UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Interactions between a web site and its customers : a relationship building approach Kumar, M. S. N. 2003

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2003-860302.pdf [ 22.77MB ]
Metadata
JSON: 831-1.0099744.json
JSON-LD: 831-1.0099744-ld.json
RDF/XML (Pretty): 831-1.0099744-rdf.xml
RDF/JSON: 831-1.0099744-rdf.json
Turtle: 831-1.0099744-turtle.txt
N-Triples: 831-1.0099744-rdf-ntriples.txt
Original Record: 831-1.0099744-source.json
Full Text
831-1.0099744-fulltext.txt
Citation
831-1.0099744.ris

Full Text

Interactions Between a Web Site and Its Customers: A Relationship Building Approach By Nanda Kumar B.Eng., Anna University, India, 1994 P.G.D.B.A., Narsee Monjee Institute of Mgt. Studies, India, 1997 A Thesis Submitted In Partial Fulfillment Of The Requirements For The Degree Of Doctor of Philosophy In The Faculty Of Graduate Studies Sauder School of Business We accept this thesis as conforming to the required standard © Nanda Kumar, 2003 The University of British Columbia In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of The University of British Columbia Vancouver, Canada Date DE-6 (2/88) Abstract This research makes a case for treating an electronic commerce web site as a social actor and argues that LT-enabled support for personalization systems and virtual communities has a significant impact on the perceived communication characteristics of a web site. This research studied the impact of two communication characteristics - adaptiveness and connectivity of a web site. Adaptiveness indicates the extent to which a web site adapts itself to support the needs of its customers whereas connectivity refers to the ability of a web site to link potential customers with other visitors. Further, synthesizing research from communication, media choice and technology adoption literatures, this thesis proposes social presence as the experiential construct and perceived usefulness as the utilitarian construct that will mediate the relationship between communication characteristics of a medium and customer loyalty. A survey and a laboratory experiment were conducted to test the linkages proposed above. The survey showed that while adaptiveness impacted on both social presence and perceived usefulness, connectivity had an impact only on social presence and an insignificant impact on perceived usefulness. Social presence did not influence perceived usefulness while both social presence and perceived usefulness had a significant impact on customer loyalty. By filtering Amazon.com real-time, the lab experiment was conducted to specifically study the causal impact of a web site's support for personalization and virtual communities. Support for personalization had a strong impact on adaptiveness, whereas support for consumer reviews had strong effect on connectivity. Interestingly, support for ii personalization had a weaker, but significant effect on connectivity and support for consumer reviews had a similar effect on adaptiveness. Data from the experiment was also analyzed using mediation analyses as well as partial least square analysis to show that the general pattern of results observed were consistent across methods thus improving the confidence in the research model proposed. This research by proposing social presence as a crucial experiential predictor of customer loyalty has addressed an important gap that exists in our current understanding of web users' behavior. It also makes a key contribution by empirically showing that a web site's support for IT-enabled personalization and virtual communities do have a significant impact not only on the perceived communication characteristics, but also on customer loyalty through social presence and perceived usefulness. ii i T A B L E O F C O N T E N T S A B S T R A C T i l INDEX O F APPENDICES V i INDEX O F FIGURES V l i INDEX O F T A B L E S VIII A C K N O W L E D G E M E N T S IX 1. INTRODUCTION 1 1.1 R E S E A R C H O B J E C T I V E S 2 2. T H E I N T E R P L A Y B E T W E E N C O M M U N I C A T I O N C H A R A C T E R I S T I C S O F T H E MEDIA AND I N T E R F A C E DESIGN DECISIONS 5 2.1 C O M M U N I C A T I O N C H A R A C T E R I S T I C S O F T H E M E D I A 5 2 . 2 I M P A C T O F T E C H N O L O G I E S U S E D A N D I N T E R F A C E D E S I G N D E C I S I O N S M A D E 1 1 2 .2 .1 P E R S O N A L I Z A T I O N S Y S T E M S ; 1 2 2 . 2 . 2 V I R T U A L C O M M U N I T I E S ' 1 7 2 .3 G E N E R A L R E S E A R C H M O D E L : A R O A D M A P 1 8 3. S O C I A L P R E S E N C E : T H E O R E T I C A L B A C K G R O U N D AND D E V E L O P M E N T 20 3.1 R E L A T I O N A L C O M M U N I C A T I O N 2 0 3 .2 D o P E O P L E T R E A T C O M P U T E R S A S S O C I A L E N T I T I E S ? - T H E O R Y O F S O C I A L R E S P O N S E 2 0 3 .2 P R E S E N C E : D I F F E R E N T C O N C E P T U A L I Z A T I O N S 2 2 3.3 S O C I A L P R E S E N C E 2 5 3 .4 S O C I A L P R E S E N C E : F R O M C O N N E C T I O N W I T H O T H E R U S E R S T O C O N N E C T I O N W I T H T H E W E B SITE : 2 7 4. R E S E A R C H M O D E L AND H Y P O T H E S E S D E V E L O P M E N T 31 4 . 1 : R A T I O N A L E B E H I N D T H E R E S E A R C H M O D E L 3 1 4 . 2 H Y P O T H E S E S 3 4 4 .3 R E S E A R C H M E T H O D O L O G Y : A B R I E F I N T R O D U C T I O N 3 7 5. STUDY 1: S U R V E Y 42 5.1 P R O C E D U R E S 4 2 5 .2 R E S U L T S A N D D I S C U S S I O N 4 4 5.2 .1 I N I T I A L R E S E A R C H M O D E L : 4 5 5 .2 .2 R E S E A R C H M O D E L : A F T E R M O D I F I C A T I O N 4 9 6. STUDY 2: E X P E R I M E N T 56 6.1 P R O C E D U R E S 5 6 6 .2 R E S U L T S A N D D I S C U S S I O N 6 1 6 .2 .1 I M P A C T O F W E B S I T E I N T E R F A C E F E A T U R E S O N C O M M U N I C A T I O N C H A R A C T E R I S T I C S 6 4 iv 6.2.2 MEDIATION ANALYSIS 73 6.2.3 PARTIAL LEAST SQUARES ANALYSIS 81 7. CONCLUSIONS 85 7.1 THEORETICAL CONTRIBUTIONS 87 7.2 LIMITATIONS 88 7.3 IMPLICATIONS FOR PRACTICE AND RESEARCH 91 REFERENCES 94 APPENDICES 99 v Index of Appendices Appendix 1: Operationalization of Constructs and Original List of Items 99 Appendix 2: Questionnaire and Instructions used in Study 1 100 Appendix 3: Questionnaire and Instructions used in Study 2 .' 114 Appendix 4: Instructions to Research Assistants for Study 2 130 Appendix 5: Code Used To Filter Amazon.com for Experimental Condition 1 134 Appendices 6.1 - 6.3: Amazon.com - Screenshots 160 Appendices 7.1-7.11: Experimental Manipulation - Screenshots 164 vi Index of Figures Figure 1: Overview of Personalization Techniques 14 Figure 2: General Research Model 19 Figure 3: Research Model Tested... 31 Figure 4: Conceptual Model Tested Using Study 1 38 Figure 5: Conceptual Model Tested Using Study 2 40 Figure 6A: Research Model - Before Modification 47 Figure 6B: Research Model - After Modification 48 Figure 7: Interaction Pattern for Adaptiveness 67 Figure 8: Interaction Pattern for Connectivity 67 Figure 9: Research Model Using PLS (Study 2) 82 vii Index of Tables Table 1: Media Characteristics Developed in Previous Research 5 Table 2: Recasting Social Presence to Connote a Broader Meaning 29 Table 3: Web Sites Used in Study 1 42 Table 4A: Discriminant Validity - Pair-wise Chi Square Tests 46 Table 4B: Estimates of Composite Reliability and Variance 49 Table 4C: Fit Indices for the Research Model: Before Modification 49 Table 5A: Discriminant Validity - Pair-wise Chi Square Tests 51 Table 5B: Estimates of Composite Reliability and Variance 51 Table 5C: Fit Indices for the Research Model 52 Table 6: Test Results 53 Table 7: Experimental Manipulations 56 Table 8A: Manipulation Check Scales and Their Reliabilities 61 Table 8B: Descriptive Statistics for Manipulation Check Variables.. 62 Table 8C: A N O V A Table for Manipulation Check 63 Table 9A: Descriptive Statistics for Perceived Communication Characteristics 64 Table 9B: Reliability Values for Constructs Used in the Study 65 Table 10: A N O V A Table for Perceived Communication Characteristics 65 Table 11: A N O V A Table for Other DVs (IV-Personalization Support) 71 Table 12: A N O V A Table for Other DVs (IV-Consumer Review Support) 71 Table 13: Descriptive Statistics for Other DVs (IV-Personalization Support) 72 Table 14: Descriptive Statistics for Other DVs (IV-Consumer Review Support) 72 Table 15: Social Presence As A Mediator Between Communication Characteristics and Loyalty 75 Table 16: Social Presence As A Mediator Between Adaptiveness and Loyalty 76 Table 17: Social Presence As A Mediator Between Connectivity and Loyalty 76 Table 18: Social Presence As A Mediator Between Communication Characteristics and Perceived Usefulness 77 Table 19: Social Presence As A Mediator Between Adaptiveness and Perceived Usefulness 78 Table 20: Social Presence As A Mediator Between Connectivity and Perceived Usefulness 78 Table 21: Perceived Usefulness As A Mediator Between Communication Characteristics and Loyalty 79 Table 22A: Estimates of Composite Reliability and Variance 83 Table 22B: Sqaure-root of A V E and Correlation Between The Latent Constructs 83 viii Acknowledgements I would like to take this opportunity to thank Dr. Izak Benbasat for helping me understand the process of research and for always having the time. I am still learning and Izak's guidance over the last six years taught me not to lose sight of the forest for trees. I would also like to thank Dr. Jai-Yeol Son and Dr. Peter Darke for spending time giving me valuable insights on this work. I also need to acknowledge the support I received from Dr. Yair Wand and Dr. Carson Woo during my tenure here as a Ph.D. student. And, of course this acknowledgement would not be complete without the mention of my fellow Ph.D. students Andrew, Arturo, Dongmin, Dorit, Genevieve, Oona, Paul and Tim. Thank you. Finally, I would like to thank my mother without whom none of this would have been possible and Natalia who helped me see a world full of possibilities. i x 1. Introduction The last few years has seen an explosion of e-commerce and a preponderance of web ' sites for selling products and enhancing corporate image. Even though the initial hype surrounding the dot.com industry has given way to more modest realities (Gartner 2000), shopping on the Internet is here to stay and will grow. In spite of the major emphasis on reaching customers via web presence, the art of web design has remained just that... an art. Most of the design guidelines for generating web interfaces concentrate predominantly on facilitating efficient communications between the user and the web site. We argue that a second dimension of information exchange that emphasizes the relationship-based' aspects of communication, such as, creating a positive company image, trust, and a favorable browsing experience, becomes more important in a virtual environment as the web interface is the key contact point between an organization and its customers for e-commerce related activities. Though researchers have attempted to address and understand how the emergence of new media with multiple capabilities impacts on the way a firm sells its products on-line, the interesting research questions in this area span several fields that still need significant research. In the past few years, there has been considerable research done on what makes an e-commerce site successful. A wide range of predictor variables - such as interactiveness of the web site (Ghose et al. 1998; Palmer 2002; Zhu et al. 2002), navigation and content (Davern et al. 2000; Lohse et al. 1998; Palmer 2002), information control (Ariely 2000), interactive decision aids (Haubl et al. 2000), download times and responsiveness (Dellaert et al. 1999; Voss 2000; Weinberg 2000), and even background color and 1 pictures (Mandel et al. 1999) - have been investigated in the e-commerce context. Most of these studies look at the effect of the aforementioned predictor variables on several important issues such as customer's overall satisfaction, trust, loyalty, decision quality and intention to buy/re-visit the web site. Hoffman and Novak (Hoffman et al. 1996) investigated flow as an experiential construct that positively correlated with a compelling on-line experience for fun, non-work related activities. Shopping can be a high involvement activity. A major obstacle to consumers buying on-line is the inability of most of the shopping sites to engage the users cognitively as well as emotionally. Our research takes a relational rather than a transactional approach and argues that web sites (and hence organizations) can cultivate and develop meaningful and rich relationships with its customers especially for B2C activities. In this era of new retail, shoppers have become guests, shopping has become an experience and malls have become entertainment centers with communities. Realizing the value of "shopping as experience", many on-line stores have begun efforts to enhance the positive experiences (e.g. increasing their involvement, enhancing their positive feelings) for their customers. This research will specifically investigate the effect of two such efforts: personalization and support for virtual communities. 1.1 Research Objectives This research makes a case for treating the web site as a social entity and proposes the construct Social Presence (SP) to capture the relational aspect of the communication between a web site and its visitors. This research focuses on social presence in a specific context - online shopping. The objectives of this research is to investigate specifically the 2 effect of web sites' support for personalization systems and virtual communities on the perceived communication characteristics of the media, and to investigate if and how social presence mediates the relationship between specific communication characteristics of the media and a user's evaluation of the web site. To accomplish this objective, the thesis reviews previous research work to show that communication characteristics of a medium are perceived, and can vary within that medium, depending on how the medium is configured. This research also proposes social presence (SP) as the construct of choice to capture the relational aspect of communication between a web site and its visitors. Social presence refers to the degree to which a medium allows a user to establish personal connection with other users (Short et al. 1976). Chapter 2 reviews the theoretical work on communication characteristics of the media and argues that enabling technologies used to configure a web site will have an impact on the communication characteristics of the media. It also briefly introduces the general research model proposed by this research that acts as a roadmap for the rest of the thesis. Chapter 3 then introduces the theoretical work that supports our contention that a web site can be treated as a social entity. It further reviews the work done on social presence and lays the groundwork for the development of the research model proposed in Chapter 3. Chapter 4 further expands on this general research model and fully develops the specific hypotheses to be tested in this research. Chapter 5 discusses the research methodology and the results for Study 1, while Chapter 6 discusses the research methodology and the results for Study 2. Chapter 7 concludes the thesis by reviewing the salient findings of 3 this work, discusses the significance of the findings and explores further avenues research suggested by the results of this thesis. 2. The Interplay Between Communication Characteristics of the Media and Interface Design Decisions 2.1 Communication Characteristics of the Media In the past two decades, many theories of media choice and effects (Daft et al. 1986; Dennis et al. 1999; Fulk et al. 1993; Short et al, 1976; Walther 1992) have developed a set of media characteristics and investigated their impact on media choice and media use. The early theories (Daft and Lengel, 1986; Short, Williams and Christie, 1976) on media choice and use viewed these media characteristics as invariant and assumed that all these characteristics are salient to the individual. The latter theories (Fulk and Boyd, 1991; Walther, 1992; Dennis and Valacich, 1999) acknowledged the impact of social influence and recognized media characteristics as perceived characteristics. Table 1 reviews the list of communication characteristics of media that have been developed earlier. Table 1: Media Characteristics Developed in Previous Research Daft and Lengel (1986), Daft, Lengel and Trevino (1987) - Media Richness Theory a) Immediacy of Feedback: The extent to which a medium allows users to give rapid feedback on the communications they receive. b) Multiple Cues: The number of cues - such as text, verbal cues or non-verbal cues - available through which information can be communicated. c) Language Variety: The range of meaning that can be conveyed with language symbols (numeric information to natural language). d) Personal Focus: The extent to which a sender can personalize the message to suit the needs and the current situation of the receiver. Sproull (1991), Valacich et al., (1993) - Additional List of Characteristics for the Electronic Media 5 a) Mul t ip l e Addressabili ty: The extent to which a message can be forwarded to several persons at the same time. b) External ly Recorded Memory : The extent to which a message can be stored for further processing. c) Concurrency: The capacity of media to support distinct communication episodes without distracting from any other episodes that may be occurring simultaneously. Dennis and Valac ich , (1999) - M e d i a Synchronicity Theory: Refined L i s t of M e d i a Characteristics a) Immediacy of Feedback: The extent to which a medium allows users to give rapid feedback on the communications they receive. b) Symbol Variety: The number of ways in which information can be communicated. This subsumes 'multiple cues' and 'language variety' developed by Daft and Lengel (1986). c) Parallel ism: The number of simultaneous conversations that can exist effectively. Similar to 'concurrency' developed by Valacich et ah, (1993) and 'multiple addressability' developed elsewhere (Please see Dennis and Valacich (1999) for constructs they looked at before arriving at this list of characteristics). d) Rehearsability: The extent to which the sender can rehearse or fine-tune the message before sending. This similar to 'editability' developed elsewhere. e) Reprocessability: It is the extent to which a message can be reexamined or processed again within the context of the communication event. This is similar to 'externally recorded memory' developed by Sproull (1991). Burgoon et a l . , (2000): Burgoon et al., (2000) identified an extensive set of properties intrinsic to Face-to-Face (FtF) communication that might be retained, supplemented, amplified or suppressed in HCI and C M C formats. They further argued that those properties individually and/or collectively would account for observed differences, in cognitions, communications and outcomes observed across mediated and non-mediated, human-human and human-computer interaction. These properties are: a) Modal i ty : The extent to which a medium can support symbol variety to present 6 rich information (similar to symbol variety). b) Synchronicity: Whether the interaction occurs in real-time or with a time delay (similar to immediacy of feedback). c) Contingency: The extent to which a person's queries and responses are dependent on the prior responses of the participating entity (some similarities to 'personal focus'). d) Participation: The extent to which senders and receivers are actively engaged in the interaction. e) Identification: The extent to which the participants are fully or partially identified or anonymous. f) Mediation: Whether the communication format is mediated or not. g) Propinquity: Whether the participants are in the same location or geographically dispersed. h) Anthropomorphism: The degree to which the interface simulates or incorporates humanlike characteristics. Dov Te'eni (2001): The Three dimensions of Media Richness a) Interactivity: The potential for immediate feedback from the receiver. It is manifested by simultaneous, synchronous, and continuous exchange of information (some similarities to 'immediacy of feedback'). b) Adaptiveness: The potential to personalize a message to a particular receiver (similar to'personal focus'). c) Channel Capacity: the potential to transmit a high variety of cues and languages (similar to 'multiple cues', 'language variety' and 'symbol variety') Most of these characteristics, with the exception of those by Burgoon et al., are in some way associated with media richness theory. As Dennis and Kinney (Dennis et al. 1998) pointed out, media richness theory has, in the past, been confused with media characteristics. 7 ' While a few studies have attempted to measure media richness by asking about cues, feedback, language, and personal focus, this is a serious confound. Cues etc., are theorized to affect richness; they are not part of richness. Media richness is defined as "the ability of information to change understanding within a time interval" (Daft and Lengel, 1986, p.560)' Dennis and Kinney (1998) persuasively argue that researchers may benefit by examining fundamental characteristics of media (such as immediacy of feedback and not media richness) to understand performance effects. We agree with Dennis and Kinney (1998) and use 'media characteristics' as a starting point in our research model. This research does not attempt to investigate nor apply media richness theory in the web-shopping context. Rather, this research seeks to portray web site as a valid social actor and attempts to characterize the relationship that emerges between a web site and its visitors. Dennis and Valacich (1999) argued that the communication capabilities of the same medium could vary depending on the way it is configured. Previous research has also shown that media characteristics are perceived (by the user) and the characteristics that are salient to a particular individual may not be salient to others. Thus, we argue that it is more appropriate to use the term 'perceived communication characteristics' rather than 'media characteristics'. This research would be focusing on the impact of some of these perceived communication characteristics within the same medium and not across media. Table 1 also demonstrates the inter-relatedness of these 'communication characteristics' across taxonomies. For example, synchronicity is quite similar to immediacy of feedback. Modality is similar to symbol variety and multiple cues. Even the characteristics 8 developed within the same theoretical framework seem to be interrelated. For example, manipulating multiple cues would have an effect on immediacy of feedback (Dennis et al. 1998) as audio/video cues inherently tend to give some immediate feedback when compared to pure text. This research also introduces connectivity - the ability of a medium to bring together people who share cornmon interests or goals - as one of the communication characteristics to the list of properties summarized in Table 1. While constructs similar to adaptiveness have been defined and used in previous research under various labels (Table 1 - personal focus, contingency, adaptiveness), connectivity is introduced as a media characteristic for the first time in this research. One of the major objectives of this research is to investigate the impact of personalization and support for consumer reviews. This research argues that adaptiveness and connectivity are the pertinent communication characteristics impacted by IT-enabled personalization of web sites and support for consumer reviews and thus the focus of this research would be restricted to these two communication characteristics. Consistent with the spirit of the definitions in Table 1, we define adaptiveness as the ability of a medium to adapt itself based on the requirements and the situation of the user. Connectivity is the extent to which and the ease with which individuals who share common goals and interests find each other. This characteristic is one of the more important qualities that distinguish Internet1 from other traditional media as this allows 1 This research views Internet as the physical network of networks; World Wide Web, e-mail, ftp and other programs as applications that use Internet. 9 spontaneous relationships to be developed among spatially and temporally distributed participants. While Face-to-Face (FtF) communication may have many advantages over computer mediated communication (some of which can be overcome as participants get to know each other over time), FtF communication has a decided disadvantage over other communication formats vis-a-vis connectivity. The ease with.which web as a medium amplifies and supplements this often overlooked characteristic, could partially help offset some other limitations that web may face in facilitating relationships among entities. While organizations recognize the importance of supporting team members working in different locations using different media and are willing to invest time and resources to, acquire associated group support systems, individuals who share a common affinity for the music of 'Miles Davis' may not be willing to do the same if this would cost them too much money and time. But, the advent of Internet has made it very easy for individuals who share common interests to come together on-line. Usenet newsgroups were the earliest examples of such groups and the World Wide Web now makes it even easier to form sophisticated virtual communities. Web sites can act as conduits through which such relationships can be forged. Sometimes, these interactions are transactional in nature (a user browsing through a recommendation by another user who bought the same product). Other times, the relationship blossoms over a period of time (virtual communities at Yahoo.com, MSN.com). In short, the Internet in general and the World Wide Web in particular, has made this quantum leap from 'virtual teams' to 'virtual communities' possible. This research argues that if a social actor (including a web site) helps another social actor (visitor to a web site) connect with others who share similar 10 interests or goals (virtual community), then this would foster a positive relationship between those actors (web site and the visitor to that site). 2.2 Impact of Technologies Used and Interface Design Decisions Made Companies are beginning to employ a wide variety of technologies to build meaningful relationships with their customers, in anticipation of the emergence of web interface as a major point of contact between companies and customers. While web interface may not be the only point of contact that customers use, companies are aware of the advantages of using the web to cater to the needs of the customers as much as possible. This concept of self-service by customers runs the whole gamut right from the pre-purchase stage all the way to the post-purchase stage (Cenfetelli et al. 2003). This concept of 'self-service' not only reduces costs for the company, but also increases customer satisfaction by addressing the transactional as well as the relational needs of the customer. The personalization systems used by several leading B2C retailers such as Amazon.com and Landsend.com are excellent examples of customers' transactional and relational needs being met especially at the pre-purchase stage. Dennis and Valacich (1999) argue that the same medium could possess different communication capabilities depending on how it is configured. The versatility of web allows different configurations to be used to support the avowed goals of a web site. Different levels of support for these technologies and the associated web interface design decisions, will affect specific communication characteristics of the web as a medium. For example, a decision on what combination of text, images, audio and video to include, 11 affects the modality (symbol variety) of the medium. A decision to include instant messenger systems affects the synchronicity (immediacy of feedback) of the medium. Decisions made on the type of personalization systems (a wide variety of solutions available based on cost and sophistication required - intelligent agents, collaborative filtering, content filtering, and constraint based recommendation systems) to be used in a web site influences, the adaptiveness (personal focus, contingency) of the medium; decision to include support for some form of consumer reviews will enhance the connectivity of the medium. This research specifically investigates the effect of two such IT-enabled web site design decisions made by companies: support for personalization systems and support for virtual communities. 2.2.1 Personalization Systems Most of the technologies and tools that companies use to manage their relationship with their customers usually fall under the banner of Customer Relationship Management (CRM) System. Even though personalization is just one piece of the C R M pie, it is a very crucial piece as effective personalization significantly enhances the ability of the organization to initiate a discourse with its customers to the point where any and all of these dialogues are seamlessly integrated with the database's historical and transactional information. Based on the data stored in these databases and recent history (the pages customers viewed in the last session), web sites automatically attempt to improve their organization and presentation of content. These web sites armed with a host of appropriate tools - including intelligent agents, recommendation engines and the like -12 attempt to anticipate the context of the interaction with their customers and personalize each customer's shopping experience (Andre et al. 2002; Billsus et al. 2002). Personalization is a process of providing special treatment to a repeat visitor to a web site by providing information and applications matched to the visitor's interests, roles and needs (Chiu 2000; Cingil et al. 2000). Personalization is needed to successfully manage customer relationships, promote the right product the customer is interested in, and manage content. Most of the advanced personalization might require sophisticated data mining techniques and the ability to display dynamic content without seriously compromising system resources (dynamic display of content will usually mean increased download time). There are three well-known techniques for personalization. Rules-based personalization modifies the content of a page based on specific set of business rules. Cross-selling is a classic example of this type of personalization. The key limitation of this technique is that these rules must be specified in advance. Personalization that uses simple filtering techniques determine the content that would be displayed based on predefined groups or classes of visitors and is very similar to personalization based on rules-based techniques. Personalization based on content-based filtering analyzes the contents of the objects to form a representation of the visitor's interest (Chiu 2000). This would work well for products with a set of key attributes. For example, a web site can identify the key attributes of movies (VHS, DVD) such as drama, humor, violence etc., and can recommend movies to its visitors based on similar content. Personalization based on 13 collaborative filtering offers recommendations to a user based on the preferences of like-minded peers. To determine the set of users who have similar tastes, this method collects users' opinion on a set of products using either explicit or implicit ratings (Chiu 2000). Please see Figure 1 for an illustration of how a web site could use all three personalization methods to best serve the customer. Collection of Visitor Information Implicit Profiling Explicit Profiling Legacy Data Personalization Processes Three Filtering Techniques: Simple, Content Based and Collaborative Filtering Business Rules Dynamic Assembly and Display of Content as Web Page Figure 1: Overview of Personalization Techniques (Based on Chiu et al., 2000) An intelligent way to make the web site adaptive is to use not only the information provided by the user (such as rating the music and of course log-in information), but also information that could be collected based on the click-stream trail left behind by the user. These two different sources of collecting information about the consumer are known as 14 explicit and implicit profiling. As the name implies, explicit profiling collects information about a user by directly asking him/her information about self and product likes and dislikes. This information is collected over a period of time and is stored in the customer database as a profile. Typically, the user would need to log-in in order for the web site to access the profile and provide personalized content. Even though cookies can be used to store this information on a user's hard disk, companies prefer to use the log-in approach as this allows the web site to identify the unique visitor (cookies won't help if the computer is shared within a family or if the customer accesses the web site from a different web site - say from the office). Implicit profiling typically tracks the actual behavior of the customer while browsing the web site. This method of collecting information is transparent to the user. While less intrusive, this method of collecting information has implications for the user's privacy. Typically, information is collected about the pages the consumer visited, the products he/she looked at and the time user spent on these pages. If a (brick and mortar) company has good information systems, the data from explicit and implicit profiling can be merged with the off-line customer information (see legacy user data in Figure 1) to effectively present a seamless web interface to the customer. Ideally, a company should use all sources of information it has about the customer. However, when a user visits a shopping web site (even a repeat user), it would be unsound business practice to expect the user to log-in every time to access personalized content. Hence, a good web site would use implicit profiling and make a few 15 assumptions about the likes and dislikes of the customer to provide adaptive content to the customer. For example, if a customer visits a specific product page, it is a good idea to assume that the customer is interested in that particular product and provide content personalized to that user's need. Of course, in most cases, even if the user logs in, the web site may have little else other than previous purchase history if the user has not provided any specific information on the products he/she likes. The level and extent of personalization offered by the web site will have an effect on the communication characteristics of the media. This research argues that different levels of support provided for personalization will specifically impact on the adaptiveness (similar to contingency used by Burgoon et al., 2000) of the web site. This is best illustrated by discussing a real life example using Amazon.com. Appendices 6.1 to 6.3 include three screen shots that show the different ways Amazon.com attempts to personalize the experience of the customer. When the user enters the web site, he is invited to log in if desired. Once the user logs in, Appendix 6.1 shows the web page that is dynamically created by Amazon.com. This page recommends products to the user based on past purchase history and on the explicit ratings provided by the user to a set of select items. Appendix 6.2 shows the product page for a book the user is interested in. The column on the left hand side of this page shows the associated related content about the product that is displayed on this page. Appendix 6.3 shows the page tailor-made for the user based on his recent browsing history and past purchase history. Of course, the scenario described above assumes that the user logged into the web site at the outset. An intelligent web site can still adapt its content in its product page by assuming that the user is interested in the 16 product he/she is browsing. Accordingly, the product page shown in screen shot 6.2 can be personalized even with out an explicit log-in by the user. If the same user were to shop for the book that he is interested in a physical store, he might have approached the sales clerk (or even a friend he had taken along for the shopping trip) for help locating the product. Now, when he mentions to his friend that he is interested in this specific book, music or movie, then it is possible to imagine a conversation happening along the lines discussed above. Of course, the above discourse with the web site is limited by the need for a shared context. The conversation will not be totally indeterminable in terms of context and content and may not move along in any arbitrary direction as is possible in a conversation with a friend. But, this research argues that there are enough cues in the discourse initiated by the personalization system of Amazon.com that is enough to give the user the impression that the conversation is contingent within that shared context. 2.2.2 Virtual Communities To enhance the relationship with the customers, companies can also provide support for virtual communities, as this will facilitate access to free-flowing and unstructured information beyond what is provided by the computer agents (Jones 1997; Preece 2001; Preece 2002). For example, companies can aggregate the opinions of consumers on a particular product and present them to a new user who is browsing that product page. Depending on the level of support provided by the web site, the new user can also get in touch with another consumer he/she might identify with, as is the case with Amazon.com. 17 Recent research that elicited participant's beliefs about the goals that can be achieved through Internet found that the participants attached considerable importance to better social relations and new friendships (Capozza et al. 2003). A recent study by Brown et al., (Brown et al. 2002) shows that community features create value for a shopping web site. Their study showed that community users accounted for about one-third of the visitors to the e-tailing sites surveyed and that they also generated two-thirds of the sales (2000 transactions worth one million dollars). Practioners have long argued that having a vibrant community in the form consumer reviews is crucial for the success of e-commerce web sites such as Amazon.com and Ebay.com (Brown et al. 2002; Kirkpatrick 2002). This research argues that providing support for consumer reviews facilitates formation of one type of virtual community. High level of support (user rating and information about the user) for consumer reviews increases connectivity.by giving the user an opportunity to express his opinions and by facilitating formation of informal peer groups/communities. 2.3 General Research Model: A Road map This research attempts to investigate the effect of support for different levels of personalization and different levels of consumer reviews (one form of virtual community) on the relationship that companies develop with its customers. The communication characteristics of the medium can be manipulated by the interface design decisions made and underlying technologies used. This in turn affects social presence and subsequently the customer loyalty (See General Research Model in Figure 2). For example, based on the book that a customer is interested in, Amazon.com recommends similar books (simple filtering/rule-based) and books bought by users with similar tastes (collaborative 18 filtering). This type of personalization effort is hypothesized to increase adaptiveness afforded by the web site. This in turn will positively influence social presence and then, user's evaluations of a web site. Technology,: Interface Design Decisions Customer Loyalty Figure 2: General Research Model This research model also proposes that the relationship between the perceived communication characteristics and evaluation criteria will be mediated by perceived usefulness (PU). To support this part of the model, the research borrows from previous research on Theory of Planned Behavior (Ajzen 1991) and Technology Acceptance Model (Davis 1989) that posited specific relationships between attitudes, belief structures, behavioral intent and actual usage. Our research model includes beliefs (perceived usefulness, social presence) and behavioral intent (loyalty). The research model will be fully developed in Chapter 4. With help from the next Chapter that discusses the nature of social presence, the theoretical support and arguments provided in Chapter 4 will also clarify the reasons why social presence is added as an additional belief structure. Perceived Communication Characteristics ' .(Adaptiveness, . Connectivity) - -SP PU 19 3. Social Presence: Theoretical Background and Development 3.1 Relational Communication Researchers in the Communication field have drawn a conceptual distinction between the content and the relational aspects of communication (Watzlawick et al. 1967). Any given interaction can be analyzed in terms of the content of the message exchange (verbal and non-verbal) and in terms of what it reveals about the nature and structure of the relationship between two participants. The relational aspect of communication can be used to characterize and measure the extent of interpersonal relationships between participants (Burgoon et al. 1987; Dillard et al. 1999; Kumar et al. 2001; Kumar et al. 2002; Walther 1992; Watzlawick et al. 1967). We propose that social presence could be used as a proxy to characterize the relational dimension of communication between human participants. This raises two important questions: • Is it appropriate to draw parallels between "relationship among humans" and "relationship between a web site and its users"? • What is the nature of the construct Social Presence (SP)? Is it appropriate to use social presence in the context of relationship between a web site and its users? 3.2 Do People Treat Computers as Social Entities? - Theory of Social Response Reeves, Nass and their colleagues at the Center for the Study of Language and Information at Stanford have shown that even experienced users tend to respond to computers as social entities (Nass et al. 1995a; Nass et al. 1995b; Nass et al. 1994). These studies indicate that computer users follow social rules concerning gender stereotypes and politeness, and that these social responses are to the computer as a social entity and 20 not to the programmer. When explicitly asked by the researchers, most users consistently said that social responses to computers were illogical and inappropriate. Yet, under appropriate manipulation, they responded to the computer as though it were a social entity. This, in fact, is the essence of the 'Theory of Social Response' (Moon 2000; Reeves et al. 1997). Youngme Moon (2000, p325), in a recent paper, argues that "More specifically, when presented with a technology possessing a set of characteristics normally associated with human behavior — such as language, turn taking, and interactivity — humans respond by exhibiting social behaviors and making social attributions. Consequently, many of the same social conventions that guide interpersonal behavior are also evident in human-computer interaction, even when the conventions no longer make 'rational' sense in this different context" She further argues that, "The theory of social response also leads to some interesting insights regarding how people generate the psychological boundaries associated with these relationships. The evidence suggests that these boundaries are structured by the physical and behavioral characteristics of the computer. For example, in studies involving text based computer interfaces with unremarkable features, researchers have^ found that users tend to psychologically orient themselves to the box... In short, it appears that the most salient cue - whether it be the box, the voice (voice based output) or the agent (computer interfaces with text based or graphics based computer agents) - is also likely to become the relational target of the social response" We believe that there is value in conceptualizing the web site as a social actor and that the web site can be equated to the 'agents' mentioned above in terms of source orientation. There are several points-of-contact between a web site and its users that will result in responses by the users not unlike the way they would respond to a social interaction. Having established the notion that "inter-personal" interactions between a web site and its users may not be so far-fetched after all, we now turn to the other important issue: To what extent does social presence represent the "relational" aspect of communication between a web site and its users? 21 3.2 Presence: Different Conceptualizations Researchers from a wide array of fields (such as communication, computer science, psychology, information systems, organizational behavior and cognitive science) have studied presence, even though the conceptualizations made different assumptions about the construct and the domain it is situated in (from the psychophysical to psychological). Lombard and Ditton (Lombard et al. 1997) identified six different conceptualizations of presence: Presence as "Socia l Richness": This is the conceptualization (social presence) that is familiar to most of the researchers in Management Information Systems and Organizational Behavior field. Social Presence is the degree to which a medium allows a user to establish a personal connection with others (Short et al. 1976). It has been predominantly used to study media selection. We argue that this construct has great potential if it can be used to represent the relational aspect of the communication between a web site and its customers. Presence as "Rea l i sm" : Presence is conceptualized as the degree to which a medium can produce realistic representations of the "things" one is interested in. This conceptualization has been widely used by researchers in computer graphics and human factors field. This can further be sub-divided into two orthogonal categories: perceptual and social realism. For example, the movie "Star Wars" may be high in perceptual realism but low in social realism. On the other hand, the usual animation techniques used 22 in serials like "Simpsons" are low in perceptual realism, but high in social realism. The Japanese animation (anime) is an example of a representation that is high in perceptual realism. Presence as "Transportat ion": Sensations of "you-are-there", "it-is-here" and "we-are-together" are the centerpiece of this conceptualization. Virtual Reality is the classic example of "you are there" sensation and should ideally lead to sensations of being transported to the world generated by the computer environment. The second kind of conceptualization is directly linked to the sensation that television viewers might feel when watching TV. Usually, the viewers feel as though the events are being brought into their drawing room (Millerson 1969; Reeves 1991). The third kind of sensation "we are together" is similar to the tele-presence concept used to describe video conferencing applications. Presence as "Immersion": This conceptualization seeks to measure the extent to which a virtual environment immerses a user perceptually as well as psychologically. Measures have been developed to measure perceptual immersion (Biocca et al. 1993) and psychological immersion (Heeter 1995). Perceptual immersion usually emphasizes psychophysical responses while psychological immersion primarily emphasizes user's involvement at a psychological level. Presence as "Social Ac to r wi thin M e d i u m " : This conceptualization of presence addresses social responses of media users to entities within a medium. Typical examples 23 include: responses of users to television personalities even though this relationship is one-sided, responses of users to interpersonal distance cues from across the medium, and responses of users to virtual actors (e.g.: Microsoft's office assistant, Virtual raising of pets through a key chain). Presence as " M e d i u m as a social actor": This conceptualization is similar to presence as "social actor within a medium", but this presence captures social responses of media users to the cues provided by a medium. Studies have shown that users respond to social cues exhibited by the medium and respond to the medium as though it were a social entity (Nass et al. 1995a; Nass et al. 1995b; Nass et al. 1994). Lombard and Ditton (1997) propose that the essential element for conceptualizing presence is the idea of perceptual illusion of non-mediation2. However, we argue that 'illusion of non-mediation' is an important issue only for immersive environments such as virtual reality. When two social actors communicate in a mediated environment, these actors get used to this mediated environment over time and ascribe characteristics to the media that might increase or decrease the richness of the medium. E-mail (Markus, 1994) and instant messenger systems are two classic examples of this phenomenon. Hence, we argue that for cases that involve facilitation of communication between two social actors, it is more elegant and parsimonious to concentrate purely on psychological rather than psychophysical dimensions. We argue that it is possible to use social presence (presence as social richness) as an experiential construct that captures the relationship between a 2 According to Lombard and Ditton, an "illusion of non-mediation" occurs when a person fails to perceive or acknowledge the existence of a medium in his/her communication environment and responds as he/she would if the medium were not there. 24 web site and its customers. The other conceptualizations are equally important, but they do not fit well within the context we are interested in (shopping in a business to consumer web site) and are outside the scope of the paper. 3.3 Social Presence Social presence refers to the degree to which a medium allows a user to establish personal connection with other users (Short et al. 1976). A high presence medium is rated toward the sociable, warm and personal end of the continuum. The theory further postulates that the level of social presence needed by a particular communication task determines the use of a medium. The theoretical foundation of Social Presence draws from Daft and Lengel's (1986) Media Richness Theory, and Short, Williams and Christie's (1976) Social Presence Theory. Carlson and Davis (1998) characterize richness of a medium as the capacity for a medium to convey rich information, ability to give immediate feedback, variety of communication cues available, language variety attainable and personalization of the medium. Researchers have long been interested in media selection: given the choice, which media would one choose to accomplish a certain task? (Burke et al. 1999; Carlson et al. 1998; Daft et al. 1986; Daft et al. 1987; Fulk et al. 1993; Markus 1994; Rice et al. 1983). For example, one might prefer to use email to convey unpleasant information. Researchers have used Media Richness and Social Presence theories to rank the media in order of increasing media richness and social presence respectively (Daft et al. 1987; Dennis et al. 1999; Dennis et al. 1998; Fulk et al. 1993; Sproull et al. 1986; Valacich et 25 al. 1993). Media Richness theory postulates that media selection depends on the equivocality and uncertainty of the task at hand. Social Presence theory postulates that selection of media is based on the degree to which social presence is necessary for a particular communication task. These two theories together have been grouped under "Trait Theories of Media Selection" because of the similarity of their approaches to media selection (Carlson et al. 1998). However, empirical evidence so far has not supported the claims made by either of these theories especially in the case of media selection. One of the reasons for this 'lack of strong support' is the insufficient and still evolving understanding about the usage of newer media. Media selection and use is also a function of social influences: users over time ascribe certain characteristics to the media that might increase or decrease the richness of the medium. For example Markus (Markus 1994) showed that e-mail, traditionally thought of as a lean medium, can be used for richer communication when the social processes surrounding media use define it as a rich medium. Dennis and Valacich (1999) argue that there is no such thing as absolute ranking of the media in terms of increasing information richness or social presence. Further, they also point out that one medium could possess different levels of communication capability depending on how that medium is. configured and used. The versatility of the web allows different configurations to be used to support the avowed goals of a web site (from selling products and information to enhancing corporate image). 26 We view the web as a dynamic distributed communication network disseminating information and capable of global reach. We view its primary purpose as similar to that of any other media - dissemination of information and communication. The most widely used operationalization measures social presence with four 7-point semantic differential scale items: warm/cold, sensitive/insensitive, personal/impersonal, sociable/unsociable (Karahanna et al. 1999; Short et al. 1976). Gefen and Straub (Gefen et al. 2003) have adopted this construct for usage in a web site shopping context and this study uses the measure developed by them. Gefen and Straub (2003) use the label Social Presence -Information Richness (SPIR) for this construct and use this construct in the manner consistent with the original definition proposed by Short and Christie (1976). Even though social presence has been used by Gefen et al., (2003) in the web-shopping context, the assumptions underlying this construct and its gradual evolution in meaning over time, which has resulted in its use in the shopping context, has not been made explicit in previous research. The next section will explore the assumptions underlying usage of the construct social presence and will propose expanding the contexts where social presence can be used. 3.4 Social Presence: F r o m Connection W i t h Other Users To Connection W i t h The W e b site The conceptualization of presence in this research focuses on the underlying structure of relationships facilitated by the medium. In the context of e-commerce, a web site that is high in social presence is more likely to facilitate a positive and mutually beneficial relationship between itself and its visitors. The participating social entities, the medium, and the way the medium is configured influences social presence. The advent of the web 27 has blurred the distinction between the media and the interface of the web site. It is possible for an organization to make web interface design decisions that would affect the characteristics of web as a medium for that web site. Web as the platform of communication affords a range of channels through which information can be conveyed. Web sites use a combination of multimedia and hypertext to integrate and provide interactive access to both static and dynamic content. The web is more versatile than most other communication platforms as it is an integrated medium with multiple capabilities - synchronous communication (chats, instant messaging, video conferencing), asynchronous communication (web based mail, message boards, web based group support tools with different levels of personalization), memory, storage and retrieval of information, control and security access. These multiple capabilities and versatility allow each web site to be configured in many different ways and this would have an impact on the perceived communication characteristics afforded by that web site. There are varied and multiple number of ways to configure and use the web. Any web site can offer a number of tools that support a variety of the functions described above at different levels. This research argues that social entities can configure the media according to their preferences; examples include voice mail settings (modifiable by the organization as well as users), web site design decisions made by the organization, customization by the visitor to a site (e.g. myCNN.com, my.yahoo.com). It is also important to understand that content of the message exchange will affect the structure of relationship between participating entities and vice-versa. 28 Table 2 shows the difference between Social Presence (SP) used in traditional and more recent contexts. The second column of Table 2 shows that the use of social presence has evolved over the years to include synchronous and asynchronous groups as well as a wide variety of settings including shopping on the web. However, the emergence of web as the new medium and its multiple capabilities suggests a need for rethinking the way social presence is used. Table 2: Recasting Social Presence to Connote a Broader Meaning Social Presence (Traditional) Social Presence (New Contexts) Synchronous Groups Synchronous & Asynchronous Groups Organizational Settings Org. Settings & More (e.g. Shopping) Entities Involved - two or more persons and a medium of interaction ^ Entities - Should they be humans? (Web site as a valid social actor) Virtual Teams ^ Virtual Teams & Virtual Communities Are the users connected through the medium pre-determined? • Modify to take into account the new media capabilities. Manner in which the web can bring together people with similar goals and interests (connectivity?) The last three rows of the Table 2 (as shown by the arrows) indicate other contexts in which social presence will be used in this research. Further, making a case for treating web sites as valid social actors (Chapter 3.2) strengthens- the case for using social presence to measure the relational aspect of communication between two or more social entities separated by time and space. 29 This work argues that it would be beneficial to de-link the construct from the media as the distinction between the web as a medium and the web interface becomes blurred. Instead, this research seeks to capture the underlying structure of relationship that emerges between participating social entities. Figure 2 shows social presence of the web site as one of the central constructs that can mediate the relation between communication characteristics of the media (influenced by a combination of technology and web interface decisions) and the evaluation of the web site. In short, social presence is used as an experiential construct that captures-the relational aspect of communication between a web site and its users. 30 4. Research Model and Hypotheses Development 4.1: Rationale Behind The Research Model Figure 3 shows the specific research model tested in this research. This research synthesizes three streams of literature - communication, media choice and theory of planned behavior - to propose a model where social presence and perceived usefulness are conceived of as primary mediators of relationship between communication characteristics of a web site and customer loyalty. The enabling technologies used to configure a web site are in turn theorized to impact on communication characteristics of the media. This research specifically looks at the interplay between a web site's support for LT-enabled personalization systems and virtual communities on adaptiveness and connectivity. Figure 3: Research Model Tested Traditionally, intention to engage in a behavior has been an important construct that predicts actual behavior better than either attitude towards a behavior or a set of beliefs 31 about that behavior (Ajzen 1991; Davis 1989; Venkatesh et al. 2003). Intention to engage in a behavior has in turn been known to be predicted either by attitude towards that behavior as in the theory of planned behavior (Ajzen 1991) or by a set of beliefs as in the technology acceptance model (Davis 1989). The technology acceptance model (TAM) used perceived usefulness and ease of use as the set of belief structures that are crucial to predicting intent to adopt a particular IT system. Belief structures in general are considered to be the cognitive component of attitudes (Taylor et al. 1995; Thompson et al. 1991; Triandis 1977) and are also known as evaluative affects. T A M uses belief structures and ignores the other components of attitudes for the sake of parsimony. This research uses customer loyalty as a behavioral intent construct (intent to return to the web site regularly) and consistent with T A M theory posits perceived usefulness as a predictor of customer loyalty. However, this research omits ease of use in the research model and proposes social presence as the other belief structure that predicts loyalty. Previous research has also shown evidence for the argument that "TAM's inherently parsimonious nature has excluded other critical belief constructs that are necessary to fully capture and mediate the influence of all external variables on user acceptance and subsequent usage behavior" (Hubona et al. 2003). This research argues that while perceived usefulness assesses the utilitarian aspect of the communication between a web site and its customers, social presence assesses the relational aspect of the communication between that web site and its customers. Chapter 3 presents arguments in support of treating a web site as a social actor and using social presence to capture the relational aspect of communication between a web site and its customers. We believe that social 32 presence is a more appropriate construct to use in the web-shopping context than ease of use. Moreover, the efficacy of ease of use in predicting behavioral intent has mixed empirical support (Venkatesh et al. 2003) and in general, ease of use has an indirect effect on behavioral intent through perceived usefulness. Previous research has also used affective component of attitude with like/dislike connotations as one of the predictors of behavioral intent (Bandura 1986; Compeau et al. 1995; Goodhue et al. 1995; Thompson et al. 1991; Triandis 1977). This affective component could be considered as one of the alternate candidates to using social presence to capture the relational aspect of communication. Empirical support for the predictive effect of affect on behavioral intent has been mixed (Compeau 1999; Compeau et al. 1995; Taylor et al. 1995; Thompson et al. 1991; Venkatesh et al. 2003). Considering the mixed support for the effect of affect, we believe that using social presence as a belief structure, which has similar connotations to affect in expressing the experiential aspect of the relationship, would be more promising. While social presence and perceived usefulness are proposed as experiential and utilitarian aspect of communication between a web site and its customers respectively, the front end of the model proposes that perceived communication characteristics of the web site would have a significant impact on social presence and perceived usefulness. This is consistent with the assertion of Dennis and Kinney (1998) who argued that using fundamental communication characteristics of the media would help researchers to better understand performance effects. 33 4.2 Hypotheses Recent research on media use and media selection has argued that there is no such thing as absolute ranking of media in terms of its communication richness (Dennis et al. 1999; Fulk et al. 1993; Walther 1992) and that communication capabilities of the same media can differ depending on how it is configured (Dennis et al. 1999; Kumar et al. 2001). We argue that support for different levels of personalization and different levels of consumer reviews will have an impact on the communication capabilities of a web site and will affect adaptiveness and connectivity afforded by that web site. A web site that offers support for IT-enabled personalization systems will be perceived as being more adaptive when compared to a web site with no support for personalization systems. Similarly, a web site that offers support for consumer reviews (a form of virtual community) will be perceived as being more connective (bringing people together) when compared to a web site with no support for consumer reviews. H l a : Higher levels of support for personalization features by a web site will lead to higher levels of perceived adaptiveness. H l b : Higher levels of support for virtual communities by a web site will lead to higher levels of perceived connectivity. Previous research has shown that a number of measures that can be construed as similar to adaptiveness - interactivity, empathy and responsiveness - have been closely associated with web site success and satisfaction (Devaraj et al. 2002; Palmer 2002; Zahedi et al. 2003). Previous research has also shown that virtual communities can be 34 influential sources of customer information and can create significant (Bickart et al. 2001; Brown et al. 2002) value for companies. Hence, by designing web sites that the user perceives as being more adaptive and connective would lead to development of a positive relationship between the web site and its visitors. Chapter 3.2 made a case for treating web sites as valid social actors (Moon 2000; Nass et al. 1995a; Nass et al. 1995b; Nass et al. 1994). When treating a web site as a social actor, communication literature offers plenty of support for the assertion that if a social actor improves one's communication capabilities, the relational aspect of communication (social presence) between the communication partners would also improve (Burgoon et al. 2000; Burgoon et al. 1987; Dillard et al. 1999; Watzlawick et al. 1967). In Chapter 4.1, we argued that social presence and perceived usefulness could be construed as the relational and utilitarian aspects of this relationship. We hypothesize that increased levels of adaptiveness and connectivity will lead to greater levels of social presence (H2a and H2c). We also believe that adaptiveness and connectivity afforded by a web site will directly influence levels of usefulness perceived by a user (H2b, H2d). For example, web sites personalizing content to specific user's needs not only helps increase the feeling that the web site is socially present, but also increases the usefulness of the content by organizing the most relevant content. H2a: Higher levels of adaptiveness afforded by a web site will lead to greater social presence. 35 H2b: Higher levels of adaptiveness afforded by a web site will lead to higher levels of perceived usefulness. , H2c: Higher levels of connectivity afforded by a web site will lead to greater social presence. H2d: Higher levels of connectivity afforded by a web site will lead to higher levels of perceived usefulness. In general it is argued that there is an underlying psychological connection between perceiving a medium as high on social presence and its usefulness across a wide variety of communication tasks (Gefen et al. 2003; Rice et al. 1983; Sherblom 1988; Steinfield 1986). Previous research also shows that social presence is an antecedent of perceived -usefulness (Gefen et al. 2003; Karahanna et al. 1999; Straub et al. 2002). Hence, we also hypothesize that social presence will have a positive impact on perceived usefulness (H3). H3: Higher levels of perceived social presence will lead to greater levels of perceived usefulness of that web site. Consistent with the technology acceptance model, where belief structures predict behavioral intent, this research model hypothesizes that the belief structures used in this research - social presence and perceived usefulness - will have a positive impact on customer loyalty. Customer loyalty indicates the intent of the user of a web site to return to the web site. While there is plenty of support for the significant impact of perceived 36 usefulness on behavioral intent to use the system (Compeau 1999; Davis 1989; Gefen et al. 2003; Karahanna et al. 1999; Straub 1994; Thompson et al. 1991; Venkatesh et al. 2003), support for the impact of social presence on behavioral intent is limited to showing that the relationship is mediated by perceived usefulness (Gefen et al. 2003; Karahanna et al. 1999; Straub 1994). Yet another stream of research, borrowing from the flow literature, has shown that shopping enjoyment can be a significant predictor of behavioral intention along with perceived usefulness (Hoffman et al. 1996; Koufaris 2002). Past research has also found mixed support for using affective component of attitude as a predictor of behavioral intent (Compeau 1999; Compeau et al. 1995; Taylor et al. 1995; Thompson et al. 1991; Venkatesh et al. 2003). This research argues that the belief structure social presence covers similar grounds in trying to express the relational component of communication between a web site and its users. Hence, this research proposes a direct impact of social presence on customer loyalty. H4a: Higher levels of perceived social presence will lead to an increase in customer loyalty. H4b: Higher levels of perceived usefulness will lead to an increase in customer loyalty. 4.3 Research Methodology: A Brief Introduction This research will utilize a survey and an experiment in order to fully test the sets of hypotheses developed in the previous section. The survey will first empirically test the hypotheses H2a-H4b (See the linkages shown in Figure 4). The survey will be conducted first to show that perceived communication characteristics do impact on social presence and perceived usefulness, thus underscoring the importance and relevance of these 37 communication characteristics. The results of the survey will further help show that the two belief structures social presence and perceived usefulness do have an impact on the behavioral intent customer loyalty thus validating one of the central arguments of the proposed research model. The validation of the central arguments posited by the research model will pave way for an experiment that sets out to show that IT-enabled support for technologies will impact on the perceived communication characteristics positively and through them on the other dependent variables studied in the model (social presence, perceived usefulness and customer loyalty). Figure 4: Conceptual Model Tested Using Study 1 The difficulty of measuring the impact of online personalization using any type of research methodology has been well documented (Padmanabhan et al. 2001). Using a survey methodology to evaluate on-line personalization suffers from the lack of a base web site for comparison. The other option is to investigate the impact of personalization on the rare occasion just before a web site is adding personalization systems to its arsenal of features so that a longitudinal study can be undertaken. A researcher needs serendipity 38 to accomplish this as well as access to that web site. Constructing an experimental web site that replicates the personalization systems used by the big shopping sites is no easy task. This is especially true when trying to construct web sites with consumer reviews as well as personalization systems. For example, collaborative filtering system used by both CDNOW and Amazon is dynamic in nature and generates recommendations based on other users who have visited the web site. The quality of these recommendations are highly dependent on the number of visitors and will definitely be better if a substantial number of users had'already visited the web site and made browsing and purchasing decisions. Developing a web site with rich virtual community content also suffers from a similar problem. This research takes a novel approach to alleviate this problem. The content of Amazon.com was filtered real-time to manipulate the support offered by Amazon.com for different levels of personalization and consumer reviews. The filtered content was then displayed to the participants in the experiment. This experiment will specifically focus on the linkage between personalization and virtual community features used in a web site and the two communication characteristics of the media - adaptiveness and connectivity (Ffla and Hlb). The experiment will also investigate whether support for personalization systems have an impact not only on adaptiveness but also on connectivity and vice-versa (Figure 5). This research argues that the way a web site is designed (how-to-provide-these-features) will play a major role in creating this effect. This assertion is also consistent with the findings from previous 39 research that argue that communication characteristics of a medium are perceived (Dennis et al. 1999; Fulk et al. 1993; Walther 1992) and that the way a medium is configured will impact on the perceptions these characteristics (Dennis et al. 1999). The data from the experimental study will also be used to reaffirm whether social presence and perceived usefulness mediate the relationship between perceived communication characteristics and customer loyalty. Figure 5: Conceptual Model Tested Using Study 2 In addition, using a survey and an experiment to test different portions of the model helps accomplish two things: Firstly, the experiment by its very nature will allow for careful control of features used in a web site and hence will be useful in examining the causal links between interface features and perceived media characteristics - adaptiveness and connectivity. Secondly, the experiment will be used to collect data about the other dependent variables (social presence, perceived usefulness and customer loyalty) and this in turn will help us test the conceptual model proposed in Figure 4 using the two sets of 40 data from two different methods. This allows us to compare the conceptual model across methods thus providing more confidence with the validity of the model. Hence this research utilizes two studies to empirically test the full research model (Figure 3). 41 5. Study 1: Survey 5.1 Procedures For this study, the web sites and tasks were carefully chosen such that they reflect the variance that is possible when designing web sites with different levels of adaptiveness and connectivity. For example, web sites such as Amazon.com, Chapters.ca, Bn.com and CDNow.com vary widely in the level of personalization they offer. It is also possible that communities within some web sites thrive more than others. Around the time the study was conducted (Nov 2001 - Jan 2002), the four web sites we chose fit roughly into a 2x2 matrix as shown in Table 3. Table 3: Web Sites Used in Study 1 Low Support for Consumer Reviews High Support for Consumer Reviews Low Personalization Chapters.ca Bn.com (Barnes and Noble) High Personalization CDNow.com Amazon.com We arrived at the above matrix after a careful examination of the above web sites and comparison of their features. However, it needs to be emphasized that the 2x2 matrix shown above is intended only to offer the readers a general sense of the features offered by each web site. Each web site used in this study is a fully functional, commercial web site selling a wide array of consumer goods. Beyond this base functionality, there can be a lot of differences in the way a web site implements and uses its personalization and virtual community systems. Often times, this is a technology intensive process that may require huge capital investment as is the case with Amazon.com. A good web site such as Chapters.ca chooses not to offer both the more expensive personalization features as 42 well as the relatively inexpensive virtual community features. Even among web sites that do offer virtual community features, there is a huge difference in terms of vibrancy of the communities. Amazon.com seems to have more participants in their virtual community than both CDNow.com and BN.com. These three sites also differ widely in the way they implement virtual community features with Amazon.com being the strongest and CDNow being the weakest in terms support for virtual communities (Chapters.ca has no support for virtual community features). In the study, the participant was asked to go through the following steps: • Each participant was randomly assigned to one of the four web sites and was asked to go through the purchase decision of three CDs (one CD as a gift for a friend and two CDs for self)-• The participant was instructed to browse the web site and locate information for a specific CD title (chosen to reflect the level of personalization and consumer reviews). For example, even though a web site may provide support for consumer reviews, it is possible for a user to encounter a product page for which there are no reviews. The CD title was carefully chosen so that it more or less reflects the features of the web site one has been assigned to. The participant was also instructed to browse the home page and get comfortable with the layout of the web site. This helps the participant to get familiar with the web site and the various features available with the web site. • The participant was given a scenario and was asked to buy a CD as a gift for a friend. The participant was told that his friend recently came across a particular 43 CD (Al Green's Greatest Hits) that he/she liked very much and that the participant must try to buy a CD that is similar to that CD. • The participant was then asked to shop for two CD titles of his/her choice for self. The participant was told that he/she had a 33% chance of winning one of the two CDs chosen for self (to help improve involvement). The participants were also paid an honorarium of $15 for participating in the study. • Then, the participant was asked to fill out a survey to help empirically test the hypotheses presented in the previous section. The survey used in this study is presented as Appendix 2. This survey also includes two pages of detailed instructions given to the participants to help them familiarize themselves with the web sites as well as to shop for the three CDs. For this study, we recruited 150 participants to shop in different web sites. All participants recruited were undergraduate and graduate students from the University of British Columbia. Of the respondents 78 were women and 72 were men. 5.2 Results and Discussion Chapter 5.2.1 will start with an analysis of measurement properties of the different constructs used in the study. Using LISREL software (version 8.53) and with the data collected from the survey, measurement properties of the constructs such as reliability, discriminant validity and convergent validity are analyzed and the research model is run without any modification. Then, Chapter 5.2.2 will selectively drop items to improve fit and discuss the results of the model. 44 5.2.1 Initial Research Model The items for measures adaptiveness and connectivity were developed by us based on our characterization of these constructs in Chapter 2. We used previously validated scales for the constructs perceived usefulness and social presence (Gefen et al. 2003; Karahanna et al. 1999). The items for customer loyalty were adopted from measures used for purchase intention and loyalty in previous research (Gefen et al. 2003). The original items used to measure each construct in this research are listed in Appendix 1. First, the model was run on an as-is basis with out any modifications and the results were examined. Figure 6A shows the loading coefficients associated with the measurement model and the structural model pictorially. Convergent validity is the extent to which multiple items that measure the same construct agree with each other. This is assessed by examining whether the factor loading coefficients that relate each item with the construct of interest are significant (t-values > 1.96; p<0.05). Factor loading coefficients of all items were significant with t-values exceeding 1.96, thus signifying good convergent validity. Discriminant validity is the extent to which measures of different constructs are distinct from each other. In this research, discriminant validity is assessed using chi-square difference test (Venkatraman 1989). The assessment is done by examining and specifying two different models using each pair of constructs. The first model (Model A) constrains the correlation between these constructs as 1 suggesting that all the items for both constructs measure the same factor. The second model frees the correlation between these constructs to be estimated by the program (LISREL 8.53). Discriminant validity is supported if there is a significant difference between the chi-square measures of both 45 models. Table 4A reports the results of these pair-wise chi-square tests. All 10 chi-square differences are significant demonstrating strong support for discriminant validity. Table 4A: Discriminant Validity - Pair-wise Chi Square Tests Model A: CONSTRAINED, 2 F A C T O R MODEL Model B: UNCONSTRAINED, 2 F A C T O R MODEL CHI-SQ A CHI-SQ B CHI-SQ Diff Sig 1 A D P _ C O N 416.19 217.23 198.96 0.0000 2 S P _ A D P 185.89 98.57 87.32 0.0000 3 S P _ C O N 191.75 131.10 60.65 0.0000 4 S P _ P U 345.61 66.76 278.85 0.0000 5 S P _ L O Y 324.94 93.41 231.53 0.0000 6 P U _ A D P 303.32 148.47 154.85 0.0000 7 P U _ C O N 434.40 145.48 288.92 0.0000 8 P U _ L O Y 847.53 138.35 709.18 0.0000 , 9 LOY_ADP 468.48 157.35 311.13 0.0000 10 L O Y _ C O N 417.91 143.95 273.96 0.0000 Table 4B reports the composite reliability and the variance extracted. The recommended values are 0.7 for reliability and 0.5 for variance explained (Hair et al. 1995). The results from Table 4B show that the variance explained for connectivity is low at 0.34 for the as-is model with all original items included in the analysis. Further, the fit indices from Table 4C for the research model show that all fit indices are within acceptable range with the exception of GFI and AGFI. To improve the low variance explained for connectivity and the moderate values for fit indices, the constructs in the model will further be analyzed to identify possible problems with individual items. 46 Table 4B: Estimates of Composite Reliability and Variance Constructs Reliability Explained Variance Adaptiveness 0.87 0.50 Connectivity 0.77 0.34 Perceived Usefulness 0.93 0.74 Loyalty 0.94 0.70 Social Presence 0.85 0.53 Table 4C: Fit Indices for the Research Model: Before Modification Research Model Desired Levels RJV1SEA 0.069 <0.1 NFI 0.93 >0.9 NNFI 0.97 >0.9 CFI 0.97 >0.9 IFI 0.97 >0.9 RFI 0.93 >0.9 RMR (Std) 0.077 <0.08 GFI 0.75 >0.9 AGFI 0.71 >0.8 X2 / d f 1.74 Between 1 and 5 5.2.2 Research Model: After Modification Each construct and the corresponding items were examined separately using LISREL to assess the value of the item factor loadings. In general, values greater than 0.6 are recommended for building reliable constructs. As a result of this analysis four items from connectivity (CON1, CON2, CON6, CON7) and one item from social presence (SP3) were eliminated. The dropping of items as a result of this procedure is considered 49 acceptable as this procedure examines unidimensionality as well as the residual variance of the individual items that do not overlap, both of which are ignored by a traditional factor analysis and/or reliability analysis (Gerbing et al. 1988). In the next stage, the model was run again using LIS REL and three more items were dropped to improve the fit of the model. Two of these three items were from the construct loyalty (LOY2 and LOY5) and the third item was from the construct perceived usefulness (PU4). Figure 6B shows the results of this analysis'. While dropping of items at this stage gives rise to the possibility that the results are empirically driven, an examination of the structural coefficients for both models (Figures 6A and 6B) suggests that the model is relatively stable (the values do not change much in the two models and the pattern of results remain the same). Further, we will use the data from Study 2 (see next Chapter) to examine whether the general pattern of results from this study is replicated. Table 5A shows that all chi-squared difference tests are significant thus demonstrating strong support for discriminant validity. Factor loading coefficients of all items were significant with t-values exceeding 1.96 (measurement model in Figure 6B), thus signifying good convergent validity. Table 5B reports the composite reliability and the variance extracted for this model. These estimates exceed the recommended values of 0.7 for reliability and 0.5 for variance explained (with the exception of connectivity whose variance was marginally below the recommended value at 0.45; but this value for connectivity is higher than the 0.34 value before the offending items were dropped). The combined results of composite reliability and variance suggest that each construct is indeed internally consistent. The fit indices for the research model are reported in Table 50 5C. The results show improved fit indices when compared to the fit indices of the model before modification. Table 5A: Discriminant Validity - Pair-wise Chi Square Tests Model A: CONSTRAINED, 2 F A C T O R MODEL Model B: UNCONSTRAINED, 2 F A C T O R MODEL CHI-SQ A CHI-SQ B CHI-SQ Diff Sig 1 A D P _ C O N 231.85 111.14 120.71 0.0000 2 S P _ A D P 170.42 85.25 85.17 0.0000 3 S P _ C O N 71.33 33.27 38.06 0.0000 4 S P _ P U 256.07 36.57 219.50 0.0000 5 S P _ L 0 Y 231.52 37.07 194.45 0.0000 6 P U _ A D P 270.43 105.44 164.99 0.0000 7 P U _ C O N 188.97 37.92 151.05 0.0000 8 P U _ L O Y 442.94 121.12 321.82 0.0000 9 L 0 Y _ A D P 415.93 85.98 329.95 0.0000 10 L O Y _ C O N 149.61 19.03 130.58 0.0000 Table 5B: Estimates of Composite Reliability and Variance Constructs Reliability Explained Variance Adaptiveness 0.87 0.50 Connectivity 0.76 0.45 Perceived Usefulness 0.93 0.74 Loyalty 0.94 0.71 Social Presence 0.85 0.58 The fit indices for the research model shows that all the fit indices with the exception of GFI are within the acceptable range. %2 / df value is also in the accepted range for the research model (However, the use of %2 / df as a fit index is in a decline with the 51 emergence of much better indices such as RMSEA and NFI). Even though the fit index for GFI is just beyond the acceptable range, considering the value of this index (0.84) and the fact that other indices are well within the recommended range, the model is considered to have good fit. This approach of examining all the indices together to arrive at a conclusion about model fit is justified and is necessary as 'the rules about when an index indicates a good fit to the data are highly arbitrary' (Kelloway 1998). In fact, GFI, AGFI and RMR used above have no known sampling distribution to justify a set cut-off value (Kelloway 1998). Further, AGFI is a derivative of GFI and AGFI for this research model is 0.80 suggesting good fit. Table 5C: Fit Indices for the Research Model Research Model: Before Modification Research Model After Modification Desired Levels RMSEA 0.069 0.052 <0.1 NFI 0.93 0.95 >0.9 NNFI 0.97 0.98 >0.9 CFI 0.97 0.98 >0.9 IFI 0.97 0.98 >0.9 RFI 0.93 0.94 >0.9 RMR (Std) 0.077 0.063 <0.08 GFI 0.75 0.84 >0.9 AGFI 0.71 0.80 >0.8 X 2 / d f 1.74 1.41 Between 1 and 5 Table 6 reports the support for the hypotheses developed in Chapter 4. Figure 6B shows the structural coefficients of the research model so that the individual hypotheses can be compared against the results. For this research model, all the hypotheses are supported with the exception of H2d (link between connectivity and perceived usefulness) and H3 52 (link between social presence and perceived usefulness). Adaptiveness had a positive impact on both social presence (t = 4.91) and perceived usefulness (t = 6.69) supporting H2a and H2b respectively. Connectivity had a positive impact on social presence (t = 5.96) supporting H2c, but did not have any impact on perceived usefulness (H2d). Social presence affected Customer Loyalty (t = 4.81) positively supporting hypothesis H4a. Perceived Usefulness also had a positive impact on Customer Loyalty (t = 5.63) supporting hypothesis H4b. There was also a significant correlation between Adaptiveness and Connectivity (t = 8.22) suggesting a link that will be fleshed out more in depth in the experimental study. Table 6: Test Results Hypothesis # Hypothesis Support H2a Higher levels of adaptiveness afforded by a web site will lead to greater social presence. Yes H2b Higher levels of adaptiveness afforded by a web site will lead to higher levels of perceived usefulness. Yes H2c Higher levels of connectivity afforded by a web site will lead to greater social presence. Yes H2d Higher levels of connectivity afforded by a web site will lead to higher levels of perceived usefulness. No H3 Higher levels of perceived social presence will lead to greater levels of perceived usefulness of that web site. No H4a Higher levels of perceived social presence will lead to an increase in customer loyalty. Yes H4b Higher levels of perceived usefulness will lead to an increase in customer loyalty. Yes The pattern of structural coefficients for this model (Figure 6A and 6B) confirms the presence of two distinct mediating influences: social presence as the experiential 53 construct and perceived usefulness as the utilitarian construct. The results also show that while adaptiveness shows a significant direct impact on both social presence and perceived usefulness, connectivity has a significant direct impact on only social presence. This suggests that connectivity influences the experiential component for a web site's customers much more than the utilitarian construct. Connectivity refers to the extent to which a web site links a shopper to other shoppers of similar interests. Originally, it was hypothesized that connectivity in addition to having an impact on social presence (H2c) by virtue of bringing together people of similar interests will also have an impact on perceived usefulness (H2d) as a shopper may perceive this as a useful function of a web site. The results show that at least in this context (shopping for a high involvement items - music CDs), the shoppers seem to attach a premium on the experiential aspect of the web site's attempts to link people together rather than on the utilitarian side. Further, the presence of adaptiveness as the other, independent variable with a strong influence on perceived usefulness might have had an effect of diminishing the impact of connectivity on perceived usefulness. This also reinforces the assertion put forth earlier in the thesis (Chapter 2) about the importance of the need to study each communication characteristic of a medium on its own. Some communication characteristics may have a bigger impact on the experiential component while others may have significant impact on the utilitarian component. Previous research has examined the impact of social presence/media richness at an aggregated level rather than attempting to examine the impact of individual communication characteristics on social presence/media richness (Dennis et al. 1998). 54 Studying the communication characteristics on their own helps us better understand the nature of these characteristics and helps organizations configure their web sites according to their fine-grained needs and requirements. The results of the model also show that social presence does not mediate the relationship between perceived communication characteristics and perceived usefulness. Rather, the results suggest that (at least) these two media characteristics directly influence social presence as well as perceived usefulness. The results also show support for social presence and perceived usefulness having a significant positive influence on loyalty. Hence organizations by increasing the adaptiveness and connectivity of their web site can not only increase the experiential (social presence) and utilitarian (perceived usefulness) components, but can also improve customer loyalty thus directly impacting on their bottom-line. 55 6. Study 2: Experiment 6.1 Procedures The content of Amazon.com was filtered in real time to generate the experimental web sites. As shown in Table 7, the study offered two levels of support for personalization -No support or High Support (personalization based on simple filtering/rule based techniques and collaborative filtering) and provided support for two levels of consumer review: no reviews or reviews with information about the rater included. Table 7: Experimental Manipulations Low Support for Consumer Reviews (LoCR) High Support for Consumer Reviews (HiCR) Low Personalization (LoPER) Condition 1 Condition 2 High Personalization (HiPER) Condition 3V Condition 4 To the best of our knowledge, this is the first study to tackle the problem in evaluating the impact of personalization and assign causality using a controlled experiment. The filtering of content is based on the assumption that Amazon.com does provide high level of support for personalization as well as virtual communities. Amazon.com is one of the B2C retailers who are well known for its industry leading efforts in personalization. For example, based on the book that a customer is interested in, Amazon.com recommends similar books (simple filtering/rule based) and books bought by users with similar tastes (collaborative filtering). Amazon.com also offers excellent support for virtual communities (especially, consumer reviews) and has a rich user base that actively 56 contributes to the 'Reviews' section. Thus, the assumption that Amazon.com offers high level of support for personalization and virtual community is justified. Filtering is done by intercepting the dynamic content sent by Amazon.com in response to an http request. When the shopper types in the web site address (URL) and requests a specific page from Amazon.com, the shopper is sent a page that is often created dynamically on the fly by Amazon.com. This page is then intercepted by a program that looks for a combination of specific key words to remove pre-planned content. The filtering process targets only the product pages of Amazon.com (for example, page that contains information about a particular music CD). The program takes advantage of the way this product page is structured and the specific combination of keywords used by Amazon.com (For example, the keyword combination "People who bought music by this artist also'bought" followed by a set of recommendations). Please see Appendix 7 series for screenshots of these manipulations. Appendix 5 shows the sample code written in ASP (Active Server Page) to filter content from Amazon.com to obtain the experimental condition 1. The code written for other conditions are not included as the underlying logic for these programs is similar for all conditions. Appendix 7.1 shows the screenshot of the first page participants see in all experimental manipulations. The content of Amazon.com is filtered in such a way that the top level tabs show only links related to music and the participants start with the page shown in this screenshot - New and Future Releases - as the default homepage (Amazon.com has a more generic home page and includes navigation tabs and content relating to different 57 product categories). While the subsequent pages the participant browses may occasionally include information about products other than CDs in the form of advertisements, the participant was encouraged to shop just for the CDs (See participant instructions below). The participant had access to sections such as "Search Music" and "Browse Styles" within the music portion of the web site. Appendices 7.2, 7.3, 7.4 and 6.5 show the screenshots of these other sections. For condition 1 (LoPER and LoCR), the product page at Amazon.com is filtered to an extent where only the basic product information is shown. The participants can also listen to samples of selected tracks provided by Amazon.com. This is a web interface that provides shoppers with a good base-level functionality and can be considered as an interface that offers low support for personalization and low support for consumer review features. Appendix 7.6 shows a sample screenshot of this product page. At the other end of the spectrum (For condition 4 - HiPER and HiCR), Amazon.com is filtered to an extent where a shopper has access to more features both in terms of support for personalization as well as support for consumer reviews. Appendix 7.7 shows a sample screenshot of this product page. More specifically, the shopper can view the recommendations generated by Amazon.com on a product page (For example, "Customers who bought this title also bought", "Customers who bought titles by A l Green also bought") as shown in Appendices 7.7 and 7.9. Upon clicking the appropriate hyperlink ("Explore similar items", Explore similar artists"), they can get more product recommendations as shown in Appendices 7.10 and 7.11. Appendix 7.8 shows part of 58 the product page that demonstrates how this condition supports customer review features. Appendices 7.7, 7.8 and 7.9 together are spatially arranged together to reflect the way the actual product page looked for this condition: the sections of the product page are displayed in the following order - basic product information, personalization (part 1), audio samples, support for consumer reviews, and more personalization (part 2). Conditions 2 and 3 are variants of the scenario described above. The product page for condition 2 (LoPER, HiCR) is similar to the base interface condition 1 with support for consumer reviews added in (similar to the one shown in Appendix 7.8). The product page for condition 3 (HiPER, LoCR) is similar to the base interface condition 1 with support for personalization added in (similar to the screenshots shown in Appendices 7.7, 7.9, 7.10 and 7.11. Of course, condition 4 described in the previous paragraph is similar to base interface condition 1 with support for personalization as well as consumer reviews features added in. The experiment was conducted as per the steps outlined below: • The research assistants were trained in conducting the experiment and were given a handout with a set of instructions to ensure uniformity in procedures. The questionnaire used in this study is presented as Appendix 3 and the instructions to the research assistants are presented in Appendix 4. • The research assistants, programmer and the author were collectively responsible for continually monitoring the experimental web site to ensure that the filtering process does indeed work as intended for all four conditions. 59 • The participant was asked to briefly surf the web sites of the two different web sites selling music CDs (CDNow.com and BN.com) to get a rough sense of a typical B2C web site. It is hoped that this will sensitize a user with limited experience in on-line shopping to the rich possibilities available beyond a simple catalogue web site. This step usually took about 15-20 minutes. The RA was available to answer any questions that the participant may have in this step. • All participants completed two tasks using the base interface represented by condition 1 (low support for personalization as well as consumer reviews). The two tasks were similar to those used in the previous study - CD for a friend and a CD for self. The participants also completed a short questionnaire that measured the participants' perceived adaptiveness and connectivity for this interface. • The research assistant then randomly assigned each participant to one of the four conditions. The participants then completed two more tasks (similar to the tasks used in the previous step) using the assigned interface for this condition. The participants completed a similar questionnaire that measured the participants' perceived adaptiveness and connectivity for this interface. The experiment was conducted with 60 participants. All participants recruited were undergraduate and graduate students from the University of British Columbia. Of the respondents 31 were women and 29 were men. To increase involvement and external validity, participants were told that they had a 33% chance to keep one of the CDs they shop for themselves (in addition to a honorarium of $15). A N O V A was used to assess the impact of a web site's support for personalization and consumer review features on 60 adaptiveness and connectivity (Hla and Hlb) and to examine the interaction effects. The experimental manipulation will help increase the confidence with which a causal connection can be inferred between the interface features and perceived communication characteristics. 6.2 Results and Discussion To check for the validity of experimental manipulations, self reported data was collected on perceived support for consumer reviews (3 items) and perceived support for personalization (3 items). These items asked the participants specifically about the interface features of the web site that were manipulated. Table 8 A reports the items used and the reliabilities of these two scales and 8B the descriptive statistics. Table 8A: Manipulation Check Scales and Their Reliabilities Construct Reliability Items Used Self-Reported Support for CR Self-Reported Support for PER 0.92 This web site provides me with product reviews This web site provides me with product reviews by other users This web site allows me to review a product 0.91 When I visit a product page (for ex, 'Thriller' by Michael Jackson), this web site recommends products that might potentially interest me When I visit a product page (for ex, 'Thriller' by Michael Jackson), this web site shows me other similar items purchased by other users This web site recommends products that might potentially interest me 61 Table 8B: Descriptive Statistics for Manipulation Check Variables D V Consumer Reviews (CR) Personalization (PER) Mean (7 pt. scale) Std. Deviation N Self-Reported Support for CR Low Support for PER Low Support for CR 2.93 1.94 15 High Support for CR 6.40 .84 15 Total 4.66 2.29 30 High Support for PER Low Support for CR 4.02 1.58 15 High Support for CR 6.37 .85 15 Total 5.19 1.73 30 Total Low Support for CR 3.47 1.82 30 High Support for CR 6.38 .83 30 Total 4.93 2.03 60 • Self-Reported Support for PER Low Support for CR Low Support for PER 3.33 2.05 15 High Support for PER 6.17 .66 15 Total 4.75 2.08 30 High Support for CR Low Support for PER 3.68 1.63 15 • High Support for PER 6.28 1.12 15 Total 4.98 1.90 30 Total Low Support for PER 3.51 1.83 30 High Support for PER 6.23 .91 30 Total 4.87 1.98 60 Table 8C reports the results of this manipulation check using ANOVA. The results show that experimental manipulations were successful. 62 Table 8C: ANOVA Table for Manipulation Check Source (IV) Manipulation Check Sum of Squares Df Mean Square F Sig Support for Consumer Reviews (CR) Self-Reported Support for CR Self-Reported Support for PER 127.14 0.82 127.14 0.82 65.79 0.38 0.000 0.539 Support for Personalization (PER) Self-Reported Support for CR 4.26 1 4.26 2.2 0.143 Self-Reported Support for PER 111.17 1 111.17 51.76 0.000 CR * PER Self-Reported Support for CR 4.64 1 4.64 2.40 0.127 Self-Reported Support for PER 0.22 1 0.22 0.11 0.748 Error Self-Reported Support for CR 108,21 56 1.93 Self-Reported Support for PER 120.27 56 2.15 The next sub-section 6.2.1 will specifically investigate impact of these different experimental conditions on perceived communication characteristics (adaptiveness and connectivity) thus testing hypothesis Hla and Hlb. Then, sub-section 6.2.2 will conduct a series of multiple regressions to further explore how social presence and perceived usefulness mediate the relationship between communication characteristics and customer loyalty as well as communication characteristics and perceived usefulness. 63 6.2.1 Impact of Web Site Interface Features on Communication Characteristics Table 9 A shows the means for the dependent variables adaptiveness and connectivity. Table 9B reports the Cronbach's alpha values for the scales used in the study (same as the items used in Study 1) suggesting good reliability. Table 9A: Descriptive Statistics for Perceived Communication Characteristics D V Personalization (PER) Consumer Reviews (CR) Mean (7 pt. scale) Std. Deviation N A D P Low Support for PER Low Support for CR 2.74 1.18 15 High Support for CR 4.11 1.35 15 Total 3.43 1.43 30 High Support for PER Low Support for CR 4.49 1.02 ' 15 High Support for CR 4.91 0.76- 15 Total 4.71 0.91 30 Total Low Support for CR 3.62 1.40 30 High Support for CR 4.51 1.15 30 Total 4.07 1.35 60 C O N Low Support for PER Low Support for CR 2.30 1.06 15 High Support for CR 4.78 .93 15 Total 3.54 1.60 30 High Support for PER Low Support for CR 3.75 1.25 15 High Support for CR 4.73 1.40 15 Total 4.24 1.40 30 Total Low Support for CR 3.03 1.35 30 High Support for CR 4.76 1.17 30 Total 3.89 1.53 60 64 Table 9B: Reliability Values for Constructs Used in the Study Construct Reliability Adaptiveness 0.89 Connectivity 0.78 Social Presence 0.89 Perceived 0.97 Usefulness Loyalty 0.94 Table 10 shows the results of the statistical analysis from the experiment (Study 2). Figures 7 and 8 show the interaction patterns for adaptiveness and connectivity. Table 10: ANOVA Table for Perceived Communication Characteristics Source (IV) Dependent Variables Sum of Squares Df Mean Square F S ig Support for Personalization ADP 18.93 1 18.93 19.98 0.000 (PER) CON 7.35 1 7.35 5.34 0.025 Support for Consumer ADP 12.13 1 12.13 9.84 0.003 Reviews (CR) CON 45.07 1 45.07 32.75 0.000 CR * PER ADP 1.57 1 1.57 2.77 0.101 CON 8.44 1 8.44 6.13 0.016 Error ADP 67.91 56 1.21 CON 77.07 56 1.38 The results of the analysis suggest that support for personalization features does indeed have a positive impact on adaptiveness supporting HI a. Similarly, support for consumer 65 reviews (a form of virtual community) positively affects connectivity thus supporting Hlb. In addition, support for personalization has a positive impact on connectivity while support for consumer reviews has a positive impact on adaptiveness. Moreover, there is an interaction effect between support for personalization and consumer reviews. For connectivity this interaction is significant, but it is not significant for adaptiveness. Figures 7 and 8 illustrate this interaction effect graphically. While the original research model (Figure 3) included only two hypotheses (HIa and Hlb) on the impact of web interfaces, the intent there was to identify primary features that may have the most impact on adaptiveness and connectivity of a shopping web site respectively. However, this does not necessarily preclude a web site's personalization features from having an impact on connectivity nor its consumer review features on adaptiveness (see Figure 5). The results show that support for personalization not only had an impact on adaptiveness, but also on connectivity. Similarly, support for consumer reviews had on impact on connectivity as well as adaptiveness. To a large extent, this impact depends on the way these features are implemented in a web site. For example, CDNow's support for consumer reviews was implemented in the form of bulletin boards and this information was not immediately available on a product page. However, Amazon.com implemented its consumer review features by displaying the reviews prominently on the product page. More information about the reviewer can be found by clicking on the name of the reviewer (if the reviewer makes his profile public). 66 L o w C R H i g h C R CR Figure 7: Interaction Pattern for Adaptiveness Figure 8: Interaction Pattern for Connectivity When a web site supports consumer reviews on a product page, this can be viewed as an effort by that web site to build a vibrant community (as measured by connectivity). But, the same effort can also lead to a shopper perceiving comments about the product (by other users) being placed on the product page as relevant information about the product, thus leading to the perception that the web site is also adaptive (in catering to the pertinent needs of the shopper who may learn more about the product by reading the consumer reviews) in addition to being connective. This effect is also consistent with the findings from previous research that argue that communication characteristics of a medium are perceived (Dennis et al. 1999; Fulk et al. 1993; Walther 1992) and that the way a medium is configured will impact on the perceptions these characteristics (Dennis et al. 1999). Hence, we argue that Amazon.com's implementation of support for consumer review features is more likely to have an impact on adaptiveness than the implementation of CDNow.com (whose consumer review features are not integrated into the product page). This Study used Amazon.com's web site to filter content live as it is considered one of the leading e-retailers in terms of implementing personalization features as well as consumer review features for shoppers with a very vibrant virtual community (whose consumer review features are integrated into the product page). Hence, it is hardly surprising that, in our experimental conditions, support for virtual communities not only impacts on connectivity but also on adaptiveness. As the results from Table 10 show, higher levels of support for consumer reviews leads to a strong effect on connectivity (F-32.75) as well as weaker, but significant effect on adaptiveness (F=9.83). 68 On a similar vein, a web site may offer support for personalization in a manner similar to Amazon.com where support for personalization is often implemented using explanations such as "Customers who shopped for this item also shopped for", "Customers who bought items by this artist also bought" • followed by a set of recommendations (see Appendix 7.7 and 7.9). These types of explanations often act as a portal from where more recommendations can be generated (see Appendix 7.10). This may very well be a deliberate strategy by leading web sites such as Amazon.com to inculcate a strong sense of (credible and vibrant) community within the shopping web site. As a side note, one of the future projects of this author is to check if explanations using phrases such as "you may also like/Amazon.com recommends" instead of "customers who shopped for this item also shopped for" makes a significant difference for connectivity (and the perception of a presence of a vibrant community). Hence, the results from Table 10 indicating a similar effect for personalization make sense. A web site's support for personalization not only leads to higher levels of adaptiveness (F=19.98), but also to higher levels of connectivity (F=5.34). In the light of the above discussion, it is also clear that when a web site offers low support for personalization, high support for consumer reviews might help alleviate the absence of an expensive personalization system if the consumer reviews are integrated into the product page (thereby indicating presence of an interaction effect). This implementation of support for consumer reviews may result in shoppers perceiving the web site with almost no support of personalization, but with a vibrant community as 69 being more adaptive. Similarly, a web site that offers excellent support for personalization can overcome the absence of a vibrant community by offering the product recommendations with the right explanations such as 'customers who bought this item also bought' followed by good recommendations (thereby indicating presence of an interaction effect). In this study, this interaction effect between personalization and consumer reviews was significant only for connectivity. Even though the interaction effect for adaptiveness was not significant, the marginal means for this interaction did move in the right direction as shown in Figure 7. For the web interface with low support for personalization, offering high support for consumer reviews not only increased the adaptiveness perceived by the shopper (mean rose from 2.74 to 4.11), but also closed the gap in terms of perceived adaptiveness when compared to a web site that offered only high support for personalization (4.49). A very similar, but significant interaction pattern can be observed for connectivity (See Figure 8). For a web interface offering low support for consumer reviews, addition of support for personalization resulted in an increase in perceived connectivity (the mean value rose from 2.3 to 3.75). Tables 11 and 12 show the one-way A N O V A results for the other dependent variables -Social Presence, Loyalty and Perceived Usefulness. Table 11 shows the group differences for personalization support while Table 12 shows the group differences for consumer review support. 70 Table 11: ANOVA Table for Other DVs (IV-Personalization Support) DV Sum of Squares df Mean Square F Sig. SP Between Groups 5.70 1 5.70 4.39 .041 With in Groups 75.39 58 1.30 Total 81.10 59 L O Y Between Groups 12.83 1 12.83 10.49 .002 With in Groups 70.95 58 1.22 Total 83.78 59 P U Between Groups 26.40 1 26.40 16.50 .000 With in Groups 92.80 58 1.60 Total 119.20 59 Table 12: ANOVA Table for Other DVs (IV-Consumer Review Support) DV Sum of Squares df Mean Square F Sig. SP Between Groups 22.20 1 22.20 21.86 .000 With in Groups 58.89 58 1.01 Total 81.10 59 L O Y Between Groups 8.62 1 8.62 6.65 .012 With in Groups 75.16 58 1.29 Total 83.78 59 P U Between Groups 14.01 1 14.01 7.72 .007 Within Groups 105.191 58 1.81 Total 119.207 59 Tables 13 and 14 show the descriptive statistics for the same set o f dependent variables. The results show that both independent variables impacted on each o f the dependent variables further under l in ing the importance o f p rov id ing support for personal izat ion and consumer reviews. 71 Table 13: Descriptive Statistics for Other DVs (IV-Personalization Support) N Mean (7 pt. scale) Std. Deviation SP Low Support for Per 30 3.64 1.18 High Support for Per 30 4.26 1.10 - Total 60 3.95 1.17 LOY Low Support for Per 30 3.81 1.21 High Support for Per 30 4.73 .98 Total 60 4.27 1.19 PU Low Support for Per 30 3.96 1.47 High Support for Per 30 5.29 1.03 Total 60 4.62 1.42 Table 14: Descriptive Statistics for Other DVs (IV-Consumer Review Support) N Mean (7 pt. scale) Std. Deviation SP Low Support for CR 30 3.34 1.05 High Support for CR 30 4.55 .95 Total 60 3.95 1.17 LOY Low Support for CR 30 3.89 1.04 High Support for CR 30 4.65 1.22 Total 60 4.27 1.19 PU Low Support for CR 30 4.14 1.46 High Support for CR 30 5.10 1.22 Total 60 4.62 1.42 The next sub-section will analyze whether social presence mediates the relationship between perceived communication characteristics and customer loyalty as well as communication characteristics and perceived usefulness. This analysis will also help in comparing the efficacy of research model across multiple methods (survey in Study 1 and experiment in Study 2). 72 6.2.2 Mediation Analysis , Study 1 tested the research model (Figure 6) that posited social presence as the experiential construct and perceived usefulness as the utilitarian construct using LISREL. Because the sample size for Study 2 was 60, data collected about the other dependent variables in this study will be used to test the mediating relationships by conducting a series of multiple regressions (Baron et al. 1986). Consistent with the research model posited in Study 1, the following three mediating influences were tested: a) Social presence mediating the relationship between perceived communication characteristics and loyalty, b) Social presence mediating the relationship between perceived communication characteristics and perceived usefulness, c) Perceived usefulness mediating the relationship between perceived communication characteristics and loyalty. To test each of the mediating relationships, three separate regressions need to be run (Baron et al. 1986). For all subsequent mediation analyses, the first equation regresses the independent variables on the mediator variable. Then, the second equation regresses the independent variable on the final outcome variable. To show that there is a mediation effect, the beta coefficients for the two regressions above must be significant (Baron et al. 1986). In addition, the mediator variable is now introduced along with the two independent variables to predict the final outcome variable in the third regression equation. Full mediation is said to occur if the beta coefficients of both independent variables become insignificant, while the beta coefficient of the mediator variable is 73 significant. Partial mediation is said to occur if the independent variables do not drop out of the equation, but rather the introduction of the mediator variable causes a drop in the value of the beta coefficients of these independent variables (Baron et al. 1986). Table 15 reports the results of three regressions run to test whether social presence mediates the relationship between perceived communication characteristics and loyalty. The results in Table 15 show that the beta coefficients for the first two regression equations are significant thus satisfying the initial conditions while testing for mediation effect. The results from the third regression equation indicate that social presence mediates the relationship only when p is set to a more liberal 0.1 criteria (p=0.10). The results further show that connectivity totally drops out of the equation (beta coefficient insignificant) suggesting a strong mediation effect for social presence between connectivity and loyalty. On the other hand, the beta coefficient for adaptiveness is the largest in the equation suggesting weak mediation. effect for social presence. The beta coefficients for adaptiveness as well as connectivity drop in value when social presence is introduced thus suggesting that social presence does mediate the relationship between communication characteristics and loyalty even though this mediation effect is stronger for connectivity than it is for adaptiveness. The results from this mediation analysis in Study 2 are consistent with the results from LISREL analysis in Study 1, thus improving our confidence in the overall research model. To understand the impact of adaptiveness 74 and connectivity on social presence, the mediation analyses are shown for each of these communication characteristics individually as shown in Tables 16 and 17. Table 15: Social Presence As A Mediator Between Communication Characteristics and Loyalty Regressions Dependent Variable R 2 (adjusted) F-Value Standardized coefficients t-value Regression #1 SP 0.567 36.67"* ADP 0.283 2.75*** CON 0.569 5.53*** Regression #2 Loyalty 0.477 27.95*** ADP 0.466 4.12*** CON 0.329 2.91*** Regression #3 Loyalty 0.493 20,14*** ADP 0.398 3.36 CON 0.193 1.40 SP 0.239 1.67* p<0.01; p<0.05; p<0.1 Table 16 shows that beta coefficient for adaptiveness still remains the largest in the regression equation even with the introduction of SP thus lending support to our contention that SP is a weak mediator of relationship between adaptiveness and loyalty. On the other hand, as the results from Table 17 show, the beta coefficient for SP is the largest when it is introduced alongside connectivity in predicting loyalty. This shows that SP mediates the relationship between connectivity and loyalty more strongly than the relationship between adaptiveness and loyalty. 75 Table 16: Social Presence As A Mediator Between Adaptiveness and Loyalty Regressions Dependent Variable R 2 (adjusted) F-Value Standardized coefficients t-value Regression #1 SP 0.347 ' 32.34*** ADP 0.598 5.69*** Regression #2 Loyalty 0.410 42.02*** ADP 0.648 6.48*** Regression #3 Loyalty 0.485 28.74*** ADP 0.434 3.72*** SP 0.357 _ „ „*** 3.06 p<0.01; p<0.05; p<0.1 Table 17: Social Presence As A Mediator Between Connectivity and Loyalty Regressions Dependent Variable R 2 (adjusted) F-Value Standardized coefficients t-value Regression #1 SP 0.518 64.48*** CON 0.726 8.03*** Regression #2 Loyalty 0.334 30.54*** CON 0.587 5.53*** Regression #3 Loyalty 0.402 20.81**v CON 0.295 2.01 SP 0.403 2.76 p<0.01; p<0.05; p<0.1 76 Table 18 shows the results of the mediation analysis for SP as a mediator between communication characteristics and perceived usefulness. The results from Table 18 show that SP does not mediate the relationship between communication characteristics and perceived usefulness as the beta coefficient is not significant. The results suggest that adaptiveness seems to have a biggest impact on perceived usefulness (beta coefficient largest in the regression equation). Table 18: Social Presence As A Mediator Between Communication Characteristics and Perceived Usefulness Regressions Dependent Variable R 2 (adjusted) F-Value Standardized coefficients t-value Regression #1 SP 0.567 36.67*** ADP 0.283 2.75 CON 0.569 5.53*** Regression #2 PU 0.684 64.89*** ADP 0.750 8.53*** CON 0.136 1.55 Regression #3 PU 0.681 42.91*** ADP 0.731 7.77 CON 0.096 0.88 SP 0.070 0.61 p<0.01; p<0.05; p<0.1 To understand the individual impact of these communication characteristics, mediation analysis was conducted for each of these characteristics on their own and the results are shown in Tables 19 and 20. 77 Table 19: Social Presence As A Mediator Between Adaptiveness and Perceived Usefulness Regressions Dependent Variable R 2 (adjusted) F-Value Standardized coefficients t-value Regression #1 SP 0.347 32.34*** A D P 0.598 5.69*** Regression #2 P U 0.676 *** 124.38 A D P 0.826 11.15*** Regression #3 P U 0.682 64.23 A D P 0.749 8.17 SP 0.129 1.41 p<0.01; p<0.05; p<0.1 Table 20: Social Presence As A Mediator Between Connectivity and Perceived Usefulness Regressions Dependent Variable R 2 (adjusted) F-Value Standardized coefficients t-value Regression #1 SP 0.518 64.48 C O N 0.726 8.03*** Regression #2 P U 0.293 25.44*** C O N 0.552 5.04*** Regression #3 P U 0.348 16.77*** C O N 0.282 1.85* SP 0.372 ** 2.44 p<0.01; ** p<0.05; * p<0.1 78 The results from Table 19 show that social presence does not mediate the relationship between adaptiveness and perceived usefulness. On the other hand, the results from Table 20 show that the relationship between connectivity and perceived usefulness is strongly mediated by social presence. Even though connectivity has a direct impact on perceived usefulness, it is weak as suggested by the smaller beta coefficient in regression equation 3 (Table 20). Table 21: Perceived Usefulness As A Mediator Between Communication Characteristics and Loyalty Regressions Dependent Variable R 2 (adjusted) F-Value Standardized coefficients t-value Regression #1 PU 0.684 64.89"* ADP 0.750 8.53*** CON 0.136 1.55 Regression #2 Loyalty 0.477 27.95*** ADP 0.466 4.12 CON 0.329 2.91 Regression #3 Loyalty 0.632 34.77*** ADP -0.07 -0.49 CON 0.232 2.39 PU 0.714 4.99*** p<0.01; p<0.05; p<0.1 Table 21 shows the results of the mediation analysis for perceived usefulness as the mediator between communication characteristics and loyalty. The results provide support for the claim that perceived usefulness is a stronger mediator for adaptiveness 79 than social presence. The third regression equation in Table 21 shows that adaptiveness totally drops out of the equation when perceived usefulness is introduced (beta coefficient is very small and is insignificant). On the other hand, connectivity stays in the equation suggesting a weaker mediation effect for perceived usefulness. When contrasted with the results from Table 15, which shows the opposite effect for social presence: Social presence is a stronger mediator of connectivity and loyalty (connectivity drops out of the third regression equation when social presence is introduced along with adaptiveness and connectivity to predict loyalty). Table 15 also shows that adaptiveness stays in the equation thus suggesting a very weak mediation effect for social presence. These results from Tables 15, 16, 17, 18, 19, 20 and 21 when taken together broadly support the general pattern of results observed in Study 1. The link between connectivity and perceived usefulness was not significant in Study 1. This is supported by a weaker relationship observed between connectivity and perceived usefulness when social presence is introduced in the equation (Tables 18 and 20). When the effect of adaptiveness and connectivity is regressed on perceived usefulness, social presence drops out of the equation in Study 2 (Table 18) as was the case in Study 1 where the relationship was not significant. In both studies, adaptiveness seems to predict perceived usefulness more strongly than connectivity does, whereas connectivity has a stronger relationship with social presence than with perceived usefulness. While the results of the mediation analyses and LISREL analysis are not directly comparable, the general pattern of results confirms the validity of the research model posited. 80 6.2.3 Partial Least Squares Analysis The data from Study 2 was also analyzed using partial least squares (PLS) to test the research model (Figures 4, 6A and 6B). The main objective of this analysis is to affirm that the general pattern of results holds true holistically as shown from the LISREL analysis in study 1 (and in a disaggregated manner by the mediation analyses in Study 2). PLS, like LISREL, is a structural equation modeling technique and PLS is chosen over LISREL in this case because of sample size constraints. PLS can be used to analyze data with small sample sizes and can be used in this case. Wynne Chin suggests "5-10 times the scale with the largest number of formative indicators or 5-10 times the largest number of structural paths directed at a particular construct in a structural model" (Chin et al. 1999). He further adds that the sample size requirements drop considerably for constructs with reflective indicators, as is the case in this model. Figure 9 shows the results of the PLS analysis pictorially. Table 22A shows the composite reliability values and the average variance extracted (AVE) for the constructs used in the study demonstrating strong reliability. Loading coefficients of all items on their respective constructs were significant with t-values exceeding 1.96 (measurement model in Figure 9), thus suggesting good convergent validity. Table 22B shows the square root of A V E on the diagonal and the correlation between the latent constructs on the off-diagonal position of the matrix displayed. The square root of the A V E of every construct exceeds the correlation between that construct and all other constructs thus signifying good discriminant validity. 81 Table 22A: Estimates of Composite Reliability and Variance Constructs Reliability Explained Variance Adaptiveness 0.95 0.74 Connectivity 0.94 0.79 Perceived Usefulness 0.97 0.86 Loyalty 0.94 0.79 Social Presence 0.92 0.73 Table 22B: Square-root of AVE and Correlation Between the Latent Constructs Constructs Adaptiveness Connectivity Perceived Usefulness Loyalty Social Presence Adaptiveness 0.86 Connectivity 0.57 0.89 Perceived Usefulness 0.83 0.75 0.93 Loyalty 0.69 0.55 0.81 0.89 Social Presence 0.61 0.55 0.60 0.62 0.85 The results from Figure 9 show that the general pattern of results holds true for this analysis as well. All hypotheses are supported with the exception of H2d and H3 (see Table 6 that shows the test results for Study 1). However, the results need to be interpreted cautiously as a direct comparison of results from LISREL analysis is not advisable. The assumptions underlying both methods are different (Barclay et al. 1995; Chin 1998; Chin et al. 1999) and in fact Chin compares both methods with the quote "Partial Least Squares is to LISREL as Principal Components Analysis is to Common Factor Analysis" (Chin 1995). The main intent here is to show that the general pattern of 83 results is consistent with those from previous analyses and hence PLS analysis is appropriate in this context in the light of small sample size fortius study. 84 7. Conclusions This research investigated the impact of perceived communication characteristics on social presence, perceived usefulness and customer loyalty. A lab experiment also examined the impact of a web site's support for personalization features and consumer review features on these perceived communication characteristics of the media (adaptiveness and connectivity). The results from Study 1 show that social presence acts as a mediator between adaptiveness and loyalty as well as connectivity and loyalty. Perceived usefulness acts as a mediator between adaptiveness and loyalty. The hypotheses suggesting social presence acting as a mediator between communication characteristics and perceived usefulness was not supported. Instead, adaptiveness had a stronger effect on perceived usefulness whereas connectivity showed a stronger relationship with social presence and vice versa. In fact, the link suggesting a direct relationship between connectivity and perceived usefulness was also rejected. These results suggest that a better way of examining these relationships may be to look at the communication characteristics individually and determine whether the experiential or the utilitarian component predominates for that communication characteristic in a specific task setting. This research argues that web sites in trying to develop meaningful relationships with its customers must focus on improving the perceived communication characteristics of the media. This research explored the impact of two such communication characteristics - adaptiveness and connectivity. 85 The second study utilizing a lab experiment set out to show how a web site's support for personalization features and consumer reviews will have an impact on these two perceived communication characteristics of media. The results unambiguously showed that these web interface design decisions did make the web site more adaptive and connective. This result, when juxtaposed with the results from the first study indicates that appropriate web interface design decisions will eventually lead to higher levels of consumer loyalty. The controlled conditions used in the experiment allows for the causal claim that higher levels of support for personalization features and consumer review features led to higher levels of adaptiveness, connectivity, social presence, perceived usefulness and loyalty. The interaction effect between support for personalization and consumer review features (Figures 7 and 8) also gives rise to interesting possibilities for designing web sites. This shows that web sites by choosing to provide support for just one of these features can simultaneously positively impact both adaptiveness and connectivity. For example, a web site that offers support for just personalization features not only makes itself adaptive, but also more connective. Similarly, a web site that provides support just for consumer reviews may lead to the web site being perceived as highly connective as well as adaptive. The direction of means in Figures 7 and 8 clearly illustrate this argument. What this means for web sites in terms of practical implications will be explored in the sub-section 7.3. The mediation analyses and'the PLS analysis from Study 2 support the general pattern of results from Study 1 thus providing triangulation across research methods increasing confidence in the general research model proposed. 86 7.1 Theoretical Contributions This research also made a number of new theoretical contributions while reiterating support for established as well as not-so-established extant theories. This work reviewed the perceived communication characteristics of the media developed by previous research (Burgoon et al. 2000; Daft et al. 1986; Dennis et al. 1999; Sproull et al. 1986; Te'eni 2001; Valacich et al. 1993) and introduced connectivity as an important communication characteristic that was missing in previous literature. Following Dennis and Kinney's (1999) arguments for treating communication characteristics of a medium as the fundamental building blocks to understand performance effects of that medium, this thesis separated the two perceived communication characteristics of a web site -adaptiveness and connectivity - from the experiential construct social presence and showed that they do help in understanding the communication capabilities of that web site. Adaptiveness had an impact on both the experiential and the utilitarian aspects of relationship between a web site and its users. On the other hand, connectivity had an impact only on the experiential aspect of the relationship. This thesis also adds further evidence to the media use research that posits that the same communication medium can possess different levels of social presence by configuring the medium differently (Dennis et al. 1999; Fulk et al. 1993). This is empirically shown in a new context (for a web site) by systematically varying a web site's support for personalization and consumer reviews. While past research has shown support for treating computers as social actors (Moon 2000; Nass et al. 1995a; Nass et al. 1995b; Nass et al. 1994; Reeves et al. 1997), this research extends the theory of social response 87 by making a case for treating a web site as a valid social actor. This work also uncovers the assumptions underlying the construct social presence and makes a case for using social presence as an experiential construct to capture the "relational" aspect of the communication between a web site and its users. These arguments for treating the web site as a valid social actor and social presence as an experiential construct pave the way for one of the key contributions of this work. This research, by synthesizing theories from communication, media use, theory of social response and theory of planned behavior, proposes an extension to the technology acceptance model to understand customer behavior better. This research argues that social presence can be used as the construct of choice to represent the relational aspect of communication between a web site and its users while perceived usefulness can be used as the utilitarian construct. Consistent with previous literature, perceived usefulness does predict a customer's intent to return to a web site as measured by customer loyalty. While previous studies have shown mixed support for the use of affect (attitude), this study offers new evidence for the impact of social presence (belief structure) on customer loyalty. 7.2 Limitations While this research investigated the impact of two communication characteristics -adaptiveness and connectivity - further research is needed to study the impact of other communication characteristics (see Table 1 for a list of these communication characteristics). Further, the research model was tested in one context involving a high involvement product (music CDs). Hence, the associated research model should be 88 tested in other contexts to check for potential moderator effects. For example, the level of involvement associated with a particular product may play a moderating role on the relationship between the experiential construct social presence and customer loyalty. Study 1 used university students as participants in the survey and the results as such generalize to this target population. While students are an attractive market segment for music CDs, care should be exercised in interpreting the results of the study across the board. Even though, the participants were offered the chance to win one of the music CDs they chose as an incentive to increase their involvement and to ensure external validity, the survey was not conducted in a real setting where customers were actually shopping in an online store. However, since the main intent of the study was to explore customer behavior in the pre-purchase stage, the employed methodology is acceptable as an alternative to the ideal. Study 2 used support for specific features of personalization and consumer reviews by one e-retailer - Amazon.com - in an experimental setting. As this research argued, there are many ways of implementing support for these features as web site interface design decisions. Amazon.com was chosen specifically because it is recognized as one of the leading e-retailers that serves as an exemplar for other on-line stores in terms of implementing support for personalization and consumer reviews. There is ample scope for further research that focuses on the details of implementing support for these features in specific ways to improve the communication capabilities of a web site. 89 Another limitation of this experimental study is that the results may be specific to the way Amazon.com (and hence this study) chose to implement support for personalization and for consumer reviews. As the results suggest, there is an element of commonality between the way support for personalization and support for consumer reviews are implemented. Amazon.com uses the explanation "other customers who bought this product also bought xxx" to recommend other similar items that a customer may be interested in thus resulting in support for personalization. There is an element of "other customers" influencing the personalized recommendations that are being made by the web site. Thus, there is a common theme of "other customers' views and recommendations" running through the implementations of both the support for personalization as well as support for consumer reviews. Similarly, Amazon.com chose to integrate support for consumer reviews on the product page of the web site. This integration leads to a customer perceiving reviews by other customers as relevant information on the product he/she is interested in, thus lending an element of 'adaptiveness' to the way support for consumer reviews is implemented. Hence, the two treatments (support for personalization and consumer reviews) are not completely orthogonal and are an artifact of the way these features are implemented in Amazon.com. While Amazon.com is a leading company in terms of implementing support for personalization as well as consumer reviews, the results of the study should be interpreted with caution keeping in mind that "specific implementations of technologies" also have a role to play. This observation suggests possible avenues of 90 future research in the different ways a web site can choose to implement these technologies some of which will be touched upon in the next section. 7.3 Implications for Practice and Research In practice, it is advantageous for the web sites to offer some form of support for personalization or virtual community as this makes the web site to be perceived as more adaptive and connective. This will facilitate better communication between the web site and the shoppers, thus leading to higher levels of social presence and perceived usefulness. Ultimately, this will result in increased customer loyalty. Companies do understand that in practical terms it takes a lot more money and effort to acquire a new customer than to keep an existing customer and the results of this study throws new light on the role of support for personalization and consumer reviews in increasing customer loyalty. Good personalization systems can be very expensive to set up. Enabling an e-commerce web site with the necessary tools to build a vibrant community costs little (especially when compared to personalization systems) as the community members provide the content. The results of the second study offers evidence that web sites by providing support for consumer reviews could not only increase connectivity but also reap the benefits in terms of increased adaptiveness despite offering very little personalization. However, the ways and means of implementing support for personalization and virtual communities deserve further research. 91 This research implemented the virtual community support by incorporating the consumer reviews on the product page (as Amazon.com does). There are other web sites such as CDNow that implement the virtual community features in a different manner. The virtual community content at CDNow is not integrated into the product page and is implemented using an elaborate bulletin board system. Future research needs to explore these implementations in finer detail to figure out if they make a difference in terms of perceived connectivity and adaptiveness. Similarly, there are many different ways of implementing a personalization system (and the subsequent set of recommendations). Even after a set of recommendations have been arrived at, the way of introducing these recommendations to the shopper may yet make a difference. What is more effective in introducing these sets of recommendations - the explanation "Customers who bought this item also bought" or the phrase "You may also like"? It would be interesting to apply the principles set forth in the Elaboration Likelihood Model (Petty et al. 1986) to investigate to what extent the content (message) being delivered and the way it is being delivered (peripheral or central route to persuasion) impacts on the constructs studied in this research. Future research should investigate how specific implementations of personalization systems and virtual communities can use these "explanations" to persuade customers about the efficacy and credibility of the message. This study only provides a starting point in looking at the way personalization systems and virtual cornmunity features can be built into an e-commerce web site. Future research should examine the 92 implementation of these features in finer detail. This will help the organizations understand more in depth the trade-offs involved in providing different types and levels of personalization and virtual communities. 93 References Ajzen, I. "The Theory of Planned Behavior," Organizational Behavior and Human Decision Processes (50:1) 1991, pp 179-211. Andre, E. , and Rist, T. "From Adaptive Hypertext to Personalized Web Companions," Communications of the ACM (45:5) 2002, pp 43-46. Ariely, D. "Controlling the Information Flow: Effects on Consumers' Decision Making and Preferences," Journal of Consumer Research (27:2) 2000, pp 233-248. Bandura, A. Social Foundations of Thought And Action: A Social Cognitive Theory Prentice Hall, Englewood Cliffs, NJ, 1986. Barclay, D., Thompson, R., and Higgins, C. "The Partial Least Squares (PLS) Approach to Causal Modeling: Personal Computer Adoption and Use as an Illustration," Technology Studies (2:2) 1995, pp 285-309. Baron, R.M., and Kenny, D.A. "The Moderator-Mediator Variable Distinction in Social Psychological Research: Conceptual, Strategic and Statistical Considerations," Journal of Personality and Social Psychology (51:6) 1986, pp 1173-1182. Bickart, B., and Schindler, R. "Internet Forums As Influential Sources of Consumer Information," Journal of Interactive Marketing (15:3) 2001, pp 31-40. Billsus, D., Brunk, C.A., Evans, C , Gladish, B., and Pazzani, M . "Adaptive Interfaces for Ubiquitous Web Access," Communications of the ACM (45:5) 2002, pp 34-38. Biocca, F., and Delaney, B. "Immersive Virtual Reality Technology," in: Communication in the Age of Virtual Reality, F. Biocca and M . Levy (eds.), Lawrence Erlbaum Associates, Hillsdale, NJ, 1993, pp. 57-124. Brown, S., Tilton, A., and Woodside, D. "Online Communities Pay," McKinsey Quarterly:(\9,\) 2002, pp 17. Burgoon, J.K., Bonito, J.A., Bengtsson, B., Cederberg, C , Lundeberg, M . , and Allspach, L. "Testing the Interactivity Model: Communication Processes, Partner Assessments, and the Quality of Collaborative Work,". Journal of Management Information Systems (16:3) 2000, pp 33-56. Burgoon, J.K., and Hale, J.L. "Validation and Measurement of the Fundamental Themes of Relational Communication," Communication Monographs (54:1) 1987, pp 19-. 41. Burke, K., and Chidambaram, L. "How Much Bandwidth is Enough? A Longitudinal Examination of Media Characteristics and Group Outcomes," MIS Quarterly (23:4) 1999, pp 557-580. Capozza, D., Falvo, R., Robusto, E. , and Orlando, A. "Beliefs About Internet: Methods of Elicitation and Measurement," Papers on Social Representation (12:1) 2003, pp 1.1-1.14. Carlson, P.J., and Davis, G.B. "An Investigation of Media Selection Among Directors and Managers: From "Self to "Other" Orientation," MIS Quarterly (22:3) 1998, pp 335-362. Cenfetelli, R.T., and Benbasat, I. "Measuring Information Technology Mediated Customer Service: A Functional Perspective," Working Paper, Sauder School of Business, University of British Columbia 2003. Chin, W.W. "Partial Least Squares Is To LISREL As Principal Components Analysis Is To Common Factor Analysis," Technology Studies (2:2) 1995, pp 315-319. 94 Chin, W.W. "Issues and Opinions on Structural Equation Modeling," MIS Quarterly (22:1) 1998, ppvii-xvi. Chin, W.W., and Newsted, P.R. "Structural Equation Modeling Analysis With Small Samples Using Partial Least Squares," in: Statistical Strategies For Small Sample Research, R. Hoyle (ed.), Sage Publications, 1999, pp. 307-341. Chiu, W. "Web Site Personalization," IBM High-Volume Web Site Team, WebSphere Software Platform. Cingil, I., Dogac, A., and Azgin, A. "A Broader Approach to Personalization," Communications of the ACM (43:8) 2000, pp 136-141. Compeau, D. "Social Cognitive Theory And Individual Reactions To Computing Technology: A Longitudinal Study," MIS Quarterly (23:2) 1999, pp 145-158. Compeau, D., and Higgins, C A . "Application of Social Cognitive Theory To Training For Computing Skills," Information Systems Research (6:2) 1995, pp 189-211. Daft, R.L., and Lengel, R.H. "A Proposed Integration among Organizational Information Requirements, Media Richness, and Structural Design," Management Science (32) 1986, pp 554-571. Daft, R.L., Lengel, R.H., and Trevino, L. "Message Equivocality, Media Selection, and Manager Performance," MIS Quarterly (11:3) 1987, pp 355-366. Davern, M.J., and Te'eni, D. "Content Versus Structure in Information Environments, A Longitudinal Analysis of Web Site Preferences," Proceedings of the International Conference on Information Systems, Brisbane, Australia, 2000. Davis, F.D. "Perceived Usefulness, Perceived Ease of Use, and User Acceptance of Information Technology," MIS Quarterly (13:3) 1989, pp 319-339. Dellaert, B., and Khan, B.E. "How Tolerable is Delay? Consumers' Evaluations of Internet Web Sites after Waiting," lournal of Interactive Marketing (13:1) 1999, pp 41-54. Dennis, A., and Valacich, J. "Rethinking Media Richness - Towards a Theory of Media Synchronicity," Proceedings of the 32nd Hawaii International Conference on System Sciences, Hawaii, 1999. Dennis, A.R., and Kinney, S.T. "Testing Media Richness Theory in the New Media: The Effects of Cues, Feedback, and Task Equivocality," Information Systems Research (9:3) 1998, pp 256-274. Devaraj, S., Fan, M . , and Kohli, R. "Antecedents of B2C Channel Satisfaction and Preference: Validating e-Commerce Metrics," Information Systems Research (13:3) 2002, pp 316-333. Dillard, J.P., Solomon, D.H., and Palmer, M.T. "Structuring the Concept of Relational Communication," Communication Monographs (66) 1999, pp 49-65. Fulk, J., and Boyd, B. "Emerging Theories of Communication in Organizations," lournal of Management (17:2) 1993, pp 407-446. Gartner "Business to Commerce: Statistics," Gartner Research; URL: http://press.gartner.com/press gartner/quickstats/b2c.html, 2000. Gefen, D., and Straub, D.W. "Managing User Trust in B2C e-Services," e-Service lournal (2:2) 2003, pp 7-25. Gerbing, D.W., and Anderson, J.C. "An Updated Paradigm for Scale Development Incorporating Unidimensionality and Its Assessment," Journal of Marketing Research (25) 1988, pp 186-192. 95 Ghose, S., and Dou, W. "Interactive Functions and Their Impacts on the Appeal of Internet Presence Sites," Journal of Advertising Research) 1998, pp 29-43. Goodhue, D.L., and Thompson, R.L. "Task Technology Fit and Individual Performance," MIS Quarterly (19:2) 1995, pp 213-236. Hair, J.F., Anderson, L .E. , Tatham, L.L. , and Black, W.C. Multivariate Data Analysis Prentice Hall, Englewood Cliffs, N.J., 1995. Haubl, G., and Trifts, V. "Consumer Decision Making in Online Shopping Environments: The Effects of Interactive Decision Aids," Marketing Science (19:1) 2000, pp 4- • 21. Heeter, C. "Communication Research on Consumer Virtual Reality," in: Communication in the Age of Virtual Reality, F. Biocca and M . Levy (eds.), Lawrence Erlbaum Associates, Hillsdale, NJ, 1995, pp. 191-218. Hoffman, D.L., and Novak, T.P. "Marketing in Hypermedia Computer-Mediated Environments: Conceptual Foundations," Journal of Marketing (60:3) 1996, pp 50-68. Hubona, G.S., and Burton-Jones, A. "Modeling the User Acceptance of E-mail," Proceedings of the 36th Hawaii International Conference on System Sciences, IEEE Computer Society, Hawaii, 2003. Jones, Q. "Virtual-communities, virtual settlements & cyber-archaeology: A theoretical outline," Journal of Computer Mediated Communication (3:3) 1997. Karahanna, E. , and Straub, D.W. "The Psychological Origins of perceived Usefulness and Ease-of-use," Information & Management (35) 1999, pp 237-250. Kelloway, K.K. Using LISRELfor Structural Equation Modeling Sage, Thousand Oaks, 1998. Kirkpatrick, D. "Esther Dyson: Living The Networked Life," Fortune (145:11) 2002, pp 168-172. Koufaris, M . "Applying the Technology Acceptance Model and Flow Theory to Online Consumer Behavior," Information Systems Research (13:2) 2002, pp 205-223. Kumar, N., and Benbasat, I. "Shopping as Experience and Web Site as a Social Actor: Web Interface Design and Para-Social Presence," Proceedings of the International Conference on Information Systems, New Orleans, 2001. Kumar, N., and Benbasat, I. "Para-Social Presence: A Re-conceptualization of 'Social Presence' to Capture the Emerging Relationship Between a Web Site and Her Visitors," 35th Annual Hawai'i International Conference On System Sciences (HICSS), IEEE, Hawai'i, 2002. Lohse, G., and Spiller, P. "Electronic Shopping," Communications of the ACM (41:7) 1998, pp 81-87. Lombard, M . , and Ditton, T.B. "At the Heart of It All: The Concept of Presence," Journal of Computer Mediated Communication (3) 1997. Mandel, N., and Johnson, E.J. "Constructing Preferences Online: Can Web Pages Change What You Want?," in: Working Paper, MIT Ecommerce Forum, Boston, 1999. Markus, L. "Electronic Mail as the Medium of Managerial Choice," Organization Science (5:4) 1994, pp 502-527. Millerson, G. The Technique of Television Production Hastings House, New York, 1969. Moon, Y. "Intimate Exchanges: Using Computers to Elicit Self-Disclosure from Consumers," Journal of Consumer Research (26) 2000, pp 323-339. 96 Nass, C , Lombard, M . , Henriksen, L. , and Steur, J. "Anthropocentrism and Computers," Behavior and Information Technology (14:4) 1995a, pp 229-238. Nass, C , Moon, Y., Fogg, B.J., Reeves, B., and Dryer, C. "Can Computers be Human Personalities?," International Journal of Human-Computer Studies (43:2) 1995b, pp 223-239. Nass, C , and Steur, J. "Voices, Boxes and Sources of Messages: Computers and Social Actors," Human Communication Research (19:4) 1994, pp 504-527. Padmanabhan, B., and Yang, Y. "On Evaluating Online Personalization," Proceedings of the Workshop on Information Technology and Systems, New Orleans, 2001, pp. 35-41. Palmer, J.W. "Web site usability, design, and performance metrics," in: Information Systems Research, 2002, pp. 151-168. Petty, R.E., and Cacioppo, J.T. Communication and Persuasion: Central and Peripheral Routes to Attitude Change Springer-Verlag, New York, 1986. Preece, J. "Sociability and Usability: Twenty Years of Chatting Online," Behavior and Information Technology (20:5) 2001, pp 347-356. Preece, J. "Supporting Community and Building Social Capital," Communications ofthe ACM (45:4) 2002, pp 37-39. Reeves, B. "Being There: Television as Symbolic Versus Natural Experience," in: Unpublished Manuscript, Stanford University, Institute for Communication Research, Stanford, 1991. Reeves, B., and et al. The Media Equation CSLI Publications, Stanford, CA, 1997. Rice, R.E., and Case, D. "Electronic Message Systems in the University: A Description of Use and Utility," Journal of Communications (33:1) 1983, pp 131-152. Sherblom, J. "Direction, Function, and Signature in Electronic Mail," Journal of Business Communication (25:4) 1988, pp 38-53. Short, J., Williams, E. , and Christie, B. The Social Psychology of Telecommunication John Wiley & Sons, New York, 1976. Sproull, L.S., and Kiesler, S. "Reducing Social Context Cues: Electronic Mail in Organizational Communication," Management Science (32:11) 1986, pp 1492-1512. Steinfield, C.W. Computer-Mediated Communications in an Organizational Setting: Explaining Task-Related and Socio-Emotional Uses Newbury Park, CA, 1986, pp. 777-804. Straub, D.W. "The Effect Of Culture On LT Diffusion: E-mail And F A X In Japan And The U.S.," Information Systems Research (5:1) 1994, pp 23-47. Straub, D.W., Hoffman, D.L., Weber, B.W., and Steinfield, C. "Measuring e-commerce in Net-enabled organizations: An introduction to the special issue;," in: Information Systems Research, 2002, pp. 115-124. Taylor, S., and Todd, P.A. "Understanding Information Technology Use: A Test Of Competing Models," Information Systems Research (6:4) 1995, pp 144-176. Te'eni, D. "Review: A cognitive-affective model of organizational communication for designing LT," MIS Quarterly (25:2) 2001, pp 251-312. Thompson, R.L., Higgins, C.A., and Howell, J.M. "Personal Computing: Toward A Conceptual Model Of Utilization," MIS Quarterly (15:1) 1991, pp 124-143. Triandis, H.C. Interpersonal Behavior Brooke/Cole, Monterey, CA, 1977. 97 Valacich, J.S., Paranka, D., and Nunamaker, J.F. "Communication Concurrency and the New Media: A New Dimension for Media Richness," Communication Research (20:2) 1993, pp 249-276. Venkatesh, V., Morris, M.G. , Davis, G.B., and Davis, F.D. "User Acceptance Of Information Technology: Toward A Unified View," MIS Quarterly (27:3) 2003, p forthcoming. Venkatraman, N. "Strategic Orientation of Business Enterprises: The Construct, Dimensionality, and Measurement," Management Science (35:8) 1989, pp 942-962. Voss, C. "Developing ah eService Strategy," Business Strategy Review (11:1) 2000, pp 21-33. Walther, J.B. "Interpersonal Effects in Computer-Mediated Interaction," Communication Research (19:1) 1992, pp 52-90. Watzlawick, P., Beavim, J.H., and Jackson, D.D. Pragmatics of Human Communication W.W.Norton, New York, 1967. Weinberg, B.D. "Don't Keep Your Internet Customers Waiting Too Long at the (Virtual) Front Door," Journal of Interactive Marketing (14:1) 2000, pp 30-39. Zahedi, F .M. , and Lu, Y. "Website Personalization for Relationship Building: A Conceptual Framework," Ninth Americas Conference on Information Systems, Tampa, Florida, 2003, pp. 2256-2264. Zhu, K., and Kraemer, K. "e-Commerce Metrics For Net-enhanced Organizations: Assessing the Value of e-Commerce to Firm Performance in the Manufacturing Sector," Information Systems Research (13:3) 2002, pp 275-295. 98 Appendices Appendix - 1 Operationalization of Constructs and Original List of Items Construct Item Labels Items Adaptiveness ADP1 ABC.com personalized my web shopping experience. ADP2 I feel that ABC.com must have been designed for individuals like me. ADP3 The pages displayed by ABC.com seemed to be tailored to my needs. ADP4 I feel as though ABC.com is custom-made for me. ADP5 I feel that ABC.com personalized its offerings based on my requirements. ADP6 I feel that my interactions with ABC.com are not at all personalized. ADP7 I felt that ABC.com did not adapt itself to serve my personal needs. Connectivity CON1 ABC.com never gave me a chance to interact with other visitors to its web site. CON2 ABC.com exposed me to opinions of other visitors to its web site. CON3 ABC.com created a sense of community. CON4 ABC.com gives me an opportunity to meet people with similar interests. CON5 ABC.com lets me meet others with similar tastes. CON6 I feel like I know the other users who visited ABC.com. CON7 I did not identify with the opinions of the visitors I met at ABC.com CON8 I felt linked to the other users of ABC.com. Perceived Usefulness Loyalty Social Presence PU1 ABC.com is useful in shopping for CDs. PU2 ABC.com improves my performance in shopping for CDs. PU3 ABC.com enables me to shop for CDs faster. PU4 ABC.com enhances my effectiveness in CD shopping. PU5 ABC.com makes it easier to shop for CDs. PU6 ABC.com increases my productivity in shopping for CDs. LOY1 I am very likely to buy CDs from ABC.com. L O Y 2 I would recommend ABC.com to my friends and relatives. L 0 Y 3 I would seriously contemplate buying from ABC.com. L 0 Y 4 I am likely to make future purchases from ABC.com. L 0 Y 5 I would be shopping at ABC.com again. L O Y 6 I would return to ABC.com to make purchases. SP1 There is a sense of human contact in the web site SP2 There is a sense of personalness in the web site SP3 There is a sense of sociability in the web site SP4 There is a sense of human warmth in the web site SP5 There is a sense of human sensitivity in the web site 99 Appendix - 2: Questionnaire and Instructions used in Study 1 Study on Web Shopping Behavior November 2001 Dr. Izak Benbasat Canada Research Chair in Information Technology Management Faculty of Commerce and Business Admin. University of British Columbia Vancouver, BC, V 6 T 1Z2 izak.benbasat@ubc.ca Nanda Kumar Ph.D. Student MIS Division Faculty of Commerce and Business Admin. University of British Columbia Vancouver, BC, V 6 T 1Z2 nanda.kumar@commerce.ubc.ca 100 G E N E R A L INSTRUCTIONS Welcome to the web shopping survey conducted by the MIS Division at the Faculty of Commerce, University of British Columbia. This survey is designed to evaluate the overall shopping experience provided by an on-line shopping web site. The questions do not evaluate you, only your opinion of the web site. As such there are no right or wrong answers. This survey is divided into three sections. The first section (section-A) primarily consists of questions on your general background. The second section (section - B) consists of questions about the web site where you will shop for CDs today (you will complete this section after the shopping trip). The last section consists of a few questions on the features of the web site you noticed while shopping for CDs. It is very important to answer all of the questions included in the questionnaire, without leaving out a single question. If you are not sure of the answer to a question, please give us your best opinion. All information provided by you is kept confidential. Thank you for your participation! Note: In the following pages, you may be asked to respond to a set of statements about your recent web shopping experience. Please circle a number that indicates the extent to which you agree with each statement. The correct way of responding to these statements is shown below. Examples: Correct way of marking: 1 Strongly Disagree 2 Disagree 3 Mildly Disagree 4 Neither Agree Nor Disagree ildly Agree 6 Agree 7 Strongly Agree Incorrect way of marking: Do not circle between numbers 1 2 3 4 5 6 7 Strongly Disagree Mildly Neither Mildly Agree Strongly Disagree Disagree Agree Nor Agree Agree Disagree Incorrect way of marking: Do not circle two numbers 1 2 3 JLr^ ' 6 7 Strongly Disagree Mildly NetTher Mildly Agree Strongly Disagree Disagree Agree Nor Agree Agree Disagree 101 S E C T I O N - A Date: Name: Gender: Telephone: Age: Major at U B C : Year of Study: Email: Instructions: Please circle one item that is closest to your experience. 1. How many hours per week do you use a personal computer? (Ex: IBM compatible/ Mac) a) Less than 1 hour d) 11 - 20 hours b) 1-5 hours e) More than 20 hours c) 6-10 hours 2. How long have you been using the Internet? (Ex: read news, e-mail, general browsing) a) Never d) 1-2years b) Less than 6 months e) More than 2 years c) 6-12 months 3. How many hours per week do you spend on the Internet ? a) Less than 1 hour d) 11-20 hours b) 1 - 5 hours • e) More than 20 hours c) 6-10 hours 4. In the past 12 months, how many times have you made a purchase on-line? a) Never d) 5-10 times b) Once e) More than 10 times 1 0 2 c) 2- 4 times 5. In the past 12 months, about how much money have you spent shopping on-line? a) None d) $500-$J,000 b) <$100 e) >$1000 c) $101-$500 6. How long have you been visiting Amazon.com? (General browsing included) a) Never d) 1-2 years b) Less than 6 months e) More than 2 years c) 6 - 12 months 7. How many hours per week do you spend at Amazon.com ? (General browsing included) a) None d) 6 - 10 hours b) Less than 1 hour e) More than 10 hours c) 2-5 hours 8. In the past 12 months, how many times have you made a purchase on-line at Amazon.com? a) Never d) 5 -10 times b) Once e) More than 10 times •c) 2-4 times 9. In the past 12 months, about how much money have you spent shopping at Amazon.com? a) None d) $500-$],000 b) <$100 e) >$1000 c) $101-$500 Please circle a number that indicates the extent to which you agree with each statement. Strongly Strongly Disagree Agree 103 j 10") I feci com fori able with using a mouse and 1 computer applications If) I am comfortable browsing the internet 1 2 3 4 5 6 7 (read email, check news, general browsing) 12) 1 am comfortable with shopping on-line 1 2 3 4 5 6 7 13) I am familiar with Ama/on.com 1 2 3 1 5 6 7 14) I \isit Ama/.on.coin regukirlv 1 2 3 4 5 6 7 E n d of Section - A If you have questions or if you do not understand the instructions please ask the Research Assistant to clarify. Otherwise, proceed to the next page... 104 Shopping for CDs - Instructions o Please follow instructions. o You have up to 70 minutes to complete these tasks. The time limit provided for each task is only a suggestion. But, please ensure that you spend at least the minimum time suggested for each task. For example, please spend at least 10 minutes on task 1. o Remember that this is a shopping trip... There is no right or wrong way of doing things... Only your way! So, relax and enjoy. You will be given specific instructions and you will be asked to shop for CDs from an on-line shopping store - Amazon.com (e.g. shop for a gift for a friend who likes a certain type of music, shop for a CD by your favorite artist). Please keep in mind that you don't need to actually purchase CDs, but we do expect you to look for information about the CDs at Amazon.com. Once you choose/find the CDs you'd like to buy, place them in the shopping cart (available in the web site). Please ask the R.A. for assistance if you are not clear about the terminology or the instructions. . Task - 1: Getting to know the web site (10 - 25 minutes). • Please open your favorite browser (Internet Explorer/Netscape Navigator) if you have not done so already. Please visit the home page of Amazon.com (Type in the URL: http://www.Amazon.com in the address bar; Please ask R.A. for help if you are not sure how to do this) • Try to locate information about the following CD: Title - Thriller; Artist - Michael Jackson. > The easiest way to do this is to use the search function (search by Artist Name/CD title or use both). > Open the page that contains information about this CD (product page). > Please spend some time looking at the content of this page. • Return to the home page and spend some time trying to familiarize yourself with the web site (lay out of the web site, the type of products sold, the way information is organized etc.) > Please remember that the objective is to familiarize yourself with the web site. There is no specific way to do this other than browsing and navigating your way around the web site. > Pay particular attention to the features and services provided by the web site to support and enhance your shopping needs and experience. > For some services you wish to use, the web site will require you log in (you may wish to create a new user account if you don't have one already). Creating a user account is easy for these shopping web sites. All you need to do is to enter an email id, your name and password. Feel free to use your personal email or ask the R.A. for a temporary email id. Task -2: Shopping for a gift for your friend (10 - 20 minutes) 105 A close friend of yours, recently told you that he bought the CD titled "Greatest Hits" by A l Green and loved it. He confessed that this was the first time he bought this type of music (Soul/R&B) and is absolutely enthralled by A l Green. Your friend's birthday is coming up in 15 days and you would like to give him a thoughtful gift for his birthday. So, your intention is to buy a CD similar to A l Green's 'Greatest Hits', but from another artist in order to pleasantly surprise him. You will use this shopping web site to look for more information when buying this gift for your friend. Please spend enough time so as to make sure that you choose a suitable CD as a gift for your friend's birthday. Use the space below to answer the following questions about your shopping for the gift 1) Name of the CD you chose: 2) Name of the artist (for the CD you chose): 3) Other Artists you considered (at least three): 4) Reasons for your choice: Task -3 : Shopping for yourself (5 - 1 0 minutes) Please search for CDs of your favorite artists. You will be shopping for a CD that you would like to buy (1 in 5 chance of winning this CD). You are familiar with this type of music and have listened to one or more songs by these artists before. Please make sure that you visit product pages for at least three CDs. Use the space below to answer the following questions about your shopping for the gift 1) Name of the CD you chose: 2) Name of the Artist: Task -4: Shopping for yourself - II (5 - 1 0 minutes) In this case shop for a CD by an artist that you are not too familiar with. You've probably heard a song by this artist somewhere; and perhaps, your friend mentioned something positive about this artist. You will use this shopping web site to look for music by this artist and also explore other music that might interest you. Use the space below to answer the following questions about your shopping for the gift 1) Name of the C D you chose: 2) Name of the Artist: End of Shopping Tasks Turn to the next page only after you've completed all tasks. 106 SECTION - B In the following pages, you will be asked to respond to a set of statements about your recent shopping experience. Please circle a number that indicates the extent to which you agree with each .statement. The questions do not evaluate you, only your opinion of the web site. As such there are no right or wrong answers. Reminder - What the numbers mean (see page 2): l Strongly Disagree 2 Disagree 3 Mildly Disagree 4 Neither Agree Nor Disagree 5 Mildly Agree 6 Agree 7 Strongly Agree S.No • > 2) ; 3) 4) • 6) (JIB 8) Strongly Disagree Strongly Agree 10) i l l ) Ama/on.com is likable. Amazon.com failed to keep me involved while I was shopping. Ama/on.com has ;ideqLiale knowledge to manage ils business on the Internet I feel as though Amazon.com is custom-made for me. Ama/on.com enhances my effectiveness in C D shopping I would return to Amazon.com to make purchases Ama/on.com created u sense of closeness Amazon.com keeps me totally absorbed in my interactions with it. 1 was completely interested in what I was doing while browsing Ama/on.com Amazon.com personalized my web shopping experience. Ama/on.com had no clue as to what I really 3 4 5 ft 7 3 4 5 6 7 3 4 5 6 7 3 4 5 6 7 4 5 6 7 4 5 6 7 107 Strongly Strongly Disagree Agree 12) I would be shopping at Axnazon.com again 1 2 3 4 5 6 7 13) Amay.on.com is pleasant. 1 2 3 4 5 6 7 14) Amazon.com is incompetent 1 2 3 4 5 6 7 15) There is a scn.se of sociability in 1 2 3 4 5 6 7 16) Amazon.com is useful in shopping for CDs 1 2 3 4 5 6 7 17) I feel lhal Ania/on.com must have been I 2 3 4 5 6 7 designed for individuals like me. IS) Amazon.com holds my allenlion 1 2 ; I ^ l1)) Ania/on.com is unfriendly. 1 2 3 4 5 h 7 20) Ama/on.com uvatcd a sense of distance. 1 2 3 4 5 6 7 21) Ama/on.com has the ability to handle sales I 2 3 4 5 6 7 transactions on Ihe lniernci securely 22) 1 did not get treated right by Ama/on.com 1 2 3 4 5 6 7 23) Ama/on.com excites my curiosity. 1 2 3 4 5 6 7 24) Amazon.com increases my productivity in 1 2 3 4 5 6 7 shopping for CDs 25) Amazon.com lets me meet others with 1 2 3 4 5 6 7 26) The pages displayed by Amazon.com 1 2 3 4 5 6 7 seemed to be tailored to my needs. 27) There is a sense of human sensitivity in I 2 3 4 5 6 7 2.8) Amazon.com did not understand m\ needs 1 2 3 4 5 6 7 2l); Ama/on.com improves my performance in 1 2 3 4 5 6 7 30) Amazon.com exposed me to opinions of 1 2 3 4 5 6 7 other visitors to its web site. 31) Promises made by Ama/on.com are likely lo I 2 3 4 5 6 7 108 S. No 32) 34) 35) 36) j j j j l l 38) 39) 40) 41) 42) 44) 45) 46) 47) 48) 49) 50) I have positive feelings about Amazon.com Ama/on.com cares for mc. I did not identify with the opinions of the visitors I met at Amazon.com 1 was deeply in\ol\ed in my interactions while shopping at Ama/on.com Ama/on.com understood what I wanted. Ama/on.com enables mo to shop for CDs I fell close to Ama/on.com Ama/on.com may be "bending the facts". Amazon.com created a sense of community. I can count un Ama/on.com to consider how its actions may affect me. I dislike Amazon.com I found Ama/on.com to bo \ cry detached in its interactions with mc I feel that Amazon.com personalized its offerings based on my requirements. I trust Ama/on.com to keep my best interests in mind. Amazon.com knows me well. There is a sense of human warmth in Ama/on.com I felt that Amazon.com was aloof in its interactions with me I would seriously contemplate buying from Ama/on.com Strongly Disagree 1 Strongly Agree 2 3 3 3 3 3 3 3 4 4 4 4 4 4 4 4 4 4 4 4 4 6 7 6 7 6 7 6 7 ft 7 E H f l i 6 7 fi 7 6 7 6 7 ft 7 () 7 6 7 ft 7 Amazon.com aroused my imagination. 6 7 109 S. No 51) 1 tell linked to the other users ol Ama/on.com Strongly Disagree JSfBlfi IIS H I 1111 Strongly Agree 52) I felt that Amazon.com did not adapt itself to serve my personal needs. i 2 3 4 5 6 7 53) Ama/on.com's intentions towards ils customers arc benevolent. H0H ItlfSSI IBS ' • 5' ISf NMH 54) I feel like I know the other users who visited Amazon.com i 2 3 4 .5 6 7 55) Ama/on.com docs not know my desires al 2 • H ilfllll H i ^^^^^^^^^ 56) Amazon.com acts sincerely in dealing with its customers i 2 3 4 5 6 7 57, Ama/on.com understood my goals. JJjfjJJl MR IIS 58) I am very likely to buy CDs from Amazon.com i 2 3 4 5 6 7 59) . 1 feel that Ama/on.com puts customers' intetesis be foie its own. KM Bill HMk 60) Amazon.com gives me an opportunity to meet people with similar interests. i 2 3 4 5 6 7 61) 1 was treated fairly by Ama/on.com M l IIP 62) I found Amazon.com to be very detached from me i 2 3 4 5 6 7 63) Ama/on.com w;is very imperson.il i n ils dealings with me. iMHi jjljjj H i tin 64) Amazon.com might be withholding crucial information from me. i 2 3 4 5 6 7 65) Amazon.com will over-charge their customers (considering the overall level of scr\ ices / products provided). MHR Ifelii M l RH H I i l l ;" '7..;-,>r 66) I am likely to make future purchases from Amazon.com i 2 3 4 5 6 7 110 S.No f»7) There is a sense of human conlacl in Strongly Disagree Rilfl Bill i l l ! s III Strongly Agree 68) Amazon.com never gave me a chance to interact with other visitors to its web site. i 2 3 4 5 6 7 W) 1 feel ihiit my interactions with Ania/on.com are not at all personali/cd. iiiiil EffK fjjll ... 5 118 Vr'-:.:'7 v,;': •'• 70) Amazon.com is fun. i 2 3 4 5 6 7 71) Ama/on.com makes il easier to shop for NN0H iflllll BR Bill 5 72) Amazon.com is honest with their customers i 2 3 4 5 6 7 73) Ama/on.com does no! ha\e sufficient expertise and resources lo do business on the IBB §|§j d 74) Amazon.com understood what I was trying to do. i 2 3 4 5 6 7 75) I would recommend Ama/on.com to my friends and relatives ^ ^ ^ ^ 111 831 jjfjjj 76) There is a sense of personalness in Amazon.com i 2 3 4 5 6 7 End of Section - B Close the browser (Internet Explorer/Netscape Navigator). Turn to the next page only after you've completed this section. m SECTION-C Instructions: Please circle a number that indicates the extent to which you agree with each statement. Please do not visit the site again to answer these statements. The questions do not evaluate you, only your opinion of the web site. S.No Features/Services Strongly Strongly Disagree Agree 1) Ama/on.com sells books 1 2 3 4 5 6 7 2) Amazon.com sells CDs 1 2 3 4 5 6 7 3) Ama/on.com provides me with product 1 2 3 4 5 6 7 8) When I visit a product page (for ex, 'Thriller' by Michael Jackson), Amazon.com shows me other similar items purchased by other users 3) Ama/on.com uses these preferences to recommend relevant products 14) If I visit again, Amazon.com will remember my preferences 4) Amazon.com provides me with product 1 2 3 4 5 6 7 reviews by other users 5) Ama/on.com allows me to review a product 1 2 3 4 5 6 7 6) Amazon.com lets me rate music 1 2 3 4 5 6 7 7) When 1 \isit a product page (for ex, "Thriller" by Michael Jackson). Ama/on.com I ? ^ 4 5 f recommends products that might potentially 4 5 6 7 v) Ama/on.com recommends products that might I 2 3 4 5 6 7 potentially interest me 10) Amazon.com allows me to create a wish list of 1 2 3 4 5 6 7 products 1 1) Ama/on.com allows me enter my personal I 2 3 4 5 6 7 information (address, preferred credit card etc) 12) Amazon.com allows me enter my personal 1 2 3 4 5 6 7 preferences (my favorite music, rating CDs) 112 End of Section - C T h a n k y o u for pa r t i c ipa t i ng i n the study. P lease do not d i scuss the s tudy w i t h other f r iends as they m a y be pa r t i c ipa t i ng i n the s tudy as w e l l . 113 Appendix - 3: Questionnaire and Instructions used i n Study 2 Study on Web Shopping Behavior Fall 2002 Dr. Izak Benbasat Canada Research Chair in Information Technology Management Faculty of Commerce and Business Admin. University of British Columbia Vancouver, BC, V6T 1Z2 izak.benbasat@ubc.ca Mr. Nanda Kumar Ph.D. Student MIS Division Faculty of Commerce and Business Admin. University of British Columbia Vancouver, BC, V6T 1Z2 nanda.kumar@commerce.ubc.ca 114 GENERAL INSTRUCTIONS Welcome to the web shopping survey conducted by the MIS Division at the Faculty of Commerce, University of British Columbia. This survey is designed to evaluate your overall shopping experience provided by an on-line shopping web site. The questions do not evaluate you, only your opinion of the web site. As such there are no right or wrong answers. This survey is divided into three sections. The first section (section-A) primarily consists of questions on your general background. The second section (section - B) consists of questions about the web site where you will shop for CDs today (you will complete this section after the shopping trip). The last section consists of a few questions on the features of the web site you visited while shopping for CDs. It is very important to answer all of the questions included in the questionnaire, without leaving out any questions. If you are not sure of the answer to a question, please enter your best opinion. All information provided by you will be kept confidential. Thank you for your participation! Note: In the following pages, you may be asked to respond to a set of statements about your recent web shopping experience. Please circle a number that indicates the extent to which you agree or disagree with each statement. The correct way of responding to these statements is shown below. Correct way of marking: 1 2 3 Strongly Disagree Mildly Disagree Disagree Examples: 4 Neither Agree Nor Disagree © Mildly Agree 6 Agree 7 Strongly Agree Incorrect way of marking: Do not circle between numbers Strongly Disagree 2 Disagree 3 Mildly Disagree 4 Neither Agree Nor Disagree o Mildly Agree 6 Agree 7 Strongly Agree Incorrect way of marking: Do not circle two numbers Strongly Disagree 2 Disagree 3 Mildly Disagree Agree Nor Disagree ildly Agree 6 Agree 7 Strongly Agree 115 SECTION-A Date: Name: Gender: Telephone: Age: Major at U B C : Year of Study: Email: Instructions: Please circle one item that is closest to your experience. 7. How many hours per week do you use a personal computer? (Ex: IBM compatible or Mac) a) Less than 1 hour d) 11-20 hours b) 1 - 5 hours e) More than 20 hours c) 6-10 hours 2. How long have you been using the Internet? (Ex: read news, e-mail, general browsing) a) Never d) 1-2years b) Less than 6 months e) More than 2 years c) 6 —12 months 3. How many hours per week do you spend on the Internet? a) Less than 1 hour d) 11-20 hours b) 1-5 hours e) More than 20 hours c) 6- 10 hours 4. In the past 12 months, how many times have you made a purchase on-line? a) Never d) 5-10 times b) Once e) More than 10 times c) 2- 4 times . 116 5. In the past 12 months, approximately how much money have you spent shopping on-line? a) None d) $50041,000 b) Less than $100 . e) More than $1000 c) $101-$500 6. How long have you been visiting Amazon.com? (General browsing included) a) Never d) 1-2 years b) Less than 6 months e) More than 2 years c) 6 — 12 months 7. How many hours per week do you spend at Amazon.com? (General browsing included) a) None d) 6 -10 hours b) Less than 1 hour e) More than 10 hours c) 2-5 hours 8. In the past 12 months, how many times have you made a purchase on-line at Amazon.com? a) Never d) 5 - 10 times b) Once e) More than 10 times c) 2- 4 times 9. In the past 12 months, about how much money have you spent shopping at Amazon.com ? a) None d) $500-$ 1,000 b) Less than $100 e) More than $1000 c) $101-$500 Please circle a number that indicates the extent to which you agree with each statement. Strongly Disagree Strongly Agree 1 0 ) 1 leel comfortable with using a mouse and 1 ijjgljjj ^ ^ ^ ^ 117 11) I am comfortable browsing the internet 1 2 3 4 5 6 7 (read email, check news, general browsing) 12) I am comfortable with shopping on-line 1 2 3 4 5 6 7 13) I am familiar with Ama/on.com I 2 3 4 5 6 7 14) I visit Ama/on.com regularly 1 2 3 4 5 6 7 End of Section - A If you have questions or if you do not understand the instructions please ask the Research Assistant to clarify. Otherwise, proceed to the next page... 118 Shopping for C D s - Instructions • Please follow instructions shown below. • You have up to 90 minutes to complete these tasks. The time limit provided for each task is only a suggestion. But, please ensure that you spend at least the minimum time suggested for each task. For example, please spend at least 20 minutes on task 1. • Remember that this is a shopping trip... There is no right or wrong way of doing things... Only your way! So, relax and enjoy. • You have a 33% chance of winning a CD. In the following pages, you will be asked to shop for a CD for yourself twice (Task B and Task D). If you win, we will randomly select one of these two CDs as your prize. You will be given specific instructions and you will be asked to shop for CDs from an on-line shopping store - Amazon.com (e.g. shop for a gift for a friend who likes a certain type of music, shop for a CD by your favorite artist). Please keep in mind that you don't need to actually purchase CDs, but we do expect you to look for information about the CDs at Amazon.com. Please ask the research assistant (R.A.) for assistance if you are not clear about the terminology or the instructions. Practice Tasks: Getting to know shopping web sites (20 - 30 minutes). • Please visit the home page of CDNow.com (Click on File>Open and then, type in the URL: http://www.CDNow.com in the address bar; Please ask the R.A. for help if you are not sure how to do this) • Try to locate information about the following CDs: Title - Thr i l l e r ; Ar t i s t - Michae l Jackson > The easiest way to do this is to use the search function (search by Artist Name/CD title or use both). > Open the page that contains information about this CD (product page). > Please spend some time looking at the content of this page. Pay particular attention to the features and services provided on the product page by the web site to support and enhance your shopping needs and experience. > Please read the list below and tick the type of features provided by the web site that you noticed while you were browsing the web site. Show this list to the RA. • Artist Information • Editorial Reviews of Albums and/or Artists • Album Information • Consumer Reviews of Albums and/or Artists • Track Listings • Message Board • Audio Samples of Track Listings • Ability to browse music from similar genre/style • Information on other related/similar artists • Other features not mentioned on this list 119 • Repeat the above steps for Barnes and Noble.com (www.bn.com) • Artist Information • Editorial Reviews of Albums and/or Artists • Album Information • Consumer Reviews of Albums and/or Artists • Track Listings • Message Board • Audio Samples of Track Listings • Ability to browse music from similar genre/style • Information on other related/similar artists • Other features not mentioned on this list • Now ask the R.A. to take you to our shopping site. Repeat the above steps for this web site as well. The objective here is to familiarize yourself with the web site before you start shopping. Please remember that the R.A. would be happy to assist you if you have any questions about the shopping experience and the features of the web site. • Artist Information • Editorial Reviews of Albums and/or Artists • Album Information • Consumer Reviews of Albums and/or Artists • Track Listings • Message Board • Audio Samples of Track Listings • Ability to browse music from similar genre/style • Information on other related/similar artists • Other features not mentioned on this list Task - A : Shopping for a gift for your friend (10 - 20 minutes) A close friend of yours, recently told you that he bought the CD titled "Greatest Hits" by Al Green and loved it. He confessed that this was the first time he bought this type of music (Soul/R&B) and is absolutely enthralled by Al Green. Your friend's birthday is coming up in 15 days and you would like to give him a thoughtful gift for his birthday. So, your intention is to buy a CD similar to Al Green's 'Greatest Hits', but from another artist in order to pleasantly surprise him. You will use this shopping web site to look for more information when buying this gift for your friend. Please spend enough time so as to make sure that you choose a suitable CD as a gift for your friend's birthday. Use the space below to answer the following questions about your shopping for the gift: 1) Name of the CD you chose: 2) Name of the artist (for the CD you chose): 3) Other Artists you considered (list at least three): 120 4) Circle the number that indicates the extent to which you agree with the following statement: a) I am familiar with this genre of Music - Soul/R&B: l Strongly Disagree 2 Disagree 3 Mildly Disagree 4 Neither Agree Nor Disagree 5 Mildly Agree Agree 6 7 Strongly Agree b) I am familiar with the music of this artist - A l Green: l Strongly Disagree 2 Disagree 3 Mildly Disagree 4 Neither Agree Nor Disagree 5 Mildly Agree 6 Agree 7 Strongly Agree 5) Reasons for your choice of CD: Task - B : Shopping for yourself (5 -10 minutes) Please search for CDs of your favorite artists. You will be shopping for a CD that you would like to buy for yourself. You are familiar with this type of music and have listened to one or more songs by these artists before. Alternately, you may also shop for CDs by an artist that you are not too familiar with. You've probably heard a song by this artist somewhere; and perhaps, your friend mentioned something positive about this artist. You will use this shopping web site to look for music by this artist and also explore other music that might interest you. Use the space below to answer the following questions about the CD you chose for yourself: a) Name of the CD you chose: b) Name of the Artist: Ask the RA for further instructions. Please don't turn to the next page. 121 Task - C : Shopping for a gift for your friend (10 - 20 minutes) A close friend of yours, recently told you that he bought the CD titled "The Best of James" by James and loved it. He confessed that this was the first time he bought this type of music (Alternative Rock) and is absolutely enthralled by James. Your friend's birthday is coming up in 15 days and you would like to give him a thoughtful gift for his birthday. So, your intention is to buy a CD similar to James' 'The Best of James', but from another artist in order to pleasantly surprise him. You will use this shopping web site to look for more information when buying this gift for your friend. Please spend enough time so as to make sure that you choose a suitable CD as a gift for your friend's birthday. Use the space below to answer the following questions about your shopping for the gift: 1) Name of the CD you chose: 2) Name of the artist (for the CD you chose): 3) Other Artists you considered (list at least three): 4) Circle the number that indicates the extent to which you agree with the following statement: a) I am familiar with this genre of Music - Alternative Rock: Strongly Disagree 2 Disagree 3 Mildly Disagree 4 Neither Agree Nor Disagree 5 Mildly Agree 6 Agree 7 Strongly Agree b) I am familiar with the music of this artist - James: Strongly Disagree 2 Disagree 3 Mildly Disagree 4 Neither Agree Nor Disagree 5 Mildly Agree 6 Agree 7 Strongly Agree 5) Reasons for your choice of CD: 122 Task - D : Shopping for yourself (5 -10 minutes) Please search for CDs of your favorite artists. You will be shopping for a CD that you would like to buy for yourself. You are familiar with this type of music and have listened to one or more songs by these artists before. Alternately, you may also shop for CDs by an artist that you are not too familiar with. You've probably heard a song by this artist somewhere; and perhaps, your friend mentioned something positive about this artist. You will use this shopping web site to look for music by this artist and also explore other music that might interest you. Use the space below to answer the following questions about the CD you chose for yourself: a) Name of the C D you chose: b) Name of the Artist: Ask the RA for further instructions. Please don't turn to the next page. 123 <U X> 3 3 <u oo CU O 3 • SH 'EH cu O H cu • s 'BH o X! oo -4—> 3 CU O CU 3 O > . 3 O 3 CU -4—» CS oo cS O •4—» T J 3 O & OO CU 5-1 oo cS CU X> 3 O OO*" CU W) cs O H 3 & O 0) X) CU X> CU - 3 3 O O H O tH 3 O 3 O s" o >> B "cd _ 3 > CU 4-* o 3 O -a 3 o • ^H oo <D 3 CD - 3 H 3 CD 1u £ x; <D CD 2H bi CS ' 3 O > X3 CD Is o 3 CD X CD CD X! ^ H OO CD -4—» cS CD O - in 3 CS - 3 CD OO 3 CS W) 3 O X! (50 o 3 i CD tH CD X3 -HH X! o 3 OO C N DI C8 cu t / 3 c « CU £ s-cu X! 3 s CU x: I ! -CU T5 C E cu o K is oo OO < bo T3 y 2 60 VI S < O z u bo TI- z < a =5 bo s s CN Q x> 8 — oo Q a U H PQ fi CZ3 H >> CU S M 2 «! CO en i* c w o 2 J cn CS S M 2 co x> $ WO t-C CJ] 2 J co Q o M i l CN WI9m •n CO r-i X) is ' o SJ o o 'C .•JS CD J. CL. • 3 O -in CN CM CD > CS X3 CD oo 1 - 1 3 <D s -^CD oo '3 1 X) ^3 CD "5 oo 3 11 ^3 •S <D , 3 CD CN CN r-i CD '7 t . > T3 '2 I' X! • C . x: ^ "D ^ 3 c O H '"5 y: CD bp '3-' & . CD O CD O c<~, >n CN cn CN 3 CD CD X) CD X! •4—» •a 3 O 6 « B cn cn O ~ 3 § e .— — xi £ £ 3 S C j- " 3 = -c . 2 V'.ooT 3 — c: cn CN m cn CN •8 '5 '^ 3 > cU oo .a S 8 3 £ 2 CS XS C ~ CD i—I 'oo CM T , cn c 1 vb"'" - . C O - -C 3 v " O .. T J O •x: '-» V-.T3 <= ).i'.tiSt"e3 ' .*"••=! " 3 ^ c s > D H x: !C -'-tl.,. O .CD-T J a CS u H <Z3 CS H < g o g 9 WD WO S-e oc o es cc Q •3D g S W 2 «! cc cc Q r-i in CN "5 •5 y. > 00 > c o H .5 3 o vo cn (N VD c o OH o 1) o £ > „ s-i 22 <u ON in 7 ? • ' y T3 ca £ o v o >n CN v o in CN O C / J & 2 2 « 5 * ^  a •§ C / J rt Oj <*> • Si J= g rt v o >n CN v o . 1) fi "J ''• 3 o a T> -•5 b y ~ i— y CN VO m cn CN VO in CN o X X O to .S -8 rt ^ 13 3 (U .=5 rt > C N llfil m . 'cn, > r- i ...O '-. a o 'E . —• c 7 O r~ V — v ' y. f l ' y. ^ ' > r."0:";'<U VO in CN v o m CN X co 3 <U •s LO CM SECTION-C In the following pages, you may be asked to respond to a set of statements about your recent shopping experience at the web site you visited to complete Tasks C and D. Please circle a number that indicates the extent to which you agree with each statement. The questions do not evaluate you, only your opinion of the web site. As such there are no right or wrong answers. Reminder - What the numbers mean (see page 2): l Strongly Disagree 2 Disagree 3 Mildly Disagree 4 Neither Agree Nor Disagree 5 Mildly Agree 6 Agree 7 Strongly Agree S. No Stronglj Disagree Stronglv Agree s i This web site is likable. IJjjjIP 2 HK Bill 6 HHl 2) This web site failed to keep me involved while I was shopping. 1 2 3 4 5 6 7 •3) This web site enhances my effecli\eness in CD shopping BBS WSSB iiiii fill Bill b HAS 4) I would return to this web site to make purchases i 2 3 4 5 . 6 7 5) This web site created a sense of closeness IRNi i MM Hi jjj§jjj| 6 |NH 6) This web site keeps me totally absorbed in my interactions with it. 1 2 3 4 5 6 7 7) I was completely interested in what I was doing while browsing this web site SHN PHI o HRI 8) This web site had no clue as to what I really wanted. i 2 3 4 5 6 7 1 would be shopping al this web site IBM 2 Bill b 10) This web site is pleasant. i -> 4 5 7 11) There is a sense of sociability in this web HiH -> IBB Hi Hi H IHH 126 S. No Strongly Disagree Si rough-Agree 12) This web site is useful in shopping for CDs 1 2 3 4 5 6 7 13) This web site holds my attention 2 B i l l H i Hn 6 H9R 14) This web site is unfriendly. 1 2 3 4 5 6 7 15) This web site created a sense of distance. HHI f§fj§ IBl fjfjjj 5 H 16) This web site excites my curiosity. i 2 3 4 5 6 7 17) 'fhis web site increases my productivity in shopping for CDs 2 (jfjj pill (Hi H 18) There is a sense of human sensitivity in this web site i 2 3 4 5 6 7 ls>) This web site did not understand my HHI •111 Mil 20) This web site improves my performance in shopping for CDs i 2 3 4 5 6 7 21) I have positive feelings about this web H i Pllll 7 22) I was deeply involved in my interactions while shopping at this web site i 2 3 4 5 6 7 23) This web site understood what I wanted. HHI H "7 24) This web site enables me to shop for CDs faster i 2 3 4 5 6 7 25) 1 felt close to this web site H i K 2 .•) HR HU in #" 26) I dislike this web site i 2 3 4 5 6 7 27) I found this web site to be very detached in its interactions with me HHI 2 H i •Pli (i IMS 28) This web site knows me well. i 2 3 4 5 6 7 29) There is a sense of human warmth in this 2 H i HB 6 30) I felt that this web site was aloof in its interactions with me i 2 3 4 5 6 7 127 S. No Strongly Disiigrw Strongly Agree 31), 1 would seriously contemplate buying from this web siie SON 2 lliifillll Ml US 32) This web site aroused my imagination. i -> 4 5 6 i 33) This web site does no! know my desires fill Bill R9I Bfljij 6 X, 34) This web site understood my goals. i 2 3 4 5 6 7 35) IJjjijiljK I am very likely to buy CDs from this jfipl Blip 5 (•> 36) I found this web site to be very detached. from me i 2 3 4 5 6 7 37) : This web site was very impersonal in its dealings with me. WKKKk jlslif Kill its!! 6 . ? 38) I am likely to make future purchases from this web site i 2 3 4 5 6 7 39) IBliliilil There is a sense of human contact in this -> SHI Bill Itllll 40) This web site is fun. i 2 3 4 5 6 7 41) This web site makes it easier to shop for 2 Bill ml WOHt 42) This web site understood what I was trying to do. i 2 3 4 5 6 7 43) 1 would recommend this web site to my friends and relatives 2 pill J§|j 44) There is a sense of personalness in this web site i 2 3 4 5 6 7 Close the browser (Internet Explorer/Netscape Navigator). T u r n to the next page o n l y after y o u ' v e c o m p l e t e d this sec t ion . 128 Please circle a number that indicates the extent to which you agree with each statement about the web site you visited to complete Tasks C and D. Please do not visit the site again to answer these statements. The questions do not evaluate you, only your opinion of the web site. S.No Features/Services Strongly Strongly _ Disagree Agree 1) This Web Silo sells books 1 2 3 4 5 6 7 2) This web site sells CDs 1 2 3 4 5 6 7 3) This web site provides me with product I 2 3 4 5 6 7 4) This web site provides me with product 1 .2 3 4 5 6 7 reviews by other users 5) This web site allows mc to review-a product 1 2 3 4 5 6 7 6) This web site lets me rate music 1 2 3 4 5 6 7 7) When I visit a product page (t'orcx, Thriller' by Michael Jackson), this web site 1 3 4 5 n 7 recommends products thai might potenlially 8) When I visit a product page (for ex, 'Thriller' by Michael Jackson), this web site shows me other similar items purchased by other users 4 5 6 7 9) This web site recommends products that might I 2 3 4 5 6 7 potentially interest mc 10) This web site allows me to create a wish list of products I I) This web site allows mc enter my personal information (addicss, preferred credit card etc) 12) This web site allows me enter my personal preferences (my favorite music, rating CDs) 13) This web site uses those preferences to recommend relevant products 14) If I visit again, this web site will remember my preferences Thank you for participating in the study. Please do not discuss the study with other friends as they may be participating in the study as well. 129 Appendix - 4: Instructions to Research Assistants for Study 2 Adaptiveness and Virtual Communities: An Experimental Study Using Real-time Filtering of Amazon.com Directives to Research Assistants Materials for the Participant: • Consent Form • Survey: This has the following components: > The first part of the survey (Background Information) > Instructions: Practice and Tasks A and B > Survey (Continued) > Tasks C and D > Rest of the Survey • Participant Payment Sheet Note: The RA should read the instructions for participants carefully to understand what the participants are required to do. This also helps the RAs understand the purpose of the study. If you have any questions about the instruction, get in touch with Nanda Kumar at nanda@commerce.ubc.ca Preparations Just Before the Participant Arrives: • Open Internet Explorer (IE) 6.x. We will use only IE v6.0 or higher in the study to standardize the browser interface. • Make sure that the Address bar is N O T visible (View>ToolBars>Address bar) • Empty Temporary Internet files (Tools>Options>General>Delete Cookies, Delete Files (check delete all offline content)) • Clear History (Tools>Options>General>Clear History) • Clear Forms and Passwords (Tools>Options>Content>Auto completoClear Forms, Clear Passwords). Ensure that all options (web addresses, forms, usernames and passwords on forms) are unchecked in the last screen. • Ensure that privacy is set to default and security is set to medium (Tools>Options>security, privacy). • Ensure that the test web site is working • Bookmarks are in place (Test Web Site - Four Bookmarks for the four conditions, CDNow.com, BN.com, Borders.com and Amazon.ca) 130 After the Participant Arrives: • Welcome the participant • Seat the participant in an area away from the computer (The monitor should be out of view) that he/she will be using. • Ask the participant to read and sign the Consent Form. • Get the participant to fill out Background Survey. Practice: • Stay with the participant for the duration of the practice. • Explain to the participant that he will be looking at two different shopping sites to understand how shopping sites enhance the shopping experience of the visitors. Ask the participants to focus more on the features on the product pages. Give them the following example: if you are interested in a particular CD, what kind of information and features shopping sites generally provide to make the shopping experience more convenient and pleasant. • Take the participant to the PC. Use IE 6.x to open the CDNow.com site (Since, the address bar will be invisible, you will need to click on File>Open and then type in the URL) • Ask the participant to read the instructions and ask him to browse the first web site (CDNow.com). Once the participant spends enough time on the product page and lists the features provided on the web site, review the list with him/her. Ensure that the participant has at least a few of the features provided on the product page (For example, for CDNow, the participant should at least understand that the web site provides artist information, album information, track listings, related artists among others). Please explain to the participant that it's okay if the participant thinks if some or all of the features are unimportant/useless/important/helpful... The purpose of writing the list is to recognize the features provided by the web site irrespective of whether the participant thinks if they are useful or not. • Make the participant repeat these steps for http://www.bn.com • Ensure that participants understand at least some of the minimal features of bn.com web site (In this case, ensure that they see the Consumer reviews section among others). Actual Task: Now the participant is ready to do the practice on the experimental web site and then continue with the tasks. Since, we are filtering the content of Amazon.com real time, not all content is enabled and the site might be a bit slow at times. We will explain this to the participants using the argument outlined below: 131 Open the bookmarked web sites borders.com and Amazon.ca, and let the participants inspect the homepage. Explain to the participants that Amazon.com in the past has collaborated with other companies and its subsidiaries to share parts of its content from the original web site Amazon.com. For example, borders.com is a physical store in U.S. with several branches in various cities in the U.S. Amazon.com helps Borders.com with on-line content and in returns shares revenues from sales. Similarly, the recently launched amazon.ca borrows selected content from Amazon.com to sell few categories of products (Books, Music, Movies). We at UBC are helping Amazon.com evaluate its content and the features it offers specifically in the music category. The participant will be looking at the Music section from Amazon.com and will shop for CDs. He will then be asked questions about his attitudes towards the web site. Since the content is valuable, this content is coming through several layers of security and authentication to UBC's server here. Hence, the site could be a bit slow at times. Please remember that we are interested in the content in this shopping trip and not the speed. The speed is the result of security arrangements and this won't be a problem in the actual site. So, please imagine that speed won't be a problem in actual situations. Now, load the experimental web site (using the direct bookmark to condition - 1) and briefly explain the lay-out of the web site. Briefly explain to the participant that the site is divided into five sections and explain their purpose. Inform the student that two other sections have been disabled to let the student focus on these five sections in the music site. If by chance, the participant finds himself in any other part of Amazon.com (DVD for example), request him to use the back button to get back to the music site. Now, ask the participant to repeat what he/she did at CDnow.com and BN.com (search for Michael Jackson's Thriller) with this web site. Then, write down the features that he noticed at this (experimental - never use this word in front of the participants) site. The RA will check the list to ensure that the participant has noticed most of the features. Again, never use the word personalization or virtual communities in front of the participant. Talk to the subject about the features of the web site using the terminology used by the web site. Now, the participant is ready to shop for CDs as instructed in the Tasks section. The RA may now leave the participant on his/her own. Of course, let the participant know that you would be happy to help if he/she had any questions. Once the participant finishes the tasks A and B. ask him to fill out the survey of 15 items (see survey - Under Column "After Tasks A and B"). — Note: The participant will have two more tasks (Tasks C and D) left to finish. The web site where the participant conducts these tasks will depend on prior assignment by the RA. There are four experimental conditions and the first participant will be assigned to condition - 1 (same as before), the second to the second condition and so on. Of course, the fifth participant will be assigned to condition 1 and the cycle will start again. If the participant is assigned to condition-1, then ask the participant to simply proceed to the next page and finish the tasks (C and D). Then, ask the participant to respond to the same questions in the space under the column "After Tasks C and D". 132 If the participant is assigned to one of the other three conditions, then explain to the participant that he will be shown a different site (in terms of content) by Amazon.com and that he will shop for the remaining two CDs (Tasks C and D) on this web site. Now, ask the participant to search for Michael Jackson's Thriller. Request him to pay attention to the product page and list the features in a separate sheet of paper. Discuss the list with him and then let him proceed to the last two tasks. After these tasks are completed, ask the participant to respond to the same questions in the space under the column "After Tasks C and D". Once this part is completed, the participant completes the rest of the survey. For this part, the participant should base the answers on shopping experience he had while completing tasks C and D. Then pay the participant and thank him for his time. Get his signature on the payment sheet and let him know that the lucky draw for the winners will be made in three months. 133 Appendix 5: Code Used To Fi l ter Amazon.com for Exper imental Condi t ion 1 <!-#include file="incl.asp"—> <!—#include file="incl2.asp"~> <!— #include file="incl3.asp"—> <% Set htobj = Server.CreateObject("ASPHTTP.conn") dim url dim b_url dim strResult strResult ="" ' — change this one. b_url="http://ecom.commerce.ubc.ca/amazon3/pl.asp" htobj .UserAgent = Request.ServerVariables("HTTP_USER_AGENT") htobj. Accept = Request.ServerVariables("HTTP_ACCEPT") htobj .ContentType =Request.ServerVariables("CONTENT_TYPE") htobj.RequestMethod=Request.ServerVariables("REQUEST_METHOD'') 'htobj.Proxy = "192.168.0.57:80" if Request.Form.Count <> 0 then iCount=0 for each objltem in Request.form iCount= iCount + 1 postext = postext + objltem + "=" + Request.form(objltem) if iCount <> Request.Form.Count then postext = postext + chr(38) Next End If htobj.PostData = postext sUrl=Request("ur") 'assign the URL you pass in to htobj htobj. url = sUrl 'get the content of thte htobj and store it in strResult. strResult = htobj. geturl 'if InStr(sUrl,"ilm-redirect") <> 0 Then 'if InStr(sUrl,"obido") = 0 Then if InStr(strResult, "pnm") <> 0 then 134 'startPoint = InStr(slM, "http://www Amazon.com/exec/obidos/ASIN/' ') 'endPoint = InStr(sUrl, "BOOO") 'pl=startPoint 'p2=endPoint 'length = p2 - p i + 1 'sUrl = Mid(sUrl,80,10) 'sUrl=b_url+"?ur="+sUrl response.redirect(sUrl) 'response.write(sUrl) ' varHREFArray = htObj.GetHREFs ' in tHREFArrayLimit = UBound(varHREFArray) -1 ' For I = 0 to intHREFArrayLimit ' Response.Write varHREFArray(I) & "<br>" & V B C r L F ' Next Else ' — F U N T I O N C A L L S T O Delete Content strResult = removeAverageCustomerReviewWithStars(strResult) '"Remove Y o u may also like these items (you might also enjoy) strResult = remove YouMayAlsoLikeTheseltems(strResult) ' removeRateThisItemBoxOnLeft ' Step 1 strResult = removeRateThisItemBoxOnLeft 1 (strResult) ' Step 2 strResult = removeRateThisItemBoxOnLeft2(strResult) ' Step 3 strResult = removeRateThisItemBoxOnLeft3(strResult) 'Step 4, k i l l the Rate this item blue image on top of the box actMatchl5="<img src="+chr(34)+"http://g-images.amazon.com/images/G/01/detail/rate-this-item-eyebrow-musi.gif'+chr(34)+" width=155 height=19 alt="+chr(34)+"Rate This Item"+chr(34)+"xbr>" actResultl5="" 135 strResult=replace(strResult,actMatchl5,actResultl5) '**** Change the link for the top Amazon logo to going to the default page actMatch9="<a href=/exec/obidos/subst/home/redirect.html/ref=nh_music/" actResult9="<a href=http://ecom.commerce.ubc.ca/amazon3/pl.asp?ur=http://www.amazon.com/exec/obidos/tg/ browse/-/465672/ref=m_mh_mn_nf" strResult=replace(strResult,actMatch9,actResult9) i * * * * * * * * * Beginning O F removeilm-redirectLink ********** '— in New & Future Releases '— k i l l those links that wi l l lead to redirect to default pages due to U R L checking 'strResult = removeilmredirectLink(strResult) '— K i l l "Used Price: $11.25" '— In "Hot New & Future Releases" section Do while Instr(strResult, "Used Price</a>:") <> 0 strResult = removeUsedPriceDollars(strResult) Loop ' K i l l See A l l New From $12.50 strResult = removeSeeAHNewFromDollars(strResult) ' K i l l See A l l Used From $12.50 strResult = removeSeeAllUsedFromDollars(strResult) ' — no Buy used actMatch9="buy used" actResult9="" strResult=replace(strResult,actMatch9,actResult9) ' — no Buy Collectable 136 actMatch9="buy collectible" actResult9=" " strResult=replace(strResult,actMatch9,actResult9) '**** Remove Your Gold Box image on upper right corner strResult = remove YourGoldBox(strResult) Remove More Buying Choices strResult = removeMoreBuyingChoices(strResult) strResult = removeMoreBuyingChoicesO(strResult) strResult = removeMoreBuyingChoicesl (strResult) strResult = removeMoreBuyingChoices2(strResult) ' delete ready to buy image actMatchl5="<img src="+chr(34)+"http://g-images.amazon.com/images/G/01/detail/buybox/ready-to-buy-02.gif'+chr(34)+" width=190 height=18 alt="+chr(34)+"Ready to Buy?"+chr(34)+">" actResultl5="" strResult=replace(strResult,actMatchl5,actResultl5) 'delete more choice image actMatchl5="<img src="+chr(34)+"http://g-images. amazon. com/images/G/01/detail/buybox/more-buying-choices-01. gif'+chr(34)+" width= 190 height^ 17 border=0>" actResultl5="" strResult=replace(strResult,actMatchl5,actResultl5) ' delete the little blue corner on the bottom left actMatchl5="<img src="+chr(34)+"http://g-images.amazon.eom/images/G/01/detail/buybox/bl-01 .gif'+chr(34)+" width=6 height=6>" actResultl5=" " strResult=replace(strResult,actMatchl5,actResultl5) ' delete the little blue corner on the bottom right actMatchl5="<img src="+chr(34)+"http://g-images.amazon.com/images/G/0l/detail/buybox/br-01 .gif '+chr(34)+" width=6 height=6>" actResultl5=" " strResult=replace(strResult,actMatchl5,actResultl5) ' change light blue table color to white, so it is transparent 137 actMatchl5="#99CCFF" actResultl5="#FFFFFF" strResult=replace(strResult,actMatchl5,actResultl5) '— nO Get it for less! actMatch9="Get it for less!" actResult9="" strResult=replace(strResult,actMatch9,actResult9) ' D E L E T E THE BUT IT USED IMAGE actMatchl5="<img src="+chr(34)+"http://g-images.amazonxom/images/G/01/detail/preorder-buy-it-used-button-detail-page.gif width=91 height=20 align="+chr(34)+"absmiddle"+chr(34)+" border="+chr(34)+"0"+chr(34)+">" actResultl5=" " strResult=replace(strResult,actMatchl5,actResultl5) ' D E L E T E THE B U Y BOTH NOW IMAGE actMatchl5="<input type=image name="+chr(34)+"submit.add-to-cart-with-promo"+chr(34)+" value="+chr(34)+"Buy Both Now"+chr(34)+" border=0 alt="+chr(34)+"Buy Both Now"+chr(34)+" src=http://g-images.amazonxonVimages/G/01A>uttonsA>uy-both.gif width=122 height=25 >" actResultl5="" strResult=replace(strResult,actMatchl5,actResultl5) strResult = removeRateThisItemLink(strResult) ' No Free Downloads actMatch9="Free Downloads</a>&nbsp;&nbsp;|&nbsp;&nbsp;" actResult9=" </a>" • strResult=replace(strResult,actMatch9,actResult9) 'No Essential CDs actMatch9="Essential CDs</a>&nbsp;&nbsp;|&nbsp;&nbsp;" actResult9="</a> " strResult=replace(strResult,actMatch9,actResult9) '**** Remove See All New strResult = removeSeeAHNew(strResult) ' — Remove So you would like to table rows after the SEARCH strResult = removeSoYouWouldLikeToTableRows(strResult) 138 ' — Remove ListMania ! table rows after the SEARCH strResult = removeListManiaTableRows(strResult) ' — Remove Rate Album You Own Link in the left navi bar strResult = removeRateAlbumYouOwnLink(strResult) ' remove the "So you would like to Box on the bottome left corner of each product page strResult = removeSoYouWouldLikeToBox(strResult) ' remove the "You may also like box on the bottom left corner of the search result page strResult = remove YouMayAlsoLikeTableRows(strResult) ' This function remove "Our Customers' Advice" section in product page. This shows only 'in certain CDs ( such as "Be Not Nobody", http://www.amazon.com/exec/obidos/ASIN/B0000646TK/) ' For A L L Conditions strResult = RemoveOurCustomerAdvice(strResult) ' — remove What's Your Advice? strResult = RemoveWhatlsYourAdvice(strResult) ' — Remove the Browse the Music in .... section strResult = RemoveBrowseMusicIn(strResult) ' — Remove the popular in and all its links in the upper part of product page strResult = removePopularlnLinks(strResult) "******* Function to remove View online books link Under the top Nav Bar '**** Two functions : Top and Bottom !! strResult = removeViewOnlineBookLinkTop(strResult) strResult = removeViewOnlineBookLinkBottom(strResult) '—Remove add to shopping cart blue box table Step 1,2 strResult = removeAddToShoppingCartl (strResult) strResult = removeAddToShoppingCart2(strResult) 139 '-- Helper function 1,2,3 for above '— Helper 1 strResult = removeAddToShoppingCartHelperl (strResult) '-- Helper 2 'strResult = removeAddToShoppingCartHelper2(strResult) '-- Helper 3 - to kill the Guaranteed picture 'actMatchl5="<img src=''+chr(34)+"http://g-images.amazon.com/images/G/01/detail/shopping-with-us-is-safe.gif'+chr(34)+" width=150 height=25 align=center border=0>" 'actResultl5=" " 'strResult=replace(strResult,actMatchl5,actResultl5) '-Remove the Buy and Sell Used Item Table Step 1,2 strResult = removeBuyAndSellUsedltemTablel (strResult) strResult = removeBuyAndSellUsedItemTable2(strResult) '—Remove the Add to Wish list and Wedding Registry Table Step 1,2 strResult = removeWeddingRegistryl (strResult) strResult = remove WeddingRegistry2(strResult) ****** R e r n ove editorial review Link strResult = removeEditorialReviewLink(strResult) '***** Remove customer review Link strResult = removeCustomerReviewLink(strResult) -***** Remove rate this item strResult = removeRateThisItemLink(strResult) >***** Remove See more by this artist strResult = removeSeeMoreByThisArtistText(strResult) '***** Remove All Album by artist Name Link strResult = removeAHAlbumByLink(strResult) ****** Remove Free Downloads by Artist Name LINK strResult = removeFreeSongsDownloadsBy(strResult) ****** Remove Essential Recording By Artist Name LINK strResult = removeEssentialRecordingsByLink(strResult) 140 '***** Remove Customers Also Bought Text (left navi bar) strResult = removeCustomersAlsoBoughtText(strResult) -***** Remove These Albums Link (left navi bar ) strResult = removeTheseAlbumsLink(strResult) t ***** Remove These other items Link (left navi bar) strResult = removeTheseOtherltemsLink(strResult) ' ***** Remove Share Your Thoughts T E X T (left navi bar) strResult = removeShareYourThoughtsText(strResult) ' ***** Remove write a review LINK (left navi bar ) strResult = remove WriteAReviewLink(strResult) ****** Remove Check Purchase Circles LINK (left navi bar ) strResult = removeCheckPurchaseCirclesLink(strResult) ****** R e m 0 v e E-Mail a Friend About this item LINK (left navi bar ) strResult = removeEmailAFriendAboutThisItemLink(strResult) ****** Remove So You Would Like To Link (left navi bar) strResult = removeSoYouWouldLikeToLink(strResult) .**** BEGIN Function Calls to delete text section in MAIN PAGE with <BR> '-- remove E N Y A Link on top strResult = removeEnyaOnTopLink(strResult) '—remove Used Price $9.99 strResult = removeUsedPriceSection(strResult) '—remove Popular in: strResult = removePopularln(strResult) '— remove Li + I have listened to this recording, and I want to review it. strResult = removeLi(strResult) strResult = removeLi 1 (strResult) strResult = removeLi2(strResult) 'strResult = removeLi3 (strResult) 'strResult = removeLi4(strResult) '-- remove International Sites links at the bottom strResult = removelnternationalSites(strResult) '— remove Rate this item to get Personal Recommendations strResult = removeRateThisItemToGetPersonalRecommendationSection(strResult) '-- remove Blue Box Bottom strResult = removeBlueBoxBottom(strResult) '— remove Average Customer Review Section strResult = removeAverageCustomerReviewSection(strResult) '— remove So You would like to section strResult = removeSoYouWouldLikeToSection(strResult) '—remoce Customers who bought this title also bought: strResult = deleteAlsoBought(strResult) strResult = conv_hrals(strResult) '—remoce Customers who SHOPPED for this item also shopped: strResult = deleteWhoShopped(strResult) '— remove Customers who shopped for this item also shopped '—remoce deleteEditorialReviews: strResult = deleteEditorialReviews(strResult) '—remoce deleteCustomerReviews: strResult = deleteCustomerReviews(strResult) ' '-remoce AUctions and zShops: strResult = deleteAuctions(strResult) '— remove Auctions and zShops type 2 strResult = removeAuctionAndZshopsType2(strResult) strResult = remo ve Auction AndZshopsType3 (strResult) '—remoce Listmania strResult = deleteListmania(strResult) '—remoce Free Downloads strResult = deleteFreeDownloads(strResult) '— remove Where is my stuff FORM section strResult = remove WherelsMyStuffSection(strResult) '#*****######******###*****************************^  '— Delete Words in the page '-- Delete the link on top of the page - VIEW CART WISH LIST YOUR ACCOUNT actMatchl4="<img src="+chr(34)+"http://g-images.amazon.com/images/G/01/nav/personalized/cartwish/right-topnav-default-2.gif'+chr(34)+" width=300 height=22 alt="+chr(34)+chr(34)+" USEMAP=#right_top_nav_map border=0>" actResultl4=" " strResult=replace(strResult,actMatchl4,actResultl4) '— Delete Images on the Navigation Bar on TOP********************* >**** Off Image ONLY ******************************************* '— W E L C O M E — actMatchl5="<img src="+chr(34)+"http://g-images.amazon.com/images/G/01/nav/personalized/tabs/welcome-off-whole.gif'+chr(34)+" width=60 height=26 border=0>" actResultl5=" " strResult=replace(strResult,actMatchl5,actResultl5) '— Your Store — actMatchl5="<img src="+chr(34)+"http://g-images.amazon.com/images/G/01/nav/personalized/tabs/yourstore-unrec-off-sliced.gif'-rchr(34)+" width=47 height=26 border=0>" actResultl5="" strResult=replace(strResult,actMatchl5,actResultl5) 143 '—Books — actMatchl5="<img src="+chr(34)+"http://g-images.amazonxom/images/G/01/nav/personalized/tabs/books-off-sliced.g^ width=39 height=26 border=0>" actResultl5="" strResult=replace(strResult,actMatchl5,actResultl5) '— Electronics — actMatchl5="<img src="+chr(34)+"http://g-images.amazonxom/images/G/01/nav/personalized/tabs/electronics-off-sliced.gif'+chr(34)+ width=74 height=26 border=0>" actResultl5="" strResult=replace(strResult,actMatch 15 ,actResult 15) '— D V D — actMatchl5="<img src="+chr(34)+"http://g-images.amazonxom/images/G/01/nav/personalizeaVtabs/dvd-off-sliced.gif'+chr(34)+" width=35 height=26 border=0>" actResultl5="" strResult=replace(strResult,actMatchl5,actResultl5) '— TOYS — actMatchl5="<img src="+chr(34)+"http://g-images.amazonxorn/images/G/01/nav/personalized/tabs/toys-off-sliced.gif'+chr(34)+" width=47 height=26 border=0>" actResultl5="" strResult=replace(strResult,actMatchl5,actResultl5) '— B A B Y — actMatchl5="<img src="+chr(34)+"http://g-images.amazonxom/images/G/01/nav/personalizeaVtabs/baby-off-sliced.gif'+chr(34)+'' width=40 height=26 border=0>" actResultl5="" strResult=replace(strResult,actMatchl5,actResultl5) '— VIDEO GAMES — actMatchl5="<img src="+chr(34)+"http://g-images.amazonxorr^images/G/01/nav/personalized/tabs/videogames-off-sliced.gif'+chr(34) width=73 height=26 border=0>" actResultl5="" 144 strResult=replace(strResult,actMatch!5,actResultl5) '— TOOLS and GARDENS — actMatchl5="<imgsrc="+chr(34)+"http://g-images.amazonxom/images/G/01/nav/personalized/tabs/hi-off-sliced.gif'+chr(34)+" width=62 height=26 border=0>" actResultl5="" strResult=replace(strResult,actMatchl5,actResultl5) '— Tools and Hardware — actMatchl5="<img src="+chr(34)+"http://g-images.amazon.com/images/G/01/nav/personalized/tabs/hi-off-sliced.gif'+chr(34)+" width=62 height=26 border=0>" actResultl5=" " strResult=replace(strResult,actMatchl5,actResultl5) '— Gifts — 'actMatchl5="<img src="+chr(34)+"http://g-images.amazonxom/images/G/01/nav/personalizeaVtabs/gifts-off-sliced.gif'+chr(34)+" width=45 height=26 border=0>" 'actResultl5=" " 'strResult=replace(strResult,actMatchl5,actResultl5) ' - - H E A L T H and B E A U T Y — actMatchl5="<img src="+chr(34)+"http://g-images.amazonxom/images/G/01/nav/personalized/tabs/drugstore-off-sliced.gif'+chr(34)+" width=50 height=26 border=0>" actResultl5=" " strResult=replace(strResult,actMatchl5,actResultl5) '— IN THEATHER — actMatchl5="<img src="+chr(34)+"http://g-images.amazonxoiri/images/G/01/naWpersonalized/te^ width=56 height=26 border=0>" actResultl5=" " strResult=replace(strResult,actMatchl5,actResultl5) ' — T R A V E L 145 actMatchl5="<img src="+chr(34)+"http://g-images.amazonxom/images/G/01/nav/persondized/tabs/travel-off-sliced.gif'+^  width=45 height=26 border=0>" actResultl5="" strResult=replace(strResult,actMatchl5,actResultl5) '— CARS actMatchl5="<img src="+chr(34)+"http://g-images.amazonxom/images/G/01/naW^ width=35 height=26 border=0>" actResultl5=" " strResult=replace(strResult,actMatchl5,actResult!5) '— OUTDOOR LIVING actMatchl5="<img src="+chr(34)+"http://g-images.amazon.com/images/G/Ol/nav/personalized/tabs/garden-off-sU^ width=59 height=26 border=0>" actResultl5=" " strResult=replace(strResult,actMatch 15 ,actResult 15) '— KITCHEN actMatchl5="<img src="+chr(34)+"http://g-images. amazon xorrVimagesA3/01/nav^ width=70 height=26 border=0>" actResultl5="" strResult=replace(strResult,actMatchl5,actResultl5) '—COMPUTERS—--actMatchl5="<img src="+chr(34)+"http://g-images.amazonxom/images/G/01/nav/personalized/tabs/computers-off-sliced.gif'+chr(34 width=70 height=26 border=0>" actResultl5="" strResult=replace(strResult,actMatchl5,actResultl5) 146 '— CELL PHONE actMatchl5="<img src="+chr(34)+"http://g-images. amazon xonVimages/G/01/na^  width=81 height=26border=0>" actResultl5="" strResult=replace(strResult,actMatchl5,actResultl5) '— VIDEO actMatchl5="<img src="+chr(34)+"http://g-images.amazonxom/images/G/01/n^  width=43 height=26 border=0>" actResultl5="" strResult=replace(strResult,actMatch!5,actResultl5) '— PHOTO actMatchl5="<img src="+chr(34)+"http://g-images.amazon.com/images/G/01/nav/personalized/tabs/photo-off-sliced.gi width=52 height=26 border=0>" actResultl5=" " strResult=replace(strResult,actMatchl5,actResultl5) '— SOFTWARE — actMatchl5="<img src="+chr(34)+"http://g-images.amazon.corMmages/G/01/naW^ width=58 height=26 border=0>" actResultl5=" " strResult=replace(strResult,actMatch!5,actResultl5) '— SEE MORE STORE — 147 1 actMatchl5="<img src="+chr(34)+"http://g-images.amazonxom/images/G/01/nav/personalizeoVtabs/see-more-off-slice width=70 height=26 border=0>" actResultl5="" strResult=replace(strResult,actMatchl5,actResultl5) '— Magazine Subscription — actMatchl5=Vimgsrc="4<:lHt34)V'http://g-images. amazon. com/images/G/01/nav/personalized/tabs/magazines-off-sliced.gif'+chr(34)+" width=78 height=26 border=0>" actResultl5="" strResult=replace(strResult,actMatch 15 ,actResult 15) '— Corporate Account — actMatchl5="<img src="+chr(34)+"http://g-images.amazon.coni/images/G/01/nav/personalized/tabs/corporate-off-sliced.gif'+chr(34)+'' width=70 height=26 border=0>" actResultl5="" strResult=replace(strResult,actMatchl5,actResultl5) '— Store Directory — actMatchl5="<img src="+chr(34)+"http://g-images.amazon.com/images/G/01/nav/personalized/tabs/homegarden-off-sliced.gif+chr(34)+" width=64 height=26 border=0>" actResultl5=" " strResult=replace(strResult,actMatchl5,actResultl5) '<img src="http://g-images.amazon.com/images/G/01/nav/personalized/tabs/homegarden-off-sliced.gif width=60 height=26 border=0> '— Home and Garden — actMatchl5="<img src="+chr(34)+"http://g-images.amazon.coiri/images/G/01/nav/personalizeaVtabs/homegarden-off-sliced.gif'+chr(34)+'' width=60 height=26 border=0>" actResultl5="" strResult=replace(strResult,actMatchl5,actResultl5) '— Your Store — actMatchl5="<img src="+chr(34)+"http://g-images.amazon.com/images/G/01/nav/personalized/tabs/yourstore-unrec-off-dropdown.gif'+chr(34)+" width=47 height=26 border=0>" 148 actResultl5="" strResult=replace(strResult,actMatch 15 ,actResult 15) '— Office Product— actMatchl5="<img src="+chr(34)+"http://g-images.amazon.conVimages/G/01/nav/personalized/tabs/office-products-off-sliced.gif'+chr(34)+" width=59 height=26 border=0>" actResultl5="" strResult=replace(strResult,actMatch 15 ,actResult 15) '— Apparel Top— actMatchl5="<img src="+chr(34)+"http://g-images.amazonxoni/images/G/01/apparel/coat_tab2_t.gif'+chr(34)+" width=70 height=34 border=0>" actResultl5="" strResult=replace(strResult,actMatchl5,actResultl5), '— Apparel Bottom— actMatchl5="<img src="+chr(34)+"http://g-images.amazon.com/images/G/01/apparel/coat_tab2_b-gif"+chr(34)+" width=70 height=26 border=0>" actResultl5=" " strResult=replace(strResult,actMatch 15 ,actResult 15) '— Replace Amazon Logo with Amazon Music Logo at very T O P actMatch3 l="<img src="+chr(34)+"http://g-images.amazon.com/images/G/01/associates/navbar2000/logo-no-border(l).gif'+chr(34)+" width=148 height=43 alt="+chr(34)+""+chr(34)+" USEMAP=#logo_top_nav_map border=0>" actResult31="<img src="+chr(34)+"http://g-images.amazon.com/images/G/01/v9/icons/music-new.gif '+chr(34)+" width=250 height=36 border=Oxbrxbr>" strResult=replace(strResult,actMatch31 ,actResult31) ' - Delete Image Map for F R E E D O W N L O A D S '— V I actMatch9="coords="+chr(34)+"367,0,453,27"+chr(34)+"" actResult9=" " strResult=replace(strResult,actMatch9,actResult9) 1 4 9 '- V2 actMatch9="coords="+chr(34)+"384,0,478,27''+chr(34)+MM actResult9=" " strResult=replace(strResult,actMatch9,actResult9) '- Delete Image Map for ESSENTIAL CDS actMatch9="coords="+chr(34)+"455,0,527,27"+chr(34)+"" actResult9="" strResult=replace(strResult,actMatch9,actResult9) '-- Delete Image Map for Gift Ideas actMatch9="coords="+chr(34)+"479,0,532,27"+chr(34)+"" actResult9=" " strResult=replace(strResult,actMatch9,actResult9) '-- Delete Image Map for USED CDS '- VI actMatch9=Mcoords=''+chr(34)+"529,0,590,27''+chr(34)+'n' actResult9="" strResult=replace(strResult,actMatch9,actResult9) '- V2 actMatch9="coords="+chr(34)+"532,0,588,27"+chr(34)+"" actResult9="" strResult=replace(strResult,actMatch9,actResult9) '-- Kill href for the hyperLink actMatch9=''href=http://www.amazonxom/exec/obidos/tg/browse/-/539444/ref=m_mh_mn_um'' actResult9="" strResult=replace(strResult,actMatch9,actResult9) '-- Delete Add to Wish List Orange icon at various places actMatchl5="<input type="+chr(34)+"image"+chr(34)+" value="+chr(34)+"Add to Wish List"+chr(34)+" name=''+chr(34)+"submit.add-to-registry.wishlist''+chr(34)+'' src=http://g-images.amazon.com/images/G/01/buttons/add-to-wl-yellow.gif border="+chr(34)+"0"+chr(34)+" height=20 width=113>" actResultl5="" strResult=replace(strResult,actMatch 15 ,actResult 15) '— Delete Add to Cart Orange icon at various places actMatchl5="<input type="+chr(34)+"image"+chr(34)+" value="+chr(34)+"Add to Shopping Cart"+chr(34)+" name="+chr(34)+"submit.add-to-cart"+chr(34)+" src=http://g-images.amazon.com/images/G/01/buttons/add-to-cart-yellow-short.gif border=0 width=l 13 height=20>" actResultl5="" strResult=replace(strResult,actMatchl5,actResultl5) '— Delete Preorder this item Orange icon at various places 150 actMatchl5="<input type=''+chr(34)+"image''+chr(34)+" value="+chr(34)+"Pre-order your copy now!"+chr(34)+" name=''+chr(34)+''subrm .^add-to-cart''+chr(34)+" src=http://g-images.amazon.corn/images/G/01/buttons/pre-order-gold.gif border=0 width=l 13 height=20>" actResultl5="" strResult=replace(strResult,actMatch 15 ,actResult 15) ' NOT USED ' — REPLACING THE TOP B L U E NAVI BAR WITH T E X T LINKS - — 'music-nav-default.gif 'actMatchl5="<img src="+clir(34)+',http://g-images.am 01/music-nav-default.gif '+chr(34)+" width=590 height=28 usemap="+chr(34)+"#music_nav_map"+chr(34)+" border=0>" 'actResultl5="<pxcenter><bxa href="+cru(34)+''http://www.amazonxom/exec/obidos/tg^ Home</a>&nbsp;&nbsp;</b> | <bxa href="+chr(34)+"http://www.amazonxom/exec/obidos/tg/stores/static/-/music/search"+chr(34)+">Music Search</a>&nbsp;&nbsp;</b>|<b>&nbsp;&nbsp;<a href="+clir(34)+''http://www.amazonxonVexec/obidos/tg/browse/730 Styles</a>&nbsp;&nbsp;</b>|<b>&nbsp;&nbsp;<a href="+chr(34)+"http://www. amazon.com/exec/obidos/tg/browse/-/573448"+chr(34)+">Classical</a>&nbsp;&nbsp;</b>|<b>&nbsp;&nbsp;<a luef=''+cru-(34)+,'http://www.amazon.com/exec/obidos/tg/browse/-/538588''+cru^ Sellers</a>&nbsp;&nbsp;</b>|<b>&nbsp;&nbsp;<a href="+crir(34)+"http://www.amazon.corri/exec/obidos/tg/browse/-/465672''+cr^^ & Future Releases</ax/bxbrxhr>" 'strResult=replace(strResult,actMatchl5,actResultl5) 'music-nav-search.gif 'actMatchl5="<img src="+chr(34)+"http://g-images.amazon.com/images/G/01/nav/music/06-12-01/music-nav-search.gif-+chr(34)+" width=590 height=28 usemap="+chr(34)+"#music_nav_map"+chr(34)+" border=0>" 'actResultl5="<pxcenterxbxa href="+chr(34)+''http://www.amazon.cor^ Home</a>&nbsp;&nbsp;</b> | <bxa href=''+chr(34)+"http://www.amazon.com/exec/obidos/tg/stores/static/-/music/search"+chr(34)+">Music Search</a>&nbsp;&nbsp;</b>|<b>&nbsp;&nbsp;<a ruef="+cm-(34)+''http://www.amazon.com/exec/obidos/tg^rowse/-/301668''+cru( Styles</a>&nbsp;&nbsp;</b>|<b>&nbsp;&nbsp;<a href="+chr(34)+"http://www.amazon.com/exec/obidos/tg/browse/-/573448"+chr(34)+">Classical</a>&nbsp;&nbsp;</b>|<b>&nbsp;&nbsp;<a href="+chr(34)+"http://www.amazon.com/exec/obidos/tg/browse/-/538588"+chr(34)+">Top Sellers</a>&nbsp; &nbsp ;</b> |<b>&nbsp;&nbsp ;<a 151 href="+crir(34)-i-"http://www.amazonxom/exec/obid & Future Releases</a></bxbrxhr>" 'strResult=replace(strResult,actMatch 15 ,actResult 15) 'music-nav-browse.gif 'actMatchl5="<img src=''+chr(34)+"http://g4mages.amazonxom/images/G/01/nav/music/06-12-01/music-nav-browse.gif'+chr(34)+" width=590 height=28 usemap="+chr(34)-i-''#music_nav_map',+chr(34)+'' border=0>" 'actResultl 5="<pxcenterxbxa href="+chr(34)+"http://www.amazonxom/exec/obidos/tg^rowse/75174"+chr(34)+ Home</a>&nbsp;&nbsp;</b> | <bxa href="+chr(34)+"http://www. amazon.com/exec/obidos/tg/stores/static/-/music/search"+chr(34)+">Music Search«c/a>&nbsp;&nbsp;</b>|<b><&nbsp;&nbsp;<a href="+chr(34)+"http://www.amazon.com/exec/obidos/tg/browse/-/301668"+chr(34)+">Brow Styles</a>&nbsp;&nbsp;</b>|<b>&nbsp;c&nbsp;<a href="+chr(34)+"http://www. amazon.com/exec/obidos/tg/browse/-/573448"+chr(34)+">Classical</a>&nbsp;&nbsp;<^>|<b>&nbsp;&nbsp;<a ru-ef="+chr(34)+"http://www.amazon.conVexec/obidos/tg/browse/-/538588''+chr(34)+''>Top Sellers</a>&nbsp;&nbsp;</b>|<b>&nbsp;&nbsp;<a href="+cru-(34)+"http://www.amazon.com/exec/obidos/tg/browse/-/465672"+chr(34)+ & Future Releases</ax/bxbrxhr>" 'strResult=replace(strResult,actMatch 15,actResultl 5) 'music-nav-topsellers.gif 'actMatch 15="<img src="+chr(34)+"http://g-images.amazon.com/images/G/0 l/nav/music/06-12-01/music-nav-topsellers.gif ,+chr(34)+" width=590 height=28 usemap="+chr(34)+"#music_nav_map"+chr(34)+" border=0>" 'actResult 15="<pxcenterxbxa href="+chr(34)+"http://www.amazon.com/exec/obidos/tg/browse/-/5174''+chr(34)+M>Music Home</a>&nbsp;&nbsp;</b> j <bxa href="+chr(34)+"http://www.amazon.com/exec/obidos/tg/stores/static/-/music/search"+chr(34)+">Music Search</a>&nbsp;&nbsp;</b>|<b>&nbsp;&nbsp;<a lu-ef="+chr(34)+"http://www.amazon.com/exec/obidos/tg/browse/-/301668''+clir(34)+''>B Styles</a>&nbsp;&nbsp;</b>|<b>&nbsp;&ribsp;<a href="+chr(34)+"http://www.amazon.com/exec/obidos/tg/browse/-/573448"+chr(34)+''>Classical<ya>&nbsp;&nbsp;</b>|<b>&nbsp;&nbsp;<a href="+chr(34)+''http://www.amazon.com Sellers</a>&nbsp;&nbsp;</b>|<b>&nbsp;&nbsp;<a href=''+chr(34)+"http://www.amazon.com/exec/obidos/tg/browse/-/465672"+chr(34)+''>New & Future Releases</ax/bxbrxhr>" 'strResult=replace(strResult,actMatchl 5,actResult 15) 'music-nav-releases.gif 'actMatch 15="<img src="+chr(34)+"http://g-images.amazon.com/images/G/01/nav/music/06-12-01/music-nav-releases.gif'+chr(34)+" width=590 height=28 usemap="+chr(34)+"#music_nav_map"-i-chr(34)+" border=0>" 152 'actResultl5="<pxcenterxbxa href="+chr(34)+''http:^ Home</a>&nbsp;&nbsp;</b> | <bxa href="+chr(34)+"http://www. amazon.com/exec/obidos/tg/stores/static/-/music/search"+chr(34)+">Music Search</a>&nbsp;&nbsp;</b>|<b>&nbsp;&nbsp;<a href="+chr(34)+"http://www. amazon. com/exec/obidos/tg/browse/-/301668"+chr(34)+">Browse Styles</a>&nbsp;&nbsp;</b>|<b>&nbsp;&nbsp;<a href="+chr(34)+"http://www. amazon.com/exec/obidos/tg/browse/-/573448"+chr(34)+">Classical</a>&nbsp;&nbsp;</b>|<b>&nbsp;&nbsp;<a href=''+chr(34)+"http://www.amazon.com/exec/obidos/tg/browse/-/538588''+chr(34)+''>Top Sellers</a>&nbsp;&nbsp;</b>|<b>&nbsp;&nbsp;<a href="+chr(34)+"http://www.amazon.com/exec/obidos/tg/browse/-/465672"+chr(34)+">New & Future Releases</ax/bxbrxhr>" 'strResult=replace(strResult,actMatch 15,actResultl 5) 'actMatch3 l="bgcolor=#3333cc" 'actResult31="" 'strResult=replace(strResult,actMatch31 ,actResult31) ' — END OF REPLACING THE TOP B L U E NAVI BAR WITH T E X T LINKS ' — END OF NOT USED '— END OF Delete Images on the Navigation Bar on TOP********************* Delete the view book online images on top navigation bar 'actMatchl5="<img src="+chr(34)+"http://g-images.amazon.com/images/G/01/books/inside/stripe/stripe-top.gif'-i-chr(34)+" width=589 height=22 alt="+chr(34)+""+chr(34)+" botder=0>" 'actResultl5="" 'strResult=replace(strResult,actMatchl5,actResultl5) actMatchl5="<img src="+chr(34)+"http://g-images.amazon.com/images/G/01/books/inside/stripe/stripe-bottom*height=15 alt="+chr(34)+""+chr(34)+" border=0>" actResultl5="" strResult=replace(strResult,actMatchl 5,actResultl 5) actMatch31="See larger picture" 153 actResult31="" strResult=replace(strResult,actMatch31 ,actResult31) actMatch31="Sign in" actResult31="" strResult=replace(strResult,actMatch31 ,actResult31) '-- no Your Favorites actMatch9="Your Favorites" actResult9="" strResult=replace(strResult,actMatch9,actResult9) '** Delete B L U E BOX Content *********************** actMatch 10="(Use if you're redeeming a promotional certificate or coupon.)" actResultl0=" " strResult=replace(strResult,actMatch 10,actResult 10) actMatch 18="(We'll set one up for you)" actResultl8="" strResult=replace(strResult,actMatchl8,actResultl8) Delete Share your thoughts under the B L U E BOX actMatch20="<input type="+chr(34)+"image"+chr(34)+" src=http://g-images.amazon.com/images/G/01/buttons/add-to-wishlist-yellow.gif border="+chr(34)+"0"+chr(34)+" height=20 width=101 name="+chr(34)+"submit.add-to-wishlist"+chr(34)+">" actResult20="" strResult=replace(strResult,actMatch20,actResult20) actMatch21="View my Wish List" actResult21=" " strResult=replace(strResult,actMatch21 ,actResult21) '** END of Delete Share your thoughts under the B L U E BOX 154 " These following codes delete the links at the bottom for various purposes ******** actMatch21="review</a> it." actResult21="</a>" strResult=replace(strResult,actMatch21 ,actResult21) actMatch21="I have listened to this recording, and I want to" actResult21="" strResult=replace(strResult,actMatch21 ,actResult21) actMatch21="E-mail a friend about this item</a>." actResult21="</a>" strResult=replace(strResult,actMatch21 ,actResult21) actMatch21="Write a So You'd Like to... guide</a>." actResult21="</a>" strResult=replace(strResult,actMatch21 ,actResult21) actMatch21="Check purchase circles</a>." actResult21="</a>" strResult=replace(strResult,actMatch21 ,actResult21) actMatch21="Correct</a> errors and omissions in this listing." actResult21="</a>" strResult=replace(strResult,actMatch21,actResult21) actMatch21="<li>Is there a specific product you'd like us to sell?" actResult21="" strResult=replace(strResult,actMatch21 ,actResult21) actMatch21="Tell us</a> about it.</li>" actResult21="" strResult=replace(strResult,actMatch21 ,actResult21) actMatch21="<lixa href=/exec/obidos/tg/stores/detail/" actResult21="<a href=/exec/obidos/tg/stores/detail/" strResult=replace(strResult,actMatch21 ,actResult21) actMatch21="<lixa href=/exec/obidos/guide-create/" actResult21="<a href=/exec/obidos/guide-create/" strResult=replace(strResult,actMatch21 ,actResult21) actMatch21="<lixa href=/exec/obidos/flex-sign-in/" actResult2 l="<a href=/exec/obidos/flex-sign-in/" strResult=replace(strResult,actMatch21 ,actResult21) 155 ' — Beginning of Kill Bottom Text !!! actMatch21="Text Only" actResult21="" strResult=replace(strResult,actMatch21 ,actResult21) actMatch21="Free Downloads</a>&nbsp;&nbsp;</b>|<b>&nbsp;&nbsp;" actResult21=" " strResult=replace(strResult,actMatch21 ,actResult21) actMatch21="Essential CDs</a>&nbsp;&nbsp;</b>|<b>&nbsp;&nbsp;" actResult21=" " strResult=replace(strResult,actMatch21,actResult21) actMatch21="Used Music</a>&nbsp;&nbsp;" actResult21=" " strResult=replace(strResult,actMatch21 ,actResult21) actMatch21="Help</a>&nbsp;&nbsp;|&nbsp;&nbsp;" actResult21="" strResult=replace(strResult,actMatch21 ,actResult21) actMatch21=" Shopping Cart</a>&nbsp;&nbsp;|&nbsp;&nbsp;" actResult21="" strResult=replace(strResult,actMatch21 ,actResult21) actMatch21="Your Account</a>&nbsp;&nbsp;|&nbsp;&nbsp;" actResult21="" strResult=replace(strResult,actMatch21,actResult21) • actMatch21="Sell Items</a>&nbsp;&nbsp;|&nbsp;&nbsp;" actResult21="" strResult=replace(strResult,actMatch21 ,actResult21) actMatch21="l-Click Settings" actResult21=" " strResult=replace(strResult,actMatch21,actResult21) ' — End of K i l l Bottom Text!!! ******* M U S I C H O M E P A G E O N L Y ********** ' Delete welcome message at center on top actMatch21="Hello." actResult21=" " strResult=replace(strResult,actMatch21 ,actResult21) actMatch21="to get" actResult21=" " strResult=replace(strResult,actMatch21,actResult21) actMatch21="personalized recommendations" actResult21="" strResult=replace(strResult,actMatch21 ,actResult21) actMatch21=". New customer?" actResult21="" strResult=replace(strResult,actMatch21 ,actResult21) actMatch21="Start here" actResult21=" " strResult=replace(strResult,actMatch21,actResult21) actMatch21="or find out" actResult21="" strResult=replace(strResult,actMatch21,actResult21) actMatch21="how to order" actResult21="" strResult=replace(strResult,actMatch21 ,actResult21) 'Delete A D D F A V O R I T E S button on H O M E P A G E on the left navigation bar actMatch20="<img src="+chr(34)+"http://g-images.amazon.com/images/G/01/icons/add-favorites.gif'+chr(34)+" width=69 height=15 align=absmiddle border=0 vspace=2>" actResult20="" strResult=replace(strResult,actMatch20,actResult20) 'Delete A D D T O S H O P P I N G C A R T S M A L L B U T T O N on H O M E P A G E on various search result 'Delete A D D F A V O R I T E S button on H O M E P A G E on the left navigation bar 157 actMatch20="<input type="+chr(34)+"image"+chr(34)+" value="+chr(34)+"Add to Shopping Cart"+chr(34)+" name="+chr(34)+"subrnit.add-to-cart,'+chr(34)+" src=http://g-images.amazonxom/images/G/01/buttons/add-to-cart-yellow-short.gif border=0 width=l 13 height=20>" actResult20="" strResult=replace(strResult,actMatch20,actResult20) '********* K E Y W O R D U S E D F O R A B O V E F U N C T I O N S SO R E M O V E D L A S T ! ! '— modifying the links actMatch 1=" action="+chr(34)+'Vexec" actResultl="action="+chr(34)+ b_url +"?ur=http://www.amazon.com/exec" actMatch2=" action=/exec" actResult2="action="+ b_url +"?ur=http://www.amazon.com/exec" strResult=replace(strResult,actMatchl ,actResultl ,1,-1,1) strResult=replace(strResult,actMatch2,actResult2,1,-1,1) edpos=l stpos=l dim ex 1 DEVI arrayHrefs(3000) iCount=0 do while(true) stPos=instr(stpos+l,strResult,"href',vbtextcompare) edPos=instr(stpos+1 ,strResult," ",vbtextcompare) edlPos=instr(stpos+l,strResult,">",vbtextcompare) if(stpos=0)then exit do if(edPos < edlPos) then arrayHrefs(iCount) = mid(strResult, stPos, (edPos-stPos)) else arrayHrefs(iCount) = mid(strResult, stPos, (edlPos-stPos)) end i f iCount=iCount+l loop tempstr="|" For I = 0 to icount newurl= arrayHrefs(I) i f instr(l,newurl,"href=/exec") = 1 then newurl="href=" & b_url & "?ur=http://www.amazon.com" & Mid(newurl,6,7000) elseif instr(l,newurl,"href=" & chr(34)& "/exec") = 1 then 158 newurl="href=" & chr(34)& b_url & "?ur=http://www.amazon.com" & mid(newurl,7,7000) elseif instr(l,newurl,"href=" & chr(34)& "http://") = 1 then newurl="href=" & chr(34)& b_url & "?ur=" & mid(newurl,7,7000) elseif instr(l,newurl,"href=http://") = 1 then newurl="href="& b_url & "?ur=" & mid(newurl,6,7000) elseif instr(l,newurl,"href = /exec") = 1 then newurl="href=" & b_url & "?ur=http://www.amazon.com" & Mid(newurl,6,7000) elseif instr(l,newurl,"HREF=/exec") = 1 then newurl="HREF=" & b_url & "?ur=http://www.amazon.com" & Mid(newurl,6,7000) elseif instr(l,newurl,"HREF=" & chr(34)& "/exec") = 1 then newurl="HREF=" & chr(34)& b_url & "?ur=http://www.amazon.com" & mid(newurl,7,7000) elseif instr(l,newurl,"HREF=" & chr(34)& "http://") = 1 then newurl="HREF=" & chr(34)& b_url & "?ur=" & mid(newurl,7,7000) elseif instr(l,newurl,"HREF=http://") = 1 then newurl="HREF="& b_url & "?ur=" & mid(newurl,6,7000) end if if instr(l,tempstr,arrayHrefs(I))=0 then strResult=replace(strResult,arrayHrefs(I),newurl) tempstr=tempstr & arrayHrefs(I) & "|" Next Response.Write strResult end if %> 159 m • a . j fa fl ! @ 1 1 ' ' * —-a C ; I i i .1 * 53 j ; fig j <3 J m 1 9 * £5 0> W) =-s o r--s E o o c o E-03 i _s 1 ! 1 1 I I ill m vo ® o V s a u V a. i V z 111 1/1 P 2 £ cn fe i ©i -T J 3 •J S a O o o e • 3J o a § > uf>.S U QI 'a o 3 6 * a l l a £ 5 o et "3 w g 0 ap o c a 5 u DI o o p g 3 S ui a S S S c " 1 § « c .5 jo -1 • < < & in! I E C V * • to f %V I i o 1 5 TJ E oi i cn cn cn cn OI c I * 1 e 8 s oi cE 5 i P i, Z S CUOJOJOI j o> ai 10 ul ui UJ UJ to LUUJ _i -J -i as 31 a" ii S & n S i 1 @> I i _© ii c/1 © in 23 • • TJ cu &< a 1 s I a i 0 I c o I S c Q -J _J 1 _ i LS co z : ' a i l al l - S 0 : ni i jo .9 c a i 1 § 01 o 1 > | a 1.? > s 1 i l 5 0 0" i-2 CQ 2 g a, oi eg cu _ 11 o 01 S o l , u 4 ' oi "5 t- u. I IP l l 1 1 a U g %-> a © l fi ^© -3 © U <2 cu 6X CS o fi TJ O a. • • -3 a a I B » J E 5 C •? M A _ J * 114 m I j5 u (fl u Ix, CQ ^ 6 - i M n -t «-! i s rr «^ *o r- od •8 to 3 Q. 13 g 3 3 I a -s S 1 1 J 'a cS 3 < * - E 2 c IB w a P >|I i l l 19 •8 9 "3 1 fl g ..a 1 e 3 .a "3 E--(on •1 a a.' K 3 #g * * J C o U I TT a #© e o U S-i OH •*•> u = TJ O S-i PH • • ON a* TI a i/ a a < si 1 i (ii ill -9! 9 CO £21 I 5 1 3 U" l>4 2 1 Jj -S B Q fi 5 * . l 3 e a to S < >> j* 4 8«l = | 33 3 I 1 a s . 0 a a tm I* PH fi S3 © U i-cS c #© .N 13 fi © SH CU PH CU © T5 fi cu fi< fi. I U ' 11 m 2: cj G a r--1 i i f! \ •gp - 1 1 3 C e re S S it CO 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0099744/manifest

Comment

Related Items