to 64 65 or over 3. What is the highest l evel of education you have completed? Less than High School High School / GED Some College 2 year College Degree 4 year College Degree Masters Degree Doctoral Degree Professional Degree (JD. MD)

PAGE 233

233 4. Which occupational category best describes you r employment? Management: professional or related occupations Management: business or financial operations occupations Management occupations, except farmers and farm managers Farmers and farm managers Business and financial operations Business ope rations specialists Financial specialists Computer or mathematical Architects, surveyors, cartographers, or engineers Drafters, engineering, or mapping technicians Community and social services Legal Education, training, or library Arts, design, e ntertainment, sports, or media Health diagnosing or treating practitioners & technical occupations Health technologists or technicians Health care support Fire fighting, prevention or law enforcement workers, (including supervisors) Other protective service workers (including supervisors) 5. What is your annual salary (including b onuses and commissions) in U.S. dollars? $0 $25,000 $25,001 $50,000 $50,000 $75,000 $75,001 $100,000 $100,001 $125,000

PAGE 234

234 $125,001 $150,000 $150,001 $175 ,000 $175,001 $200,000 $200,001 + 6. What is your ethnicity? White / Caucasian African American Hispanic Asian Native American Pacific Islander Other 7. Gender: Male Female 8. Ethnicity: Arabic Asian Black/African American C aucasian Hispanic/Latino Other Thank you for participation. Your opinions are extremely important to us, Again, we appreciate your valuable time.

PAGE 235

235 APPENDIX C UNIVERSITY OF FLORIDA INSTITUTIONAL REVIEW BOARD INFORMED CONSENT APPROVAL Protocol Sub mission Form UFIRB 02 Social & Behavioral Research Protocol Submission Form This form must be typed. Send this form and the supporting documents to IRB02, PO Box 112250, Gainesville, FL 32611. Should you have questions about completing this form, call 352 392 0433. Title of Protocol: Adoption of Twitter and Its Effectiveness in e WOM Principal Investigator: Hyunsang Son UFID #: N/A Degree / Title: Mailing Address: (If on campus include PO Box address): N/A Email: hyunsangson@ ufl.edu Department: College of Journalism and Communication Telephone #: N/A Co Investigator(s): UFID#: Email: Supervisor (If PI is student) : Dr. Sylvia M. Chan Olmsted UFID#: Degree / Title: P h.D., / Associate Dean for Research., / Profess or Mailing Address: N/A Email : chanolmsted@jou.ufl.edu Department: D epartment of Telecommunication College of Journ alism and Communications Telephone #: N/A Date of Proposed Research: From May 9 th 2011 to June 9 th 2011. Source of Funding (A copy of the grant proposal must be submitted with this protocol if funding is involved): Unfunded

PAGE 236

236 Scientific Purpose of the Study: The study is designed to reveal the relationship s among various factors affecting evaluation of microblogging (Twitte r) adoption and factors affecting brand related information forwarding in the Twitter among general consumers This study analyzes whether response differently depending on perceived utility, perceived ease of use and consumer demographics (age, gender, be investigated in terms of different level of perceived similarity, source credibility, product category and perceived fit between Twitter and eac h item. Therefore whether attitude toward the Twitter and evaluation of brand and message spreading intention are differently affected by different consumer characteristics ( i.e., perceived utility, perceived ease of use, demographics, perceive d similarity, source credibility, product category and perceived fit ) will be examined in this research Applying these findings in the context of Twitter adoption and implication for potential marketing tool, the proposed research seeks to examine the rol e of different consumer characteristics. This study lays the theoretical groundwork for the new media adoption studies and electronic word of mouth (eWOM) perspective. Describe the Research Methodology in Non Technical Language: ( Explain what will be d one with or to the research participant. ) For this study a survey method will be conducted. adoption and their intention to spreading message in Twitter will be analyzed. The survey will be conducted through the national consumer panel with online access for its web based questionnaires. The participants will receive a small amount of money ($4) in the exchange of their participation. A total of 300 consumers randomly selected from a national consumer panels opera ted by leading market research received an email with the link to reach stimuli and questionnaires. The sample consisted the United States based consumers who have not participated during past two weeks and they rewarded small premium ($4) through the pane l company Participants will be e mailed a link to online consent form. When they agree to participate in the survey, they will be directed to an online survey. Upon consenting to take part in the study, participants will be asked to view the questionnaire and fill out their feeling attitude toward Twitter. Subjects will be anonymous. Describe Potential Benefits: A small amount of premium ($4) will be given on behalf of the consumer panel for participating in this study. Describe Potential Risks: ( If risk of physical, psychological or economic harm may be involved, describe the steps taken to protect participant.) The project should not create any physical, psychological or economic risks. Most of the scales used in the questionnaire are routinely used by marketing and communication scholars in their research. No risk associated with the questions has been reported. Describe How Participant(s) Will Be Recruited: The national consumer panel with online access for its web based questionnaires. The part icipants will receive a small amount of money ($4) in the exchange of their participation through the consumer panel company.

PAGE 237

2 37 Maximum Number of Participants (to be approached with consent) 300 Age Range of Participants: 18 65 yrs Amount of Compensation/ c ourse credit: $ 4 Describe the Informed Consent Process. (Attach a Copy of the Informed Consent Document. See http://irb.ufl.edu/irb02/samples.html for examples of consent.) The participants will re ad the posted consent statement. When they agree to participate in the survey, they are supposed to click the given link to the survey, indicating their willingness to participate. When they decide not to participate, they are supposed to click a link to t he University of Florida official web site. Refer to the attached consent form. Subjects will be anonymous. (SIGNATURE SECTION) Principal Investigator(s) Signature: Date: Co Investigator(s) Signature(s): Date: stud ent): Date: Department Chair Signature: Date:

PAGE 238

238 Informed Consent Protocol Title: Adoption of Twitter and Its Effectiveness in e WOM Please read this consent document carefully before you decide to participate in this study. Purpose of the research study: The purpose of this study is to examine your responses to adoption of Twitter and attitude toward brand, message spreading intention in Twitter. What you will be asked to do in the study: You will be asked to indicate your thoughts and fee lings about Twitter in a brief questionnaire. Time required : Less than 30 minutes Risks and Benefits : We do not anticipate there will be any risks or direct benefits to you as a consequence of your decision to complete the survey. Compensation : Small mo netary compensation will be given on behalf of the experimenter for participating in this study. Consumer Panel Company will reach you within 2 weeks after you have finished the study. Confidentiality : Your answers from this study will be anonymous. No na mes will be used in any part of the study. Your identity will be kept confidential to the extent provided by law. Voluntary participation : Your participation in this study is entirely voluntary. There is no penalty for not participating. You can choose not to answer any question you do not wish to answer. Right to withdraw from the study : You have the right to withdraw from the study at anytime without consequence. Whom to contact if you have questions about the study: Principle Investigator : Hyunsang College of Journalism and Communications hyunsangson@ufl.edu Whom to contact about your rights as a research participant in the study : UFIRB Office IRB02 Office, Box 112250, University o f Florida, Gainesville, FL 32611 2250; phone 392 0433

PAGE 239

239 Agreement: I have read the document stating the procedures to be used and followed in this study. I have received a copy of informed consent and AGREE to participate in the study.

PAGE 240

240 LIST OF REFERENCES Aaker, D. A., & Keller, K. L. (1990). Consumer evaluations of brand extensions, Journal of Marketing, 54( 1), 27 41. Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50 (2), 179 211. Ajzen, I. (2002). Perceived behavioral control, self efficacy, locus of control, and the theory of planned behavior. Journal of Applied Social Psychology, 32 (4), 665 683. Ajzen, I. & Fishbein, M. (2005). The influence of attitudes on behavior in Allbarrac in, D., Johnson, B. T., & Zanna, M. P. (Eds.). The handbook of attitudes, Lawrence Erlbaum Associates, Mahwah, NJ. 173 221. Ajzen, I., & Fishbein, M. (1977). Attitude behavior relations: A theoretical analysis and review of empirical research, Psychologica l Bulletin, 84 (5), 888 918. Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior. Englewood Cliffs, NJ: Prentice Hall. Akhter, S. H. (2003). Digital divide and purchase intention: Why demographic psychology matters. Jour nal of Economic Psychology, 24 (3), 321 327. Aldridge, A., Forcht, K., & Pierson, J. (1997). Get linked or get lost: marketing strategy for the internet, Internet Research: Electronic Networking Applications and Policy, 7 (3), 161 169. Allsop, D. T., Bassett B. R., & Hoskins, J. A. (2007). Word of mouth research: Principles and applications. Journal of Advertising Research, 47 (4), 398 411. Ames, T. (2009). Exclusive PQ media research: Despite worst recession in decades, brands increased spending on word of m outh marketing 14.2% to $1.54 billion in 2008. PQ Media Retrieved on February 21, 2011 from http://www.pqmedia.com/about press 20090729 wommf.html Amoako Gyampah, K., & Salam, A. F. (2004). An extension of the technology acceptance model in an ERP implementation environment. Information & Management, 41 (6), 731 745. Anderson, E. W. (1998). Customer satisfaction and word of mouth. Journal of Service Research, 1 (1), 5 17. Andrew, D., No nnecke, B., & Preece, J. (2003). Electronic survey methodology: A case study in reaching hard to involve Internet users. International Journal of Human Computer Interaction, 16 (2), 185 210.

PAGE 241

241 Armitage, C. J., Conner, M., Loach, J., & Willetts, D. (1999). Dif ferent perceptions of control: applying and extended theory of planned behavior to legal and illegal drug use. Basic and Applied Social Psychology, 21 (4), 301 316. Arndt, J. (1967). Word of mouth advertising and informal communication. In D. F. Cox (Ed.). Risk taking and information handling in consumer behavior Boston: Harvard University. Asch, S. E. (1951). Effects of group pressure upon the modification and distortion of judgment. In H. Guetzkow (Eds.), Group, leadership and men Pittsburgh, PA: Carneg ie Press. Atkin, D. J., & LaRose, R. (1994). An analysis of the information service adoption literature, in J. Hanson (Eds.) Advances in Telematics 91 101. Atkin, D. J., Neuendorf, K., Jeffres, L. W., & Skalski, P. (2003). Predictors of audience interest in adopting digital television. Journal of Media Economics, 16 (3), 159. Baaren, E., Wijngaert, L. V., & Hulzer, E. (2011). Understanding technology adoption through individual and context characteristics: the case of HDTV, Journal of Broadcasting & Electr onic Media, 55 (1), 72 89. Babbie, E. (2001) The practice of social research (9 th ed.). Belmont, CA: Wadsworth/Thomson Learning. Babin, B. J., Darden, W. R. (1995). Consumer self regulation in a retail environment, Journal of Retailing, 71 (1), 47 70. Babin B. J., Darden, W. R., & Griffin, M. (1994). Work and/or fun: Measuring hedonic and utilitarian shopping value. The Journal of Consumer Research, 20 (4), 644 656. Babrow, A. S. (1987). Student motives for watching soap opera. Journal of Broadcasting and El ectronic Media, 31 (3), 309 321. Bachmann, D., Elfrink, J., & Vazzana, G. (1996). Tracking the progress of e mail versus snail mail. Marketing Research, 8 (2), 31 35. Bandalos, D. L. (1997). Assessing sources of error in structural equation models: The effec t of sample size, reliability, and model misspecification. Structural Equation Modeling, 4 (3), 177 192. Bansal, H. S., & Voyer, P. A. (2000). Word of mouth processes within a services purchase decision context. Journal of Service Research, 3 (2), 116 176. B arker, V. (2009). Older adolescents' motivations for social network site use: The influence of gender, group identity, and collective self esteem. CyberPsychology & Behavior, 12 (2), 209 213.

PAGE 242

242 Basuroy, S., Chatterjee, S., & Ravld, S. A. (2003). How critical are critical reviews? The box office effects of film critics, star power, and budgets. Journal of Marketing, 67 (4), 103 117. Batra, R., & Ahtola, O. T. (1990). Measuring the hedonic and utilitarian sources of consumer attitude. Marketing letter, 2 (2), 159 179. Bennett, S. (2011, June). Is Twitter a waste of time? All Twitter Retrieved from http://www.mediabistro.com/alltwitter/twitter waste of time infographic_b10651?utm_source=feedburner&utm_medium=feed&utm_campaign =Feed%3A+twittercism+%28Twittercism%29 Bentler, P. M. (1990). Comparative fit indexes in structural models, Psychological Bulletin, 107 (2), 238 246. Berry, L. L. (1980). Services marketing is different, Business 30(5), 24 29. Bhat, S., & Reddy, S. K. (2001). The impact of parent brand attribu te associations and affect on brand extension evaluation. Journal of Business Research, 53 (3), 111 122. Bickart, B., & Schindler, R. M. (2001). Internet forums as influential sources of consumer information. Journal of Interactive Marketing, 15 (3), 31 40. Bloch, P. H. (1995). Seeking the ideal form: product design and consumer response, Journal of Marketing, 59 (3), 16 29. Bone, P. (1995). Word of mouth effects on short term and long term product judgments. Journal of Business Research, 32 (3), 213 223. Bous h, D. M., & Loken, B. (1991). A process tracing study of brand extension evaluation. Journal of Marketing Research, 28 (1), 16 28. Boush, D., Shipp, S., Loken, B., Gencturk, E., Crockett, S., Kennedy, E., Minshall, B., Misurell, D., Rochford, L., & Stobel, J. (1987). Affect generalization to similar and dissimilar line extension. Psychology & Marketing, 4 (3), 225 241. Bowman, D., & Narayandas, D. (2001). Managing customer initiated contacts with manufactures: The impact on share of category requirements and word of mouth behavior. Journal of Marketing Research, 38 (3), 291 297. Bracken, C. C. (2005). Presence and image quality: The case of high definition television. Media Psychology, 7 (2), 191 205. Brewer, M. B., & Gaertner, S. L. (2004). Toward reduction of prejudice: Intergroup contact and social categorization. In M. B. Brewer and M. Hewston (Eds.), Self and social identity Blackwell Publishing.

PAGE 243

243 Bridges, S., Keller, K. L., & Sood, S. (2000). Communication strategies for brand extensions: Enhancing perceiv ed fit by establishing exploratory links. Journal of Advertising, 28 (2), 1 11. 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. Brown, J. J., & Reingen, P. H. (1987). Social ties and word of mouth referral behavior. Journal of Consumer Research, 14 (3), 350 362. Brown, S. P., & Stayman, D. M. (1992). Antecedents and consequences of attitude toward the ad: A meta analy sis. Journal of Consumer Research, 19 (1), 34 51. Brown, T. J., Barry, T. E., Dacin, P. A., & Gunst, R. F. (2005). Spreading the word: of mouth intentions and behaviors in a retailing context. Journal of the Academy of Marketing Science, 33 (2), 123 138. Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In K. A. Bollen., & Long, J. S. (Eds.), Testing Structural Equation Models CA: Sage Publication. Brucks, M. (1985). The effects of product class knowledge on information search behavior. Journal of Consumer Research, 12 (1), 1 16. Buunk, A., & Gibbons, F. (2007). Social comparison: The end of a theory and the emergence of a field. Organizational Behavior and Human Decision Process, 102 (1), 3 21. Buunk, B. P., & Oldersma, F. L. (2001). Social comparisons and close relationship. In G. J. Fletcher & M.S. Clark (Eds.), Blackwell handbook of social psychology: Interpersonal process (pp. 388 409). Blackwell Publishing. Cacioppo, J. T., Pe tty, R. E., & Morris, K. (1983). Effects of need for cognition on message evaluation, recall, and persuasion. Journal of Personality and Social Psychology, 45 (4), 805 818. Carr, C. (2008, May). Social influence on an organization's successful adoption of i nstant messaging. Paper presented at the meeting of the International Communication Association, Montreal, QC. Carroll, B., & Ahuvia, A. (2006). Some antecedents and outcomes of brand love. Marketing Letters, 17 (2), 79 89. Cha, J. (2007, May). Factors aff ecting attitude toward shopping for real and virtual items on social networking websites Paper presented at the meeting of the International Communication Association, San Francisco, CA.

PAGE 244

244 Cha, J. (2009a). Shopping on social networking web sites: Attitudes toward real versus virtual items. Journal of Interactive Advertising, 10 (1), 77 93. Cha, J. (2009b). Television versus the Internet: A comparative analysis of traditional and new video platforms in substitutability, perceptions, and displacement effects Unpublished doctoral dissertation, University of Florida, Gainesville. Cha, J. (2010). Factors affecting the frequency and amount of social networking site use: Motivations, perceptions, and privacy concerns. First Monday, 15 (12). Retrieved from http://firstmonday.org/htbin/cgiwrap/bin/ojs/index.php/fm/article /view/2889/2685 Chau, P. Y. K. (1996). An empirical assessment of a modified technology acceptance model. Journal of Management Information Systems, 13 (2), 185 204. Chau, P. Y. K., & Hu, P. J. (2002). Examining a model of information technology acceptance by individual professionals: An exploratory study. Journal of Management Information Systems, 18( 4), 191 229. Chan Olmsted, S. M., & Cha, J. (2007). Branding television news in a multichannel environment: An exploratory study of network news brand personality. The International Journal on Media Management, 9 (4), 135 150. Chan Olmsted, S. M., & Cha, J. (2008). Exploring the antecedents and effects of brand images for television news: An application of brand personality construct in a multichannel news environment. The International Journal on Media Management, 10 (1), 32 45. Chan Olmsted, S. M., & Chang, B. (2006). Audien ce knowledge, perceptions and factors affecting the adoption intent of terrestrial digital television. New Media & Society, 8 (5), 773 800. Chan Olmsted, S. M., Li, J. C. C., & Jung, J. (2005). The profiling of cable modem broadband consumers: Characterist ics, perceptions and satisfaction. Journal of Targeting, Measurement and Analysis for Marketing 13, 327 345. Chang, B. (2005). Factors affecting evaluation of cable network brand extension: Focusing on parent network, fit, consumer characteristics and vie wing habits (Doctoral dissertation). Available from Dissertations and Theses Database. (UMI No. 3177950). Chang, B., Chan Olmsted, S. M. (2010). Success factors of cable network brand extension: focusing on the parent network, composition, fit, consumer c haracteristics, and viewing habits. Journal of Broadcasting & Electronic Media, 54 (4), 641 656.

PAGE 245

245 Chang, B., Lee, S., & Kim, S. (2006). Exploring factors affecting the adoption and continuation of online game among college students in South Korea: Integratin g uses and gratification and diffusion of innovation approaches. New Media & Society, 8 (2), 295 319. Chatterjee, R. (2001). Online review: Do consumers use them? Advances in Consumer Research, 28 129 133. Cheema, A., & Kaikati, A. M. (2010). The effect of need for uniqueness on word of mouth. Journal of Marketing Research, 47 (3), 553 563. Cheema, A., & Papatla, P. (2010). Relative importance of online versus offline information for internet purchases: Product category and internet experience effects. Journ al of Business Research, 63 (9 10), 979 985. Chen, L., Gillenson, M. L., & Sherrell, D. L. (2002). Enticing online consumers: An extended technology acceptance perspective. Information & Management, 39 (8), 705 719. Chen, P., Wu, S., & Yoon, J. (2004). The i mpact of online recommendations and consumer feedback on sales. Proceedings of the International Conference on Information Systems 711 724. Chen, Q., & Wells, W. (1999). Attitude toward the site. Journal of Advertising Research, 39 (5), 27 38. Chen, Y., Fa y, S., & Wang, Q. (2004). Marketing implications of online consumer product reviews. Working paper, Department of Marketing, University of Florida. Chen, Y., & Xie, J. (2008). Online consumer review: Word of mouth as a new element of marketing communicati on mix. Management Science, 54 (3), 477 491. Cheung, C., & Law, R. (2009). Have the perceptions of the successful factors for travel web sites changed over time? The case of consumers in Hong Kong, Journal of Hospitality & Tourism Research, 33( 3), 438 446. Chevalier, J. A., & Mayzlin, D. (2006). The effect of word of mouth on sales: Online book reviews. Journal of Marketing Research, 43 (3), 345 354. Chiou, J., & Cheng, C. (2003). Should a company have message boards on its web sites? Journal of Interactive M arketing, 17 (3), 50 61. Chitturi, R., Raghunathan, R., & Mahajan, V. (2007). Form versus function: How the intensities of specific emotions evoked in functional versus hedonic trade offs mediate product preferences. Journal of Marketing Research, 44 (4), 70 2 714. Cho, C., & Cheon, H. J. (2004). Why do people avoid advertising on the internet? Journal of Advertising, 33 (4), 89 97.

PAGE 246

246 Cho C ., & Leckenby J. D (1999). Interactivity as a m easure of a dvertising e ffectiveness Proceedings of the 1999 Conference of t he American Academy of Advertising, Gainesville, FL: American Academy of Advertising Marilyn S. Roberts, ed s., 162 179. Chu, S., & Kim, Y. (2011). Determinants of consumer engagement in electronic word of mouth (eWOM) in social networking sites. Internati onal Journal of Advertising, 30 (1), 47 75. Chung, D., & Nam, C. S. (2007). An analysis of the variables predicting instant messenger use. New Media & Society, 9 (2), 212 234. Cialdini, R. B. (1984). Influence: How and why people agree to things. New York: W illiam Morrow. Cialdini, R. B., & Goldstein, N. J. (2004). Social influence: compliance and conformity. The Annual Review of Psychology, 55 (1), 591 621. Cialdini, R. B., & Trost, M. R. (1999). Social influence: social norms, conformity, and compliance. In D. Gilbert, S. Fishke & G. Lindzy (Eds.), The Handbook of Social Psychology (Vol. 2, pp. 151 192). Boston: McGraw Hill, Inc. home VCRs: Evolution or revolution. Communica tion Research, 15 (6), 772 780. Comrey, A, L., & Lee, H. B. (1992). A first course in factor analysis (2 nd Ed.). Hillsdale, NJ: Lawrence Erlbaum. ComScore (2007, November). Online consumer generated reviews have significant impact on offline purchase behav ior, Retrieved from http://www.comscore.com /Press_Events/Press_Releases/2007/11/Online_Consumer_Reviews_Impact_Offli ne_Purchasing_Behavior Conway, J. C., & Rubin, A. M. (1991). Psychological predictors of televisio n viewing motivation. Communication Research, 18 (4), 443 463. Daft, R. L., & Lengel, R. H. (1984). Information richness: A new approach to managerial behavior and organization design. In L. L. Cummings & B. M. Staw (Eds.), Research in organizational behavi or (Vol. 6, pp. 191 233). Greenwich, CT: JAU Press. Daft, R. L., & Lengel, R. H. (1986). Organizational information requirements, media richness and structural design. Management Science, 32 (5), 554 571 Daft, R. L., Lengel, R. H., & Trevino, L. K. (1987). Message equivocality, media selection and manager performance: Implications for information system. MIS Quarterly 11 (3), 355 366.

PAGE 247

247 Davis, F. D. (1989). Perceived usefulness perceived easy of use, and user acceptance of information technology. MIS Quarterl y, 13 (3), 319 340. Davis, F. D. (1993). User acceptance of information technology: System characteristics, user perceptions and behavioral impacts. International Journal of Man Machine Studies, 38 (3), 475 487. Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1989). User acceptance of computer technology: A comparison of two theoretical models. Management Science, 35 (8), 982 1003. Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1992). Extrinsic and intrinsic motivation to use computers in the workplace. Journ al of Applied Social Psychology, 22 (14), 1111 1132. Dean, D. H., & Biswas, A. (2001). Third party organization endorsement of products: An advertising cue affecting consumer prepurchase evaluation of goods and services. Journal of Advertising, 30 (4), 41 57 Dellarocas, C. (2003). The digitization of word of mouth: Promise and challenges of online feedback mechanisms. Management Science, 49 (10), 1407 1424. Dellarocas, C. (2004). Strategic manipulation of Internet opinion forums: Implications for consumers and firms. MIT Sloan Working Papers No. 4501 04. Dellarocas, C., Zhang, X., & Awad, N. F. (2007). Exploring the value of online product reviews in forecasting sales: The case of motion pictures. Journal of Interactive Marketing, 21 (4), 23 45. Deutch, M., & Ge rard, H. B. (1955). A study of normative and informational social influence upon individual judgment. Journal of Abnormal and Social Psychology, 51 (3), 629 636. Dimmick, J., Kline, S., & Stafford, L. (2000). The gratification niches of personal e mail and telephone. Communication Research, 27 (2), 227 248. Dimmick, J., Sikand, J., Patternson, S. (1994). The gratifications of the household telephone: Sociability, instrumentality and reassurance. Communication Research, 21 (5), 643 663. Dishaw, M. T., & Stron g, D. M. (1999). Extending the technology acceptance model with task technology fit construct. Information & Management, 36 (1), 9 21. De Matos, C. A., & Rossi, C. A. V. (2008). Word of mouth communications in marketing: A meta analytic review of the antec edents and moderators. Journal of the Academy of Marketing Science, 36 (4), 578 596.

PAGE 248

248 Dennen, S. (2011, February 1). State of the Internet: video, mobile, and social. ComsScore. Retrieved February 22, 2011, from http://www.comscore.com /Press_Events/Presenta tions_Whitepapers/2011/State_of_the_Internet_Video_Mo bile_and_Social Dhar, R., & Wertenbroch, K. (2000). Consumer choice between hedonic and utilitarian goods. Journal of Marketing Research, 37 (1), 60 71 Dholakia, U. M., Basuroy, S., & Soltysinski, K. (20 02). Auction or agent (or both)? A study of moderators of the herding bias in digital auction. International Journal of Research in Marketing, 19 (2), 115 130. Dobele, A., Tloeman, D., & Beverland, M. (2005). Controlled infection! Spreading the brand messag e through viral marketing. Business Horizons 48 (2), 143 149. ease of use instruments for decision making: A confirmatory and multigroup invariance analysis. Decision S ciences, 29 (4), 839 869. Duan, W., Gu, B., & Whinston, A. B. (2008). Do online reviews matter? an empirical investigation of panel data. Decision Support Systems, 45 (4), 1007 1016. Duhan, D. F., Johnson, S. D., Wilcox, J. B., Harrell, G. D. (1997). Influ ences on consumer use of word of mouth recommendation sources. Journal of the Academy of Marketing Science, 25 (4), 283 295. Dutton, W., Rogers, E., & Jun, S. (1987). Diffusion and social impacts of personal computers. Communication Research, 14 (2), 219 250 Dye, R. (2000). The buzz on buzz. Harvard Business Review, 78 139 146. East, R., Hammond, K., & Lomax, W. (2008). Measuring the impact of positive and negative word of mouth on brand purchase probability. International Journal of Research in Marketing, 25 (3), 215 224. East, R., Hammond, K., & Wright, M. (2007). The relative incidence of positive and negative word of mouth: A multi category study. International Journal of Research in Marketing 24 (2), 175 184. Eastin, M. S. (2002). Diffusion of e commerce : An analysis of the adoption of four e commerce activities. Telematics and Informatics, 19 (3), 251 267. Eckhardt, A., Laumer, S., & Weitzel, T. (2009). Who influences whom? Analyzing adoption. J ournal of Information Technology, 24 (1), 11 24.

PAGE 249

249 Edwards, A., Edwards, C., Shaver, C., & Oaks, M. (2009). Computer mediated word of mouth communication on RateMyProfessors.com: Expectancy effects on student cognitive and behavioral learning. Journal of Comp uter Mediated Communication, 14 (2), 368 392. Elias, T., Appiah, O., & Gong, L. (2008, August). identity on consumer attitudes: A multiple group model approach Paper Presented at the meeting of the Association for Educ ation and Journalism and Mass Communication, Chicago, IL. Elliott, W. R., & Rosenberg, W. L. (1987). The 1985 Philadelphia Newspaper strike: A uses and gratification study. Journalism Quarterly, 64( 4), 679 687. Ellison, N. B., Steinfield, C., & Lampe, C. ( Journal of Computer Mediated Communication, 12 (4), 1143 1168. Engel, J. F., Blackwell, R. D., & Miniard, P. W. (1993). Consumer behavior (8 t h ed.). Fort Worth: Dryden Press. Ernst & Young. (1999). The second annual Ernst & Young Internet shopping study Retrieved June 10, 2011, from http://www.e consultancy.com/knowledge/whitepapers /9/the/second annual Internet shopping study.html Fabrigar, L. R., Wegener, D. T., MacCallum, R. C., & Strahan, E. J. (1999). Evaluating the use of exploratory factor analysis in psychological resea rch. Psychological Methods, 4 (3), 272 299. Feick, L. F., & Higie, R. A. (1992). The effects of preference heterogeneity and source characteristics on ad processing and judgments about endorsers. Journal of Advertising, 21 (2), 9 23. Feldman, J. M., & Lynch J. G. (1988). Self generated validity and other effects of measurement on belief, attitude, intention, and behavior. Journal of Applied Psychology, 73 (3), 421 435. Festinger, L. (1954). A theory of social comparison processes, Human Relations, 7 (2), 117 140. Festinger, L., Pepitone, A., & Newcomb, T. (1952). Some consequences of de individuation in a group. Journal of Abnormal and Social Psychology, 47 (2), 382 389. Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intension and behavior: An introduction to theory and research. Reading, MA: Addison Wesley. Fisher, R. J., & Price, L. L. (1992). An investigation into the social context of early adoption behavior. Journal of Consumer Research, 19 (3), 477 486.

PAGE 250

250 Flosi, S. L. (2011, February 21). Comscore media metrix ranks top 50 U.S. web properties for January 2011. Retrieved March 7, 2011, from http://www.comscore.com/Press_Events/Press_Releases/2011/2 /comScore_Media_Metrix_Ranks_Top_ 50_U.S._Web_Properties_for_January_20 11 Forman, C., Ghose, A., & Wiesenfeld, B. (in press). Examining the relationship between review and sales: the role of reviewer identity discloser in electronic markets. Information System Research Frenzen, J., & Nak amoto, K. (1993). Structure, cooperation, and the flow of market information. Journal of Consumer Research, 20 (3), 360 375. Fulk, J., Schmitz, J., & Steinfiled, C. (1990). A social influence model of technology use. In J. Fulk, & Steinfield (Eds.), Organiz ations and Communication Technology 117 140. Newbury Park, CA: Sage Publications. Gambon, J. (1998, March 16). Keep the browser simple, stupid. Information Week, 673, 16. Gefen, D., Karahanna, E., & Straub, D. W. (2003). Trust and TAM in online shopping: An integrated model. MIS Quarterly, 27 (1), 51 90. Gefen, D., & Straub, D. (2000). The relative importance of perceived ease of use in IS adoption: A study of e commerce adoption. Journal of the Association for Information Systems, 1 (8), 1 21. Gilly, M. C., Garham, J. L., Wolfnbarger, M. F., & Yale, L. J. (1998). A dyadic study of interpersonal information search, Journal of Academy of Marketing Science, 26 (2), 83 100. Gilly, M. C., & Zeithaml, V. A. (1985). The elderly consumer and adoption of technologies Journal of Consumer Research, 12 (3), 353 357. Godes, D., & Mayzlin, D. (2004). Using online conversations to study word of mouth communication. Marketing Science, 23 (4), 545 560. Goldsmith, R. E., & Horowitz, D. (2006). Measuring motivations for online o pinion seeking. Journal of Interactive Advertising, 6 (2), 1 16. Gotlieb, J. B., & Sarel, D. (1991). Comparative advertising effectiveness: The role of involvement and source credibility. Journal of Advertising, 20 (1), 38 45. Goyette, I., Ricard, L., Berger on, J., & Marticotte, F. (2010). e WOM scale: Word of Mouth measurement scale for e services context. Canadian Journal of Administrative Sciences. 27 (1), 5 23.

PAGE 251

251 Graham, J., & Havlena, W. (2007). Finding the "missing link": Advertising's impact on word of m outh, web searches, and site visits. Journal of Advertising Research, 47 (4), 427 435. Granovetter, M (1973). The strength of weak ties. American Journal of Sociology, 78 (6), 1360 1380. Guhan Canli, Z. & Maheswaran, D. (2000). Comparative advertising in the global marketplace: the effects of cultural orientation on communication. Working Papers series no. 328, Stephen M. Ross Business School, William Davidson Institute, University of Michigan. Gulas, C., & McKeage, K. (2000). Extending social comparison: An examination of the unintended consequences of idealized advertising imagery. Journal of Advertising, 29 (2), 17 28. Hagtvedt, H., & Patrick, V. M. (2009). The broad embrace of luxury: Hedonic potential as a driver of brand extendibility. Journal of Consumer Psychology, 19 (4), 608 618. Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. (1998). Multivariate Data Analysis Upper Saddle River, NJ: Prentice Hall. Hamburger, E. (2011, June). 12 reasons you absolutely need to be on Twitter, Business Insid er, Retrieved from http://www.businessinsider.com/best reasons to use twitter 2011 6?op=1 Hanson, W. A., & Putler, D. S. (1996). Hits and misses: Herd behavior and online product popularity. Marketing Letters, 7 (4), 297 305. 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 (1), 60 75. Hart, C. W. L., Heskett, J. L., & Sasser Jr., W. E. (1990). The profitable ar t of service recovery. Harvard Business Review, 68 (4), 148 156. Haywood, K. M. (1989). Managing word of mouth communications. The Journal of Services Marketing, 3 (2), 55 67. He, F., & Mykytyn, P. P. (2007). Decision factors for the adoption of an online pa yment system by customers. International Journal of E business Research, 3 (4), 1 32. Heeter, C., & Grrenberg, B. G. (1985). Cable and program choice, Selective Exposure to Communication Zillmann, Dolf., and Bryant, J. (Eds). Hillsdale, NJ: Lawrence Erlbau m Associates, 203 224.

PAGE 252

252 Hennig Thurau, T., Gwinner, K. P., Walsh, G., & Gremler, D. D. (2004). Electronic word of mouth via consumer opinion platforms: What motivates consumers to articulate themselves on the Internet? Journal of Interactive Marketing, 18 ( 1), 38 52. Hennig Thurau, T., Malthouse, E., Friege, C., Gensler, S., Lobschat, L., Rangaswamy, A., & Skiera, B. (2010). The impact of new media on customer relationship. Journal of Service Research, 13 (3), 311 330. Hennig Thurau, T., & Walsh, G. (2003). E lectronic word of mouth: Motives for and consequences of reading customer articulations on the Internet. International Journal of Electronic Commerce, 8 (2), 51 74. Herr, P. M., Kardes, F. R., & Kim, J. (1991). Effects of word of mouth and product attribut e information of persuasion: An accessibility diagnosticity perspective. Journal of Consumer Research, 17 (4), 454 462. Heung, V. C. S. (2003). Internet usage by international travelers: Reasons and barriers. International Journal of Contemporary Hospitalit y Management, 15 (7), 370 378. Hirschman, E. C., & Holbrook, M. B. (1982). Hedonic consumption: emerging concepts, methods and propositions. Journal of Marketing, 46 (3), 92 101. Hoffman, D. L., Novak, T. P., & Peralta, M. A. (1999). Building consumer trust online, Communications of the ACM, 42 (4), 80 85. Hogg, M. A. (2004). Social categorization, depersonalization, and group behavior. In M. B. Brewer and M. Hewstone (Eds.). Self and social identity Blcakwell Publishing. Hogg, M. A., & Reid, S. A. (2006). So cial identity, self categorization, and the communication of group norms. Communication Theory, 16 (1), 7 30. Hogg, M. A., & Terry, D. J. (1995). Social identity and self categorization processes in organizational contexts. The Academy of Management Review, 25 (1), 121 140. Hogue, J. (2011). Integrating social media into the marketing mix. Holbrook, M. B., & Hirshman, E. C. (1982). The experiential aspects of consumption: Consumer fantasies, feelings, and fun. Journal of Consumer Research, 2 (9), 132 149. Hom adoption of sales technology: A multilevel analysis. Journal of the Academy of Marketing Science, 38 (2), 159 168. Homer, P. M., & Kahle, L. R. (1990). Source expertise, time of s ource identification, and involvement in persuasion: An elaborative processing perspective. Journal of Advertising, 19 (1), 30 39.

PAGE 253

253 Hong, S., Milik, M. L., & Lee, M. K. (2003). Testing configural, metric, scalar, and latent mean invariance across genders in sociotropy and autonomy using non western sample. Educational and Psychology Measurement, 63 (4), 636 654. Hong, S. Y., & Rim, H. (2010). The influence of customer use of corporate websites: corporate social responsibility, trust, and word of mouth communic ation, Public Relations Review, 36 (4), 389 391. Hong, S. Y., & Yang, S. (2009). Effects of reputation, relational satisfaction, and customer company identification on positive word of mouth intentions. Journal of Public Relations Research, 21 (4), 381 403. Hong, W., Thong, J. Y. L., Wong, W. M., & Tam, K. Y. (2002). Determinants of user acceptance of digital libraries: An empirical examination of individual differences and system characteristics. Journal of Management Information Systems, 18 (3), 97 124. Hovl and, C. I., Janis, I. L., & Kelley, H. H. (1953). Communication and persuasion. Yale University Press, New Heaven, CT. Hsu, C., & Lin, J. C. (2008). Acceptance of blog usage: The roles of technology acceptance, social influence and knowledge sharing motiv ation. Information & Management, 45 (1), 65 74 Hsu, M., Yen, C., Chiu, C., & Chang, C. (2006). A longitudinal investigation of continued online shopping behavior: An extension of the theory of planned behavior. International Journal of Human Computer Studi es, 64 (9), 889 904. Hu, L. Z., & Bentler, P. M. (1999). Cut off criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6 (1), 1 55. Huang, J., & Chen, Y. (2006). Herding in onl ine product choice. Psychology & Marketing, 23 (5), 413 428. I lieva, J., Baron, S., & Healey, N. M. (2002). Online survey in marketing research: Pros and cons. International Journal of Market Research, 44 (3), 361 367. Imkar, C., Vallen, B., & Sen, S. (2010) Uniqueness motivations in product preferences. Journal of Consumer Research, 37 (3), 443 455. Interactive Advertising Bureau. (2009, March 18). IAB social advertising best practices. Retrieved June 4, 2011, from http://www.iab.net/socialads

PAGE 254

254 Iyengar, S., Valentino, N. A. (2000). Who says what? Source credibility as a mediator of campaign advertising, in A. Lupia & M. D. McCubbins (Ed.), Elements of reason: Co gnition, choice, and the bounds of rationality Cambridge studies in political psychology and public opinion. Cambridge: Cambridge University Press. of nomological val idity. In Monroe, K. B. (Eds.), Advances in Consumer Research Association for Consumer Research, Ann Arbor, MI, 299 303. Jaffe, J. (2007). Join the conversation: How to engage marketing weary consumer with the power of community, Dialogue and Partnership. New York: John Wiley & Sons. Jansen, B. J., Zhang, M., Sobel, K., & Chowdury, A. (2009). Twitter power: Tweets as electronic word of mouth. Journal of the American Society for Information Science and Technology, 60 (11), 2169 2188. Jrvelinen, J. (2007). Online purchase intentions: An empirical testing of a multiple theory model. Journal of Organizational Computing and Electronic Commerce, 17 (1), 53 74. Jeffres, L., & Atkin, D. (1996). Predicting use of technologies for communication and consumer needs. Jo urnal of Broadcasting & Electronic Media, 40 (3), 318 330. Jiang, J. J., Hsu, M., & Klein, G. (2000). E commerce user behavior model: An empirical study. Human Systems Management, 19 (4), 265 276. Jiang, P. (2009). Consumer adoption of mobile internet servic es: An exploratory study. Journal of Promotion Management, 15 (3), 418 454. Jones, L. W., Sinclair, R. C., & Courneya, K. S. (2003). The effects of source credibility and message framing on exercise intentions, behaviors, and attitudes: An integration of th e elaboration likelihood model and prospect theory. Journal of Applied Social Psychology, 33 (1), 179 196. Joreskog, K. G. (1973). A general method for estimating a linear structural equation system. In A. S. Goldberger, & O. D. Duncan (Eds.), Structural eq uation models in the social sciences (pp. 85 112). New York: Seminar Press. Jun, J. W., Cho, C., & Kwon, H. J. (2008). The role of affect and cognition in consumer evaluations of corporate visual identity: Perspectives from the Unites States and Korea. Bra nd Management, 15 (6), 382 398. Jung, J., Chan Olmsted, S., Park, B., & Kim, Y. (2011). Factors affecting e book reader awareness, interest, and intention to use. New Media & Society, doi: 10.1177/1461444811410407

PAGE 255

255 Kamins, M. A. (1990). An integration into t advertising: when beauty may be only skin deep. Journal of Advertising, 19 (1), 4 13. Kang, J. W. (2003). media environment: An innovation adoption model Unpublished doctoral dissertation, University of Florida, Gainesville. Karahanna, E., Straub, D. W., & Chervany, N. L. (1999). Information technology adoption across time: A cross sectional comparison of pre adoption and post adoption belief. MIS Qu arterly, 23 (2), 183 214. Kaplan, A. M., & Haenlein, M. (2010). Users of the world, unite! The challenges and opportunities of social media. Business Horizons, 53 (1), 59 68. Katona, Z., Zubcsek, P. P., & Sarvary, M. (2011). Network effects and personal infl uences: The diffusion of an online social network. Journal of Marketing Research, 48 (3), 425 443. Katz, E., & Lazarsfeld, P. F. (1955). Personal Influence: the part played by people in the flow of mass communications Glencoe, New York: Free Press. Keller, E. (2007). Unleashing the power of word of mouth: Creating brand advocacy to drive growth. Journal of Advertising Research, 47( 4), 448 452. Keller, E. & Berry, J. (2006). Word of mouth: the real action is offline. Advertising Age, 77 (49), 20. Keller, E., & Fay, B. (2009). The role of advertising in word of mouth. Journal of Advertising Research, 49 (2), 154 158. Kelley, H. H. (1973). The process of causal attribution. American Psychologist, 28 (2), 107 128. Kelly, L. (2007). Beyond Buzz: The next generatio n of word of mouth marketing. New York, AMACOM. Kelman, H. C. (1961). Processes of opinion change, Public Opinion Quarterly, 25 (1), 57 78. Ki, E., Chang, B., & Khang, H. (2006). Exploring influential factors on music piracy across countries. Journal of Com munication, 56 (2), 406 426. Kim, J., & Chan Olmsted, S. (2005). Comparative effects of Organization Public relationships and Product related attributes on brand attitude. Journal of Marketing Communications, 11 (3), 145 170.

PAGE 256

256 Kim, J., & Kim, M., & Hong, S. (2009). Writing Thesis with Structural Equation Modeling Communication Books, Seoul, Korea. Kiousis, S. (2001). Public trust or mi strust? Perceptions of media credibility in the information age. Mass Communication & Society, 4 (4), 381 403. Kisielius, J., & Sternthal, B. (1984). Detecting and explaining vividness effects in attitudinal judgment. Journal of Marketing Research, 21 (1), 5 4 64. Kliatchko, J. G. (2008). Revisiting the IMC construct. International Journal of Advertising, 27 (1), 133 160. Kline, R. B. (2005). Principles and practice of structural equation modeling New York: Guilford. Ko, H., Cho, C., & Roberts, M. S. (2005). I nternet uses and gratifications: A structural equation model of interactive advertising. Journal of Advertising, 34 (2), 57 70. Korgaonkar, P. K., & Wolin, L. D. (1999). A multivariate analysis of web usage. Journal of Advertising Research, 39 (2), 53 68. Ko zinets, R. V., De Valck, K., Wojnicki, A. C., & Wilner, S. J. S. (2010). Networked narratives: Understanding word of mouth marketing in online communities. Journal of Marketing, 74 (2), 71 89. Krugman, D. (1985). Evaluating the audience of the new media, Jo urnal of Advertising, 14 (4), 14 19. Kwak, H., Lee, C., Park, H., & Moon, S. (2010, April). What is Twitter, a social network or a news media? Paper presented at the meeting of the 19 th International World Wide Web Conference, Raleigh, NC. Kwon, K. H., & Ch on, B. S. (2009). Social influences on terrestrial and satellite mobile TV adoption in Korea: Affiliation, positive self image, and perceived popularity The International Journal on Media Management, 11 (2), 49 60. Laczniak, R. N., DeCarlo, T. E., & Ramasw ami, S. N. (2001). Consumers' responses to negative word of mouth communication: An attribution theory perspective. Journal of Consumer Psychology,11 (1), 57 73. Ladhari, R. (2007). The effect of consumption emotions on satisfaction and word of mouth commun ications. Psychology & Marketing, 24 (12), 1085 1108. LaRose, R., & Atkin, D. (1988). Satisfaction, demographic, and media environment predictors of cable subscription. Journal of Broadcasting & Electronic Media, 32( 4), 403 413.

PAGE 257

257 LaRose, R. K., & Atkin, D. J (1992). Audiotext and the reinvention of the telephone as a mass medium. Journalism Quarterly, 69 (2), 413 421. LaRose, R., Lai, Y., Lange, R., Love, B., & Wu, Y. (2005). Sharing or piracy? An exploration of downloading behavior. Journal of Computer Media ted Communication 11 (1), 1 21. Lau, G., & Ng, S. (2001). Individual and situational factors influencing negative word of mouth behavior. Canadian Journal of Administrative Sciences, 18 (3), 163 178. Ledbetter, A. M., Mazer, J. P., Degroot, J. M., Meyer, K. R., Mao, Y., & Swafford, B. (2010). Attitudes toward online social connection and self disclosure as predictors of Facebook communication and relational closeness. Communication Research, 38 (1), 27 53. Lee, C., Kim, J., & Chan Olmsted, S. M. (2010). Brand ed product information search on the web: the role of brand trust and credibility of online information sources. Journal of Marketing Communications doi:10.1080/13527266.2010.484128. Lee, D., Park, J., & Ahn, J. (2001). On the explanation of factors affec ting e commerce adoption. Proceedings of the 22 nd international conference on information systems, V. Storey, S. Sarker, and J. I. Degross. (Eds.), New Orleans, Louisiana, 2001, pp. 109 120. Lee, J. (2003). Testing social information processing and techno logy acceptance model in a distance learning environment with a social network approach, (Doctoral dissertation). Retrieved from Dissertations and Theses database. (UMI No. 3087007). Lee, J., Cho, H., Gay, G., Davison, B., & Ingraffea, A. (2003). Technolog y acceptance and social networking in distance learning. Educational Technology & Society, 6 (2), 50 61. Lee, M., Rodgers, S., & Kim, M. (2009). Effects of valence and extremity of eWOM on attitude toward the brand and website. Journal of Current Issues & R esearch in Advertising, 31 (2), 1 11. Lee, M., & Youn, S. (2009). Electronic word of mouth (eWOM). International Journal of Advertising, 28 (3), 473 499. Lee, P. S. N., & Leung, L. (2008). Assessing the displacement effect of the Internet. Telematics and In formatics 25, 145 155. Lee, S. (2007). The Internet and adolescent social capital: Who benefits more from Internet use? (Doctoral dissertation). Retrieved from Dissertations and Theses database. (UMI No. 3291209).

PAGE 258

258 Lengel, R. H., & Daft, R. L. (1989). Th e selection of communication media as an executive skill. The Academy of Management Executive (1987 1989), 2 (3), 225 232. Lenhart, A., Purcell, K., Smith, A., & Zickuhr, K. (2010, February 3). Social media and mobile interent use among teens and young adul ts Retrieved June 9, 2011, from http://pewresearch.org/pubs/1484/social media mobile internet use teens millennials fewer blog Leung, L. (2001 ). College student motives for chatting on ICQ. New Media & Society, 3 (4), 483 500. Leung, L., & Wei, R. (1999). Seeking news via the pager: An expectancy value study. Journal of Broadcasting and Electronic media, 43 (3), 299 315. Levy, M. R. (1987). VCR us e and the concept of audience activity. Communication Quarterly, 35 (3), 267 275. Li, S C. S. (2004). Exploring the factors influencing the adoption of interactive cable television services in Taiwan. Journal of Broadcasting & Electronic Media, 48 (3), 466 4 83. Libai, B., Bolton, R., Sugel, M. S., Ruyter, K. D., Gotz, O., Risselada, H., Stephen, A. T. (2010). Customer to customer interaction broadening the scope of word of mouth research. Journal of Service Research, 13 (3), 267 282. Licheterstein, A., & Rosen feld, L. (1984). Normative expectations and individual decisions concerning media gratification choices. Communication Research, 11 (3), 393 413. Lin, C. (1999). Online service adoption likelihood. Journal of Advertising Research, 39 (2), 79 89. Lin, C. (200 4). Webcasting adoption: Technology, user innovativeness, and media substitution. Journal of Broadcasting & Electronic Media, 48 (3), 446 465. Lin, C. A. (1998). Exploring personal computer adoption dynamics. Journal of Broadcasting and Electronic Media, 42 (1), 95 112. Lin, C. A. (2009). Exploring the online radio adoption decision making process: cognition, attitude, and technology fluidity. Journalism & Mass Communication Quarterly, 86( 4), 884 89. Lin, C. A. (2010). Satellite radio adoption demand: Consume r beliefs, attitudes and intentions. Journal of Broadcasting & Electronic Media, 54 (2), 265 281.

PAGE 259

259 Lin, J., & Cho, C. (2010). Antecedents and consequences of cross media usage: A Journal of Broadcasting & Electronic Media, 54 (2), 316 336. Lin, J., Chuan, C., & Rivera, M. (2009, May). The interaction effects of peer and social influence: An application of the unified theory of acceptance and use of technology to messenger adoption. Paper presented at the meeting of th e International Communication Association, Chicago, IL. Lin, T. M. Y., & Fang, C. (2006). The effects of perceived risk on the word of mouth communication dyad. Social Behavior and Personality, 34 (10), 1207 1216. Lipsman, A., Mudd, G., Rich, M., & Bruich, S. (2011). The power of like: How brands reach and influence fans through social media marketing, Comscore, Retrieved July 20, 2011, from http://www.comscore.com/Press_Events /Presentations_Whitepapers /2011/The_Power_of_Like_How_Brands_Reach_and _Influence_Fans_Through_Social_Media_Marketing Liu, Y. (2006). Word of mouth for movie: its dynamics and impact on box office revenue. Journal of Marketing 70 (3), 74 89. LP, C., Lee, H., & Law, R. (2011). Profili ng the users of traveling websites for planning and online experience sharing, Journal of Hospitality & Tourism Research, 35( 3), 1 9. Lu, J., Yao, J. E., & Yu, C. (2005). Personal innovativeness, social influences and adoption of wireless internet services via mobile technology. The Journal of Strategic Information Systems, 14 (3), 245 268 Lucas, H., & Spitler, V. K. (1999). Technology use and performance: A field study of broker workstations. Decision Science, 30 (2), 291 311. Lynn, M., & Harris, J. (1997). The desire for unique consumer products: A New individual difference scale. Psychology and Marketing, 14 (6), 601 616. Madden, G., Savage, S. J. (2000). Some economic and social aspects of residential Interent usd in Austrailia. The Journal of Media Economi cs, 13 (3), 171 185. Mangold, W. G., & Faulds, D. J. (2009). Social media: The new hybrid element of the promotion mix. Business Horizons, 52 (4), 357 365. access. Communic ation Research, 14 (5), 491 511. Martinez, E., & Pina, J. M. (2010). Consumer responses to brand extensions: a comprehensive model. European journal of Marketing, 44 (7), 1184 1280.

PAGE 260

260 Mastro, D. E. (2003). A social identity approach to understanding the impact of television messages. Communication Monographs, 70 (2), 98 113. Mathieson, K. (1991). Predicting user intentions: Comparing the technology acceptance model with the theory of planned behavior. Information Systems Research, 2 (3), 173 191. Mayfiled, A. (20 08, January). What is social media? iCrossing Retrieved January 31, 2011, from www.icrossing.co.uk/.../What_is_Social_Media_iCrossing_ebook.pdf McAlister, D. T., & Erffmeyer, R. C. (2003). A content analysis of outcomes and responsibilities for consumer complaints to third party organizations. Journal of Business Research, 56 (4), 341 351. McGill, A. L., & Anand, P. (1989). The effects of vivid attributes on the evaluation of alternatives: The role of differential attention and cognitive elaboration. Journ al of Consumer Research, 16 (2), 188 196. Mendelsohn, H. (1964). Listening to radio in people, society and mass communication Lewis, A. D., & White, D. M. (Eds.). New York: Free Press, 239 249. Miller, C. C. (2011, June 28). Another try by Google to take o n Facebook. The New York Times Retrieved from http://www.nytimes.com/2011/06/29/technology /29google.html?_r=1&ref=technology Milstein, S., Chowdhury, A., Hockmuth, G., Lorica, B., & Magoulas, R. (2008). Twitter and the micro messaging revolution: Communication, connections, and immediacy Misner, I. R. (1999). with word of mouth marketing (2 nd ed.). Austin: Brand Press. Mizerski, R. W. (1982). An attribution explanation of the disproportionate influence of unfavorable information. The Journal of Consumer Research, 9 (3), 301 310. Monroe, K. (1976). The influence of price diff erences and brand familiarity on brand preference, Journal of Consumer Research, 3 (1), 42 49. Murray, K. B. (1991). A test of services marketing theory: consumer information acquisition activities. Journal of Marketing, 55 (1), 10 25. Murray, K. B., & Schla cter, J. L. (1990). The impact of services versus goods on Journal of the Academy of Marketing Science, 18 (1), 51 65.

PAGE 261

261 of mouth marketing? Social comput ing and consumer control put momentum into viral marketing. Forrester Research. Retrieved April 1, 2011, from http://www.forrester.com/rb/Research/wha ts_buzz_on_word of mouth_marketing/q/id/36916/t/2 Nasco, S. A., Kulviwat, S., Kumar, A., & Bruner II, G. C. (2008). The CAT model: Extensions and moderators of dominance in technology acceptance. Psychology & Marketing, 25 (10), 987 1005. Nielsen. (2011). Integrating social media into the marketing mix. The Nielsen Company Retrieved March 3, 2011, from http://www.nielsen.com/us/en/ins ights/events webinars/2011/integrating social media into marketing mix.html Nunnally, J. C. (1978). Psychometric theory New York, NY: McGraw Hill. Okada, E. M. (2005). Justification effects on consumer choice of hedonic and utilitarian goods. Journal of Marketing Research, 42 (1), 43 53. Okazaki, S. (2008). Determinant factors of mobile based word of mouth campaign referral among Japanese adolescents. Psychology & Marketing, 25 (8), 714 731. Okazaki, S. (2009). The tactical use of mobile marketing: How ad olescents' social networking can best shape brand extensions. Journal of Advertising Research, 49 (1), 12 26. telephone. Journalism and Mass Communication Quarterly, 72 (4), 922 933. Ozcan, K. (2004). Consumer to consumer interactions in a networked society: Word of mouth theory, consumer experiences, and network dynamics. Unpublished doctoral dissertation, University of Michigan, Ann Arbor, Michigan. Papacharissi, Z., & Rubin, A. M. (2000). Predictors of Internet use, Journal of Broadcasting & Electronic Media 44 (2), 175 196. Papadimitriou, D., Apostolopoulou, A., & Loukas, I. (2004). The role of perceived fit in Internation al Journal of Sports Marketing and Sponsorship, 6( 1), 31 48. Papies, D., & Clement, M. (2008). Adoption of new movie distribution services on the internet. Journal of Media Economics, 21 (3), 131 157. Park, C. W., Jaworski, B. J., & Maclnnis, D. J. (1986). Strategic brand concept image management. The Journal of Marketing, 50 (4), 135 145.

PAGE 262

262 Park, D., & Kim, S. (2008). The effects of consumer knowledge on message processing of electronic word of mouth via online consumer reviews. Electronic Commerce Research an d Applications 7 (4), 399 411. Park, D., 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. Park, H. S., Dail ey, R., & Lemus, D. (2002). The use of exploratory factor analysis and principal components analysis in communication research. Human Communication Research, 28 (4), 562 577. Park, N. (2010). Adoption and use of computer based voice over internet protocol p hone service: Toward an integrated model. Journal of Communication, 60 (1), 40 72. electronic courseware: An application of the technology acceptance model. Journal of Comp uter Mediated Communication, 13 (1), 163 186. Patti, C. H., & Chen, C. H. (2009). Types of word of mouth messages: Information search and credence based services. Journal of Promotion Management, 15 (3), 357 381. Pedersen, P. E. (2005). Adoption of mobile in ternet services: An exploratory study of mobile commerce early adopters, Journal of Organizational Computing and Electronic Commerce 15 (3), 203 222. Pedersen, P. E., & Nysveen, H. (2003). Using the theory of planned behavior to explain of text messaging services. Working paper. Perloff, R. M. (2010). The dynamics of persuasion: Communication and attitudes in the 21st century London: Routledge. Peter, J., Valkenburg, P., Schouten, A. (2005). Characteristics and motives of adolescents ta lking with strangers on the Internet and its consequences. Paper presented at the annual meeting of the Internet Communication Association, New York, NY. Petty, R. E., & Cacioppo, J. T. (1984) Communication and persuasion: Central and peripheral routes to attitude change New York: Springer. Petty, R. E., & Cacioppo, J. T. (1986). The elaboration likelihood model of persuasion. L. Berowitz (Ed.), Advances in Experimental Psychology (Vol. 19, pp. 124 203). New York, NY: Academic Press.

PAGE 263

263 Pew Research Center. (2005, May 10). Tracking online life: How women and men use the Internet. Pew Internet and American Life Report Retrieved on February 1, 2011, from http://www.pewinternet.org/Press Releases/2000/Tracking online life How women use the Internet to cultivate relationships with family and friends.aspx Pew Research Center. (2011, January 4). Inte news source. Pew Internet and American Life Report, Retrieved April 4, 2011, from http://people press.org/report/689/ Phelps, J. E., Lewis, R., Mobilio, L., Perry, D., & Raman, N. (2004). Viral marketing or electro nic word of mouth advertising: Examining consumer response and motivations to pass along email. Journal of Advertising Research, 44 (4), 333 348. Pingdom (2010). Study: Ages of social network users, Pingdom, Retrieved on June 6, 2011 from http://royal.pingd om.com/2010/02/16/study ages of social network users/ Pommer, M. D., Berkowitz, E. N., & Walton, J. R. (1980). UPC scanning: An assessment of shopping response to technological changes. Journal of Retailing, 56 (2), 25 44. Preece, J., Rogers, Y., & Sharp, H (2002). Interaction design: beyond human computer interaction New York, NY: J. Wiley & Sons. Prendergast, G., Ko, D., & Yeun, S. Y. V. (2010). Online word of mouth and consumer purchase intentions. International Journal of Advertising, 29 (5), 687 708. P rice, L. L., Feick, L. F., & Higie, R. A. (1989). Preference heterogeneity and coorientation as determinants of perceived informational influence. Journal of Business Research, 19 (3), 227 242. Price, L. L., Feick, L. F., & Guskey, A. (1995). Everyday marke t helping behavior. Journal of Public Policy & Marketing, 14 (2), 255 266. Putzke, J., Schoder, D., & Fishbach, K. (2010). Adoption of mass customized newspapers: an augmented technology acceptance perspective. Journal of Media Economics, 23 (3), 143 146. Raacke, J., & Bonds Raacke, J. (2008). MySpace and facebook: Applying the uses and gratifications theory to exploring friend networking sites. CyberPsychology & Behavior, 11 (2), 169 174. Radwanick, S. (2011, February 2). U.S. digital year in review: A rec ap of the year in digital media Retrieved March 03, 2011, from http://www.comscore.com /Press_Events/Presentations_Whitepapers/2011/State_of_the_Internet_Video_Mo bile_and_Social/%28language%29/eng US

PAGE 264

264 Ranaweera, C., & Prabhu, J. (2003). On the relative i mportance of customer satisfaction and trust as determinants of customer retention and positive word of mouth. Journal of Targeting, Measurement & Analysis for Marketing, 12 (1), 82 90. Raskin, J. (1997). Looking for a humane interface: Will computer ever b ecome easy to use? Communications of the ACM, 40 (2), 98 101. Ratchford, B. T., Lee, M., & Talukdar, D. (2003). The impact of the Internet on information search for automobiles, Journal of Marketing Research 40 (2), 193 209. Reibstein, D. J. (2002). What att racts customers to online store, and what keeps them coming back? Journal of the Academy of Marketing Science, 30 (4), 465 473. Reingen, P., & Kernan, P. (1986). Analysis of referral networks in marketing: methods and illustration. Journal of Marketing Rese arch, 23 (4), 370 378. Rezabakhsh, B., Bornemann, D., & Hansen, U., & Schrader, U. (2006). Consumer power: a comparison of the old economy and the Internet economy. Journal of Consumer Policy, 29 (1), 3 36. Richins, M. L., & Root Shaffer, T. (1988). The role of involvement and opinion leadership in consumer word of mouth: An implicit model made explicit. Advances in Consumer Research, 15 (1), 32 36. Rivard, S. H. (1988). Factors for success for end user computing. Communications of ACM, 31 (5), 552 561. Rogers. E. M. (1995). Diffusion of Innovations The Free Press, New York. NY. Romaniuk, J. (2007). Word of mouth and the viewing of television programs. Journal of Advertising Research, 47 (4), 462 471. Rosen, E. (2000). The anatomy of buzz: How to create word of mouth marketing. New York: Doubleday. Rosen, E. (2009). The anatomy of buzz revisited. New York: Doubleday Business. Rubin, A. M. (1981). An examination of television viewing motivation. Communication Research, 8 (2), 141 165. Rubin, A. M. (1983). Televisi on uses and gratifications: Interactions of viewing patterns and motivations. Journal of Broadcasting, 27 (1), 37 52. Rubin, A. M. (1984). Ritualized and instrumental television viewing. Journal of Communication, 34 (3), 67 73. Ryan, B., & Gross, N. (1943). The diffusion of hybrid seed corn in two Iowa communities. Rural Sociology, 8 (1), 15 24.

PAGE 265

265 Salancik, G. R., & Pfeffer, J. (1978). A social information processing approach to job attitudes and task design. Administrative Science Quarterly, 23 (2), 224 252. Sch iffan, L. G., & Kunuk, L. L. (1978). Consumer behavior Englewood Cliffts, NJ: Prentice Hall. Seiple, P. (2011, July 5). 10 terrific new Twitter inforgraphics in 2011. Message posted to http://blog.hubspot.com/blog/tabid/6307/bid/19023/10 Terrific New Twi tter Infographics in 2011.aspx Sernovitz, A. (2009). Word of mouth marketing: How smart companies get people talking. Chicago, IL: Kaplan Business Publishing. Severin, W. J., & Tankard, J. W. (2001). Communication theories: Origins, methods, and uses in th e mass media New York: Addison Wesley Longman. Sheth, J. N. (1971). Word of mouth in low risk innovations. Journal of Advertising Research, 11 (3), 15 18. Short, J., Williams, E., & Christine, B. (1976). The social psychology of telecommunications New Y ork: Wiley. Silverman, G. (2001). The power of Word of Mouth. Direct Marketing, 64 (5), 47. Silverman, G. (1997). How to harness the awesome power of word of mouth Direct Marketing 60 (7), 32 37. Simonson, I., & Nowlis, S. (2000). The role of explanations and need for uniqueness in consumer decision making: unconventional choices based on reasons. Journal of Consumer Research, 27 (1), 49 68. Skowronski, J., & Carlston, D. E. (1987). Social judgment and social memory: The role of cue diagnositicity in negativ ity, positivity, and extremity biases. Journal of Personality and Social Psychology, 52 (4), 689 699. Skowronski, J., & Carlston, D. E. (1989). Negativity and extremity biases in impression formation: A review of explanations. Psychological Bulletin 105 (1) 131 142. Smith, T., Coyle, J. R., Lightfoot, E., & Scott, A. (2007). Reconsidering models of influence: The relationship between consumer social networks and word of mouth effectiveness. Journal of Advertising Research, 47 (4), 387 397. Snook, S. C., & G orsuch, R. L. (1989). Component analysis versus common factor analysis: A monte carlo study. Psychology Bulletin, 106 (1), 148 154. Sohn, D. (2009a). Disentangling the effects of social network density on electronic word of mouth (eWOM) intention. Journal o f Computer Mediated Communication, 14 (2), 352 367.

PAGE 266

266 Sohn, D. (2009b, May). The effects of social norms on electronic word of mouth intention: A comparison of three models. Paper presented at the meeting of the International Communication Association, Chicag o, IL. Spizziri, M. (2000, July). E surveys: Consider this before you start. American Society of Business Publication Editors Newsletter Retrieved from http://www.asbpe.org /archives/2000/07esurveys.htm Steinfield, C. W., Dutton, W. H., & Kovaric, P. (1989). A framework and agenda for research on computing in the home, in Salvaggio J. L. & J. Bryant (Eds.), Media Use in the Information Age, 61 86. Hillsdale, NJ: LEA. Steyer, A., Garcia Bardidia, R., & Quester, P. (200 6). Online discussion groups as social networks: An empirical investigation of word of mouth on the Internet, Journal of Interactive Advertising, 6 (2), 51 60. Straub, D., Keil, M., & Brenner, W. (1997). Testing the technology acceptance model across cultur es: A three country study. Information & Management, 33 (1), 1 11. Subramani, M. R., & Rajagopalan, B. (2003). Knowledge sharing and influence in online social networks via viral marketing. Communications of the ACM, 46, 300 307. Sun, T., Youn, S., Wu, G., & Kuntaraporn, M. (2006). Online word of mouth (or mouse): An exploration of its antecedents and consequences. Journal of Computer Mediated Communication, 11 (4), 1104 1127. Sundaram, D. S., Mitra, K., & Webster, C. (1998). Word of mouth communications: A m otivational analysis. Advances in Consumer Research, 25 (1), 527 531. Sundaram, D. S., & Webster, C. (1999). The role of brand familiarity on the impact of word of mouth communication on brand evaluations. Advances in Consumer Research, 26 (1), 664 670. Swan J. E., & Oliver, R. L. (1989). Postpurchase communications by consumers. Journal of Retailing, 65 (4), 516 533. Sweeney, J. C., Soutar, G. N., & Mazzarol, T. (2008). Factors influencing word of mouth effectiveness: receiver perspectives. European Journal of Marketing, 42 (3), 344 364. Synder, C. R., & Fromkin, H. L. (1977). Abnormality as a positive characteristic: the development and validation of a scale measuring need for uniqueness. Journal of Abnormal Psychology, 86 (5), 518 527. Tabachnik, B. G., & F idell, L. S. (1996). Using multivariate statistics (3 rd Ed.). NY: Harper Collins.

PAGE 267

267 Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup behavior. In S. Worchel & W. G. Austin (Eds.), Psychology of intergroup relations (pp. 7 24). Chi cago, IL: Nelson Hall. Taylor, H. (2000). Does Internet research work? Comparing electronic survey results with telephone survey. International Journal of Market Research, 42 (1), 51 63. Taylor, S., & Todd, P. A. (1995). Understanding information technology usage: A test of competing models. Information Systems Research, 6 (2), 144 176. Teng, W., Lu, H., & Yu, H. (2009). Exploring the mass adoption of third generation (3G) mobile phones in Taiwan. Telecommunications Policy, 33 (10 11), 628 641. Thong, J. Y. L. Hong, W., & Tam, K. Y. (2002). Understanding user acceptance of digital libraries: What are the role of interface characteristics, organizational context, and individual differences? International Journal of Human computer Studies, 57 (3), 215 242. Tian, development and validation. The Journal of Consumer Research, 28 (1), pp. 50 66. Tong, S. T., Van Der Heide, B ., Langwell, L., & Walther, J. B. (2008). Too much of a good thing? the relationship between number of friends and interpersonal impressions on facebook. Journal of Computer Mediated Communication, 13 (3), 531 549. Trusov, M., Bucklin, R. E., & Pauwels, K. (2009). Effects of word of mouth versus traditional marketing: Findings from an Internet social networking site. Journal of Marketing, 73 (5), 90 102. Tucker, L. R., Lewis, C. (1973). A reliability coefficient for maximum likelihood factor analysis. Pyschometrika, 38 (1), 1 10. Turner, J. C. (1982). Towards a cognitive redefini tion of the social group. Social Identity and intergroup relations H. Tafjel (Eds.), Great Britain: Cambridge University Press. Twitter (2011, June 30). 200 million Tweets per day. Message posted to http://blog.twitter.com/2011/06/200 million tweets per d ay.html Venkatesh, V., & Davis, F. D. (2000). A theoretical extension of technology acceptance model: Four longitudinal filed studies. Management Science, 46 (2), 186 204. Gender, social influence, and their role in technology acceptance and usage behavior. MIS Quarterly, 24 (1), 115 139.

PAGE 268

268 Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003). User acceptance of information technology: toward a unified view. MIS Qu arterly, 27 (3), 425 478. Verhagen, T., Boter, J., & Adelaar, T. (2010). The effect of product type on consumer preferences for website content elements: An empirical study. Journal of Computer Mediated Communication, 16 (1), 139 170. Vijayasarathy, L. R. (2 004). Predicting consumer intentions to use on line shopping: The case for an augmented technology acceptance model. Information & Management, 41 (6), 747 762. Vilpponen, A., Winter, S., & Sundquist, S. (2006). Electronic word of mouth in online environment s: Exploring referral network structure and adoption behavior. Journal of Interactive Advertising, 6 (2), 71 86. Vishwanath, A. (2009). From belief importance to intention: The impact of framing on technology adoption. Communication Monographs, 72 (2), 177 2 06. Vishwanath, A., & Goldhaber, G. M. (2003). An examination of the factors contributing to adoption decisions among late diffused technology products. New Media & Society 5 (4), 547 572. Vollmer, C., & Precourt, G. (2008). Always On: Advertising, Marketin g, and Media in An Era of Consumer Control. New York: McGraw Hill. Voss, K. E., Spangengerg, E. R., & Grohmann, B. (2003). Measuring the hedonic and utilitarian dimensions of consumer attitude. Journal of Marketing Research, 40 (3), 310 320. Wangenheim, F. V., & Bayn, T. (2004). The effect of word of mouth on service switching: measurement and moderating variables. European Journal of Marketing, 38 (9), 1173 1185. Wangenheim, F. V., & Bayn, T. (2007). The chain from customer satisfaction via word of mouth r eferrals to new customer acquisition. Journal of the Academy of Marketing Science, 35 (2), 233 249. Washburn, J. H., Till, B. D., & Priluck, R. (2004). Brand alliance and customer based brand equity effects. Psychology & Marketing, 21 (7), 487 508. Webster, T. (2010, April). Twitter user in America: 2010. The Edision Research/Arbitron Internet and Multimedia Study. Edison Research Retrieved from http://www.edisonresearch.com/twitter_usage_2 010.php Webster, T. (2011, June). The social habit II. The Edision Research/Arbitron Internet and Multimedia Study 2011. Edison Research Retrieved from http://www.edisonresearch.com/home/archives/2010/06/the_social_habit_frequent_social _networkers_in_america.php

PAGE 269

269 Wei, R. (2008). Motivations for using the mobile phone for mass communications and entertainment. Telematics and Infor matics, 25 (1), 36 46. West, S. G., Finch, J. F., & Curran, P. J. (1995). Structural equation models with nonnormal variables: Problems and remedies. In R. H. Hoyle (Ed.), Structural equation modeling: Concepts, issues, and applications (pp. 56 75) Thousand Oaks, CA: Sage. Westbrook, R. A. (1987). Product consumption based affective responses and post purchase processes. Journal of Marketing Research, 24 (3), 258 270. Whyte, W. H. (1954, November). The web of word of mouth, Fortune 140 143. Wilson, W. R. (19 79). Feeling more than we can know: exposure effects without learning, Journal of Personality and Social Psychology, 37 (6), 811 821. Wimmer, R. D., & Dominick, J. R. (2006). Mass media research: An introduction Belmont, CA: Thomson, Wadsworth. Winters, T. Detlefsen., R., & Davie, W. (2010, May). Quarterly retain e commerce sales 3 rd quarter 2010. U.S. Census Bureau News 2010 Retrieved February 2, 2011, from http://www.census.gov/retail/mrts/www/data/pdf/ec_current.pdf Woodside, A. & Delozier, W. (1976). Effects of word of mouth advertising on consumer risk taking. Journal of Advertising, 5 (4), 12 19. Wu, J., & Wang, S. (2005). What drives mobile commerce? An empirical evaluation of the revised technology acceptance model. Information & Management, 42 (5), 719 729. Wuyts, S. H. K., Dekimpe, M. G., Gijsbrechts, E., & Pieters, F. G. M. (2010) The connected customers: The changing nature of consumer and business markets. New York: Routledge. Xifra, J., & Grau, F. (2010). Nanoblogging PR: The discourse on publi c relations in Twitter Public Relations Review, 36 (2), 171 174. Xingyuan, W., Li, F., & Wei, Y. (2010). How do they really help? An empirical study of the role of different information sources in building brand trust. Journal of Global Marketing, 23 (3), 2 43 252. Xue, F., & Zhou, P. (2011). The effects of product involvement and prior experience on Journal of International Consumer Marketing, 23 (1), 45 58.

PAGE 270

270 Yang, J., Morris, M. R., Teevan, J., Adamic, L. A., Ackerman, M. S. (2011, June). Culture matters: A survey study of social Q&A behavior Paper presented at the meeting of the Association for the Advancement of Artificial Intelligence, Auckland, New Zealand Yang, K. C. C. (2005). Exploring the factors affecting the adoption of mobile commerce in Singapore. Telematics and Informatics 22 (3), 257 277. adoption of Internet television in Taiwan. First Monday 11 (3). Yoon, K ., Kim, C. H., & Kim, M. (1998). A cross cultural comparison of the effects of source credibility on attitudes and behavioral intentions. Mass Communication & Society, 1 (3), 153 175. Yun, G. W., & Trumbo, C. W. (2000). Comparative response to a survey exec uted by post, E mail, & web form. Journal of Computer Mediated Communication, 6 (1), doi:10.1111/j.1083 6101.2000.tb00112. Zaichkowsky, J. L. (1985). Measuring the involvement construct. Journal of Consumer Research, 12 (3), 341 352. Zaichkowsky, J. L. (1994 ). Research notes: The personal involvement inventory: Reduction, revision, and application to advertising. Journal of Advertising, 23 (4), 59 70. Zajonc, R. (1968). The attitudinal effects of mere exposure, Journal of Personality and Social Psychology Mono graph, 9( 2), 1 27. Zeithaml, V. A. (1981). How consumer evaluation processes differ between goods and services in Donnelly, J. H., & George, W. R. (Eds), Marketing of Services American Marketing Association, Chicago, IL, 186 190. Zeithaml, V. A., Berry, L L., & Parasuraman, A. (1993). The nature and determinants of customer expectations of service. Journal of the Marketing Science, 21 (1), 1 12. Zeithaml, V. A., Parasuraman, A., & Berry, L. L. (1985). Problems and strategies in services marketing, Journal of Marketing, 49 (2), 33 46. Zhou, Y. (2008). Voluntary adopters versus forced adopters: integrating the diffusion of innovation theory and the technology acceptance model to study intra organizational adoption. New Media & Society, 10 (3), 475 496. Zhu, F., & Zhang, X. (2010). Impact of online consumer reviews on sales: The moderating role of product and consumer characteristics. Journal of Marketing, 74 (2), 133 148.

PAGE 271

271 BIOGRAPHICAL SKETCH Hyunsang Son graduated summa cum laude, ranked 1 st in his class, with a Bachelor of Arts in Mass Communication, with a minor in sociology, from Chung Ang Florida, he finished five doctoral advance seminars Brand Management, Persuasion Theory, Ex perimental Methods in Mass Communication, Social Influence on New Media Environment, and Health Communication Campaigns in addition to core courses. He also actively participated in research and successfully accepted a total of 9 papers to the following na tional conferences with his coauthors: The 97th National Communication Association (NCA) Annual Convention, 2011, the 61th International Communication Association (ICA) Annual Conference 2011, Annual Conference of 2011 American Academy of Advertising (AAA) the 14th International Public Relations (IPR) research conference, the 95th and 96th Association for Education in Journalism and Mass Communication (AEJMC), Annual Conference 2010, 2011, and the AEJMC Mid winter Conference 2009. In particular, his paper was selected as the top paper and he received the Steve Lacy Top Paper Award in Media Management and Economics Division at the 2011 AEJMC conference. His research interests include online consumer behavior, brand management, media management, and new commu nication technologies such as social media. In addition, he worked as a research assistant with Dr. Moon Lee on health communication campaigns He graduated in the fall of 2011 and received an M.A. in mass communication from the University of Florida.