UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

An investigation of virtual product experience and its effect mechanism Jiang, Zhenhui 2004

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

Item Metadata

Download

Media
831-ubc_2004-994317.pdf [ 13.17MB ]
Metadata
JSON: 831-1.0091783.json
JSON-LD: 831-1.0091783-ld.json
RDF/XML (Pretty): 831-1.0091783-rdf.xml
RDF/JSON: 831-1.0091783-rdf.json
Turtle: 831-1.0091783-turtle.txt
N-Triples: 831-1.0091783-rdf-ntriples.txt
Original Record: 831-1.0091783-source.json
Full Text
831-1.0091783-fulltext.txt
Citation
831-1.0091783.ris

Full Text

An Investigation of Virtual Product Experience and its Effect Mechanism by Zhenhui Jiang Bachelor of Engineering, Tsinghua University, 1997 Bachelor of Economics, Tsinghua University, 1997 Master of Management, Tsinghua University, 1999 A THESIS SUBMITTED IN PARTIAL FULFILMENT 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 standards THE UNIVERSITY OF BRITISH COLUMBIA June 16, 2004 © Zhenhui Jiang, 2004 Abstract A significant challenge is posed to online businesses by the inability of customers to directly feel, touch, or sample products sold over the world-wide-web. Virtual product experience (VPE) is designed to alleviate this concern, by providing interactive web interfaces that enable consumers to virtually experience online products via computer mice and keyboards. This study investigates the effects of V P E on web consumers' intentions to revisit particular websites and their intentions to complete purchases on the websites. Specifically, the study analyzes V P E by looking at its two fundamental technological characteristics: interactivity and vividness. Steuer has defined interactivity as "the extent to which users can participate in modifying the form or content of a mediated environment in real time", and defined vividness as "the representational richness of a mediated environment as defined by its formal features; that is, the way in which an environment presents information to the senses". A research model is proposed in this study. It posits that VPE affects consumers' behavioral intentions through the joint effects of vividness and interactivity. Both of these characteristics directly affect perceived diagnosticity, compatibility, and shopping enjoyment. Interactivity alone influences the perceived ease of use of websites, while perceived diagnosticity and compatibility influence perceived usefulness of the websites. Perceived usefulness, perceived ease of use and shopping enjoyment together influence consumers' attitudes toward the websites, which, in turn, affect consumers' intentions to return to the websites. The model also posits that perceived diagnosticity reduces perceived product risk and improves customers' attitudes toward products. In addition, consumers' attitudes toward websites positively affect their attitudes toward products. Perceived product risk and attitudes toward products jointly affect consumers' intentions to purchase goods from particular websites. i i The research model was investigated in a laboratory experiment, by comparing participants' reactions to four different types of web interfaces: static-image, video-without-narration, video-with-narration, and VPE. Two products were presented on the experimental websites: a sports watch and a personal digital assistant (PDA). 176 subjects were recruited from to participate in the study. They were each randomly assigned to one of the four interface conditions and asked to examine the products as if they were shopping online. Experimental data was analyzed using PLS and M A N O V A . Results have largely supported the research model. i i i Table of Contents ABSTRACT II T A B L E OF CONTENTS iv LIST OF TABLES VII LIST OF FIGURES ix ACKNOWLEDGEMENTS x CHAPTER 1 INTRODUCTION 1 CHAPTER 2 BACKGROUND: DIRECT PRODUCT EXPERIENCE AND VIRTUAL PRODUCT EXPERIENCE 4 2.1 An Analog: Direct Product Experience 4 2.2 Virtual Product Experience 6 2.3 Chapter Summary 10 CHAPTER 3 VIRTUAL PRODUCT EXPERIENCES: PRODUCTS AND TECHNOLOGY 12 3.1 Online Product Evaluation 12 3.2 Basic VPE Methods 16 3.3 Typical Types of VPE 18 3.4 Chapter Summary 22 CHAPTER 4 TECHNOLOGICAL CHARACTERISTICS OF VIRTUAL PRODUCT EXPERIENCES: INTERACTIVITY AND VIVIDNESS 24 4.1 Literature Review on Interactivity 24 4.1.1 Effects of Interactivity 25 4.1.2 Approaches to Interactivity Design 28 4.2 VPE Vividness 30 4.2.1 Effects of Vividness 30 4.2.2 Approaches to Vividness Design 33 4.3 Chapter Summary 34 CHAPTER 5 RESEARCH MODEL 35 5.1 Research Model 35 5.2 Description of Variables 36 5.3 Model Development 40 5.4 Chapter Summary 50 CHAPTER 6 RESEARCH METHODS 51 6.1 Experimental Treatments 51 6.2 Experimental Procedures 55 6.3 Questionnaire Design 57 6.4 Chapter Summary 60 CHAPTER 7 EXPERIMENT DATA ANALYSIS 61 7.1 Subject Background Information 61 7.1.1 General Background 61 7.1.2 Subject Backgrounds in Each Treatment Condition 64 7.2 Initial Check of Vividness and Interactivity Measurement 66 7.3 Manipulation Check 68 7.4 Model Testing 69 iv 7.4.1 Measurement Model 70 7.4.2 Structural Model 74 7.5 Follow-up ANOVA Tests 78 7.5.1 Overall MANOVA Results 78 7.5.2 Multiple ANOVA Analysis 79 7.6 Chapter Summary 85 CHAPTER 8 CONCLUSIONS AND DISCUSSION 87 8.1 Findings and Implications 88 8.2 Limitations ". 93 8.3 Future Research 96 APPENDIX 1 HOMEPAGE OF T H E SPORTS W A T C H WEBSITE ( V P E CONDITION) 98 APPENDIX 2 HOMEPAGE OF T H E P D A WEBSITE ( V P E CONDITION) 99 APPENDIX 3 SPORTS W A T C H SIMULATOR (NORMAL MODE) 100 APPENDIX 4 SPORTS W A T C H SIMULATOR (ALARM FUNCTION) 100 APPENDIX 5 P D A SIMULATOR (NORMAL M O D E ) 101 APPENDIX 6 P D A SIMULATOR (DATE BOOK) 101 APPENDIX 7 HOMEPAGE OF T H E SPORTS W A T C H WEBSITE (VIDEO CONDITIONS).... 102 APPENDIX 8 H O M E P A G E OF T H E P D A WEBSITE (VIDEO CONDITIONS) 103 APPENDIX 9 VIDEO DEMONSTRATION OF A L A R M FUNCTION OF T H E SPORTS W A T C H (WITHOUT NARRATION) 104 APPENDIX 10 VIDEO DEMONSTRATION OF A L A R M FUNCTION OF T H E SPORTS W A T C H (WITH NARRATION) 104 APPENDIX 11 VIDEO DEMONSTRATION OF D A T E BOOK FUNCTION OF T H E P D A (WITHOUT NARRATION) 105 APPENDIX 12 VIDEO DEMONSTRATION OF D A T E BOOK FUNCTION OF T H E P D A (WITH NARRATION) 105 APPENDIX 13 HOMEPAGE OF T H E SPORTS W A T C H WEBSITE (MULTIPLE-STATIC-IMAGES CONDITION) 106 APPENDIX 14 HOMEPAGE OF T H E P D A WEBSITE (MULTIPLE-STATIC-IMAGES CONDITION) 107 APPENDIX 15 A L A R M FUNCTION OF T H E SPORTS W A T C H (MULTIPLE-STATIC-IMAGES CONDITION) 108 APPENDIX 16 D A T E B O O K FUNCTION OF T H E P D A (MULTIPLE-STATIC-IMAGES CONDITION) 108 APPENDIX 17 ADVERTISING F O R M 109 APPENDIX 18 CONSENT F O R M 110 APPENDIX 19 SUBJECT BACKGROUND INFORMATION F O R M I l l V APPENDIX 20 GENERAL INFORMATION SHEET FOR SUBJECTS 113 APPENDIX 21 EXPERIMENTAL QUESTIONNAIRE 115 APPENDIX 22 F O R M FOR RESEARCH ASSISTANT 121 REFERENCE 123 vi List of Tables TABLE 1 Product and Attribute Categories and Online Diagnosis 14 TABLE 2 Various Basic VPE methods on Current E-Commerce Websites 17 T A B L E 3 Four Typical Types of VPE 19 TABLE 4 Subjects' Academic Background Faculties/Schools 62 TABLE 5 Internet Experience 62 TABLE 6 Daily Internet Usage 63 TABLE 7 Internet Purchase Experience 63 T A B L E 8 Amounts Spent on Previous Internet Purchases 63 TABLE 9 Gender Composition in Each Treatment Condition 64 TABLE 10 Age Composition of Different Treatment Conditions 64 TABLE 11 Subjects' Comfort with the Internet 65 TABLE 12 Subjects' Familiarity with Internet Shopping 65 TABLE 13 Subjects' Familiarity with Sports Watches 66 TABLE 14 Subjects' Familiarity with PDAs 66 TABLE 15 Factor Analysis of the Eleven Vividness Measurement Items 67 TABLE 16 Factor Analysis of Vividness and Interactivity 68 TABLE 17 Homogeneous Subsets of Interactivity 69 TABLE 18 Homogeneous Subsets of Vividness 69 TABLE 19 Loadings and Cross-Loadings of Measures 71 TABLE 20 Internal Consistency of Measurements 72 TABLE 21 Correlation of Constructs 73 TABLE 22 Hypothesis Tests 76 TABLE 23 R Square Values of Model Constructs 77 TABLE 24 Total Effects of Vividness and Interactivity 78 TABLE 25 Overall MANOVA Test 78 TABLE 26 MANOVA - Tests of Between-Subjects Effects 79 T A B L E 27 Homogeneous Subsets of Perceived Diagnosticity 80 TABLE 28 Homogeneous Subsets of Compatibility 80 TABLE 29 Homogeneous Subsets of Perceived Usefulness 81 vii TABLE 30 Homogeneous Subsets of Perceived Usefulness 81 TABLE 31 Homogeneous Subsets of Shopping Enjoyment 82 TABLE 32 Homogeneous Subsets of Attitudes toward Websites 83 TABLE 33 Homogeneous Subsets of Intentions to Return 83 TABLE 34 Homogeneous Subsets of Perceived Product Risk 84 TABLE 35 Homogeneous Subsets of Attitudes toward Products 84 TABLE 36 Homogeneous Subsets of Intentions to Purchase 84 List of Figures Figure 1 Basic VPE Methods and Their Related Virtual Product Experiences ..20 Figure 2 Research Model: Effect Mechanism of VPE 36 Figure 3 Expected Distribution of Vividness and Interactivity of Treatments Conditions 53 Figure 4 PLS Test of the Research Model 75 ix Acknowledgements I would like to express my sincere gratitude for the contributions my teachers, associates and friends, have made to my research and this thesis. First, I wish to thank my research supervisor, Dr. Izak Benbasat, for his always-energetic assistance and insightful guidance throughout my thesis development. He has eagerly devoted his time and effort to my research, from the creation of my initial ideas to the refinement of this thesis. His rigorous scholarly style and high academic standards have and will continue to deeply influence and benefit my academic career. I also wish to thank the other members of my dissertation committee for their advice to the improvement of the thesis. Dr. Kellogg Booth has contributed significantly to my experiment design. Dr. Paul Chwelos has offered his valuable insight into the model justification. Dr. Peter Darke is a great mentor and a faithful friend; he has always made himself available to me, and his expertise in consumer behavior research has substantially enriched this thesis. I also wish to thank my fellow students for their help with my thesis. In particular, Weiquan Wang, Lingyun Qiu, Dongmin Kim, Lei Zhu, and Sherrie Xiao have provided valuable suggestions and comments for developing a research model, designing experiments, and analyzing data. Lei Zhu has also shown her warm and personal attention to the writing of this thesis. Victor Wong and Kevin Chen have contributed extensive effort into designing the experimental websites. Steve Doak and David Patient have been helpful in editing this thesis. I wish to express my deepest appreciation to my parents and my brother for their persistent love and endless giving. They are forever my moral foundation. Without their consistent support, the completion of this thesis would not have been possible. Chapter 1 Introduction A N INVESTIGATION OF VIRTUAL PRODUCT EXPERIENCE AND ITS E F F E C T MECHANISM Chapter 1 Introduction Businesses operating online and the consumers who visit their sites have been concerned about realism and tangibility, given the consumers' inability to feel, touch, and sample products online as they can do in a physical shopping environment. This lack of direct experience makes customers feel less confident in the performance of the products in which they are interested, and less willing to buy the products on the Internet. For example, as Peterson et al. (1997) have noted, "Internet-based marketing would seem to be a poor substitute for traditional transaction channels, where the goods are available for inspection." Rose et al. (1999) have similarly argued that this limitation has been a major impediment to business-to-consumer electronic commerce (e-commerce). Likewise, Burke (2002) and Alba et al. (1997) have indicated that the quality of product information on the Internet has been mediocre, particularly for consumers who usually rely on physical interaction to evaluate product quality. To alleviate this concern, many online firms have employed Internet-based virtual reality (VR) technologies to allow customers to "feel, touch, and try" products on their websites (Ryan, 2001). For example, using their computer mice and keyboards, customers visiting Olympus' website (www.olympus.com) can rotate cameras three-dimehsionally and view them, from different angles; they can place themselves inside cars and obtain panoramic views of interior settings on Honda's website (www.honda.com); they can operate different functions of sports watches by pressing functional buttons on Timex' website (www.timex.com); and they can customize virtual faces wearing different cosmetics on EZface's website (www.ezface.com). A l l of these online experiences can be categorized as virtual product experiences (VPE), defined as 1 Chapter 1 Introduction web shopping experiences that allow consumers to interact with and to try products via web interfaces. By offering consumers interactive access to sample product information at their discretion, VPE likely affects the quantity and quality of product information that customers acquire online, and their style of interaction with particular websites. Product information and interaction with websites are known to bear particular importance to customers' online shopping. Notably, Jarvenpaa and Todd (1996-97) have suggested that product perception and shopping experiences are two salient factors when consumers form their attitudes towards and intentions regarding online shopping. Similarly, McKinney et al. (2002), who have developed an instrument to measure web customers' satisfaction with web sites, have found that website understandability and entertainment are dimensions of information quality and system quality, respectively, which influence web consumers' satisfaction. Website understandability concerns the quantity and quality of product information that customers can acquire, while website entertainment is related to consumer interaction with websites. Therefore, there is a need to study VPE as an emerging phenomenon when shopping online. This dissertation aims at addressing this need by continuing the research by Jiang and Benbasat (forthcoming). Jiang and Benbasat have argued that V P E technologies, such as visual control and functional control, consist of basic techniques such as direct manipulation and multimedia. They have found that both visual and functional control could increase websites' capability to covey product information and provide consumers with an enjoyable online shopping experience. However, they have not specifically investigated the effects of direct manipulation and multimedia, respectively, and they underlying functional mechanisms of VPE. Therefore, this dissertation is intended to 1) investigate the composition and characteristics of V P E technologies and to categorize different types of VPE; 2) propose a research model from an online firm' stance to illustrate how V P E changes online consumers' beliefs, attitudes and 2 Chapter 1 Introduction behavior, with a focus on the roles of two VPE technological characteristics, i.e. interactivity and vividness; and 3) to test the research model by conducting an experiment involving four interface conditions: static image, video without narration, video with narration, and VPE. This dissertation is organized as follows. Chapter 2 introduces the background of VPE by reviewing literature from prior studies of direct product experiences, the counterpart of VPE in physical shopping experiences, followed by a survey of literature on VPE. Chapter 3 discusses V P E technologies, in particular various basic V P E methods, and how V P E technologies fit different product categories to create different types of experiences. Chapter 4 focuses on the two fundamental technological characteristics of VPE, i.e. interactivity and vividness, and reviews the existing literature about them. Chapter 5, the theoretical core of this dissertation, develops the research model used in the present study, by focusing on interactivity and vividness. The model demonstrates how VPE affects consumers' intentions to return to websites and intention to purchase by identifying several mediating variables, including perceived diagnosticity, compatibility, perceived usefulness, perceived ease of use, shopping enjoyment, attitudes toward websites, attitudes toward products, and perceived product risk. Chapter 6 describes the research method used for this study, including the website design and experimental procedures. Chapter 7 reports data analysis to test the research model. The contributions and limitations of this study and promising future research directions are suggested in Chapter 8. 3 Chapter 2 Background Chapter 2 Background: Direct Product Experience and Virtual Product Experience Virtual product experience (VPE) utilizes virtual reality technology to enable online customers to feel, touch, and try simulations of products through their computers (Ryan, 2001). L i et al. (2001) have contended that VPE resembles direct (in-store) product experience, inasmuch as both are interactive in nature; however, it differs from direct product experience because it generates a sense of presence, i.e., the feeling of being with products, indirectly through a communication medium (Kumar and Benbasat, 2002). We first discuss, in Section 2.1, direct product experience and its benefits and constraints compared to indirect product experience, in the context of physical shopping environments. Following that, in Section 2.2., we review the existing literature on virtual product experience so as to understand how it simulates direct product experience. 2.1 An Analog: Direct Product Experience Direct product experience encompasses a customer's direct feeling, touching, and trial of a product, and is generally believed to be one of the most effective, albeit one of the most expensive, ways to introduce new products (Kotler, 1998). Compared to indirect product experiences, where product information is conveyed to potential customers through some type of medium (e.g. advertising, catalogs, or word of mouth) and-therefore is subject to potential loss and distortion of some details, direct product experiences provide high levels of information fidelity. Prior research has suggested that direct product experiences are beneficial for conveying product information and engaging customers for several reasons (Hoch and Deighton, 1989; Kempf and Smith, 1998). First, direct experiences involve multiple sensory cues, including sight, 4 Chapter 2 Background feel, smell, sound, and taste. Combined together, these cues form a vivid, informative, and impressive presentation of products. Second, while customers are involved in direct experiences, they are generally more motivated to evaluate products; hence their attention is aroused and they are more engaged in examining products than they would be otherwise. Third, when product information can be generated by customers themselves, it is perceived as more accurate and attributed with higher credibility and trustworthiness than when product information is acquired from other sources. Fourth, learning from direct experiences is a self-controlled activity: when customers are able to manage how they receive, store, and process product information, following their personal preferences, product information is learned more effectively. Direct product experience has been studied extensively in contrast with indirect product experience, typically advertising. A prominent example is an Information Response Model that has been proposed by Smith et al. (1982), suggesting that direct experience generates higher information acceptance, and higher order (i.e. stronger) beliefs and effects, than advertising. This model has been supported by numerous empirical studies. For example, Smith et al. (1983; 1988) and Marks et al. (1988) have found in their experiments that attitudes are generally stronger, in terms of extremity and confidence, for subjects exposed to product trials treatment than for those exposed to advertising treatment. Also, their experimental results suggest that consumer behavior is more consistent with consumer attitudes among subjects who try products directly, compared to subjects who are exposed to the products through advertising, demonstrating the greater predictive power of attitudes formed through direct experiences (Fazio, et al., 1990). Other research has suggested that the effects of direct product experience depend on the particular product attributes that are experienced. Two prominent types of attributes have been distinguished by Nelson (1974): search attributes and experience attributes. Search attributes are those whose information is conveyed better through secondhand sources (e.g. advertising, catalogs, or word of mouth), than they are through direct trials of the products. For example, 5 Chapter 2 Background price, vitamin content, and storage capacity belong to this category. Experience attributes, on the other hand, are those that can be evaluated only by direct use of the products. Examples include the taste, feel, and workmanship of the products. Based on this categorization, Wright et al. (1995) have proposed a Media Congruence Hypothesis, which suggests that "the medium that best communicates a type of product information... is most congruent with that type of information" (p. 710). In summary, prior studies have yielded substantial evidence that direct product experience is effective in conveying relevant product information, and therefore it leads to stronger beliefs and attitudes than indirect product experiences (e.g. advertising). In particular, the effects of direct experience are best demonstrated for experience attributes, rather than for search attributes. 2.2 Virtual Product Experience Online shopping provides indirect product experiences, in which web interfaces mediate the relationship between consumers and products. Usually due to the limited representation capability of interfaces, product information is lost or distorted somewhat before it reaches consumers. For example, if a product is presented in static images, then the quantity and quality of information about the three-dimensional appeal of the product might be reduced; similarly, if a customer cannot touch an online product, then the haptic sense is lost. Therefore, the challenge for interface designers is to maintain high levels of information fidelity when product information is transposed into particular interfaces for presentation to consumers. Logically, the ideal interface would be invisible (Norman, 1998). In an invisible interface, little information would be lost or distorted, and customers would be able to simulate a wide range of interactions with products, emulating direct product experiences. 6 Chapter 2 Background In the context of e-commerce website design, an approach to approximate invisibility of interfaces has been suggested through the use of virtual product experience (VPE) technology (Jiang and Benbasat, forthcoming; L i , et al., 2002; 2003; Ryan, 2001). V P E allows consumers to virtually feel, touch, and try online products by manipulating product presentations, trying product functions, and customizing products in online environments. In other words, V P E aims at simulating direct product experiences and obtaining the benefits associated with these experiences. Furthermore, VPE may be more flexible than direct product experiences, because VPE allows customers to access product information that might not be feasible in a physical shopping experience (Wann and Mon-Williams, 1996). For example, on www.viewpoint.com, consumers are able to virtually engrave their names on a ring, which is difficult, if not impossible, in a physical shopping environment. On most websites that use V P E technologies, only some pages of product presentation are displayed with virtual reality, while most other information is organized on regular H T M L pages. Therefore, VPE does not intend to simulate the entire shopping experience (Chittaro and Ranon, 2000; Jeandrain, 2001; Mass and Herzberg, 1999), but only to display products in the manner of virtual reality. In other words, online consumers are able to virtually try products, but not to "walk" virtually around a shopping mall. Several prior research endeavors have focused on examining whether V P E benefits online product demonstrations in the same manner that direct product experiences augment physical product demonstrations. For example, L i et al. (2001) have used verbal protocol analysis to trace experimental subjects' 3D product experiences, finding that their subjects were involved in examining product attributes, were curious about and eager to search for product information, tended to feel high levels of presence, and enjoyed inspecting products in 3D environments. However, this study was quite exploratory, as it did not include a control group in its experimental design. 7 Chapter 2 Background Later, L i and his colleagues conducted a series of experiments to compare 3D product experiences with 2D product experiences (Daugherty, et al., forthcoming; L i , et al., 2002; 2003). They have suggested that the effects of VPE depend on specific product categories. Three product categories relevant to VPE have been identified in their studies: geometric products, referring to those whose attributes can be fully comprehended through vision; material products, referring to those products that can only be fully recognized and assessed through physical contact; and mechanical products, referring to those whose most characteristic features must be demonstrated through product behavior. In general, they found that consumers paid closer attention to and were more engaged in evaluating products in 3D. With respect to product knowledge and attitudes toward specific brands, inconsistent experimental results have been reported in the above studies. L i et al. (2002) have argued that both product knowledge and brand attitudes are significantly heightened by 3D experiences for geometric products and material products. However, L i et al. (2003) and Daugherty et al. (forthcoming) have countered that 3D applications increase product knowledge for geometric products and mechanical products, but not for material products, and that there is no difference in terms of brand attitudes between 3D experiences and 2D experiences for all product categories. Suh and Lee (2004) have categorized product attributes, in the general context of online shopping, into two types. Virtually experiential attributes refer to those that can currently be evaluated through virtual reality; virtually non-experiential attributes are those that are evaluated better through indirect experience (e.g. online information search) or only through physical product trials (e.g. the taste of candy). Based on this categorization, they labeled products that mainly consist of virtually experiential attributes virtually-high-experiential (VHE) products and those that contain mainly virtually-non-experiential attributes virtually-low-experiential (VLE) products. Through an experiment by manipulating interface design (VR or static images) and product types (VHE or VLE) , they have found that, in general, compared to static image 8 Chapter 2 Background interfaces, V R can increase consumers' actual and perceived product knowledge, and can change their attitudes toward products and purchase intentions. In particular, the impact of V R interfaces on consumers' knowledge, attitudes toward products, and purchases will be greater for V H E products than for V L E products. Jiang and Benbasat (forthcoming) have identified two types of VPE technologies: visual control and functional control. Visual control enables online consumers to manipulate product images, e.g. to move, rotate, and magnify a product's image so as to view it from different angles and distances. Functional control, on the other hand, enables consumers to sample different functions of products. For example, consumers can press the buttons of a sports watch to activate simulations of its various functions (e.g. alarm, stopwatch). Jiang and Benbasat have investigated the separate and combined effects of visual and functional control on perceived diagnosticity and online consumers' perceptions of flow. Perceived diagnosticity is defined as the extent to which consumers believe particular shopping experiences are helpful for them to understand and evaluate products. Flow is defined as an affective state that characterizes human-computer interaction as playful and exploratory (Webster, et al., Winter 1993). Jiang and Benbasat have found that visual control and functional control increase the perceived diagnosticity of their corresponding product attributes, and both technologies enhance consumer perceptions of the overall diagnosticity of products and perceptions of flow, although functional control appears to exert a stronger influence. To summarize, these aforementioned studies have examined effects of V P E from various perspectives, including the communication of product knowledge, attitudes toward brands, product evaluation, and online shopping experiences. In general, it appears that VPE can simulate direct product experience and thus facilitate consumer access to product information, although the products should be matched with corresponding V P E technologies. 9 Chapter 2 Background However, previous studies have not illustrated how the effects of VPE on product understanding and evaluation, shopping experiences, or attitudes toward websites and brands are related to each other and can be combined to effectively predict web customers' behavior in response to VPE, e.g. their intentions to return to websites and their intentions to make purchases. Furthermore, prior empirical studies have generally compared V P E interfaces to interface conditions in which consumers can only see plain text and static product images, ignoring the fact that many websites use video clips (multimedia) to present products (Raney, et al., 2003). Recent literature has suggested that video product demonstrations are more effective than static image demonstrations in conveying product information and strengthening customers' beliefs and attitudes (Klein, 2003). Therefore, a better and more comprehensive evaluation of V P E technology would compare it to both image-based product presentations and video-based product presentations, so as to exclude the possibility that VPE may appear effective or ineffective simply due to arbitrary choices of experimental control conditions. A more serious concern arises from the restriction of most previous studies to simply testing the effects of VPE, without investigating the technological characteristics of VPE so as to understand why there are such effects. Jiang and Benbasat (forthcoming) have made a theoretical contribution by identifying the technological components of visual and functional control, namely, direct manipulation and multimedia. However, they have not tested and proved whether these two techniques individually work as predicted. Therefore, more research is warranted to dig deeper into V P E to examine its effect mechanisms. This thesis is a first step in that direction. 2.3 Chapter Summary In this chapter, we have reviewed prior theories and empirical studies on virtual product experience and direct product experience. Their important findings have been summarized and 10 Chapter 2 Background strengths highlighted. Also, we have pointed out the particular areas that have not been investigated thoroughly, and which therefore deserve further attention from researchers. In the next chapter, we will proceed to study the relationships between products and technology, in the context of VPE. 11 Chapter 3 Products and Technology Chapter 3 Virtual Product Experiences: Products and Technology This chapter addresses two fundamental aspects of virtual product experiences — the products and the technology — and the relationship between them. Section 3.1 categorizes different products and product attributes in the context of online shopping. Section 3.2 summarizes various basic technical methods that are suitable for generating virtual product experience. Section 3.3 identifies four types of VPE, and a corresponding product taxonomy. 3.1 Online Product Evaluation In physical, in-store shopping activities, not all products are suitable for direct product experience; only those for which experience attributes are determinative of consumer assessments about them are amenable to this product evaluation method (Nelson, 1974; Wright and Lynch Jr., 1995). Similarly, in online environments, not all products are suitable for V P E presentations. For example, as mentioned in Section 2.2 above, Suh and Lee (2004) have distinguished two types of products that are sold online. Virtually-high-experiential (VHE) products are evaluated primarily by their virtually experiential attributes (characteristics that consumers can currently evaluate through virtual reality (VR)); whereas virtually-low-experiential (VLE) products are primarily determined by their virtually non-experiential attributes (characteristics that can be evaluated through product search or only through direct product experience). This categorization, although indicating a fit between particular products and technology, is too preliminary to cover all product categories, and it does not provide sufficient insight to determine whether particular attributes can be evaluated by V R or not. Thus, more research is needed on this issue. 12 Chapter 3 Products and Technology According to Peterson et al. (1997) and Chaudhury et al. (2001), online products can be divided into digital and physical categories. Digital products are purely based on digital characteristics and can be fully experienced online, e.g. music can be played on the Internet and software can be tried online. However, because the trial of digital products is a real product trial, it does not belong to VPE. Physical products, on the other hand, contain non-digitizable assets, and therefore it may be impossible to evaluate them fully through web interfaces, e.g. when customers cannot feel the workmanship of a watch or feel the fabric of a shirt. Further examination of non-digital products should examine the nature of product attributes, and the extent to which product attributes can be represented online (see Table 1). In general, attributes of physical products can be categorized into search and experience attributes, as defined in Section 2.1 above, regarding direct product experiences (Nelson, 1974). Web interfaces are effective for presenting search attributes because text, hypertext, and static images can be used with technological ease to describe them, e.g. the weight and dimensions of a computer, or the cover of a book. Meanwhile it has traditionally been believed that web interfaces are not suitable for presenting experience attributes (Alba, et al., 1997; Peterson, et al., 1997), because web interfaces do not enable customers to try those attributes directly (Rose, et al., 1999). Fortunately, VPE allows customers to virtually interact with and to try some experience attributes, e.g., to view the appearance of an automobile three-dimensionally, and to activate various functions of a Personal Digital Assistant (PDA). These attributes that can be experienced virtually are called VPE-amenable attributes, while others are categorized as non-VPE-amenable attributes (see Table 2). It is obvious that products that are selected for V P E presentation must contain some VPE-amenable attributes. A simple and temporary method to identify whether an experience attribute is VPE-amenable or not in e-commerce environments involves examining whether the experiential information can be conveyed through the senses of vision and hearing. If customers can 13 Chapter 3 Products and Technology understand an experiential attribute simply by viewing and hearing it, then the attribute is VPE-amenable; otherwise, it is likely non-VPE-amenable, or difficult to represent using current V P E technologies. Several researchers (Reeves and Nass, 2000; Ryan, 2001) have noted that among the five senses (vision, hearing, touch, smell and taste), vision and hearing are most congruent with the current computing environment, i.e. visual and auditory information is most expressible online. Information for the other senses, on the other hand, is difficult to represent technologically and therefore it has not been used, and will not likely be used in the near future in typical e-commerce settings. T A B L E 1 Product and Attribute Categories and Online Diagnosis Product and Attribute Category Product or Attribute Examples Suitable Evaluation Methods Digital products Music, software Real trials on the Internet Non-digital products Desktop computers, apparel, cameras, flowers Depending on the opportunity to try out particular attributes • Search attributes Price, storage capacity, simple appearance Product information search on regular web interfaces, with text, hypertext, and static images • Experience attributes VPE-amenable Complex geometric appearance, functions Virtual Product Experience Non-VPE-amenable Taste, smell Hard to evaluate in current e-commerce environments Two comments should be made concerning the above arguments. First, the current merit of presenting visible and auditory information through web interfaces does not exclude the possibility that other senses may also be represented online. Actually, research has been undertaken to digitize touch, smell, and taste, although these efforts are quite underdeveloped from the point of view of commercialization. For example, the last decade has seen numerous efforts to use mice or some other mechanical or electronic device to transfer the sense of touch 14 Chapter 3 Products and Technology and force (Burdea, et al., 1996; Richard, et al., 1996; Salinas, et al., 2000; Yokokohji, et al., 1999; Yoshikawa and Nagura, 2001). Other efforts have attempted to deliver smell online by first transmitting digitized fragrance data over the Internet, and then generating the corresponding scent at the receiving computer, from a kiosk that contains a given set of chemical ingredients (see www.aromajet.com/comp.htm). Therefore, the use of only vision and hearing to assess the nature of experiential attributes (VPE-amenable or not) is only a temporary, simplified solution, rather than a final or definitive approach. Technological development will eventually transform more and more non-VPE-amenable attributes into VPE-amenable attributes by representing touch, smell and taste online, and by further commercializing these applications. Second, the senses can cross modalities in the virtual world. That is, information that belongs to one sensory modality could be "imaged" by using information from other sensory modalities. Biocca and Kim (2001) have labeled this cross-modality illusion as "intermodal integration." For example, the manipulation of product images by visually zooming, rotating, and shifting could offer consumers a limited haptic sensation, as suggested by L i et al. (2002). Klatzky et al. (Klatzky, et al., 1993) have similarly argued that the visual recognition of an object might "trigger the retrieval of information about its properties stored in memory," thus creating an imagined haptic exploration. Therefore, even if the sense of touch is necessary to understand products in physical shopping situations, it is possible that customers can use their vision and hearing, with their physical contact with computer mice, to generate a virtual sense of touch. Or, in other words, the sensory feeling conveyed through a VPE interface can actually be richer than seemingly allowed by vision or hearing. 15 Chapter 3 Products and Technology 3.2 Basic V P E Methods A survey of current e-commerce websites conducted for this study has indicated that various basic V P E methods have been used to present online products. In general, these methods can be categorized by the sensory modalities that they are related to: vision and hearing (Table 3). Visual simulation aims at demonstrating products' appearance information to customers. Two types of visual simulation methods are identified in the present research: those that portray static characteristics of products and those that portray dynamic characteristics of products. In the first type of simulation, products themselves are static, although consumers may inspect them in a dynamic way. Such basic V P E methods include: 1) zooming, referring to magnification or de-magnification of product images; 2) rotation, by turning products or adjusting the perspective for viewing products; 3) shifting, referring to the movement of product images across a two-dimensional plane; 4) contextualization, referring to the demonstration of products in contexts where they may be consumed or used; and 5) spatial browsing, referring to consumers' inspection of an environment as they "walk" through it. In the second type of visual simulation, products themselves exhibit particular dynamic characteristics. Two basic VPE methods can be used: 1) behavior animation, referring to the use of animation to demonstrate product behavior or functions; and 2) customization, referring to the modification of the forms or contents of products or product contexts. Auditory simulation is used to simulate sound effects in product demonstrations. There are two types of basic VPE methods. Performance sound refers to sounds associated with product performance, e.g. the sound of a watch alarm. Usage sound, on the other hand, refers to sounds associated with the use of products, e.g. the sound emitted when a key is pressed on a PDA. 16 Chapter 3 Products and Technology Table 2 summarizes the basic VPE methods permitted by commercially available technologies in current e-commerce settings. TABLE 2 Various Basic VPE methods on Current E-Commerce Websites Modality Basic VPE Methods Examples Characterizing static aspects of products • Zooming To continuously magnify the image of a watch • Rotation To turn a computer desk • Shifting To move a shelf Visual Simulation • Contextualization To place a sofa in a room • Spatial browsing To walk around in a room and inspect its interior Characterizing dynamic aspects of products • Behavioral animation To try the functions of a PDA • Customization To change the color of a sofa cover Auditory Simulation • Product performance To emit an alarm sound from a watch • Usage sound To emit a mechanical sound when a functional button is pressed virtually It should be noted that Table 2 adopts some V P E methods from L i et al. (2001), who have listed various three-dimensional product simulation methods that are beneficial and illuminating for researchers studying V P E technologies. However, some concerns arise from their work. For example, "social simulation" was listed as a three-dimensional product simulation method, but it is actually related to shoppers' behavior, rather than directly to product behavior; therefore it is excluded from Table 2. Additionally, some other V P E methods, like those for auditory simulation, are missing in L i et al.'s work. 17 Chapter 3 Products and Technology 3.3 Typical Types of V P E L i et al. (2002; 2003) have identified three types of products that are relevant to VPE, as noted above in Section 2.2: geometric products, material products, and mechanical products. They have defined geometric products as those possessing attributes that can be fully understood through vision, material products as those that can only be evaluated by touching and feeling, and mechanical products as those whose performance is best demonstrated through product behavior. The contribution of this taxonomy lies in highlighting the important roles of visual examination and behavior examination (corresponding to geometric products and mechanical products respectively) in the context of VPE. However, this taxonomy is limited by its lack of sufficient accuracy and comprehensiveness. First, L i et al.'s definition of geometric products is too broad. Obviously, not all geometric products need to be examined in VPE, such as those that have a relatively simple or standard appearance, like books and pens. Furthermore, the definition of geometric products does not consider different levels of complexity of geometric appearance. For example, the experience of inspecting a watch is much different than that of inspecting the inside of a house. Towards the same end, the category of material products, although the subject of potentially promising development in the future, is not currently supported by corresponding technology, as acknowledged by L i et al. (2002; 2003). Finally, products that need to be virtually "tried" through customization in specific contexts in which they will be used are not included in L i et al.'s taxonomy. Based on these concerns, we propose the following (non-exclusive) categorization, based on the features that characterize products, in the context of VPE: 1. Products that have a three-dimensional appearance; 2. Products that form an environment, that is, which can be viewed from the "inside"; 18 Chapter 3 Products and Technology 3. Products characterized by behavior or functionality; and 4. Products that need to be evaluated in relation to other products, or in specific or variable contexts. At least four general types of VPE can be identified, corresponding to the four types of products listed above, as feasible guidelines for designers in current e-commerce environments (Table 3). TABLE 3 Four Typical Types of VPE Virtual Product Experience Product Category Product Examples Type 1: Customers visually examine products from different distances and angles Products that have complex three-dimensional appearances Desks, jewelry, and laptop computers Type 2: Customers examine an environment panoramically from inside Products that are large enough to form an environment, and which have relatively complex appearances Rooms, cars, and gardens Type 3: Customers examine a product's behavior or function Products that are associated with behavior or functions Watches, MP3 players, and PDAs Type 4: Customers adjust product attributes or product contexts to their preferences Products that are shown and customized in particular contexts Cosmetics, apparel, and furniture The first type of VPE, corresponding to the visual control technology that has been discussed by Jiang and Benbasat (forthcoming), enables customers to virtually manipulate and visually inspect product images in three dimensions. For example, on www.officedepot.com, consumers can rotate and zoom in towards or out from a desk image, thereby examining the desk from different angles and distances. The corresponding products for this type of VPE should have three-dimensional appearance information that cannot be described by a simple image or a text description, such as desks, jewelry (www.eluxury.com), and laptops (www.sony.com). Research (Jiang and Benbasat, forthcoming; L i , et al., 2003) has found that 3D inspections can provide customers with detailed product appearance information, and that customers' 19 Chapter 3 Products and Technology autonomous manipulation of products and the continuous video product presentation format can engage customers who are interacting with the interface. As demonstrated in Figure 1, zooming, rotation, and shifting operations are the basic VPE methods best suited for this type of VPE. Basic VPE Methods (Extracted From Table 3) Four Typical Types of Virtual Product Experience Visual simulation Customers visually examine products from different distances and angles • Zooming • Rotation A product forms an environment, where customers can enter and examine the environment's interior panoramically • Shifting • Contextualization • Spatial browsing Customers examine a product's behavior or • Behavior animation function • Customization Customers adjust product attributes or product / / — * contexts to their preferences Auditory simulation • Performance sound • Usage sound / Note: the above arrows represent the specific VPE methods used in different types of VPE \ Figure! ? Basic VPE Methods and Their Related Virtual Product Experiences The second type of VPE allows customers to feel that they are situated within a particular virtual environment, and that they can navigate around the environment and examine its interior settings panoramically. Here, appropriate products are those environments that have relatively complex interior appearances. For example, on www.hyatt.com customers can enter a hotel room to view its inside panoramically. Other examples include cars (www.hondacars.com) and real estate (www.circlevision.com/realestate.html). It is expected that user-controlled interior inspections allow customers to collect detailed product information, and that the navigation behavior plus the panoramic view may conjointly generate interesting and compelling shopping 20 Chapter 3 Products and Technology experiences. As displayed in Figure 1, zooming, rotation, and spatial browsing are the basic VPE methods best suited to provide this type of VPE. Although both of the first two types of VPE allow customers to visually examine product appearances in three dimensions, the second type is different from the first. With 3D visual controls, products are manipulated or rotated against static backgrounds, while in panoramic V P E applications, products are large enough to form environments by themselves, where customers feel that they are inside the environment such that they can examine the interior panoramically. The third type of VPE, corresponding to the functional control technology identified by Jiang and Benbasat (forthcoming), allows customers to examine a product's behavior or functionality. For example, on www.timex.com, customers are able to click on a watch's functional buttons and examine its functions, such as stopwatch and alarm features. Products of this category usually contain behavioral or operational cues. Other examples are PDAs (www.palmos.com) and folding calculators (www.circlevision.com/foldingcalc.html). It is expected that live trials of product functions allow customers to understand functionality better, and stimulate customers toward a sense of control and enjoyment. As displayed in Figure 1, behavior animation, performance sound, and usage sound operations are the most amenable basic methods for this type of VPE. The fourth type of V P E allows customers to adjust product attributes or product contexts to their preferences. For example, on www.thomasville.com, customers are able to change the colors of a sofa, to assist them in selecting their preferred colors. Other websites not only allow customers to customize products, but also allow them to customize the contexts where products are to be sampled. For example, www.ezface.com offers customers an option to create virtual faces using their own digital photos, i.e. customers can personalize a virtual face, and then sample different cosmetics on the new virtual face, to give them a better understanding of how 21 Chapter 3 Products and Technology the cosmetics look on their faces rather than on a standard template. Other product examples belonging to this type of V P E include home settings (www.benjaminmoore.com) and clothes (www.landsend.com) (Ives and Piccoli, 2003). Obviously, trials of customizable features provide customers valuable opportunities to understand product-context matches, while engaging customers by enabling them to adjust the appearances of products and contexts according to their preferences. As depicted in Figure 1, contextualization and customization operations are the basic V P E methods best suited to provide this type of VPE. These four types of V P E presentations characterize the four basic interaction relationships between customers and products that are currently supported by commercial Internet-based virtual reality technologies. However, designers should also pay attention to two other issues. First, a VPE can be a hybrid of two or more types of VPEs described above. For example, on www.landsend.com customers are able to try different clothes, e.g. shirts, jackets, and pants, on a virtual model to see whether they fit the model body (i.e. the aforementioned fourth type of VPE); and they can then rotate the model to observe it from different perspectives (i.e. the aforementioned first type of VPE). Second, a particular product can be suitable for more than one type of VPE. For example, on www.honda.com customers are able to rotate a car and view its outside three dimensionally (i.e. the aforementioned first type of VPE), while they can also enter a car and examine its interior panoramically (the aforementioned second type of VPE). 3.4 Chapter Summary In this chapter, we have studied VPE with a focus on its two key elements, namely the products and the technology. We have described the matching relationship between products and V P E technologies. We have also summarized all the basic VPE methods that are used to create 22 Chapter 3 Products and Technology VPE. Finally, we have identified four typical types of V P E that are available on current e-commerce websites for web designers' choice. In the next chapter, we will investigate deeper into VPE technologies by looking at the two fundamental technological characteristics of all V P E methods: interactivity and vividness. 23 Chapter 4 Interactivity and Vividness Chapter 4 Technological Characteristics of Virtual Product Experiences: Interactivity and Vividness The technology that currently supports VPE is web-based non-immersive virtual reality (VR) (Jiang and Benbasat, forthcoming; Suh and Lee, 2004). Steuer (1992) has described virtual reality as "a real or simulated environment in which a perceiver experiences telepresence." Telepresence is defined as "being there" (Heeter, 1992), the experience of presence in an environment by means of a communication medium (Steuer, 1992). Hence, information is not transmitted directly from a sender to a receiver; instead, a mediated environment is created, with which people interact. The two major determinants of telepresence, as noted by Steuer (1992), are interactivity and vividness. Specifically, in a VPE environment, interactivity refers to consumer interaction with products; while vividness represents the representation richness of product demonstrations. In this chapter, interactivity and vividness are analyzed as two fundamental technological characteristics of VPE. Specifically, Sections 4.1 and 4.2 survey existing literature on the effects of interactivity and vividness, respectively, and discuss different methods of creating impressions of interactivity and vividness. 4.1 Literature Review on Interactivity Computer-based interactivity can be viewed from two perspectives: from an interpersonal communication perspective, and from a mechanical perspective (Coyle and Thorson, 2001; Ha and James, 1998; Hoffman and Novak, July 1996). The former focuses on computer-mediated communication between individuals on a social level, e.g. via email communication; while the latter focuses on user interactions with computers on an individual level, such as consumer 24 Chapter 4 Interactivity and Vividness interactions with web interfaces. V P E interactivity belongs to the latter category (Daugherty, et al., forthcoming). Steuer (1992, p. 84) has defined interactivity from a mechanical perspective, as "the extent to which users can participate in modifying the form or content of a mediated environment in real time." Three factors contribute to interactivity: 1) speed, which refers to how rapidly input can be assimilated into a mediated environment; 2) range, which refers to the number of actions that can be implemented at any given time; and 3) mapping, which refers to the ability of system controls to respond in a natural and predictable manner (Steuer, 1992). 4.1.1 Effects of Interactivity Based on the degree of control that users maintain in interactions via their computers, mechanical interactivity can be divided into several types. For example, Thompson and Jorgensen (1989) have proposed three approaches to analyzing interactivity: a reactive model, a proactive model, and an interactive model. In the reactive model, users adopt passive roles, and only react to system-generated stimulus information. In the proactive model, in contrast, users take on active roles; they initiate and dominate the processes and contents of their interactions. The interactive model is a combination of the reactive and proactive model, where technical systems and users construct their interactions jointly, i.e., the users view and respond to the systems, while the systems simultaneously "view" and respond to the users. In a VPE context, a product will not exhibit its behavior until it receives a stimulus from online customers; therefore the interactivity is essentially proactive. In other words, unlike video product demonstrations where customers acquire product information by watching the video and reading any accompanying text, V P E empowers customers to actively control the delivery of product information. 25 Chapter 4 Interactivity and Vividness Kettanurak et al. (2001) have argued that a high level of interactivity provides users with autonomy in determining the material they want to examine and the pace at which they want to proceed. This greater autonomy and flexibility give users a sense of control, and further they lead to more positive user attitudes toward a system and enhance user competence and self-efficacy in gathering the information. Kettanurak et al. (2001) have also contended that synchronous feedback responding to user practices can reinforce what users have already learned, thereby helping users to adjust their behavior, and likely improving their performance and understanding of the system. This in turn can lead to higher user confidence. In an experiment that asked subjects to review instructional material on three systems with different degrees of interactivity (non-, low-, and high-interactivity) and then tested their attitudes toward the system and their performance, Kettanurak et al. (2001) measured user attitudes on six dimensions: user control, content, format, feedback, ease of use, and motivation. Performance, on the other hand, was measured by the subjects' achievements and gains in learning. The results of this experiment indicate that high levels of interactivity positively influence user attitudes (with an exception for the content dimension), and that some dimensions of attitudes (i.e., user control, feedback, motivation, and format), in turn, enhance user performance. Overall, it was found that high interactivity is preferable in information systems because it is interesting, stimulating, fun, and beneficial to the formation of positive attitudes. According to Kinzie (1990), user control is central to the design of interactivity. It allows users to tailor their human-computer experiences to meet their specific needs and interests. Ariely (2000) has suggested that online information control can maximize the fit between heterogeneous and dynamic information needs and the information available, thus increasing consumer abilities to explore and understand the structure of information, as well as their confidence in their abilities. Ariely (2000) conducted two experiments (i.e., Experiments 4 and 5 in the series of experiments reported in their study) to compare people's confidence between a 26 Chapter 4 Interactivity and Vividness High-InfoControl condition and a Low-InfoControl condition. In the High-InfoControl condition, subjects were given complete freedom to choose the sequence and time to view product information; while in the Low-InfoControl condition, subjects viewed the information in a pre-defined sequence. He found subjects had better memory and knowledge of their experimental environment and reported higher confidence in the High-InfoControl condition. A number of other concurring studies have provided substantial empirical evidence regarding the effects of interactivity on users' learning and attitude formation processes. For example, in their experiment, Teo et al. (2003) created websites with three levels of interactivity by providing different functional features, such as FAQ lists, feedback forms, search engines, online guestbook, and online forums, and they tested user reactions to these websites. They found that increased interactivity enhanced website effectiveness and efficiency in delivering relevant information, and that it increased user satisfaction and users' perceptions of the value of particular websites, which in turn contributed to more favorable attitudes toward websites. Summers (1990-91) randomly assigned student subjects to an experimental group (with interactive videodisc-based instructions) and a control group (with a linear video tape-based presentation). He found no statistically significant difference in student achievement or in the time subjects used to complete the study between the two groups. However, a significant preference for the interactive mode was reported. L i et al. (2003) compared online customers' product knowledge, brand attitudes, and purchase decisions between 3D virtual experiences and T V commercials. The key difference between these two media was that the 3D virtual experiences employed user-controlled interactivity, which was absent in the T V commercials. The results of the experiment suggested that there was no difference in terms of product knowledge between the 3D condition and the T V condition. However, customers reported more positive attitudes and purchase intentions with the 3D experiences than with the TV commercials. 27 Chapter 4 Interactivity and Vividness In the context of VPE, interactivity can be illustrated by behavior in simulated product trials. It is not only an experience in which consumers can control the pace of their interaction with the interface and accordingly receive feedback, but also an essential tool, in the context of VPE, for consumers to uncover or "dig out" information "directly" from products. For example, a consumer may not know the functions of a digital watch until he "presses" its buttons and observes its functional reactions. Research on interactive online shopping suggests that if customers are provided with simulated interactive experiences with products, their assessments of the risk involved in purchasing the products will decrease, while their confidence in the product quality will increase. For example, Klein (Klein, 1998) has argued that customers tend to perceive high risks when purchasing products online. Thus, the provision of simulated experiences with products allows consumers to explore product information by themselves, and therefore induces customers to believe that the interface experience is credible. Furthermore, customers may tend to rely more on those attributes that they are able to try or experience. To summarize, despite some inconsistent empirical results, in general, prior research has suggested that a high level of interactivity with computer systems generally increases users' knowledge and understanding of the systems, it enhances users' confidence in their usage behavior and in predicting systems' reactions, and it improves users' attitudes toward the use of the systems. 4.1.2 Approaches to Interactivity Design Typically, V P E interactivity is represented as direct manipulation (Fishkin, et al., 2000; Hutchins, et al., 1986; Shneiderman, 1983). As Wann and Mon-Williams (1996) suggested, a central component of some advanced virtual environment systems is "the ability to interact, and intrinsic to this are the principles of direct manipulation" (p. 835). Jiang and Benbasat 28 Chapter 4 Interactivity and Vividness (forthcoming) also argued that direct manipulation could be embedded in visual and functional control to allow consumers to manipulate online product images, i.e. to control perspectives and distances to examine a product's appearance, and to enable consumers to try a product's various functions by pressing its functional buttons. Direct manipulation allows users to directly control the objects of their interests, and is usually associated with graphic representations. In a direct manipulation interface, "manipulating a representation can have the same effects and the same feel as manipulating the thing being represented" (p. 99, Hutchins, et al., 1986). Numerous studies (e.g. Benbasat and Todd, 1993; Eberts and Bittianda, 1993; Hutchins, et al., 1986; Morgan, et al., 1991) have demonstrated that direct manipulation shortens the psychological distance between users and their computers, because direct manipulation interfaces are easy to learn and to use, and because they can enhance users' engagement in using an information systems. V P E interactivity can also be designed by using non-direct manipulation methods, especially in the context of virtual product customization. For example, a customer can customize a sofa at www.eprevue.net by selecting different colors and materials. There, the common design practice involves customers selecting their desired colors and materials and then confirming their selections with the press of a mouse button. This approach is not direct manipulation in a strict sense, because labeled buttons cannot afford "a continuous representation of the object of interest" (p. 120, Hutchins, et al., 1986), while a standard direct manipulation interface will allow customers to directly drag colors or material over to the sofa. Notwithstanding, these non-direct manipulation methods can also elicit online customers' feelings of interactivity. 29 Chapter 4 Interactivity and Vividness 4.2 V P E Vividness Vividness is "the representational richness of a mediated environment as defined by its formal features; that is, the way in which an environment presents information to the senses" (p. 81, Steuer, 1992). Two important factors contribute to vividness: 1) sensory breadth, which refers to the number of sensory cues, and 2) sensory depth, which refers to the degree of resolution of each sensory channel. In other media theories, these two factors are characterized as "multiple cues" (in the Media Richness theory developed by Daft et al. (1986) and Daft et al. (1987)), and "fidelity" (in the Media Equation theory presented by Reeves et al. (1996)). 4.2.1 Effects of Vividness Nisbett and Ross (1980, p.45) argue that vividness is "likely to attract and hold our attention and to excite the imagination to the extent that it is (a) emotionally interesting, (b) concrete and imagery-provoking, and (c) proximate in a sensory, temporal, or spatial way." The more vivid a presentation is, the richer the information it contains. Specifically, a vivid product presentation exposes consumers to more product information cues, and likely employs more sensory channels than a pallid product presentation. Taylor and Thompson (1982) have examined the effects of vividness on consumers' judgment and decision making. They have suggested that vividness affects consumer judgment only under conditions of differential attention. In other words, if vivid information is of the same salience (i.e., receives the same attention) as pallid information, the vividness effect will not emerge. However, their study has also suggested that vividness and salience are closely related. In most cases vivid information is likely to be more salient (i.e. likely to receive more attention) than pallid information, and therefore it is able to exert a more decisive effect on judgments. 30 Chapter 4 Interactivity and Vividness Kisielius and Sternthal (1984; 1986) have used their availability-valence hypothesis to explain the effect of vividness. Here, availability refers to the ease with which presented information can be accessed from memory, while valence refers to the favorableness of information. According to this hypothesis, more vivid information is more likely to engage people in cognitive elaboration when compared to the same information with a pallid presentation, because vivid information is more interesting and prompts more thorough review and more elaborate encoding processes (Nisbett and Ross, 1980). Thus, in decision making, substantially more pathways would be available to process vivid information, likely affecting people's judgment. Furthermore, whether the greater availability of vivid information actually enhances, undermines, or exerts no effect on persuasion depends on the favorableness of information. If vividly presented information stimulates the elaboration that is favorable to the object of advocacy, it will compel the consumers to engage in greater cognitive elaboration, consequently facilitating the formation of more positive attitudes. However, i f the vividly presented information stimulates elaboration that is unfavorable to the object of advocacy, the vividness will lead to less positive attitudes. Many other studies have tested the availability-valence hypothesis and proved its validity, in addition to Kisielius and Sternthal (1984; 1986). Of particular interest to us are those where vividness is achieved via computer-based presentations. For example, Coyle and Thorson (2001) manipulated different levels of vividness on websites by adding audio and animation to web interfaces with only text, hypertext and images. They found that the increase of vividness led to more positive and enduring attitudes toward websites. L im et al. (2000) have compared the effects of vivid (multimedia-based) presentations to pallid (text-based) presentations in changing people's biases toward other people upon their first impressions of the others. They found that multimedia presentations, but not text-based presentations, could reduce the influence of first impression bias. They offered two explanations 31 Chapter 4 Interactivity and Vividness for this phenomenon. First, the rich media, multimedia, can convey information in a more concrete and less ambiguous way, thus reducing the potential for misinterpretations. Second, multimedia is vivid and contains multiple information cues, thus leading to better retention and retrieval of information. Houston et al. (1995) examined the persuasive power of presentation modes on mock jurors' decision-making in adjudicating the responsibility for a fatal accident. In an experiment, mock juries were presented evidence of an actual commercial jet crash in one of the three modes (a) a computer simulation (high vividness); (b) an audiotape and written transcript (medium vividness); and (c) an individual reading the transcript aloud (low vividness). This experiment revealed no difference in terms of memory of the accident information among the three conditions. However, it indicated that mock juries in the most vivid condition, i.e., the computer simulation condition, rated the responsibility of the airline flight crew significantly lower than those in the reading- aloud- transcript condition, which demonstrated that vividness was able to affect people's attitudinal judgment. To summarize, prior literature has largely supported the assertion that vividness can portray information more concretely, it can attract people's interest and attention, it provokes people's imagination, it can be encoded with little efforts, and it elicits more cognitive elaboration than non-vivid presentation methods. Therefore, people are likely to learn more effectively from vividly presented information than pallid information, and their judgment and attitudes toward the objects that are thus presented are likely to be significantly influenced by vividness. 32 Chapter 4 Interactivity and Vividness 4.2.2 Approaches to Vividness Design Similar to traditional marketing communication strategies, which use a variety of different methods to present products, e.g., advertising by pictorial media such as TV, catalogs, and posters, V P E employs various vivid methods to display products online. Multimedia is often used to create the sense of vividness. Multimedia has been defined by the Cognitive Technology Group at Vanderbilt (1993, p. 118) as the "linkage of text, sound, video, graphics, and the computer in such a way that the user's access to these media becomes non-linear and virtually instantaneous." Lim et al. (2000) have likewise identified two unique characteristics of multimedia, namely rich language and complementary cues. Specifically, they have argued that multimedia can bring together "the symbolic and processing capabilities of various media" and thus create "a richer symbolic system of communication" (p. 118). They have also suggested that different information cues in multimedia (e.g., verbal and nonverbal cues) do not compete with each other for limited cognitive resources, but are complementary to each other and strengthen the overall effects of the information. Prior studies have already demonstrated that multimedia can be used to generate virtual product experiences. For example, Jiang and Benbasat (forthcoming) have argued that multimedia can be used to portray products' visual appeal or behavior in response to online consumers' manipulation. In their study, sports watches were used as experimental products, presented with two interactive features in addition to text and static images. First, watches could be rotated three dimensionally. Second, the watch functions could be activated once consumers pressed the functional buttons displayed. In either case, multimedia was used to represent a watch's reactions through dynamic and smooth display changes and sound cues. Vividness can also be enhanced through user-controlled manipulations of product images, in particular, in virtual product customization. Reeves and Nass (1996, p.237) have argued that changes of visual scenes "allow a visual presentation to simulate real-world experiences that 33 Chapter 4 Interactivity and Vividness would otherwise take considerably longer to unfold and therefore would take considerably longer to benefit from." This method is often used in customizing products, and to represent product responses to customers' input. In an example discussed above, www.eprevue.net allows customers to customize the color and material of a sofa. After a customer chooses a color and a material, a product image with the selected color and material will be displayed. If the customer changes her selection, the image will be changed accordingly. Similarly, on www.landsend.com, images of clothes (e.g., shirts, jackets, and pants) are displayed with different styles and colors. Consumers can select the clothes they like and confirm their choices with a mouse click. Then the selected pieces will be tried on a virtual model. The customer can then judge whether the selected clothes fit the virtual model and match each other (e.g., whether a jacket is coordinated with a pair of pants). 4.3 Chapter Summary In this chapter, we have discussed the two technological characteristics of VPE: interactivity and vividness. Prior literature on interactivity and vividness has been reviewed and different approaches to design interactivity and vividness have been suggested. In the next chapter these theories and research findings will be used as the basis to develop a research models for the present thesis. 34 Chapter 5 Research Model Chapter 5 Research Model Virtual product experience aims at generating web-based non-immersive virtual reality (VR) environments to simulate direct product experience. According to Walsh and Pawlowski (2002) and Bhatt (2004), V R is a very promising technology for presenting products in electronic commerce. However, as Walsh and Pawlowski have further discussed, prior research on V R has predominantly focused on the technology per se, rather than on its social and behavioral impacts. Addressing this issue, the present study will examine whether V P E technologies, compared to other web-based product presentation technologies, can affect online consumers' beliefs, attitudes, behavioral intentions in their shopping activities, and if so, how. A research model is proposed in this chapter, based on interactivity and vividness, the two technological characteristics of VPE as identified in Chapter 4. Specifically, Section 5.1 presents the research model, which focuses primarily on the effects of V P E on consumer intentions to return to particular websites and intentions to purchase products. Section 5.2 describes all variables and Section 5.3 develops the corresponding hypotheses. 5.1 Research Model The research model is illustrated in Figure 2. It posits that V P E affects consumer intentions to revisit particular websites and intention to purchase products through the joint effects of vividness and interactivity. Both vividness and interactivity directly affect perceived diagnosticity, compatibility, and shopping enjoyment, while interactivity alone influences the perceived ease of use Of a website. Perceived diagnosticity and compatibility jointly affect the perceived usefulness of a website. Perceived usefulness, perceived ease of use and shopping 35 Chapter 5 Research Model enjoyment, together influences consumers' attitudes toward websites, which furthers influences consumers' intentions to return to the website. The model also demonstrates that consumers' perceptions of the diagnosticity of products affect their perceptions of risk and their attitudes toward products. In addition, consumers' attitudes toward websites also influence their attitudes toward products. Finally, product risk and attitudes toward products jointly affect consumers' intentions to purchase products. Vividness Interactivity Perceived diagnosticity Compatibility Perceived usefulness Perceived ease of use Shopping enjoyment Perceived risk Attitudes toward products Intention s to purchase ^ Attitudes toward * websites Intentions to return Figure 2 Research Model: Effect Mechanism of VPE 5.2 Description of Variables The model contains two exogenous variables, namely vividness and interactivity, and ten endogenous variables: perceived diagnosticity, compatibility, shopping enjoyment, perceived usefulness, perceived ease of use, attitudes toward websites, intentions to return to websites, perceived product risk, attitudes toward products, and intentions to purchase. Perceived usefulness, perceived ease of use, attitudes, and behavioral intention are established constructs in 36 Chapter 5 Research Model the Technology Acceptance Model (TAM) (Davis, 1989). More specifically, in the present study perceived usefulness refers to the degree to which a particular website is expected to improve online customers' abilities to accomplish their shopping goals. Perceived ease of use refers to the degree to which online customers expect that a particular website requires effort from them to use. Attitude toward websites encompass customers' overall evaluations of websites, while attitudes toward products are consumers' overall evaluations of products. Intention to return refers to the likelihood that online customers will return to a particular website, thus representing the capability of the website to generate potential future sales (Koufaris, 2002). Intentions to purchase assess the likelihood that particular customers will complete purchases when they shop on particular websites, and therefore higher intentions to purchase indicate that the websites are more likely to generate current sales. 5.2.1 Perceived Diagnosticity In their study of consumer trials of products, Kempf and Smith (1998) have used the concept of perceived diagnosticity to represent the extent to which consumers believe that a particular shopping experience is helpful to evaluate products. Perceived diagnosticity can be measured by asking consumers "how helpful would you rate this shopping experience you just had in judging the quality and performance of the products?" Kempf and Smith have found that perceived diagnosticity positively affects the cognitive evaluation of product attributes in product trial, and they have suggested that any research associated with direct product experience should include this construct. The inclusion of perceived diagnosticity is particularly important in the present study, inasmuch as V P E attempts to simulate direct product experience. In the context of e-commerce, perceived diagnosticity reflects consumer perceptions of the ability of a web interface to convey to customers relevant product information, which can assist them in understanding and 37 Chapter 5 Research Model evaluating the quality and performance of products sold online (Jiang and Benbasat, forthcoming). If V P E can increase perceived diagnosticity, then customers will feel that they are more capable of understanding products and better informed regarding their purchase decision making. 5.2.2 Compatibility Moore and Benbasat (1991) have used a construct, compatibility, to study the adoption of IT innovations. Compatibility is defined as "the degree to which an innovation is perceived as being consistent with the existing values, needs, and past experiences of potential adopters" (p. 195). Moore and Benbasat have further suggested that compatibility is very influential in users' adoption of new technology, because people tend to like and to use new technology that coincides with their existing knowledge, habits, and working environments, so they do not need to spend extra effort to adjust themselves or their habits to the technology. Jarvenpaa and Todd (1996-97) have extended the concept of compatibility to online shopping contexts. They have found that the compatibility of an online shopping experience with consumers' shopping habits and product evaluation styles in physical in-store shopping is a key factor associated with the adoption of e-commerce, in particular when online shopping experiences cannot sufficiently enable consumers to evaluate product quality. Therefore, it is likely that online consumers will prefer a V P E application if it corresponds to their habits of trying physical products. Thus, the construct of compatibility has been adopted in the present research model to represent the extent to which consumers believe their online product evaluation experiences are consistent with their existing styles, habits, and past experiences in physical shopping environments. 38 Chapter 5 Research Model 5.2.3 Shopping Enjoyment In physical in-store shopping experiences, consumers' feelings of shopping enjoyment is very important in influencing their behavior (Babin, et al., 1994; Shiv and Fedorikhin, 1999; Smith and Swinyard, 1982). This is especially true for direct product experiences, because consumers are more strongly stimulated through their senses by physical product cues than they are in indirect product experiences. For example, product trials can significantly influence consumers' emotions, which, in turn, influence their overall evaluation of products, as Kempf and Smith (1998) have observed. Similar to the research findings regarding conventional shopping scenarios, prior research in e-commerce has likewise suggested that consumers' shopping enjoyment is a key aspect of their online shopping experiences. For instance, McKinney et al. (2002) have found that the degree to which a website is visually attractive, fun and interesting is perceived as part of the website's system quality, which directly affects consumer satisfaction. Jarvenpaa and Todd (1996-97) have suggested that website playfulness is influential in forming consumers' attitudes. Similarly, Wolfinbarger and Gilly (2001) have argued that "higher playfulness associated with experiential behavior results in a more positive mood, greater shopping satisfaction, and a higher likelihood of impulse purchasing." Overall, these studies have confirmed that shopping enjoyment is important in influencing consumer behavior. 5.2.4 Perceived product risk Perceived risk is defined as consumers' perceptions of uncertainty and adverse consequences when buying products (Dowling and Staelin, 1994). Prior research has found that non-store purchasing (e.g. catalog shopping and telephone shopping) is perceived to be associated with higher risk than in-store shopping, because in non-store purchases consumers 39 Chapter 5 Research Model lose the ability to examine physical products directly and to contact service persons face-to-face (Engel and Blackwell, 1970; Peterson, et al., 1989). This logic also appears to apply in the context of online shopping (Bhatnagar, et al., 2000; Grazioli and Jarvenpaa, 2000; Tan, 1999). For example, Jarvenpaa and Todd (1996-97) have indicated that perceived risk is a potential inhibitor of consumer adoption of e-commerce. They have categorized five types of risk facing online consumers: economic risk, social risk, performance risk, personal risk, and privacy risk. Based on a survey, they have found that performance risk and personal risk are the forms of risk that are most salient to online consumers, among the five risk types. Performance risk refers to "consumers' perceptions that a product or service may fail to meet expectations," particularly when the consumers desire, but have no opportunity, to try a product or service prior to purchasing it. Personal risk "involves the possibility of harm to the consumer resulting from either a product or the shopping process," e.g. when credit card information is disclosed on the Internet. The present research focuses on consumers' perceived risk in product performance, labled as perceived product risk. 5.3 Model Development Prior research has suggested that vivid presentations can portray products more concretely and with more information cues than pallid presentation formats, because vividness involves non-verbal language and multiple sensory channels (Lim, et al., 2000; Nisbett and Ross, 1980). Therefore the more vivid a product presentation is, the richer the product information that is exposed to consumers. On the other hand, because vividness is usually associated with salience (Taylor and Thompson, 1982), vivid product presentations can attract more attention from consumers than less vivid presentations. In other words, consumers are more likely to focus on examining and understanding products that are presented vividly (1984; 1986). 40 Chapter 5 Research Model This logic has also been supported by Lim and Benbasat (2000), who have examined the effects of multimedia on perceived equivocality and perceived usefulness of information systems by comparing multimedia to a less vivid format, namely text-based presentations. Their experimental results suggest the following: 1) for an analyzable task, there is no difference between multimedia and text in terms of the level of perceived equivocality; 2) for a less-analyzable task, multimedia representation can lead to a lower level of perceived equivocality; and 3) multimedia representations may be perceived by users as more useful than text-based representations. Because consumers do not follow predetermined procedures when they are evaluating products, the product evaluation process should be regarded as a less-analyzable task. Therefore, increased vividness of product presentations should lead to a lower level of perceived equivocality, and consequently, relevant product information will be more fully and clearly represented. The overall result will be that consumers are able to understand and evaluate products better in a vivid presentation environment than in a less vivid environment. Therefore, we posit: HI: Vividness of product presentations positively affects consumers' perceived diagnosticity. As mentioned above, vividness can portray information concretely and in detail; therefore it is likely that the more vivid a product presentation is, the more consumers will perceive the presentation as "realistic" and the more their online product evaluation experiences will resemble physical product evaluation experiences. In fact, researchers have already used multimedia to simulate real products. For example, Urban et al. (1997; 1996) have used multimedia-based product presentations to predict new products' market shares before they are introduced into the market. They have found that, inasmuch as multimedia can portray products vividly and "with a 41 Chapter 5 Research Model high degree of realism," it can simulate user experiences and thus lead to better market predictions. Therefore, regarding the effects of vividness on compatibility, defined in this study as one's online product evaluation experience being consistent with his experience in physical stores, we posit: H2: Vividness of product presentations positively affects consumers' perceptions of compatibility. Higher vividness is often associated with more information cues and more sensory channels, and therefore it is generally believed to be more emotionally interesting (Nisbett and Ross, 1980). For example, Miller et al. (1997) have compared three imagery-evoking strategies in radio advertising — sound effects (the highest vividness), vivid verbal messages (a medium level of vividness), and instructions to imagine (the lowest vividness) — in influencing consumers' affective responses. They have found that sound effects exert greater influences in generating positive feelings than vivid verbal messages do, and vivid verbal messages exert greater influences than mere instructions. Chapman et al. (1999) have arrived at similar conclusions, after examining different formats of training systems: a text format, an audio format, and a video format. They have found that users experience the highest levels of engagement in the most vivid medium, i.e. video, and the lowest level of engagement in the least vivid format, i.e. text. Overall, it appears that vividness can generate an optimistic association on a person with the experience that the person is doing. Therefore, we propose: H3: Vividness of product presentations positively affects consumers' shopping enjoyment. 42 Chapter 5 Research Model Prior research on interactivity has suggested that interactivity, in general, can benefit website design by enabling users to acquire information in personalized ways, thus facilitating users' learning processes. First, users are able to examine the information content that they desire to know and can skip the information content that they already know or do not want to learn immediately. In other words, customers are able to match the information content available to their particular information needs (Ariely, 2000). Second, customers are able to control the pace of product examination, i.e. they may spend more time on the information that they are interested in or want to focus on and less time on the information that is of less interest to them (Kettanurak, et al., 2001). Supplementing these aforementioned reasons, we argue that interactivity can enhance users' attention and encourage self-regulated learning. When users interact with information systems, they must pay attention to what they are doing and what the systems' reactions are in order to properly use the systems and proceed with the interaction. Therefore, users are more aroused and motivated to learn when a system requires two-way interactivity, and consequently they are likely to understand information better. Therefore, we posit: H4: Interactivity in product presentations positively affects perceived diagnosticity. If consumers are able to interact with a product and its features, it is natural for them to believe that the online product experience is similar to, or compatible with, their physical product trials. For example, Hutchins et al. (1986) have argued that, in a direct manipulation interface, "manipulating a representation can have the same effects and the same feel as manipulating the thing being represented." Therefore, we posit: 43 Chapter 5 Research Model H5: Interactivity in product presentations positively affects consumers' perceptions of compatibility. As mentioned above, interactivity arouses computer users' attention, because they need to understand how to interact with systems and to interpret and evaluate feedback from the systems, so as to determine the appropriate way to proceed. Inasmuch as users' mental models about the system are continuously updated when they interact with interactive systems, it is likely that their arousal in using the systems will be increased, thereby further enhancing their feelings of engagement with the systems (Raney, et al., 2003). For example, Hutchins (1986) has argued that an important aspect of direct manipulation interfaces is a qualitative feeling of engagement. Raney et al. (2003) have compared web product demonstrations with interactive video advertising against those with action footage and music and those with action footage only. They have found that interactive video advertising significantly increases websites' entertainment value. Chapman (1999) has compared two types of training systems — i.e. an interactive medium, software, and a less interactive medium, video tape. He has found that the interactive computer-based training results in higher engagement than video-tape based training. Hence, we posit: H6: Increased interactivity in product presentations enhances consumers' shopping enjoyment. Because interactivity requires users' participation in interactions with particular systems, different levels of interactivity might influence the amount of effort users must devote to the processes. Higher interactivity, though benefiting users' learning and enhancing their engagement (Kettanurak, et al., 2001), might lead to a lower level of perceived ease of use, because users have to expend more effort to actively participate in interactions with the systems. 44 Chapter 5 Research Model However, if a highly interactive system is designed to match users' mental models, it is also likely that users may feel that their use of the system is effortless. For example, in a direct manipulation interface, a user's thoughts are readily translated into physical action and the system outputs are easily interpreted by the user; thus direct manipulation has been connected to ease of use in numerous studies (e.g., Davis and Bostrom, 1993; Morgan, et al., 1991). Therefore, we posit the following hypothesis without indicating the direction of the effect: H7: Interactivity of product presentations influences the perceived ease of use of websites. When consumers interact with product demonstrations, products react to the consumers' inputs in a dynamic way; therefore consumers may feel that the product demonstrations are lively and vivid. Hence, we posit: H8: Increased interactivity of product presentations enhances vividness of the presentations. In the context of e-commerce, perceived diagnosticity reflects the perceived ability of a web interface to convey to consumers relevant and useful product information that helps them understand and evaluate the quality and performance of products sold online. If a website can increase the perceived diagnosticity of products, then customers feel that they are more capable of understanding products and able to make more informed purchase decisions. Therefore, we posit: H9: Perceived diagnosticity of online products positively affects the perceived usefulness of a website. » 45 Chapter 5 Research Model Inasmuch as a major concern for current online shopping is that consumers cannot feel, touch, and try products as they do in physical shopping environments (Burke, 2002; Peterson, et al., 1997; Rose, et al., 1999), it is obvious that consumers seek websites that permit them to "try" products virtually, so they can use their entrenched styles and habits, which they use in physical shopping environments, to evaluate online products. In other words, a highly "compatible" product experience helps consumers to economize their cognitive efforts (Todd and Benbasat, 1991) when they are learning how to evaluate experience products in online environments. Overall, the higher a website's compatibility is, the more likely it is that customers will believe the website can provide them with "realistic and natural" direct product experiences to evaluate products; correspondingly, they will perceive the website to be more useful. H10: Consumers' perception of compatibility positively affects the perceived usefulness of the website. According to the Theory of Reasoned Action (Fishbein, 1975) and the Technology Acceptance Model (Davis, 1989), the perceived usefulness and the perceived ease of use of a technology can both influence users' attitudes toward the websites. Perceived usefulness partly mediates the effects of perceived ease of use on the attitudes. In summary, we posit the following: H l l : The perceived usefulness of a website positively affects consumers' attitudes toward the website. H12: The perceived ease of use of a website positively affects consumers' attitudes toward the website. H13: The perceived usefulness of a website partly mediates the effects of perceived ease of use of the website on consumers' attitudes toward the website. 46 Chapter 5 Research Model Attitude is a high-level construct that represents people's cognitive as well as affective responses. In the context of online shopping, perceived usefulness and perceived ease of use represent cognitive elements, while shopping enjoyment represents an affective element; therefore, the inclusion of shopping enjoyment as an antecedent of attitudes toward websites in the context of online shopping is justifiable (Koufaris, 2002). Empirically, McKinney et al. (2002) have argued that an enjoyable shopping experience is a significant predictor of web consumers' satisfaction. Similarly, Raney et al. (2003) have proposed a research model indicating that entertainment features improve consumer attitudes toward websites, and thereby increase consumer intentions to return to the site. The model was supported by the data from an experiment with four web conditions — featured film, interactive video, action footage with music, and action footage alone. Therefore, we hypothesize: H14: Shopping enjoyment positively affects consumer attitudes toward a website. Grazioli and Jarvenpaa (2000) have argued that consumers' perceptions of uncertainty increase when they cannot evaluate the quality of products. Therefore, it is likely that if consumers are able to understand and evaluate products better, their perceived product risk will decrease. This assertion has been supported by Dowling and Staelin (1994), who have proposed a process model for perceived risks and information searches. Their model suggests that consumers aim at lowering their perceived risks by conducting information searches. The extent of search behavior is a function of the level of perceived risk, the person' acceptable risk level, the costs and benefits of risk-reduction activities, and the ability of the person to suffer a loss. The search process is generally discontinued after people have searched general product-class information and acquired enough information to reduce the perceived risk associated with a particular product to at least an acceptable risk level. Thus, if consumers report a higher level of 47 Chapter 5 Research Model perceived diagnosticity of products, it means that they have acquired a higher level of product information for evaluating the products, hence their perceived risk will decrease. H15: Perceived diagnosticity of products negatively affects perceived product risk. According to Kempf and Smith (1998), if consumers believe that a product experience is more diagnostic, it is likely that their beliefs toward the products will be stronger and held with more confidence. Prior research has also suggested that consumers' overall evaluations of products are positively associated with the strength of their beliefs and their confidence with their own evaluations of product attributes (Smith, et al., Winter 1998; Smith, 1993; Smith and Swinyard, 1983). It follows that if advocated product information is positive1, which is often the case when online firms promote their products, a higher perceived diagnosticity will strengthen consumers' positive evaluation of products, and further lead to more positive attitudes toward products. H16: Perceived diagnosticity influences consumers' attitudes toward products. We also hypothesize that consumers' attitudes toward product presentations may positively influence their attitudes toward products. This prediction is based on two mechanisms: direct affect transfer and inferential belief formation, which have been proposed by (Kim and Allen, 1996). In their study, Kim et al. have investigated the mediational mechanisms through which unconditioned stimuli (US), such as pictures or visual images, affect consumer attitudes toward conditioned stimuli (CS), such as products. They have argued that, on the one hand, when US provokes a positive or negative affect, the systematic pairing of US and CS causes a transfer of 1 In a pre-test, we asked 12 participants to evaluate different functions of the watch and the PDA used in the study as objectively as they could (based on a 7-point Likert scale). Results have revealed that the individual evaluation of each product function and the overall evaluation of products were all positive. 48 Chapter 5 Research Model affect from US to CS (direct affect transfer); on the other hand, people might infer the performance of CS based on their beliefs about US (inferential belief formation). These two mechanisms have been clarified through two experiments, suggesting that advertisements can shape consumers' brand attitudes by affecting their affective and cognitive responses towards the advertisements (also see (MacKenzie, et al., 1986)). Hence, when online product presentations are used, if consumers form positive attitudes toward websites, the cognitive and affective responses elicited may positively influence their attitudes toward advocated products. We therefore propose: H17: Consumers' attitudes toward websites positively influence their attitudes toward products. Finally, According to the Theory of Reasoned Action (Fishbein, 1975), people's attitudes are among the direct determinants of their behavioral intentions. Therefore, we propose: H18: Consumers' attitudes toward websites positively affect their intentions to return to the websites. H19: Consumers' attitudes toward products positively affect their intentions to purchase the products from a website. However, if consumers perceive substantial risks associated with product performance, it is likely that they will hesitate to purchase the products because consumers are not certain that the products fit their particular needs and preferences (Jarvenpaa, et al., 2000). H20: Perceived product risk negatively affects customers' intentions to purchase. 49 Chapter 5 Research Model 5.4 Chapter Summary In this chapter, we have developed a research model to explain how V P E affects online consumers' beliefs, attitudes, and behavioral intentions. Specifically, interactivity and vividness are driving forces in the model, which focuses on examining the effects of V P E on consumers' intentions to return to particular websites and intentions to purchase. The research methods used to test the model is described in the next chapter. 50 Chapter 6 Research Methods Chapter 6 Research Methods The research models proposed in the present study have been tested through a laboratory experiment involving websites designed with four different product presentation formats: 1) multiple static images, 2) video without narration, 3) video with narration, and 4) virtual product experience (VPE). Section 6.1 introduces the experimental manipulations conducted in the study, i.e. the design and development of experimental websites. Section 6.2 describes the procedures for the experiment, and the design of the questionnaire used in the experiment is reported in Section 6.3. 6.1 Experimental Treatments The four types of web-based product presentation formats employed in this study's laboratory experiment were used as treatments to generate different levels of vividness and interactivity. The multiple-static-images format is a website condition where product information is represented through multiple static images combined with relevant explanatory text or hypertext descriptions. The two video formats, video-without-narration and video-with-narration, are website conditions where product information is represented by continuous video demonstrations, including dynamic visual stimuli such as product rotation, and with corresponding sound effects such as a watch's alarm, if the products are designed to emit sounds. In the V P E format, product information is represented by a VPE presentation, supplemented by text explanations of product features similar to those presented in the multiple-static-images format. In particular, in this experiment functional control was used as a typical VPE technology, as it is becoming an increasingly prevalent standard for websites to present dynamic, functional simulations of products (Jiang and Benbasat, forthcoming). 51 Chapter 6 Research Methods Specifically, the two video formats differ in that the video-without-narration condition uses text descriptions to explain product features dynamically and synchronized with the pace of video demonstrations, while in the video-with-narration condition text explanations are narrated aloud in connection with the video. Prior research and practice have already confirmed the important roles of both video conditions in multimedia design and implementation. However, according to the Dual Processing Theory (Mayer and Moreno, 1998; 2002; Moreno and Mayer, 1999; 2002), video-with-narration is more effective in provoking user memory and learning than video-without-narration, because video-with-narration conveys product explanatory information by employing audio channels, and hence facilitates people's cognitive processing. This is unlike the video-without-narration condition, where explanatory text information competes for consumers' visual processing resources when consumers are simultaneously engaged in examining video product feature demonstrations. A l l four presentation conditions are widely used in current commercial websites, with varying degrees of vividness and interactivity. The two video conditions and V P E condition should increase the level of vividness over the multiple-static-images condition, because the video conditions and the V P E condition employ continuous visual stimuli and sound effects (Coyle and Thorson, 2001). On the other hand, both video conditions should maintain the same level of interactivity as the multiple-static-images condition and this interactivity should be lower than the V P E condition, because only the VPE condition allows consumers to interact with online products. Overall, we expect to see a wide range of vividness and interactivity across these four conditions (see Figure 3). Notably, we do not have an experimental condition with high-interactivity and low-vividness. This is not because it is overlooked by the present study, but simply due to the fact that in the current e-commerce environment there are no corresponding applications available. Recall that we have predicted in Section 5.3 that increased interactivity is 52 Chapter 6 Research Methods likely to lead to a high level of vividness. Therefore, a product demonstration with high interactivity is likely to be associated with high vividness, which also justifies the absence of high-interactivity and low-vividness applications. 1 Interactivity High Virtual Product experience Low Multiple Static Images Video without narration Video with narration • Low High Vividness Figure 3 'Expected Distribution of Vividness and Interactivity of Treatments Conditions In order to increase the applicability of the present research, two products have been used in the experiment: a sports watch (the Timex Rush) and a Palm Pilot (model M515). According to Table 3, sports watches and Palm Pilots are characterized by particular functions or operational behavior, and therefore are suitable for functional control. If the validity of our research models is empirically found to be different for the two products, it means, at least, that we need to consider product type as a potential moderator. If the product difference does not make much difference in terms of the model validity, it gives us more confidence in generalizing the model to a diverse variety of products, with the awareness that products and technologies should be properly matched, as discussed above in Section 3.3. On the homepage of each product in the VPE condition, a line of hypertext connects to the corresponding V P E presentation, for example "Try out the watch with our full simulation" on 53 Chapter 6 Research Methods the watch's homepage, and "m515 simulator" on the PDA's homepage (see Appendixes 1 and 2). When the hypertext is selected, a product simulator is launched, and users can virtually operate different functions of the simulator. For example, a user may press the buttons of the sports watch to set the time, the stopwatch, the timer, or the alarm, simply by clicking on their computer mice. The watch may react according to users' inputs by changing the display or emitting sound (see Appendixes 3 and 4 for examples). A user can also use his mouse in lieu of a stylus to point to different areas of the simulated Palm Pilot screen, to add new contact addresses or appointments, to perform calculations, or to compose email messages. The Palm Pilot reacts as a real product does, i.e., it changes the screen and emits sounds when the "stylus" touches the screen (see Appendixes 5 and 6 for examples). For both the sports watch and the PDA, text explanations are also provided adjacent to the simulators, to guide users about using the different functions of the products. In the video conditions, either with or without narration, almost the same contents and layout are included on the homepage as in the V P E condition, but all product functions are hyperlinked to video files demonstrating particular functions of different functional modes (see Appendixes 7 and 8). A l l of the video files have been screen-captured (using Camtasia 1.0.0 software) when users sample different functions of the watch and the PDA in the V P E condition, therefore presenting the same information contents as the V P E condition, although the video files do not provide any interactivity. That is, users can only watch video product demonstrations, but cannot pause, fast forward or rewind the video files. Specifically, in the video-without-narration condition, text-based feature explanations are presented adjacent to a product demonstration, further explicating what the video is portraying. In other words, text explanations are scrolled and highlighted in unison with the pace of the corresponding video files. This design applies the temporal-contiguity principle for multimedia design proposed by Moreno and Mayer (1999). In the video-with-narration condition, detailed 54 Chapter 6 Research Methods text-based feature explanations are not displayed, but narrated. The narration was previously recorded using a standard North American male voice. Appendixes 9 through 12 display examples of website designs for the two video conditions. The homepages of the multiple-static-images condition are presented in Appendixes 13 and 14. The major difference from the video conditions is that product function hypertexts are linked to corresponding static images displaying the products in different functional modes. Usually, multiple images are needed to demonstrate particular functions fully. Appendixes 15 and 16 provide examples of how the multiple-static-images condition has been designed. Overall, to the best of our efforts, factual product information content has been maintained consistently across the different conditions. The only difference among treatments is in the presentation formats. 6.2 Experimental Procedures 176 subjects were recruited from a North American university campus (see Appendix 17 for advertising) and randomly assigned to the four interface conditions, with 44 in each condition. Because two products were presented in the experiment, the order by which subjects examined products was randomized, such that half of the participants in each treatment examined the sports watch first, while the other half examined the PDA first. In the experiment, subjects first read and signed a consent form (Appendix 18), which stated that each participant, upon their completion of the study, would receive $15 as a reward, and a one-in-four chance to receive a $50 bonus, based on their performance on product evaluations. We expected that this reward mechanism would effectively motivate the subjects. The subjects were also required to fill in background information forms (Appendix 19). The 55 Chapter 6 Research Methods collected background information enabled us to evaluate demographic difference in different treatment conditions. Each subject was then given a general experiment information sheet (Appendix 20) to inform him of the experiment procedures. Next, a research assistant directed the subject to the experiment website and trained him on how to examine products using the website condition to which he was assigned, so he would not have difficulty using the website. We were also concerned that subjects might need a benchmark to evaluate particular websites. This concern arises from Adaptation Theory (Helson, 1964), which suggests that people's judgments are based on 1) the sum of their past experiences, 2) the context and background of an experience, and 3) a stimulus. In the experiment, we randomly assigned subjects to different treatment conditions to ensure that the sum of subjects' past experiences were homogeneous across conditions.2 Furthermore, if a common benchmark was also provided to all subjects, we could be confident that the context and background of their experimental experiences were equivalent, leaving the differences across different conditions caused only by different treatment stimuli. Therefore, before the subjects were asked to examine products in their assigned conditions, they were shown websites that demonstrated other products, with static images and text (a sports watch: TimeX Digital Compass; and a PDA: PalmV). The subjects were asked to treat the sample websites as standard websites, and to use them as benchmarks against which to judge the experiment websites. Each subject was then directed to his assigned experiment website and asked to examine the products as if he was themselves shopping and making purchase decisions. Following that, the subject was given questionnaires to complete (see Appendix 21). 2 Demographics analysis has supported this assumption. Chapter 7 reports the results. 56 Chapter 6 Research Methods The above procedures were repeated for the two products. After subjects finished answering the questionnaires, they were debriefed and paid $15, as promised in the consent form. They were also told that the bonus payment would be delivered later, after their performances had been evaluated. The subjects' use of the websites was screen-captured and recorded, with their permission, using Camtasia 1.0.0 screen-capture software. The research assistant filled out a form to keep track of the experiment procedures (Appendix 22). 6.3 Questionnaire Design The questionnaire used in the experiment is shown in Appendix 21. A seven-point Likert Scale was used for all the measurement items. The reliability and validity of the measures are reported in the next chapter. 6.3.1 Interactivity Four items were developed to measure interactivity. 1. I am able to interact with this product. 2. The product can respond to my input on this web interface. 3. I can interact fully with this website. 4. I can acquire product information in an interactive way. 6.3.2 Vividness Eleven items were used to measure vividness. Seven of them were borrowed from Kelley et al. (1989). As we were concerned that the seven items were not sufficient to depict vividness in conditions where animation, continuous visual stimuli, and sound effects were involved, we added additional four items based on discussion with other graduate students. 1. The product on this website looks colorful. 57 Chapter 6 Research Methods 2. The product information on this website is clear. I can acquire product information on this website from different sensory channels, (developed for this study) The product demonstration on this website is lively, (developed for this study) The product information on this website is descriptive. 3. 4. 5. 6. The product information on this website is distinctive. The product demonstration on this website is animated, (developed for this study) The product information on this website is rich.. This website contains product information exciting to senses, (developed for this study) 7. 8 9. 10. The product information on this website is vague. 11. The product information on this website is concrete. 6.3.3 Perceived Diagnosticity One item was adapted from Jiang and Benbasat (forthcoming). The other two were created for this study. 1. This website is helpful for me to evaluate the product. 2. This web interface is helpful in familiarizing me with the product. 3. This web interface is helpful for me to understand the performance of the product. 6.3.4 Compatibility Three items were adopted from Moore and Benbasat (1991) to measure compatibility. 1. Evaluating the product on this website is compatible with how I evaluate products in physical stores. 2. Evaluating the product on this website fits well with the way I like to evaluate products in physical stores. 3. Familiarizing myself with the product on this website is similar to my product evaluation 6.3.5 Perceived Usefulness Four items were adapted Koufairs (2002) and Venkatesh and Davis (1996) to measure perceived usefulness. style in physical stores. 2. 3. 1. This website improves my online shopping performance. This website improves my decision making in online shopping. This website increases my online shopping effectiveness. 58 4. I find this website useful. 6.3.6 Perceived Ease of Use Chapter 6 Research Methods Four items were adapted Koufairs (2002) and Venkatesh and Davis (1996) to measure perceived ease of use. 1. I find this website easy to use. 2. Learning to use this website is easy for me. 3. My interaction with this website is clear and understandable. 4. It would be easy for me to become skillful at using this website. 6.3.7 Shopping Enjoyment Four items were adapted Koufairs (2002) to measure shopping enjoyment. 1. I find my experience with this website interesting. 2. I find my experience with this website enjoyable. 3. I find my experience with this website exciting. 4. I find my experience with this website fun. 6.3.8 Intention to Return The first three measurement items to measure intention to return were adapted from Coyle and Thorson (2001). The fourth item was developed for this study. 1. I would like to revisit this website in the future. 2. Next time I need to shop for a sports watch, I would like to use this website. 3. Next time I need to shop for a sports watch as a gift for a friend, I would like to use a website with characteristics similar to those of this website. 4. I would use websites with similar characteristics to those of this website in the future. 6.3.9 Attitudes toward Websites Four items were adapted from Grazioli and Jarvenpaa (2000) and Coyle and Thorson (2001). 1. I like shopping on this website. 2. Shopping on this website is a good idea. 3. Shopping on this website is appealing. 59 Chapter 6 Research Methods 4. I have formed a favorable impression toward this website. 6.3.10 Attitudes toward Products The first two items were borrowed from Kempf and Smith (1998), while the third one was developed for this study. 1. The product that F ve just examined is good. 2. I have formed a favorable impression toward the product that I've just examined. 3. I like the product that I've just examined. 6.3.11 Product Risk Three items were developed for this study to measure perceived product risk. 1. If I buy this product from the website, I am not sure whether or not it can meet my expectations. 2. In general, I would characterize a decision of buying the product from this website as risky. 3. I believe that I can buy the right product that I want from this website. 6.3.12 Intention to Purchase Four items were adapted from Coyle and Thorson (2001) to measure intention to purchase. 1. It is likely that I will buy this product. 2. I will purchase the product the next time I need a sports watch. 3. Suppose that a friend calls me to get my advice in his/her search for a sports watch, I would recommend him/her to buy the product. 4. I will definitely try this product. 6.4 Chapter Summary Research methods, including website design, experimental procedures, questionnaire design, have been reported in this chapter. The analysis of the data collected from the experiment will be discussed in the next chapter. 60 Chapter 7 Data Analysis Chapter 7 Experiment Data Analysis This chapter reports the analysis of data gathered through the experiment outlined above. The background information obtained from the experiment subjects is described in Section 7.1. Section 7.2 reports the manipulation check of the experiment design. Sections 7.3 tests the research model primarily by using the Partial Least Squares (PLS) approach. Section 7.4 compares the four experimental conditions using a M A N O V A approach. A summary of all of the analysis results is discussed in Section 7.5. 7.1 Subject Background Information 7.1.1 General Background Among the 176 subjects, 92 (52.3%) were female and 84 (47.7%) were male. Seven were graduates while all the rest were undergraduate students. The minimum age was 17 and the maximum age was 43, with a mean of 21.6 and a standard deviation of 4.1. The subjects were recruited from a variety of academic faculties and schools, representing very diverse backgrounds (Table 4). In general, most subjects reported that they had over four years' experience of Internet usage, and they used the Internet quite frequently (Tables 5 and 6). In addition, a large proportion (about 71%) of the subjects reported that they had previously purchased items online (Tables 7 and 8). 61 Chapter 7 Data Analysis TABLE 4 Subjects' Academic Background Faculties/Schools Faculties/Schools Frequency Percent Valid Percent Arts 58 33.0 33.0 Agricultural Sciences 3 1.7 1.7 Applied Sciences 17 9.7 9.7 Commerce 38 21.6 21.6 Education 2 1.1 1.1 Forestry 1 .6 .6 Human Kinetics 4 2.3 2.3 Journalism 1 .6 .6 Law 3 1.7 1.7 Medicine 1 .6 .6 Nursing 4 2.3 2.3 Science 33 18.8 18.8 Unclassified 11 6.3 6.3 Total 176 100.0 100.0 TABLE 5 Internet Experience Internet Use (Years) Frequency Percent Valid Percent < 1 year 0 0 0 1-2 years 0 0 0 2 ~ 4 years 21 11.9 12.1 > 4 years 153 86.9 87.9 Missing 2 1.1 Total 174 98.9 100.0 62 Chapter 7 Data Analysis TABLE 6 Daily Internet Usage Internet Usage (Daily) Frequency Percent Valid Percent < 15 minutes 3 1.7 1.7 15 minutes ~ 1 hour 33 18.8 18.9 1-2 hours 47 26.7 26.9 >2 hours 92 52.3 52.6 Missing 1 0.6 Total 175 99.4 100.0 TABLE 7 Internet Purchase Experience Purchase Experience Frequency Percent Valid Percent No purchase experience 50 28.41 28.57 Previous online purchases 125 71.02 71.43 Missing 1 0.57 Total 175 99.43 100.00 TABLE 8 Amounts Spent on Previous Internet Purchases Amount Frequency Percent Valid Percent $0 51 29.0 29.1 <$100 50 28.4 28.6 > $100 and < $500 51 29.0 29.1 >$500 and < $1,000 13 7.4 7.4 > $1,000 10 5.7 5.7 Missing 1 0.6 Total 175 99.4 100.0 63 Chapter 7 Data Analysis 7.1.2 Subject Backgrounds in Each Treatment Condition The gender composition of each treatment condition is reported in Table 9. In each of the four conditions, there were 21 males and 23 females. TABLE 9 Gender Composition in Each Treatment Condition GROUP N N (Male) N (Female) Condition 1 (Multiple-Static-Images) 44 21 23 Condition 2 (Video-Without-Narration) 44 21 23 Condition 3 (Video-With-Narration) 44 21 23 Condition 4 (VPE) 44 21 23 Total 176 84 92 The age distribution of each treatment condition is reported in Table 10. One-way A N O V A reveals no significant difference in ages across the four groups (p=.53). TABLE 10 Age Composition of Different Treatment Conditions GROUP N Mean Median Minimum Maximum Variance Condition 1 (Multiple-Static-Images) 44 21.70 20 18 35 19.75 Condition 2 (Video-Without-Narration) 44 21.18 20 17 43 21.50 Condition 3 (Video-With-Narration) 44 21.25 20.5 18 30 9.22 Condition 4 (VPE) 44 22.34 21 17 35 16.74 Total 176 21.62 20 17 43 16.73 Subjects were asked to report their comfort with using the Internet, on a 7-point Likert Scale. In general, the subjects felt very comfortable with their Internet usage (mean: 6.50, Table 11). There was no significant difference across the four conditions (p = .416). 64 Chapter 7 Data Analysis TABLE 11 Subjects' Comfort with the Internet GROUP N Mean Median Variance Condition 1 (Multiple-Static-Images) 44 6.59 7.00 .340 Condition 2 (Video-Without-Narration) 44 6.57 7.00 .344 Condition 3 (Video-With-Narration) 44 6.39 7.00 .522 Condition 4 (VPE) 43 6.44 7.00 .586 Total 175 6.50 7.00 .447 Missing 1 Subjects were also asked to report their familiarity with Internet shopping, on a 7-point Likert Scale. The mean familiarity was 4.74, suggesting that the subjects were familiar with online shopping. No difference was found across the four conditions (see Table 12) (p = .358). TABLE 12 Subjects' Familiarity with Internet Shopping GROUP N Mean Median Variance Condition 1 (Multiple-Static-Images) 44 4.77 5.00 2.66 Condition 2 (Video-Without-Narration) 44 5.05 5.00 2.65 Condition 3 (Video-With-Narration) 44 4.45 5.00 2.44 Condition 4 (VPE) 43 4.70 5.00 1.88 Missing 1 Total 175 4.74 5.00 2.41 The subjects' self-reported familiarity with sports watches was 3.71, based on 7-point Likert Scale (see Table 13). There was no significant difference across the four conditions (p= .725). Subjects' self-reported familiarity with PDA was 2.93 (based on 7-point Likert Scale, see Table 14), suggesting that customers are quite unfamiliar with PDAs. No significant difference was found across the four conditions (p = .33). 65 Chapter 7 Data Analysis TABLE 13 Subjects' Familiarity with Sports Watches GROUP N Mean Median Variance Condition 1 (Multiple-Static-Images) 44 3.73 4.00 2.53 Condition 2 (Video-Without-Narration) 44 3.50 4.00 2.07 Condition 3 (Video-With-Narration) 44 3.80 4.00 2.17 Condition 4 (VPE) 44 3.82 4.00 1.69 Total 176 3.71 4.00 2.09 TABLE 14 Subjects' Familiarity with PDAs GROUP Mean Median Minimum Maximum Variance Condition 1 (Multiple-Static-Images) 2.66 2.00 1 7 3.16 Condition 2 (Video-Without-Narration) 3.20 3.00 1 7 2.96 Condition 3 (Video-With-Narration) 2.73 2.00 1 6 2.67 Condition 4 (VPE) 3.14 3.00 1 7 3.00 Total 2.93 3.00 1 7 2.96 Overall, because the subjects' backgrounds were quite homogeneous across different experimental conditions, the background information has not been considered as potential moderators for data analysis3. 7.2 Initial Check of Vividness and Interactivity Measurement An exploratory factorial analysis (using Unweighted Least Square as an extraction method, and Varimax. as a rotation method) was first performed on the eleven items for 3 For example, we examined whether or not consumers' familiarity with products moderates the effect of interactivity on perceived diagnosticity. Partial correlation shows that interactivity is significantly correlated with perceived diagnosticity when familiarity is controlled, while, on the other hand, familiarity is not significantly correlated with perceived diagnosticity when interactivity is controlled. This means that familiarity does not play a role in the relationship between interactivity and perceived diagnosticity. 66 Chapter 7 Data Analysis measuring vividness to examine the underlying factor pattern (Table 15). The factor analysis yields three factors. The first factor includes the four items, which have been developed for this study. The other seven items, which have been borrowed from (Kelley, et al., 1989) load on two factors, with six on one factor and the other one on the second factor. A further semantic analysis of the items shows that the first factor can accurately capture the vividness manipulation in this experimental context where product images, videos, and V P E are involved, while the other two factors are suitable for measuring vividness specifically in a textual versus image-based setting as was used in Kelley et al. (1989), the source of the measurement. Therefore, the four items of the first factor have been used in the following-up analysis. As a matter of fact, these four items corresponds to the largest eigenvalue of 4.43 and explains 40.3% variance, while the second factor has an eigenvalue of 1.61 and explains 14.8% variance, the third factor has an eigenvalue of 1.01 and explains 9.23% variance. Therefore, using the first factor to measure vividness is statistically as well as conceptually appropriate for this particular experimental context. T A B L E 15 Factor Analysis of the Eleven Vividness Measurement Items Factor 1 2 3 V4 0.83 0.22 0.15 V7 0.76 0.16 V9 0.71 0.27 0.13 V3 0.64 V11 0.69 V5 0.22 0.67 V10 0.61 V8 0.36 0.56 V2 0.28 0.53 0.27 V6 0.42 0.49 0.16 V1 0.14 0.91 A further exploratory factor analysis was conducted to test whether vividness and interactivity are empirically different (Table 16). Results largely support the separation of the 67 Chapter 7 Data Analysis two constructs with the exception that the fourth item for measuring interactivity load heavily on both factors; therefore this item has been dropped from further analysis. T A B L E 16 Factor Analysis of Vividness and Interactivity Factor 1 2 INT1 0.86 0.39 INT2 0.86 0.19 INT3 0.78 0.22 INT4 0.64 0.50 V7 0.13 0.82 V4 0.39 0.76 V3 0.20 0.64 V9 0.38 0.61 7.3 Manipulation Check A M A N O V A test was then conducted to check whether the manipulation of different levels of interactivity and vividness were successful. The results indicate that the four experimental conditions significantly affect subjects' perceptions of interactivity (p< .01) and vividness (p< .01). Post Hoc multiple comparisons using the Sheffe test indicate two levels of perceived interactivity (Table 17). Subjects in the first three conditions (i.e. multiple-static-images, video-without-narration, and video-with-narration) perceived a low level of interactivity, while the V P E condition perceived a high level. Multiple comparisons also indicate two levels of perceived vividness (Table 18). Specifically, the multiple-static-images condition is associated with a low level of vividness, while the other three conditions have a high level of vividness. Overall, the manipulation check supports our expectations of the variation of perceived interactivity and vividness across different conditions. 68 Chapter 7 Data Analysis TABLE 17 Homogeneous Subsets of Interactivity N Subset for alpha = .05 GROUP 1 2 Condition 1 (Multiple-Static-Images) 44 3.12 Condition 3 (Video-With-Narration) 44 3.50 Condition 2 (Video-Without-Narration) 44 3.59 Condition 4 (VPE) 44 6.03 Sig. 0.19 1.00 TABLE 18 Homogeneous Subsets of Vividness N Subset for alpha = .05 GROUP 1 2 Condition 1 (Multiple-Static-Images) 44 3.20 Condition 3 (Video-With-Narration) 44 4.75 Condition 2 (Video-Without-Narration) 44 4.94 Condition 4 (VPE) 44 5.16 Sig. 1.00 0.12 7.4 Model Testing The research model was tested by Partial Least Squares (PLS). PLS was adopted due to "its ability to model latent constructs under conditions of non-normality and with small to medium sample sizes" (Chin, et al., 2003). PLS is a component-based structural equation modeling technique, which facilitates simultaneous tests of measurement models and structural models (Barclay, et al., 1995). In this study, the PLS Graph 3.0 software was used for PLS analysis. The model was first tested using data from the sports watch and data from the PDA individually. Because no significant difference was found between the two sets of data in terms of the 20 hypotheses presented in Chapter 5 above, data on the two products were combined with each other. The analysis described below was therefore based on the combined data set. 69 Chapter 7 Data Analysis 7.4.1 Measurement Model Barclay et al. (1995) have suggested that the assessment of a measurement model should examine: 1) individual item reliability, 2) internal consistency, and 3) discriminant validity. A general method for checking individual item reliability involves checking whether individual item loadings are above 0.6, or ideally 0.70 (Barclay, et al., 1995; Chin, 1998). The measurement items in the model generally load heavily on their respective constructs (Table 19), with loadings above 0.7, thus demonstrating adequate reliability. To examine internal consistency, composite reliability and Cronbach's alpha are reported in Table 20. Because all reliability figures are above 0.7 (Barclay, et al., 1995; Nunnally, 1978), internal consistency criteria are met. The third step to assess the measurement model is to examine discriminant validity. Off-diagonal elements in Table 21 represent correlations of all latent variables, while the diagonal elements are the square roots of the Average Variance Extracted (AVE) of the latent variables. For adequate discriminant validity, the A V E of any latent variable should be greater than the variance shared between the latent variable and other latent variables (Barclay, et al., 1995). In other words, the diagonal elements should be greater than corresponding off-diagonal elements. Data presented in Table 21 meet this condition. 70 Chapter 7 Data Analysis TABLE 19 Loadings and Cross-Loadings of Measures VIV INT DIAG CMPT USE EOU ENJOY ATTWB IR RISK ATTP IP VIV3 0.74 0.37 0.38 0.33 0.38 0.11 0.40 0.30 0.32 -0.10 0.27 0.16 VIV4 0.89 0.59 0.52 0.46 0.52 0.21 0.66 0.44 0.55 -0.23 0.34 0.36 VIV7 0.82 0.36 0.46 0.35 0.40 0.24 0.46 0.34 0.42 -0.14 0.23 0.23 VIV9 0.83 0.53 0.46 0.44 0.51 0.24 0.65 0.55 0.55 -0.25 0.32 0.42 INT1 0.63 0.95 0.51 0.37 0.47 0.08 0.59 0.42 0.49 -0.30 0.28 0.31 INT2 0.47 0.92 0.36 0.31 0.36 -0.02 0.49 0.34 0.42 -0.22 0.25 0.35 INT3 0.46 0.88 0.36 0.25 0.32 0.07 0.36 0.27 0.34 -0.22 0.11 0.23 PD1 0.50 0.47 0.91 0.54 0.63 0.55 0.49 0.55 0.60 -0.45 0.52 0.32 PD2 0.53 0.36 0.90 0.56 0.66 0.60 0.50 0.60 0.65 -0.44 0.58 0.34 PD3 0.47 0.41 0.87 0.50 0.62 0.45 0.49 0.50 0.57 -0.44 0.44 0.36 CMPT1 0.49 0.37 0.58 0.92 0.57 0.38 0.52 0.58 0.61 -0.43 0.48 0.42 CMPT2 0.42 0.32 0.56 0.92 0.55 0.35 0.52 0.61 0.63 -0.47 0.49 0.50 CMPT3 0.40 0.23 0.45 0.86 0.43 0.25 0.43 0.50 0.48 -0.42 0.37 0.37 PU1 0.51 0.48 0.55 0.47 0.88 0.28 0.61 0.68 0.65 -0.53 0.40 0.47 PU2 0.48 0.35 0.58 0.52 0.91 0.32 0.60 0.73 0.67 -0.53 0.46 0.52 PU3 0.45 0.35 0.55 0.47 0.88 0.41 0.54 0.60 0.64 -0.47 0.37 0.40 PU4 0.50 0.31 0.77 0.54 0.81 0.58 0.55 0.64 0.72 -0.49 0.57 0.39 PEOU1 0.26 0.03 0.60 0.32 0.46 0.88 0.26 0.40 0.41 -0.32 0.43 0.24 PEOU2 0.15 -0.01 0.51 0.28 0.34 0.91 0.19 0.32 0.32 -0.28 0.39 0.16 PEOU3 0.29 0.17 0.52 0.38 0.42 0.86 0.33 0.40 0.45 -0.37 0.37 0.28 PEOU4 0.15 -0.04 0.45 0.30 0.37 0.86 0.24 0.38 0.37 -0.32 0.40 0.13 EN1 0.60 0.45 0.59 0.50 0.69 0.35 0.90 0.74 0.76 -0.52 0.55 0.56 EN2 0.62 0.49 0.55 0.56 0.65 0.33 0.95 0.76 0.76 -0.50 0.59 0.57 EN3 0.67 0.51 0.45 0.51 0.56 0.20 0.93 0.67 0.65 -0.40 0.50 0.58 EN4 0.64 0.55 0.47 0.49 0.60 0.23 0.95 0.68 0.67 -0.44 0.50 0.57 AW1 0.45 0.34 0.53 0.59 0.68 0.35 0.70 0.94 0.75 -0.63 0.63 0.61 AW2 0.52 0.42 0.56 0.61 0.69 0.41 0.73 0.95 0.78 -0.66 0.67 0.62 AW3 0.41 0.33 0.57 0.56 0.73 0.42 0.64 0.91 0.74 -0.67 0.56 0.61 AW4 0.50 0.34 0.61 0.58 0.74 0.42 0.76 0.91 0.78 -0.63 0.69 0.58 IR1 0.54 0.41 0.44 0.54 0.60 0.27 0.74 0.73 0.81 -0.44 0.58 0.71 IR2 0.45 0.39 0.62 0.52 0.69 0.38 0.66 0.76 0.89 -0.56 0.58 0.61 IR3 0.49 0.36 0.63 0.56 0.70 0.48 0.62 0.70 0.89 -0.54 0.59 0.53 IR4 0.53 0.46 0.70 0.64 0.74 0.46 0.65 0.70 0.92 -0.55 0.58 0.52 RISK1 -0.11 -0.16 -0.32 -0.30 -0.39 -0.23 -0.24 -0.41 -0.28 0.74 -0.27 -0.26 RISK2 -0.03 -0.13 -0.29 -0.29 -0.33 -0.23 -0.20 -0.42 -0.32 0.74 -0.21 -0.22 RISK3 -0.30 -0.31 -0.51 -0.51 -0.59 -0.38 -0.60 -0.73 -0.67 0.90 -0.55 -0.61 AP1 0.38 0.26 0.54 0.48 0.49 0.41 0.56 0.64 0.63 -0.40 0.93 0.62 AP2 0.29 0.22 0.50 0.45 0.48 0.43 0.52 0.62 0.60 -0.50 0.92 0.56 AP3 0.33 0.20 0.57 0.47 0.50 0.44 0.52 0.68 0.63 -0.46 0.95 0.57 IP1 0.34 0.33 0.23 0.40 0.37 0.08 0.57 0.51 0.51 -0.38 0.51 0.87 IP2 0.32 0.35 0.35 0.46 0.42 0.14 0.53 0.59 0.60 -0.48 0.58 0.91 IP3 0.31 0.23 0.43 0.42 0.52 0.31 0.51 0.62 0.61 -0.54 0.57 0.84 IP4 0.32 0.21 0.26 0.37 0.44 0.28 0.50 0.50 0.60 -0.37 0.48 0.81 71 Chapter 7 Data Analysis T A B L E 20 Internal Consistency of Measurements Variables Composite Reliability Cronbach's Alpha Vividness 0.89 0.83 Interactivity 0.94 0.90 Perceived Diagnosticity 0.92 0.87 Compatibility 0.93 0.80 Perceived Usefulness 0.93 0.89 Perceived Ease of Use 0.93 0.90 Shopping Enjoyment 0.97 0.95 Attitudes toward Products 0.95 0.93 Attitudes toward Websites 0.96 0.94 Perceived product risk 0.84 0.81 Intentions to Return 0.93 0.90 Intentions to Purchase 0.92 0.88 Another criterion for adequate discriminant validity requires that loadings of indicators on their respective latent variables be higher than loadings of other indicators on the latent variables and the loadings of these indicators on other latent variables. The loadings and cross-loadings presented in Table 18 demonstrate adequate discriminant validity. 72 Chapter 7 Data Analysis TABLE 21 Correlation of Constructs Vividness Interactivity Diagnosticity Compatibility Usefulness Ease of Use Enjoyment Attitude (Product) Attitude (Website) Risk Intention Return Intention Purchase Vividness 0.82 Interact 0.58 0.92 Diagnosticity 0.56 0.46 0.90 Compatibility 0.49 0.35 0.60 0.90 Usefulness 0.56 0.43 0.71 0.58 0.87 Ease of Use 0.25 0.05 0.60 0.37 0.46 0.88 Enjoyment 0.68 0.54 0.55 0.55 0.67 0.30 0.93 Attitude (Product) 0.36 0.24 0.58 0.50 0.52 0.46 0.57 0.93 Attitude (Website) 0.51 0.39 0.61 0.63 0.77 0.43 0.76 0.69 0.93 Risk -0.23 -0.27 -0.50 -0.49 -0.58 -0.37 -0.50 -0.49 -0.70 0.80 Intention Return 0.57 0.46 0.68 0.64 0.78 0.45 0.76 0.66 0.83 -0.60 0.88 Intention Purchase 0.38 0.33 0.38 0.48 0.51 0.24 0.61 0.63 0.65 -0.52 0.68 0.86 73 Chap te r 7 Da ta Ana lys is 7.4.2 Structural Model Two resampling methods are available in PLS Graph 3.0: Jackknife and Bootstrap. By default, each Jackknife resample "consists of n-1 cases, where n equals the total number of cases in your original sample," while Bootstrap employs "100 resamples with each sample consisting of the same number of cases as your original sample set" (PLS Graph 3.0 manual). Compared to Jackknife, Bootstrap takes more time but is more efficient in assessing the significance of PLS parameters estimates because Jackknife can be considered as an approximation of Bootstrap (Chin, 1998). Therefore Bootstrap resampling was performed on the structural model to examine significance levels of path coefficients4. Figure 4 exhibits the results. Both the interactivity and the vividness of product demonstrations significantly enhance perceived diagnosticity and shopping enjoyment. Vividness, alone, affects compatibility, but interactivity does not. The effect of interactivity on perceived ease of use is not significant either. Perceived diagnosticity and compatibility are significant predicators of perceived usefulness. Perceived usefulness, perceived ease of use and shopping enjoyment jointly improves consumers' attitudes toward websites, which, in turn, enhance their intentions to return. Contrary to T A M , perceived usefulness is not a mediator between perceived ease of use and attitudes toward websites. 4 As a matter of fact, Bootstrap and Jackknife do not demonstrate significant difference for data analysis in the present study. 74 Chapter 7 Data Analysis Risk R2=0.488 Note: Solid lines indicate that the corresponding path coefficients are significant (p < .05), while dotted lines indicate insignificant effects (p > .05). Figure 4 PLS Test of the Research Model Consumers' attitudes toward websites also significantly affect their attitudes toward products. Perceived diagnosticity positively affects consumers' attitudes toward products (assuming that the product information itself is positive), but negatively affects consumers' perceptions of product-related risk. Both product risk and attitudes toward the products affect consumers' intentions to purchase the products from the websites. A summary of relevant hypothesis-testing is presented in Table 22. 75 Chapter 7 Data Analysis TABLE 22 Hypothesis Tests Hypothesis Path Coefficient Supported? HI a: Vividness -> (+) Perceived diagnosticity 0.44 Yes H2: Vividness -> (+) Compatibility 0.43 Yes H3: Vividness (+) Shopping enjoyment 0.55 Yes H4a: Interactivity (+) Perceived diagnosticity 0.21 Yes H5: Interactivity -> (+) Compatibility 0.10 No H6: Interactivity (+) Shopping enjoyment 0.22 Yes H7: Interactivity -> Perceived ease of use 0.05 No H7: Interactivity -> (+) Vividness 0.58 Yes H9: Perceived diagnosticity (+) Perceived usefulness 0.55 Yes H10: Compatibility (+) Perceived usefulness 0.24 Yes H l l : Perceived usefulness (+) Attitudes toward websites 0.41 Yes H12: Perceived ease of use (+) Attitudes toward websites 0.11 Yes H13: Perceived ease of use -> (+) Perceived usefulness 0.05 No H14: Shopping enjoyment -> (+)Attitudes toward websites 0.46 Yes H15: Perceived diagnosticity -> (-) Perceived product risk -0.50 Yes H16: Perceived diagnosticity Attitudes toward products 0.24 Yes HI7: Attitudes toward websites (+) Attitudes toward products 0.54 Yes HI8: Attitudes toward websites (+) Intentions to return 0.83 Yes H19: Attitudes toward products (+) Intentions to purchase 0.49 Yes H20: Perceived product risk -> (-) Intention to purchase -0.29 Yes The predictive power of a PLS model can be measured by examining the R Square values of its endogenous variables. The model explains a significant amount of the variances of all endogenous variables except perceived ease of use. The R Squares of these constructs are reported in Table 23. 76 Chapter 7 Data Analysis TABLE 23 R Square Values of Model Constructs Endogenous Variables R Square Perceived diagnosticity 0.34 Compatibility 0.25 Shopping enjoyment 0.49 Perceived usefulness 0.55 Perceived ease of use <0.01 Attitudes toward websites 0.71 Risk 0.25 Attitudes toward products 0.51 Intentions to return 0.68 Intentions to purchase 0.46 Since the effects of interactivity is partly mediated by vividness, the total effects of interactivity are calculated by adding the direct effects to the indirect effects through vividness, as exhibited in Table 24. In terms of the degree of substantial effects, both vividness' and interactivity exert comparable effect sizes on perceived diagnosticity, compatibility, and shopping enjoyment (total effects accounted for by vividness and interactivity: 0.44 versus 0.46 for perceived diagnosticity, 0.43 versus 0.35 for compatibility, and 0.55 versus 0.54 for shopping enjoyment), though interactivity has significantly smaller direct effects on perceived diagnosticity and shopping enjoyment and insignificant direct effect on compatibility. Perceived usefulness and shopping enjoyment exhibit similar amounts of influence on intentions to return (path coefficient: 0.41 and 0.46), and both are much greater than the influence of perceived ease of use. Attitudes toward products has a much larger effect on intentions to purchase than perceived product risk (path coefficient: -0.29 versus 0.49). 77 Chapter 7 Data Analysis TABLE 24 Total Effects of Vividness and Interactivity Vividness Interactivity Perceived diagnosticity 0.44 0.46 =(0.21+0.58*0.44) Compatibility 0.43 0.35 =(0.10+0.58*0.43) Shopping enjoyment 0.55 0.54 =(0.22+0.58*0.55) 7.5 Follow-up ANOVA Tests Partial Least Squares analysis has generally supported the proposed research model, thereby clarifying the underlying mechanisms of V P E effects on web consumers' beliefs, attitudes, and behavioral intentions, when compared to other online product presentation methods. A follow-up M A N O V A has also been conducted to compare the four presentation formats. We expect the results further enhance the practical significance of this research by directly revealing the relative effectiveness of these presentation formats. 7.5.1 Overall MANOVA Results The results from the M A N O V A test are significant (Table 25), indicating that, overall there are significant difference across the four conditions in terms all of the variables. TABLE 25 Overall MANOVA Test Value F Hypothesis df Error df Sig. Pillai's Trace .79 4.86 36 489.000 <0.01 Wilks's Lambda .32 6.26 36 476.420 <0.01 Hotelling's Trace 1.81 8.02 36 479.000 <0.01 Roy's Largest Root 1.61 21.86 12 163.000 <0.01 78 Chapter 7 Data Analysis Tests of between-subjects effects (Table 26) suggest that experimental treatments significantly affect all modeled variables except perceived ease of use, attitudes toward products, perceived product risk, and intentions to purchase. TABLE 26 MANOVA - Tests of Between-Subjects Effects Dependent Variable Type III Sum of Squares df Mean Square F Sig. Vividness 24.60 3.00 8.20 13.50 <0.01 Interactivity 233.54 3.00 77.85 77.20 <0.01 Perceived Diagnosticity 19.63 3.00 6.54 12.11 <0.01 Compatibility 21.39 3.00 7.13 7.24 <0.01 Perceived Usefulness 25.46 3.00 8.49 13.48 <0.01 Perceived Ease of Use 1.20 3.00 0.40 0.88 0.45 Enjoyment 47.68 3.00 15.89 16.58 <0.01 Attitudes toward Products 3.46 3.00 1.15 1.32 0.27 Attitudes toward Websites 21.62 3.00 7.21 8.07 <0.01 Perceived Risk 7.12 3.00 2.38 2.63 0.05 Intentions to Return 28.03 3.00 9.34 9.44 <0.01 Intentions to Purchase 8.63 3.00 2.88 2.32 0.08 7.5.2 Multiple ANOVA Analysis A one-way A N O V A test with multiple comparisons using the Scheffe test was conducted on each variable. The analysis of vividness and interactivity has been presented in Section 7.3 (see Tables 17 and 18). Regarding vividness, the two video conditions and the V P E condition are not significantly different from each other, but all are significantly higher than the multiple-static-images condition. Regarding interactivity, the multiple-static-images, video-without-narration and video-with-narration conditions are on the same level, but are significantly lower than the V P E condition. A N O V A analysis of other variables is reported below. 79 Chapter 7 Data Analysis 7.5.2.1 Perceived Diagnosticity A N O V A results reveal three levels of perceived diagnosticity (Table 27). Subjects in the multiple-static-images condition perceived the lowest level of diagnosticity. Those in the video-without-narration and video-with-narration conditions perceived higher levels of diagnosticity, and there were no significant differences between these two groups of people. Subjects in the V P E condition perceived the highest level of diagnosticity, but this diagnosticity was not significantly different from that in the video-without-narration condition. TABLE 27 Homogeneous Subsets of Perceived Diagnosticity GROUP N Subset 1 2 3 Condition 1 (Multiple-Static-Images) 44 5.13 Condition 3 (Video-With-Narration) 44 5.57 Condition 2 (Video-Without-Narration) 44 5.71 5.71 Condition 4 (VPE) 44 6.06 Sig. 1.00 .85 .18 7.5.2.2 Compatibility The compatibility reported by the experiment subjects is presented in Table 28. There are no significant differences among the multiple-static-images, video-without-narration, and video-with-narration conditions, nor are there any significant differences among the video-without-narration, video-with-narration, and VPE conditions. However, the V P E condition exhibits significantly higher compatibility than the multiple-static-images condition. TABLE 28 Homogeneous Subsets of Compatibility GROUP N Subset 1 2 Condition 1 (Multiple-Static-Images) 44 4.15 Condition 2 (Video-Without-Narration) 44 4.66 4.66 Condition 3 (Video-With-Narration) 44 4.68 4.68 Condition 4 (VPE) 44 5.13 Sig. 0.10 0.18 80 Chapter 7 Data Analysis 7.5.2.3 Perceived Usefulness The perceived usefulness of the experiment websites is reported in Table 29. Perceived usefulness is not significantly different between the video-without-narration and video-with-narration conditions. Both conditions involve significantly higher levels of perceived diagnosticity than the multiple-static-images condition does, and lower levels of perceived diagnosticity than the V P E condition does. TABLE 29 Homogeneous Subsets of Perceived Usefulness GROUP N Subset 1 2 3 Condition 1 (Multiple-Static-Images) 44 4.51 Condition 3 ( Video-With-Narration ) 44 4.99 Condition 2 ( Video-Without-Narration) 44 5.01 Condition 4 (VPE) 44 5.58 Sig. 1.00 1.00 1.00 7.5.2.4 Perceived Ease of Use There is no significant difference in the perceived ease of use of the websites across the four conditions (Table 30). TABLE 30 Homogeneous Subsets of Perceived Usefulness GROUP N Subset 1 Condition 1 (Multiple-Static-Images) 44 5.99 Condition 3 (Video-With-Narration) 44 6.12 Condition 4 (VPE) 44 6.13 Condition 2 (Video-Without-Narration) 44 6.23 Sig. .46 81 Chapter 7 Data Analysis 7.5.2.5 Shopping Enjoyment The A N O V A testing on shopping enjoyment is listed in Table 31. The video-without-narration and video-with-narration conditions exhibit no significant difference in the shopping enjoyment reported by experiment subjects. However, both values are significantly lower than the shopping enjoyment in the V P E condition. The video-without-narration condition generates a significantly higher level of shopping enjoyment than the multiple-static-images condition, but the video-with-narration condition does not. TABLE 31 Homogeneous Subsets of Shopping Enjoyment GROUP N Subset 1 2 3 Condition 1 (Multiple-Static-Images) 44 3.59 Condition 3 (Video-With-Narration) 44 4.08 4.08 Condition 2 (Video-Without-Narration) 44 4.28 Condition 4 (VPE) 44 5.04 Sig. 0.14 0.82 1.00 7.5.2.6 Attitudes toward Websites Subjects' attitudes toward the websites are reported in Table 32. The multiple-static-images, video-with-narration, and video-without-narration condition exhibit no significant differences in attitudes, nor is there any significant difference between the video-without-narration condition and the VPE condition. However, the V P E condition is associated with significantly more positive attitudes toward websites than the multiple-static-images and video-with-narration conditions. 82 Chapter 7 Data Analysis TABLE 32 Homogeneous Subsets of Attitudes toward Websites GROUP N Subset 1 2 Condition 1 (Multiple-Static-Images) 44 4.20 Condition 3 (Video-With-Narration) 44 4.50 Condition 2 (Video-Without-Narration) 44 4.68 4.68 Condition 4 (VPE) 44 5.17 Sig. .14 .12 7.5.2.7 Intentions to Return The intentions of subjects to return to the websites are reported in Table 33. Subjects in the VPE condition appeared significantly more likely return to the websites that they used in the experiment than those in the other three conditions. Intentions to return in the multiple-static-images, video-with-narration, and video-without-narration conditions are not significantly different from each other. TABLE 33 Homogeneous Subsets of Intentions to Return GROUP N . Subset 1 2 Condition 1 (Multiple-Static-Images) 44 4.39 Condition 3 (Video-With-Narration) 44 4.73 Condition 2 (Video-Without-Narration) 44 4.88 Condition 4 (VPE) 44 5.49 Sig. .15 1.00 7.5.2.8 Perceived product risk There is no significant difference in perceptions of risk across the four conditions (Table 34). 83 Chapter 7 Data Analysis TABLE 34 Homogeneous Subsets of Perceived Product Risk GROUP N Subset 1 Condition 4 (VPE) 44 3.05 Condition 1 (Multiple-Static-Images) 44 3.52 Condition 2 (Video-Without-Narration) 44 3.53 Condition 3 (Video-With-Narration) 44 3.58 Sig. .08 7.5.2.9 Attitudes toward Products The experiment subjects did not report any difference in their attitudes toward products across the four conditions (Table 35). TABLE 35 Homogeneous Subsets of Attitudes toward Products GROUP N Subset 1 Condition 1 (Multiple-Static-Images) 44 4.95 Condition 3 (Video-With-Narration) 44 5.05 Condition 2 (Video-Without-Narration) 44 5.13 Condition 4 (VPE) 44 5.33 Sig. .31 7.5.2.10 Intentions to Purchase There is no significant difference in the subjects' reports of their intentions to complete purchases, across the different treatment conditions (Table 36). TABLE 36 Homogeneous Subsets of Intentions to Purchase N Subset GROUP 1 Condition 3 (Video-With-Narration) 44 3.70 Condition 1 (Multiple-Static-Images) 44 3.82 Condition 2 (Video-Without-Narration) 44 3.91 Condition 4 (VPE) 44 4.29 Sig. .11 84 Chapter 7 Data Analysis 7.6 Chapter Summary PLS tests have generally confirmed that vividness and interactivity are the driving forces of the persuasion process involved in web-based product presentations. Based on this finding, online firms are informed on how to exert their influence on consumers by manipulating the styles of their product presentations. Specifically, vividness and interactivity are able to influence web consumers' beliefs (e.g. their perceptions of the diagnosticity, compatibility, usefulness, ease of use, and risk inherent in particular websites), their affective responses (e.g. shopping enjoyment), their attitudes (e.g. their attitudes toward websites and products), and their behavioral intentions (e.g. their intentions to return and to purchase items). When all of these beliefs, affects, attitudes, and behavioral intentions are combined, they form a general picture of the effect mechanisms involved in VPE. Follow-up A N O V A tests have been used to compare the four experimental treatments, demonstrating that the product presentation formats exert strong causal influences on online consumers' beliefs, attitudes, and behavioral intentions. Since V P E is associated with high levels of vividness and interactivity, it is most effective in enhancing consumers' perceptions of and feelings toward websites, such as their perceptions of the diagnosticity, compatibility, and usefulness of websites, their shopping enjoyment, and their intentions to return to the websites. On the other hand, inasmuch as multiple-static-images is associated with low levels of vividness and interactivity, it affects the above variables to the least degree. The fact that no significant difference in vividness or interactivity has been detected between the video-without-narration and video-with-narration conditions helps explain why these two conditions cannot make a difference for the above variables. A l l of these findings lend substantial support to the PLS analysis results. 85 Chapter 7 Data Analysis It is also interesting that variables that are directly related to products, such as the risk perceived by consumers, consumers' attitudes toward products, and their intentions to complete purchases, are not influenced by experimental treatments. This suggests that online customers are unlikely to diverge from their perceptions of products simply because of different product presentation formats. However, possible doubts may arise from the apparent contradiction between the PLS analysis results and the A N O V A results. The PLS models suggest that perceived product risk and attitudes toward products are influenced by perceived diagnosticity and attitudes toward websites, which are driven by vividness and interactivity, while vividness and interactivity are directly affected by the experimental manipulation. This concern is valid, but it is not decisive, because statistically speaking, the A N O V A analysis demonstrates that the variances of dependent variables can be explained by variances in experimental treatments, while PLS analysis demonstrates that the variances of endogenous variables can be explained by the variances of exogenous variables, i.e. vividness and interactivity. Although experimental treatments lead to different levels of vividness and interactivity, the variances of vividness and interactivity cannot be substituted for the treatment variances. Therefore, the PLS results are best used to explain underlying effect mechanisms, while the A N O V A results are best used to explain the effectiveness of different product presentation formats. 86 Chapter 8 Conclusions and Discussion Chapter 8 Conclusions and Discussion This study investigates virtual product experience and its effect mechanism. Chapter 1 has introduced the motivation behind this study. Chapter 2 has reviewed prior literature on direct product experience, i.e., the counterpart of VPE in physical in-store shopping. It is expected that the benefits and constraints of direct product experience could shed light on VPE. Prior research on V P E has also been reviewed in Chapter 2, discussing previous findings as well as areas that deserve further research efforts. In particular, it is evident that new research is appropriate to examine how VPE technologies work, rather than simply testing the observable effects of such technologies. Chapter 3 has described the relationships between various product categories and V P E technologies, informing a taxonomy for online product presentations, and categorization of various basic V P E methods and corresponding product categories. Chapter 4 has elaborated on the two technological characteristics of VPE, i.e. interactivity and vividness, and specific methods to design them. Chapter 5 has presented a research model, aiming at depicting the effects of V P E on consumers' beliefs, attitudes, and behavior intentions. Chapter 6 has described the research method of this study, including website design, experimental procedures, and questionnaire design. Chapter 7 has reported the findings of this study. In particular, analysis of the data through PLS methods provides support for the validity of two research model. This final chapter concludes with a summary of the findings from this study, and a discussion of the contributions it has made and the limitations that exist. Section 8.1 discusses the results and their implications, Section 8.2 points out several limitations, and future research avenues are suggested in Section 8.3. 87 Chapter 8 Conclusions and Discussion 8.1 Findings and Implications As discussed in Chapter 2, prior studies on VPE have predominantly treated its support technology as a black box. The effects of VPE have been examined by comparing it to other interface conditions (Li, et al., 2002; 2003; Suh and Lee, 2004), an experimental approach that is effective for evaluating a technology as a whole, but that sheds little light on the functional processes of the technology. Addressing this concern, the present study has been developed from the two primary technological characteristics of VPE: interactivity and vividness. Interactivity is "the extent to which users can participate in modifying the form or content of a mediated environment in real time" (Steuer, 1992, p. 84), while vividness is "the representational richness of a mediated environment" (Steuer, 1992, p. 84). The analytic approach taken by the present study is appropriate for studying new information technology, because it allows researchers to directly examine the internal mechanisms of a technology, thereby accurately explaining and predicting the effects of the technology. This approach extends the research performed by Jiang and Benbasat (forthcoming), who have argued that visual and functional control technologies are implemented by integrating direct manipulation and multimedia techniques. Based on this assertion, they have explained the effects of visual and functional control theoretically. However, they have not empirically tested whether the two techniques (direct manipulation and multimedia) work as they have been predicted. Inasmuch as direct manipulation is a particular design method to develop interactivity, and multimedia is a separate design method used to augment vividness, the significant effects of interactivity and vividness identified in the present study have lent substantial support to Jiang and Benbasat's perspective. In this study, vividness has significant direct effects on perceived diagnosticity, compatibility, and shopping enjoyment, while interactivity has significant direct effects only on 88 Chapter 8 Conc lus ions and Discussion perceived diagnosticity and shopping enjoyment, and the amounts of effects are relative small. However, after considering the indirect effects through vividness, the total effects of interactivity can reach a similar magnitude as those of vividness. This relative effectiveness of vividness and interactivity may be one concern when designing the corresponding techniques, such as multimedia and direct manipulation of VPE. Another concern may be the efforts that are need in designing those techniques. In the present technological environment, the design of vividness is generally better supported and requires less effort than the design of interactivity. For example, many current web development tools, such as Macromedia's Flash, Apple's QuickTime, and Sun Microsystem's Java 3-D APIs, simplify the design of web applications, such as online product demonstrations. Usually, these software tools enable users to easily create vivid and rich media by including animation, MP3 audio or high-quality videos, or other similar design elements. In contrast, these tools offer limited support for interactivity design. For example, although they enable designers to insert standard interactivity on a graphic interface with relative ease, such as tools to display new contents or to control video and sound, designers must write application scripts on their own if they desire special interactivity with online products. Inasmuch as users' interactions with online product demonstrations are disparately heterogeneous across different products and their behavior is arbitrary and generally unpredictable, it is unfeasible to develop special interactivity for different product designs and embed this interactivity behavior into software development tools. Since it is relatively easy to design vividness, it is not unusual to see that website designers are keen on adding video, audio, and image animation to their product presentation. However, it is possible that unnecessary increases of vividness might not be positive, but on the contrary, may turn out to have a negative effect on consumers' evaluation of products and websites. For example, adding inordinate animation or sound effects may distract people from capturing relevant product information and examining products (Geissler, et al., 2001; Zhang, 89 Chapter 8 Conclusions and Discussion 2000). Therefore, designers should be cautioned that the relative ease of design of vividness should not lead to the overuse of its corresponding technologies. The T A M model has been tested and verified in numerous Information Systems studies. The next promising step to supplement these prior studies could involve examination of the determinants of exogenous variables in T A M , such as perceived usefulness and perceived ease of use (Venkatesh, 2000). From this perspective, the present study contributes to Information Systems research by identifying two antecedents of perceived usefulness of websites — perceived diagnosticity and compatibility — in the specific context of online product demonstrations. Perceived diagnosticity refers to consumers' beliefs regarding the extent to which a website enables them to understand and evaluate products. Compatibility concerns consumers' beliefs of the extent to which their existing styles, habits or experiences in physical shopping can be utilized in online shopping experiences. Both of these constructs are influenced by interactivity and vividness, but they characterize two conceptually different aspects of online consumer behavior. Regarding perceived ease of use, the present study demonstrates that it is not affected by interactivity and does not affect perceived usefulness, but it does affect web consumers' attitudes toward websites. These results might be explained by the experiment subjects' relatively high levels of Internet experience. Hence, interaction with those particular experimental websites did not pose any challenges to the subjects, and therefore their reported perceived ease of use was generally high5 and did not reveal significant variance. Or, more generally, it can be inferred that most current e-commerce websites are designed for the general population, and therefore perceived ease of use may not be an impediment at all. For example, Koufaris (2002) has reported similar findings, after integrating T A M and the flow theory to predict web consumers' 5 Data analysis reveals that perceived ease of use is reportedly high for all conditions (average=6.12 based on a 7-point scale), and there is no significant difference across different treatment conditions (p=.45). 90 Chapter 8 Conclusions and Discussion intentions to return to a website. He has found that the perceived ease of use of a website is not affected by and does not affect any other variables. The results of the present study have also supported Koufaris (2002), by demonstrating that both perceived usefulness and shopping enjoyment significantly affect consumers' intentions to return to a website. Closer examination of the corresponding path coefficients in PLS analysis reveals that, compared to perceived usefulness, shopping enjoyment has a similar magnitude of effectiveness on intentions to return and attitudes toward websites, thus indicating shopping enjoyment is at least as influential as, if not more than, perceived usefulness in presenting online products. This finding further suggests that entertainment features should be included as an important consideration in e-commerce website design (McKinney, et al., 2002). The present research has also found that consumers' attitudes toward websites significantly influence their attitudes toward products. This finding confirms prior marketing research on affect transfer and inferential belief formation (Kim and Allen, 1996), and suggests that consumers may perceive products differently simply because they like or dislike a website. The practical implication for online firms is that, when presenting products online, the interactivity and vividness of product demonstrations not only directly influences consumers' perceptions of web presence and their attitudes toward it, but also indirectly influences consumers' perceptions of products and, in turn, affects their intentions to complete purchases. Therefore, it is possible that consumers may perceive a relatively inferior product favorably simply because the product presentation is attractive; or, on the other hand, consumers may perceive a quality product unfavorably simply because they are not satisfied with the website design. The practical value of this research is also increased through its comparison of four different web-based product demonstrations (i.e. multiple-static-images, video-without-narration, video-with-narration, and VPE) on consumers' beliefs, attitudes, and intentions. It offers direct 91 Chapter 8 Conclusions and Discussion advice to web designers regarding their options for product presentation formats. To the best of our knowledge, most studies have compared VPE to image-based product presentations (Jiang and Benbasat, forthcoming; L i , et al., 2002; 2003; Suh and Lee, 2004), while ignoring the fact that video-based product demonstrations are widely used in current e-commerce websites (Raney, et al., 2003), and over-looking previous observations that video-based demonstrations are more effective and persuasive than image-based product demonstrations (Klein, 2003). Therefore, an objective and comprehensive evaluation of V P E should compare it against video demonstrations. Addressing this need and going beyond it, this study has identified two types of video-based product demonstrations — those with narration (video-with-narration), and those without narration (video-without-narration) — and included them in the experimental design. Therefore, the comprehensive comparison of four experimental treatments in this study ensures web designers that the V P E benefits found in this study are not the results of arbitrary choices of experimental control groups. A N O V A results suggest that the V P E condition is, in general, more influential than the other three presentation formats in enhancing website-related or shopping experience-related variables, such as perceived diagnosticity, compatibility, perceived usefulness, shopping enjoyment, attitudes toward websites, and intentions to return to websites. The two video conditions are not significantly different from each other. Both of them are more influential on perceived diagnosticity and perceived usefulness than the multiple-static-images condition, while they are as influential as the multiple-static-images condition in influencing attitudes toward websites and intentions to return to the websites. Overall, the predominant advantages of VPE over other presentation formats lead us to conclude that VPE is the most effective among these four conditions in improving consumer interaction with products and websites and generating future website visits. 92 Chapter 8 Conclusions and Discussion On the other hand, the four conditions do not exhibit significantly different results in variables that are directly related to the perception or evaluation of products themselves, such as perceived product risk, attitudes toward products, and intentions to purchase goods. This result suggests that subjects' product-related beliefs are not affected by the experiment treatments. Readers should be cautious about confusing this conclusion with the aforementioned relationship between attitudes toward websites and attitudes toward products. The PLS analysis result that consumers' attitudes toward websites will affect their attitudes toward products is more general on the theoretical level in predicting relationships between these two constructs, while the absence of A N O V A differences in product-related variables only demonstrates that the variance of these four treatment conditions cannot sufficiently explain those variables. 8.2 Limitations The generalizability of the findings of this study is augmented by the consideration of two separate products (i.e. a sports watch and a PDA) in the experiment. Inasmuch as no significant difference has been found between these two products, in terms of model support, the data related to them were combined together. The combined dataset justifies more confidence in the use of the model to predict the effects of online product presentation technologies on consumer behavior. However, it should be pointed out that the particular VPE technology that was chosen in the experiment is functional control, and the products chosen both have functionality as one of their main features. Therefore, the generalizability of the findings should be best applied to this particular type of VPE, which allows customers to try various functions of online products. It is likely that the relative importance of interactivity and vividness may change for other types of VPEs. 93 Chapter 8 Conclusions and Discussion In this study, the website download speed is not considered as a potential moderator. In the experimental settings, the server was local and the download speed was fast enough (less than three seconds per page), and subjects did not report any feeling of "waiting too long." However, in real online shopping experiences, consumers may encounter significant delays when downloading complex product presentations, such as VPE files, video files, or even multiple product images. Recent studies have demonstrated that website download speed affects consumer perceptions of website usability (Palmer, 2002), and has further unfavorable effects on consumers' performance, attitudes, and behavioral intentions (Galletta, et al., 2004). Therefore, it is very likely that if the waiting time exceeds their tolerance levels, consumers will give up. Or, even if consumers persist in waiting, their attitudes toward or satisfaction with the websites may be unfavorably affected. Another limitation of this study is that we do not examine the effects of the novelty of product presentations on consumers' perception of websites and products. Prior research has found that novelty may arouse people's attention, stimulate their active information processing, and enhance their exploratory behavior (Dabholkar and Bagozzi, 2002; Hirschman, 1980). Therefore, if VPE is novel to online consumers, it is likely that their beliefs, attitudes, and behavioral intentions will be improved or enhanced simply due to the novelty of the V P E interfaces. On the other hand, if the novelty of product presentations decreases, consumers' beliefs, attitudes, and behavioral intentions might be undermined. For example, Edwards and Gangadharbatla (Edwards and Gangadharbatla, 2001) have found that consumers' purchase intentions decrease with the increase of their familiarity with 3D product experience. In this study, we tested subjects' reactions to the treatment conditions immediately after they experienced the experimental treatments without measuring subjects' perception of the novelty of their assigned interface conditions. We believe that a longitudinal study is appropriate to test 94 Chapter 8 Conclusions and Discussion the changes in consumers' beliefs, attitudes, and behavioral intentions while they become familiar with VPEs. We are also concerned that subjects' demographic characteristics may affect the generablizability of the findings to other consumers. In this study, all subjects were recruited from the same university, and their backgrounds, in terms of gender, age, internet usage, prior online purchase experience, and familiarity with sports watches and PDAs, were quite homogeneous across the different experimental conditions. This homogeneity explains why subjects' background is not considered as a moderator in the research model. However, the background information may also arguably limit the generalizability of this study, because the experiment subjects are, on average, young and generally familiar with Internet usage and Internet purchasing, and thus may not be representative of the general population. It is likely that young people with high levels of Internet experience tend to like and explore emerging Internet technologies, such as V P E and other multimedia applications. Therefore, their reactions to the experimental treatments may be different from other consumer segments. The research model proposed investigates the topic from sellers' perspective, rather than consumers'. In other words, this research aims at understanding the functional mechanism of how VPE affects consumers' intentions to return to particular websites and intentions to shop on those websites. However, consumers' behavioral intentions are driven by consumers' beliefs and attitudes. Some beliefs may be of less interest to consumers than to online firms. For example, consumers try products in order to understand products, therefore are more interested their actual understanding of products than their perceived understanding. However, the construct perceived diagnosticity rather than actual diagnosticity is used in the research model because, from online sellers' perspective, it is consumers' beliefs such as perception of diagnosticity that influence their behavioral intention. Therefore, the results of the present study are most illuminating for online firms. 95 Chapter 8 Conclusions and Discussion 8.3 Future Research This study has categorized four types of VPEs and identified their corresponding product categories. For products with three dimensional appeal, VPE allows consumers to visually examine them three dimensionally; for products that form an environment, V P E enables consumers to enter and walk through the environment so as to examine the interior settings panoramically; for products that are associated with particular behavior or functionality, VPE allows consumers to try product behavior and functions; for products that need to be evaluated in relation to other products within specific contexts, VPE allows consumers to adjust or to customize product attributes or the context. Because the research models for this study were tested by using only the third type of VPE, in which consumers can sample product functions, the generalizability of the research findings is very limited and even problematic. Therefore, it would be interesting and meaningful to examine whether the proposed models are valid for other types of VPEs. Future studies should be conducted in this area. It would also be promising to incorporate intelligent agents into V P E design. Most current implementations of VPE aim at providing online consumers with only V P E simulations, while overlooking the importance of providing dynamic usage guidelines, which often results in that consumers sample product features too arbitrarily and not in a proper order thus weakening their capability to understand products thoroughly. In our experiment, textual explanatory information and guidelines were provided beside the products to facilitate subjects' use of the simulators. This method worked well as subjects could follow the usage procedures. However, the capability of such text-based guidelines was still limited due to their over-rigidness. For example, if subjects did not exactly follow the guidelines in the experiment, the simulators would react differently than what was stated in the guidelines, and then the subjects might get lost. Therefore, it is desirable to embed in a VPE application an intelligent agent that can 96 Chapter 8 Conclusions and Discussion automatically detect users' behavior and provide dynamic and contextualized guidelines. Furthermore, when consumers are trying a particular product feature on a simulator, they usually desire experts' advice on how to correctly evaluate that feature. This dynamic advising support is not supported by current V P E technologies, but is promising if intelligent agents are involved in the V P E design. Therefore, we recommend future work to propose guidelines on how to apply intelligent agent technologies to VPE. This study has compared V P E to video-based and image-based product demonstrations and has found that VPE, in general, is more beneficial in improving consumers' product understanding, shopping enjoyment, their attitudes toward websites and intentions to return to those websites. Future research could compare the effectiveness of VPE to direct product experience. On the one hand, in physical shopping environments consumers are exposed directly to products and therefore can acquire product information from multiple sensory channels, including physical feel and touch, which is not available in current VPE. On the other hand, when trying online products using VPE, consumers may be aroused by interactive and vivid web interfaces and therefore have more interest in products presented and the examination of the products. In fact, in our experiment some subjects who were shown the real products after they finished the main experiment reported that they would prefer the products presented on the websites to those real products. Therefore, it would be practically as well as theoretically interesting to explore and justify the potential differences between V P E and direct product experiences. 97 Appendix Appendix 1 Homepage of the Sports Watch Website (VPE Condition) IAAE Get more details herefrom our Hints and Highlights quick instruction hnnkiat 4 4 3 5 2 R u s h mMEmmm The Rush series of watches from Timex is grounded in technology with a sleek, streamlined design. This new product line of athletic wrisfrvatches is created exclusively for fitness enthusiasts. People of all ages have been quick to embrace Rush's thoughtful combination of useful functions and ease-of-operation. The "user friendly" approach allows people to focus their attention away from watch operations and onto more important things, like their workout. The Rush series has incorporated creative new features and eliminated unnecessary functions. (<^7iy out the watcIT^yi our full simulation. Features • Set Time User friendly instructions for setting current time and date • Timer 10pre-set timers from 1 minute to 1 hour with hallway turnaround alert, • Stopwatch Peeking to time of day and "shortcut' to Pulse mode while stopwatch is running. • Alarm Clpck Basic alarm clockfunctipns. • Pulse timer Allows quick calculation of heart rate based on the length of time required foryou to count ten "pulse beats." Great for a quick check during a workout! • Indiglo Niyln-Liglrt 3 seconds of Indiglo Night-Light each time it is pressed. Other Features • On-screen mes sages and prompts during set mode for time and alarm. • Bargraph progress indicator gives quick estimation of time. • All-day Indiglo • 3 seconds of Indiglo IMight-Light each time it is pressed. • Black resin strap • Yellow button accents • Get more details here from our Hints and Highlights quick instruction booklet. CAD $55.00 Free 1 st c lass shipping This item usually ships within 24 hours. 98 Appendix Appendix 2 Homepage of the PDA Website (VPE Condition) Products m515 HANDHELD Hand he ids • Accessories; Solutions » t h e fus i >n o f s t y l e a n d p o w e r , A stylish marriage of form and function, the Palm™ m515 handheld is perfect for today's tech-savvy professional. Whether you're managing contracts or your calendar, composing email or a Word-compatible doc, you'll get brilliant color, simplicity and sophistication in one ultra-slim design, Brilliant, adjustable color display. A new, brighter color display offers outstanding readability with more than 65,000 colors and an adjustable backlight for maximum viewing control. Plenty of room. With 16MB of internal memory, you can store hefty applications, PowerPoint presentations and even video clips with room to spare. Loaded with powerful programs. Bonus software lets you view Word, Excel and PowerPoint files right on your handheld, send email (ISP account and data-enabled phone or modem, not included, required), read eBooks, view video clips and photos, browse Web content offline and more. Expandable to grow with your needs. Plug in stamp-sized MultiMediaCard and SD expansion cards (sold separately) to instantly add software applications, additional memory, backup capabilities, eBooks, large databases and more without taking up built-in memory space. Or, use the Palm Universal Connector to add optional I-didn't-know-you-could-do-that peripherals such as a wireless LAN module, or a portable keyboard. Customizable with software. Visit Software Connection to browse the thousands of applications, in categories from business to games, available for immediate download. Many are free. *399cAD Product Details Size 4.5" x 3,1" x ,5" Weight 4.9 oz. Memory 16MB Battery Rechargeable Display 65,000+ colors Expansion Expansion Card Slot Universal Connector Palm OS© Version 4.1 HotSync® USB Cradle Wireless w/Accessories IClick on the link below to try our product simulator. c m515 Simulator Ins t ruc t i ons PDA Interface Off-screen Buttons Shortcut Keys Operating System Address Book Calculator Card Info Clock Date Book Expense HotSvnc Mail Memo Pad Mote Pad Preferences Security To-Do List Welcome Graffiti 99 Appendix 3 Sports Watch Simulator (Normal Mode) Appendix 100 Appendix Appendix 5 PDA Simulator (Normal Mode) Appendix 6 PDA Simulator (Date Book) 101 Appendix 7 Homepage of the Sports Watch Website (Video Conditions) TIME> Appendix 44352 Rush The Rush series of watches from Timex is grounded in technology with a sleek, streamlined design. This new product line of athletic wristwatches is created exclusively for fitness enthusiasts. People of all ages have been quickto embrace Rush's thoughtful combination of useful functions and ease-of-operation. The "user friendly" approach allows people to focus their attention away from watch operations and onto more important things, like their workout. The Rush series has incorporated creative new features and eliminated unnecessary functions. CjLe.itutes j V i d e o D e m o s t t a t i o n s j • S e t T i m e Userfriendly Instructions for setting current time and date • T i m e r 10 pre-set timers from 1 minute to 1 hour with halfway turnaround alert.. • S t o p w a t c h Peeking to time of day and "shortcut' to Pulse mode while stopwatch is running. • A l a r m C l o c k Basic alarm clock functions. • P u l s e T i m e r Allows quick calculation of heart rate based on the length of time required for you to count ten "pulse beats." Great for a quick check during a workout! • Imliiilo Nmlit-Liultt 3 seconds of Indiglo Night-Light each time it is pressed. Other Features • On-screen messages and prompts during set mode for time and alarm. • Bargraph progress indicator gives quick estimation of time. • All-day Indiglo. • Black resin strap • Yellow button accents. • 50-meter water resistant. • Get more details here from our Hints and Highlights quick instruction booklet. C A D $55.00 Free 1st class shipping This item usually ships within 24 hours. 102 Appendix 8 Homepage of the P D A Website (Video Conditions) Appendix Products ITl515 Handheld* t Accessories » Solutions *> HANDHELD sion o f s t y l e a n d p o w e r . A stylish marriage of form and function, the Palm™ m515 handheld is perfect for today's tech-savvy professional. Whether you're managing contracts or your calendar, composing email or a Word-compatible doc, you'll get brilliant color, simplicity and sophistication, in one ultra-slim design. Brilliant, adjustable color display. A new., brighter color display offers outstanding readability with more than 65,000 colors and an adjustable backlight for maximum viewing control. Plenty of room. With 16MB of internal memory, you can store hefty applications, PowerPoint presentations and even video clips with room to spare. Loaded with powerful programs. Bonus software lets you view Word, Excel and PowerPoint files right on your handheld, send email (ISP account and data-enabled phone or modem, not included, required), read eBooks, view video clips and photos, browse Web content offline and more, Expandable to grow with your needs. Plug in stamp-sized MultiMediaCard and SD expansion cards (sold separately) to instantly add software applications, additional memory, backup capabilities, eBooks, large databases and more without taking up built-in memory space, Or, use the Palm Universal Connector to add optional I-didn't-know-you-could-do-that peripherals such as a wireless LAN module, or a portable keyboard, Customizable with software. Visit Software Connection to browse the thousands of applications, in categories from business to games, available for immediate download, Many are free, Product Details Size 4,5" x 3.1" x ,5" Weight 4,9 oz. Memory 16MB Battery Rechargeable Display 65,000+ colors Expansion Expansion Card Slot Universal Connector Palm OS© Version 4,1 HotSync© USB Cradle Wis*e=less w/Accessories PDA Interface Off-screen Buttons Shortcut Kevs Operating System Address Book Calculator Card Info Clock Date Book Expense Graffiti HotSync Mail Memo Pad Note Pad Preferences Security To-Po .List Welcome 103 Appendix Appendix 9 Video Demonstration of Alarm Function of the Sports Watch (Without Narration) Alarm Clock Basic alarm clock functions Instructions Press M O D B P U L 5 E until you get to Alarm mode. Press START/STOP to turn on/oft the Alarm Press SET/CLEAR to begin setting. Press MODE.'PULSE to switch focus among hour, minute, second, Onvof, and AM/PM, Press START/STOP to increase hour, minute, or second, change on/off or AM/PM. Press SET/CLEAR to confirm the alarm setting. Appendix 10 Video Demonstration of Alarm Function of the Sports Watch (With Narration) Alarm Clock Baste alarm clock function*. 104 Appendix Appendix 11 Video Demonstration of Date Book Function of the PDA (Without Narration) Palm 515 - date-book - Microsoft Internet Explorer Write down personal schedule Set alarm for reminder To set a reminder for an event click on detail for more option. A window will popup and from there, you can change trie date and time of the event. Make sure to check the alarm checkbox if you want a reminder for that event Display weekly schedule Appendix 12 Video Demonstration of Date Book Function of the PDA (With Narration) Palm 515 - date book - Microsoft Internet Explorer Write down personal schedule Set alarm for reminder Display weekly schedule 105 Appendix Appendix 13 Homepage of the Sports Watch Website (Multiple-Static-Images Condition) IME: Get more details here from our Hints and Highlights quick instruction booklet. 4 4 3 5 2 Rush The Rush series of watches from Timex is grounded in technology with a sleek, streamlined design. This new product line of athletic wristwatches is created exclusively for fitness enthusiasts. People of all ages have been quick to embrace Rush's thoughtful combination of useful functions and ease-of-operation. The "user friendly" approach allows people to focus their attention away from watch operations and onto more important things, like theirworkout. The Rush series has incorporated creative new features and eliminated unnecessary functions. Creatures (Image Preview) • Set Time User friendly instructions for setting current time and date • Timer 10 pre-set timers from 1 minute to 1 hour with halfway turnaround alert. • Stopwatch Peeking to time of day and "shortcut' to Pulse mode while stopwatch is running. • Alarm Clock Basic alarm clock functions. • Pulse timer Allows quick calculation of heart rate based on the length of time required for you to count ten "pulse beats" Great for a quick check during a workout! • liuliulo Niijht Litjht 3secondsof Indiglo Night-Light each time it is pressed. Other Features • On-screen messages and prompts during set mode for time and alarm. • Bargraph progress indicator gives quick estimation of time. • All-day Indiglo. • Black resin strap. • Yellow button accents. • 50-meter water resistance. • Get more details here from our Hints and Highlights quick instruction booklet. C A D $55.00 Free 1st class shipping This item usually ships within 24 hours. 106 Appendix Appendix 14 Homepage of the PDA Website (Multiple-Static-Images Condition) m515 HANDHELD Handheld* > Accessaries Solutions Products the of style and power. A stylish marriage of form and function, the Pa lm ' " m515 handheld is perfect for today's tech-savvy professional. Whether you're managing contracts or your calendar, composing email or a Word-compatible doc, you'll get brilliant color, simplicity and sophisticatic in one ultra-slim design. Brilliant, adjustable color display. A new, brighter color display offers outstanding readability with more than 65,000 colors and an adjustable backlight for maximum viewing control. Plenty of room. With 16MB of internal memory, you can store hefty applications, PowerPoint presentations and even video clips with room to spare. Loaded with powerful programs. Bonus software lets you view Word, Excel and PowerPoint files right on your handheld, send email (ISP account and data-enabled phone or modern, not included, required), read eBooks, view video clips and photos, browse Web content offline and more. Expandable to grow with your needs. Plug in stamp-sized MultiMediaCard and SD expansion cards (sold separately) to instantly add software applications, additional memory, backup capabilities, eBooks, large databases and more without taking up built-in memory space. Or, use the Palm Universal Connector to add optional I-didn't-know-you-could-do-that peripherals such as a wireless LAM module, or a portable keyboard. Customizable with software. Visit Software Connection to browse the thousands of applications, in categories from business to games, available for immediate download. Many are free. *399CAD Product Details Size 4.5" x 3.1" x .5" Weight 4.9 oz. Memory 16MB Battery Rechargeable Display 65,000+ colors Expansion Expansion Card Slot Palm am HotSyncig Wireless Universal Connector Version 4.1 USB Cradle w/Accessories ICIIck on one of the inks below for image ^reviews. PDA Interface Off-screen Buttons SjToj^j_t_Ke_is Operating System Address Book Calculator Card Info Clock Date Book Expense Graffiti HsSSiflS. Mail Memo Pad Note Pad Preferences Security To-Do l ist Welcome 107 Appendix Appendix 15 Alarm Function of the Sports Watch (Multiple-Static-Images Condition) Using the Alarm Clock Instill ctions Prass MODE/PULSE until you get to Alarm mode. Press START/STOP to turn on/off the Alarm. Press S E T / C L E A R \ o begin setting. Press MODS/PULSE to switch focus among hour, minute, second, on/off, and AM/PM. Press START/STOP to increase hour, minute, or second, change on/off or AM/PM. Press SET/CLEAR to confirm the alarm setting. Appendix 16 Date Book Function of the PDA (Multiple-Static-Images Condition) I TIMS 5/20/0.3 5:00 pm^ 1 Lunch with Susan 6:00 pm 17 / (J i\ ill 1 ni.-ii 112:00 .. y { J 1 1 1 £08 1 4:00 1 M O Write tlown peisonal schedule • P r e s s the Date Book icon in the Menu page to launch the Date Book application. • The Date Book allows you to input your daily schedules and set an alarm for reminder. Set .ii.in,i for reminder To set a reminder for an event, cl ick on detail for more option. A window will popup and from there, you can change the date and time of the event. Make sure to check the alarm checkbox if you want a reminder for that event. Display weekly schedule View your schedule in a weekly display by cl icking the E button 108 Append ix 6. Have you bought any P D A by yoursel f? D Y e s D N o If not, do you intend to buy a P D A in the near future? D Y e s Q N o Section 3: Background on Internet Shopping 1. How often do you use the Internet each D A Y , on ave rage? • < 1 5 minutes Q 1 5 minutes ~ 1 hour D 1 ~ 2 hours D > 2 hours 2 . How long have you been using the Internet? • < 1 year Q 1 ~ 2 years Q2 ~ 4 years D > 4 years 3. I a m comfortable with us ing the Internet. 1 2 3 4 5 6 7 Not familiar at all Moderately familiar Very familiar 4. Have you ever made any purchase on the Internet? D Y e s D N o If s o , how much did you spend Internet shopp ing in the past 12 months? • zero rj<$100 • between $100 and $500 • between $500 and $1,000 • > $1,000 5. I am famil iar with Internet shopp ing . 1 2 3 4 5 6 7 Not familiar at all Moderately familiar Very familiar 112 Appendix Appendix 20 General Information Sheet for Subjects General Information Sheet In this study you will examine two products on the Internet, a sports watch and PDA (Personal Digital Assistant). For each product, please follow these steps: 1. The research assistant will first direct you to a website that shows the product information. 2. Examine the product on the website as if you were considering whether you would purchase this product or not. Read the product information available on the website. Note that you can explore more information by clicking on the hyperlinked text. 3. After you finish examining the product on that website, call the research assistant. She/he will then direct you to another website where the same product is presented in a different way. 4. Continue examining the product on this second website as if you were considering whether you would purchase this product or not. Read the product information available on the website. Note that you can explore more information by clicking on the hyperlinked text. After you finish examining the product on the second website, call the research assistant. She/he will give you a questionnaire to complete. Note that the purpose of showing you the first website is to provide you with a benchmark. However, you should answer all questions based on the second website. 5. Please answer carefully all questions based on your true feelings and beliefs. Your answers will be used for us to decide how to assign the $50 bonus (Note: it is guaranteed that one-fourth of all participants will receive the bonus). 6. After answering questions please call the research assistant. She/he will then give you a real product for you to examine. 7. Examine the real product as i f you were considering whether you would purchase this product or not. 8. After examining the real product, call the research assistant. She/he will give you another questionnaire. 9. Please answer all questions carefully. The above steps will be repeated for both products (the sports watch and the PDA). Please also note that before these steps there is a short (5 minutes) training session for each product. Two points that you should note: • A l l the brands and trademarks you will see in the website do not indicate any relationship between the companies and us. Therefore, your true feelings and thoughts are desired. 113 Appendix • During the test, please do all the activity as if you were seriously evaluating the products in a real purchasing mode. The research assistant is now going to help you through the training session. We appreciate your cooperation. 114 Appendix Appendix 21 Experimental Questionnaire The following questions ask you about your feelings and attitudes toward the web interface and your behavior intentions when you used the website. All questions are based on a 1-7 scale. For example, four persons a, b, c, and d have the following feelings toward the web interface they used: 1. Person a found the interface not attractive, and her feeing was very strong; 2. Person b found the interfaces neither attractive nor non-attractive, i.e. her feeling was neutral: 3. Person c found the interface not attractive, but her feeing was mild: 4. Person dfound the interface attractive and her feeling was fairly strong. Their responses are shown below: Q: The web interface is attractive. Strongly disagree © 1 Neutral © © t t © t Strongly agree 7 a c That is, person a would select 1, person b would select 4, person c would select 3, while person d would select 6. P l e a s e a n s w e r the f o l l o w i n g q u e s t i o n b a s e d o n the p r o d u c t d e m o n s t r a t i o n o n the w e b s i t e y o u ' v e just s e e n . 1. The product on this website looks colorful. 1 2 3 4 Strongly disagree Neutral 2. The product information on this website is clear. 1 2 3 4 5 6 Strongly disagree Neutral 3. I can acquire product information on this website from different sensory channels. 1 2 3 4 5 6 Strongly disagree Neutral 4. The product demonstration on this website is lively. 1 2 3 4 5 6 Strongly disagree Neutral Strongly agree Strongly agree Strongly agree Strongly agree 5. I am able to interact with this product. 1 2 3 Strongly disagree 4 Neutral 5 6 7 Strongly agree 6. The product information on this website is descriptive. 1 2 3 4 5 Strongly disagree Neutral 7. The product can respond to my input on this web interface. 1 2 3 4 5 Strongly disagree Neutral Appendix Strongly agree Strongly agree 8. The product information on this website is distinctive. 1 2 3 4 Strongly disagree Neutral 9. The product demonstration on this website is animated. 1 - 2 3 4 Strongly disagree Neutral 10. I can interact fully with this website. 1 2 3 Strongly disagree 4 Neutral 11. The product information on this website is rich. 1 2 3 4 Strongly disagree Neutral 12. This website contains product information exciting to senses. 1 2 3 4 Strongly disagree Neutral 13. The product information on this website is vague. 1 2 3 4 Strongly disagree Neutral 14. The product information on this website is concrete. 1 2 3 4 Strongly disagree Neutral 15.1 can acquire product information in an interactive way. 1 2 3 4 Strongly disagree Neutral 16. This website is helpful for me to evaluate the product. 1 2 3 4 Strongly disagree Neutral 7 Strongly agree Strongly agree Strongly agree Strongly agree Strongly agree 7 Strongly agree Strongly agree Strongly agree Strongly agree 116 Appendix 17. This web interface is helpful in familiarizing me with the product. 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree 18. Evaluating the product on this website is compatible with how I evaluate products in physical stores. 1 2 3 4 5 . 6 7 Strongly disagree Neutral Strongly agree 19. Evaluating the product on this website fits well with the way I like to evaluate products in physical stores. 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree 20. This web interface is helpful for me to understand the performance of the product. 1 2 ' 3 4 5 6 7 Strongly disagree Neutral Strongly agree 21. I am capable of judging this product. 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree 22. Familiarizing myself with the product on this website is similar to my product evaluation style in physical stores. 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree 23. I can manipulate the functions of this product online as if I were testing the real product. 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree 24. This website improves my online shopping performance. 1 2 3 4 5 6 7 Strongly disagree Neutral ' Strongly agree 25. This website improves my decision making in online shopping. 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree 26. I find this website easy to use. 1 2 . 3 4 5 6 7 Strongly disagree Neutral Strongly agree 117 Appendix 27. This website increases my online shopping effectiveness. 1 Strongly disagree 4 Neutral Strongly agree 28. I find this website useful. 1 2 Strongly disagree 4 Neutral Strongly agree 29. Learning to use this website is easy for me. 1 2 3 4 5 6 Strongly disagree Neutral 30. My interaction with this website is clear and understandable. 1 - 2 3 4 5 6 Strongly disagree Neutral 31. It would be easy for me to become skillful at using this website. 1 2 3 4 5 6 Strongly disagree Neutral 32. I like the product that I've just examined. 1 2 3 4 5 6 Strongly disagree Neutral 33. The product that I've just examined is good. 1 2 3 4 5 6 Strongly disagree Neutral 34. I have formed a favorable impression toward the product that I've just examined. 1 2 3 4 5 6 Strongly disagree Neutral 35. I like shopping on this website. 1 2 3 4 5 6 Strongly disagree Neutral 36. Shopping on this website is appealing. 1 2 3 4 5 6 Strongly disagree Neutral 37. Shopping on this website is a good idea. 1 2 3 4 5 6 Strongly disagree Neutral Strongly agree Strongly agree 7 Strongly agree Strongly agree 7 Strongly agree Strongly agree Strongly agree Strongly agree 7 Strongly agree 118 Appendix 38.1 have formed a favorable impression toward this website. 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree 39. It is likely that I will buy this product. 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree 40. I would like to revisit this website in the future. 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree 41. Next time I need to shop for a sports watch, I would like to use this website. 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree 42. I will purchase the product the next time I need a sports watch. 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree 43. Suppose that a friend calls me to get my advice in his/her search for a sports watch, I would recommend him/her to buy the product. 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree 44. I will definitely try this product. 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree 45. In general, I would characterize a decision of buying the product from this website as sound. 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree 46. If I buy this product from the website, I am not sure whether or not it can meet my expectations. 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree 47. I would use websites with similar characteristics to those of this website in the future. 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree 48. Next time I need to shop for a sports watch as a gift for a friend, I would like to use a website with characteristics similar to those of this website. 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree m Appendix 49. In general, I would characterize a decision of buying the product from this website as risky. 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree 50. I believe that I can buy the right product that I want from this website. 1 2 3 4 5 Strongly disagree Neutral 7 Strongly agree 51.1 find my experience with this website interesting. 1 2 3 4 Strongly disagree Neutral 6 7 Strongly agree 52. I find my experience with this website enjoyable. 1 2 3 4 Strongly disagree Neutral 6 7 Strongly agree 53. I find my experience with this website exciting. 1 2 3 4 Strongly disagree Neutral 6 7 Strongly agree 54. I find my experience with this website fun. 1 2 3 4 Strongly disagree Neutral Strongly agree 120 Appendix Appendix 22 Form for Research Assistant Participant Name: No: Group: Sequence: . Sports watch P D A 1. Assign subjects randomly to different conditions. 2. Fill in the webconsent form and then the background information form. 3. A brief introduction of the study. 4. Show subjects corresponding base conditions (Sports Watch Compass for 3 minutes, PDA Palm V for 5 minutes). 5. Train subjects with corresponding products and interface conditions. 6. Tell subjects that they will be directed to a website to demonstrate an experimental product and that they are expected to do as if they were shopping online, evaluating a product, and deciding whether or not to purchase the product. Time: Time: 7. Press 1-9 and show subjects experiment products in their assigned inieil.iee Time: Time: X. After subjects finish examining products, press F10. and sa\e the video file under C:/Jack/Screen Capture/. **• **• lJ. Before answering questions, show them the base condition again and restate that the base condition is a regular website, neither good or bad. lust use it as a BhNCIIMARK. Time: Time: 10. Then ask subjects to move a different desk and give them questionnaires to answer. Tell them that those objective questions are not going to be marked on an individual base, but on a group base, so there is no need to pay too much extra attention and efforts on remember product details. The key is to do what they would do in a real shopping mode. Time: Time: 11. Note down the time when subjects finish answering questions. Time: Time: 12. Show subjects real products with corresponding product specifications. Time: Time: 13. After they finish examining products, give subjects another page of questionnaire to answer. Time: Time: 14. After they finish answering questions, ask subject to move back to the computer desk. 15. Do the above activities from 3 - 1 4 again for the second product. Always remember to press F9 and F10 buttons to keep track of subjects' behavior. Time: Time: 16. Ask subjects to fill in two online open-end questions. 121 Appendix Time: Time: 17. Note down the time when subjects finish answering the open-end questions. 18. Pay them $15 and let them to sign the payment form. 122 References Reference: Alba, J., Lynch, J., Weitz, B., Janiszewski, C , Lutz, R., Sawyer, A. , and Wood, S. "Interactive Home Shopping: Consumer, Retailer, and Manufacturer Incentives to Participate in Electronic Marketplaces," Journal of Marketing (61), 1997, pp. 38-53. Ariely, D. "Controlling the Information Flow: Effects on Consumers' Decision Making and Preferences," Journal of Consumer Research (27), 2000, pp. 233-248. Babin, B.J., Darden, W.R., and Griffin, M . "Work and/or Fun: Measuring Hedonic and Utilitarian Shopping Value," The Journal of Consumer Research (20:4), 1994, pp. 644-656. Barclay, D., Thompson, R., and Higgins, C. "The Partial Least Squares (PLS) Approach to Cusal Modeling: Personal Computer Adoption and Use as an Illustration," Technology Studies (2:2), 1995, pp. 285-309. Benbasat, I., and Todd, P. "An Experimental Investigation of Interface Design Alternatives: Icon vs. Text and Direct Manipulation vs. Menus," International Journal of Man-Machine Studies (38:3), 1993, pp. 369-402. Bhatnagar, A. , Misra, S., and Rao, H.R. "On Risk, Convenience, and Internet Shopping Behavior," Communications of the ACM, (43:11), 2000, pp. 98 - 105. Bhatt, G. "Bringing Virtual Reality for Commercial Web Sites," International Journal of Human-Computer Studies (6), 2004, pp. 1-15. Biocca, F., Kim, J., and Choi, Y . "Visual Touch in Virtual Environments: An Exploratory Study of Presence, Multimodal Interfaces, and Cross-Modal Sensory Illusions," Presence (10:3), 2001, pp. 247-265. Burdea, G., Richard, P., and Coiffet, P. "Multimodal Virtual Reality: Input-Output Devices, Systems Integration, and Human Factors," International Journal of Human-Computer Interaction (8:1), 1996, pp. 5-24. Burke, R.R. "Technology and the Customer Interface: What Consumers Want in the Physical and Virtual Store," Journal of the Academy of Marketing Science (30:4), 2002, pp. 411-432. Chapman, P., Selvarajah, S., and Webster, J. "Engagement in Multimedia Training Systems," Proceedings of the 32nd Hawii International Conference on System Science, 1999. Chaudhury, A. , Mallick, D.N., and Rao, H.R. "Web Channels in E-Commerce," Communications of the ACM (44:1), 2001, pp. 99-104. Chin, W.W. "The partial least squares approach for structural equation modeling," In Modern methods for business research, G. A. Marcoulides (ed.) Lawrence Erlbaum, Mahwah, NJ, 1998, pp. 295-336. 123 References Chin, W.W., Marcolin, B.L. , and Newsted, P.R. "A Partial Least Squares Latent Variables Modeling Approach for Measuring Interaction Effects: Results from a Monte Carlo Simulation Study and an Electronic-Mail Emotion/Adoption Study," Information Systems Research (14:2), 2003, pp. 189-217. Chittaro, L. , and Ranon, R. "Virtual Reality Stores for 1-to-l E-commerce," Proceedings of the CHI2000,2000. Cognition and Technology Group at Vanderbilt Learning Technology Center "Examining the Cognitive Challenges and Pedagogical Opportunities of Integrated Media Systems: Toward a Research Agenda," Journal of Special Education Technology (XIL2), 1993, pp. 118-124. Coyle, J.R., and Thorson, E. "The Effects of Progressive Levels of Interactivity and Vividness in Web Marketing Sites," Journal of Advertising (30:3), 2001, pp. 65-77. Dabholkar, P.A., and Bagozzi, R.P. "An Attitudinal Model of Technology-Based Self-Service: Moderating Effects of Consumer Traits and Situational Factors," Journal of Consumer Research (30:3), 2002, pp. 184-201. Daft, R.L., and Lengel, R.H. "Organizational Information Requirements, Media Richness and Structural Design," Management Science (32:5), 1986, pp. 554-571. Daft, R.L., Lengel, R.H., and Trevino, L .K . "Message Equivocality, Media Selection, and Manager Performance: Implications for Information Systems," MIS Quarterly (11:3), 1987, pp. 355-367. Daugherty, T., L i , H. , and Biocca, F. "Experiential Ecommerce: A Summary of Research Investigating the Impact of Virtual Experience on Consumer Learning," In Online Consumer Psychology: Understanding and Influencing Consumer Behavior in the Virtual World, C. Haugtvedt, K. Machleit and R. Yalch (eds.), Lawrence Erlbaum Associates, Mahwah, NJ, forthcoming. Davis, F.D. "Perceived Usefulness, Perceived Ease of Use, and User Acceptance of Information Technology," MIS Quarterly (13:3), 1989, pp. 319-339. Davis, S.A., and Bostrom, R.P. "Training end users: An experimental investigation of the roles of the computer interface and training methods," MIS Quarterly (17:1), 1993, pp. 61-85. Dowling, G.R., and Staelin, R. "A Model of Perceived Risk and Intended Risk-handling Activity," Journal of Consumer Research (21:1), 1994, pp. 119-134. Eberts, R.E., and Bittianda, K.P. "Preferred Mental Models for Direct Manipulation and Command-Based Interfaces," International Journal of Man-Machine Studies (38:5), 1993, pp. 769-785. Edwards, S.M., and Gangadharbatla, H. "The Novelty of 3D Product Presentations Online," Journal of Interactive Advertising (2:1), 2001. 124 References Engel, J.F., and Blackwell, R.D. "Perceived Risk in Mail-Order and Retail Store Buying," Journal of Marketing Research (7:000003), 1970, pp. 364-369. Fazio, R.H., Zanna, M.P., and Cooper, J. "Direct experience and attitude-behavior consistency: An information processing analysis," Personality and Social Psychology Bulletin (4), 1990, pp. 48-52. Fishbein, M.a.A., I. Belief Attitude, Intention and Behavior: An introduction to Theory and Research, Addison-Wesley, Reading, MS, 1975. Fishkin, K.P., Gujar, A. , Harrison, B.L. , Moran, T.P., and Want, R. "Embodied user interfaces for really direct manipulation," Communications of the ACM (43:9), 2000, pp. 74 - 80. Galletta, D.F., Henry, R., McCoy, S., and Polak, P. "Web Site Delays: How Tolerate Are Users?" Journal of Association for Information Systems (5:1), 2004, pp. 1-28. Geissler, G., Zinkhan, G., and Watson, R.T. "Web Home Page Complexity and Communication Effectiveness," Journal of the Association for Information Systems (2:Article 2), 2001. Grazioli, S., and Jarvenpaa, S.L. "Perils of Internet Fraud: An Empirical Investigation of Deception and Trust with Experienced Internet Consumers," IEEE Transactions on Systems, Man, and Cybernetics (30:4), 2000, pp. 395-410. Ha, L. , and James, E.L. "Interactivity reexamined: A baseline analysis of early business Web sites.," Journal of Broadcasting & Electronic Media (42:4), 1998, pp. 456-473. Heeter, C. "Being There: The Subjective Experience of Presence," Presence: Teleoperators and Virtual Environments (1:2), 1992, pp. 262-271. Helson, H. Adaptation-Level Theory, Harper & Row, New York, 1964. Hirschman, E.C. "Innovativeness, Novelty Seeking, and Consumer Creativity," Journal of Consumer Research (7:3), 1980, pp. 283-295. Hoch, S.J., and Deighton, J. "Managing What Consumers Learn from Experience," Journal of Marketing (53:2), 1989, pp. 1-20. Hoffman, D.L., and Novak, T.P. "Marketing in hypermedia computer-mediated environments: Conceptual foundations.," Journal of Marketing, (60: 3), July 1996, pp. 50-68. Houston, J.M., Joiner, C.L., Uddo, F., Harper, C , and Stroll, A . "Computer Animation in Mock Juries' Decision Making," Psychological Reports (76), 1995, pp. 987-993. Hutchins, E.L., Hollan, J.D., and Norman, D.A. "Direct Manipulation Interface," In User Centered Design: New Perspectives on Human Computer Interaction, D. A. Norman and S. Draper (eds.), Lawrence Erbaum Associates, Hillsdale, NJ, 1986, pp. 87-124. Ives, B. , and Piccoli, G. "Custom Made Apparel and Individualized Service at Lands' End," Communications of the AIS (11), 2003, pp. 79-93. 125 References Jarvenpaa, S.L., and Todd, P.A. "Consumer Reactions to Electronic Shopping on the World Wide Web," International Journal of Electronic Commerce (1:2), 1996-97, pp. 59-88. Jarvenpaa, S.L., Tractinsky, N . , and Vitale, M . "Consumer Trust in an Internet Store," Information Technology and Management (1:1-2), 2000, pp. 45-71. Jeandrain, A. -C. "Consumer Reactions in a Realistic Virtual Shop : Influence on Buying Style," Journal of Interactive Advertising (2:1), 2001. Jiang, Z., and Benbasat, I. "Virtual Product Experience: Effects of Visual and Functional Control of Products on Perceived Diagnosticity and Flow in Electronic Shopping," Journal of Management Information Systems, forthcoming. Kelley, C.A., Gaidis, W.C., and Reingen, P.H. "The Use of Vivid Stimuli to Enhance Comprehension of the Content of Product Warning Messages," The Journal of Consumer Affairs (23:2), 1989, pp. 243-266. Kempf, D.S., and Smith, R.E. "Consumer Processing of Product Trial and the Influence of Prior Advertising: A structural Modeling Approach," Journal of Marketing Research (35), 1998, pp. 325-337. Kettanurak, V . N . , Ramamurthy, K., and Haseman, W.D. "User attitude as a mediator of learning performance improvement in an interactive multimedia environment: an empirical investigation of the degree of interactivity and learning styles," International Journal of Human-Computer Studies (54:4), 2001, pp. 541-583. Kim, J., and Allen, C T . "An investigation of the mediational mechanisms underlying attitudinal conditioning.," Journal of Marketing Research (33:3), 1996, pp. 318-328. Kinzie, M . "Requirements and benefits of effective interactive instruction: learner control, self-regulation, and continuing motivation," Educational technology, research and development (38:1), 1990, pp. 5-21. Kisielius, J., and Sternthal, B. "Detecting and Explaining Vividness Effects in Attitudinal Judgments," Journal of Marketing Research (XXI), 1984, pp. 54-64. Kisielius, J., and Sternthal, B. "Examining the Vividness Controversy: An Availability-Valence Interpretation," The Journal of Consumer Research (12:4), 1986, pp. 418-431. Klatzky, R.L., Lederman, S.J., and Matula, D.E. "Haptic Exploration in the Presence of Vision," Journal of Experimental Psychology: Human Perception and Performance. (19:4), 1993, pp. 726-743. Klein, L.R. "Evaluating, the Potential of Interactive Media through a New Lens: Search versus Experience Goods," Journal of Business Research (41:3), 1998, pp. 195-203. Klein, L.R. "Creating Virtual Product Experiences: the Role of Telepresence," Journal of Interactive Marketing (17:1), 2003, pp. 41-55. 126 References Kotler, P. Marketing Management, Prentice-Hall, Englewood Cliffs, NJ, 1998. 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. "Para-Social Presence and Communication Capabilities of a Web Site: A Theoretical Perspective," e-Service Journal (1:3), 2002, pp. 5-24. L i , H. , Daugherty, T., and Biocca, F. "Characteristics of Virtual Experience in Electronic Commerce: A Protocol Analysis," Journal of Interactive Marketing (15:3), 2001, pp. 13-30. L i , H. , Daugherty, T., and Biocca, F. "Impact of 3-D Advertising on product Knowledge, Brand Attitude, and Purchase Intention: The Mediating Role of Presence," Journal of Advertising (31:3), 2002, pp. 43-57. L i , H . , Daugherty, T., and Biocca, F. "The Role of Virtual Experience in Consumer Learning," Journal of Consumer Psychology (13:4), 2003, pp. 395-407. Lim, K . H . , and Benbasat, I. "The Effect of Multimedia on Perceived Equivocality and Perceived Usefulness of Information Systems," MIS Quarterly (24:3), 2000, pp. 449-471. Lim, K . H . , Benbasat, I., and Ward, L . M . "The Role of Multimedia in Changing First Impression Bias," Information Systems Research (22:2), 2000, pp. 449-471. MacKenzie, S.B., Lutz, R.J., and Belch, G.E. "The Role of Attitude Toward the Ad as a Mediator of Advertising Effectiveness: A Test of Competing Explanations," Journal of Marketing Research (JMR), (23:2,), 1986, pp. 130-143. Marks, L.J . , and Kamins, M . A . "The Use of Product Sampling and Advertising: Effects of Sequence of Exposure and Degree of Advertising Claim Exaggeration on Consumers' Belief Strength, Belief Confidence, and Attitudes," Journal of Marketing Research (25:3), 1988, pp. 266-281. Mass, Y . , and Herzberg, A. "VRCommerce - Electronic Commerce in Virtual Reality," Proceedings of the E-COMMERCE 99, Denver, Colorado, 1999, pp. 103-109. Mayer, R.E., and Moreno, R. "A Split-Attention Effect in Multimedia Learning: Evidence for Dual Processing Systems in Working Memory," Journal of Educational Psychology (90:2), 1998, pp. 312-230. Mayer, R.E., and Moreno, R. "Aids to Computer-Based Multimedia Learning," Learning and Instruction (12), 2002, pp. 107-119. McKinney, V. , Yoon, K., and Zahedi, F . M . "The Measurement of Web-Customer Satisfaction: An Expectation and Disconfirmation Approach," Information Systems Research (13:3), 2002, pp. 296-315. 127 References Miller, D.W., and Marks, L.J. "The Effects of Imagery-Evoking Radio Advertising Strategies on Affective Responses," Psychology and Marketing (14:4), 1997, pp. 337-360. Moore, G.C., and Benbasat, I. "Development of an Instrument to Measure the Perceptions of Adopting an Information Technology Innovation," Information Systems Research (2:3), 1991, pp. 192-222. Moreno, R., and Mayer, R.E. "Cognitive Principles of Multimedia Learning: The Role of Modality and Contiguity," Journal of Educational Psychology (91:2), 1999, pp. 358-368. Moreno, R., and Mayer, R.E. "Learning Science in Virtual Reality Multimedia Environments: Role of Methods and Media," Journal of Educational Psychology (94:3), 2002, pp. 598-610. Morgan, K., Morris, R.L., and Gibbs, S. "When Does a Mouse Become a Rat? Or.. .Comparing the Performance and Preference in Direct Manipulation and Command Line Environment," Computer Journal (34:3), 1991, pp. 265-271. Nelson, P. "Advertising as Information," Journal of Political Economy (83), 1974, pp. 729-754. Nisbett, R., and Ross, L. "Assigning weights to data the "vividness criterion"," In Human Inference: Strategies and shortcomings of social judgment, Prentice-Hall, Inc., Englewood Cliffs, New Jersey 07632, 1980. Norman, D.A. The invisible computer: why good products can fail, the personal computer is so complex, and information appliances are the solution, MIT Press, Cambridge, Mass., 1998. Nunnally, J. Psychometric theory, McGraw-Hill, New York, 1978. Palmer, J.W. "Web Site Usability, Design, and Performance Metrics," Information System Research (13:2), 2002, pp. 151-167. Peterson, R.A., Albaum, G., and Ridgway, N . M . "Consumers Who Buy From Direct Sales Companies," Journal of Retailing (65:2), 1989, pp. 273-286. Peterson, R.A., Balasubramanian, S., and Bronnenberg, B.J. "Exploring the Implications of the Internet for Consumer Marketing," Journal of the Academy of Marketing Science (25:4), 1997, pp. 329-346. Raney, A . A . , Arpan, L . M . , Pashupati, K., and Bril l , D.A. "At the Movies, On the Web: An Investigation of the Effects of Entertaining and Interactive Web Content on Site and Brand Evaluation," Journal of Interactive Marketing (17:4), 2003, pp. 38-53. Reeves, B. , and Nass, C. The Media Equation: How People Treat Computers, Televisions, and New Media Like Real People and Places, Cambridge University Press, Cambridge, 1996. Reeves, B. , and Nass, C. "Perceptual Bandwidth," Communications of the ACM (43:3), 2000, pp. 65-70. 128 References Richard, P., Birehent, G., Coiffet, P., Burdea, G., Gomez, D., and Langrana, N . "Effects of Frame Rate and Force Feedback on Virtual Object Manipulation," Presence (5:1), 1996, pp. 95-108. Rose, G., Khoo, H. , and Straub, D.W. "Current Technological Impediments To Business-To-Consumer Electronic Commerce," Communications of theAIS (l:Article 16), 1999. Ryan, C. "Virtual Reality in Marketing," Direct Marketing (63:12), 2001, pp. 57-62. Salinas, E.-L., Rassmus-Grohn, K., and Sjostrom, C. "Supporting Presence in Collaborative Environments by Haptic Force Feedback," ACM Transactions on Computer-Human Interaction (7:4), 2000, pp. 461-476. Shiv, B. , and Fedorikhin, A. "Heart and Mind in Conflict: The Interplay of Affect and Cognition in Consumer Decision Making," Journal of Consumer Research (26:3), 1999, pp. 278-292. Shneiderman, B. "Direct Manipulation: a Step Beyond Programming Languages," IEEE Computer), 1983, pp. 57-69. Smith, P.W., Feinberg, R.A., and Burns, D.J. "An examination of classical conditioning principles in an ecologically valid advertising context," Journal of Marketing Theory and Practice; Statesboro; (6:1), Winter 1998, pp. 63-72. Smith, R.E. "Integrating Information from Advertising and Trial: Processes and Effects on Consumer Response to Product Information," Journal of Marketing Research (30:2), 1993, pp. 204-219. Smith, R.E., and Swinyard, W.R. "Information Response Model: An Integrated Approach," Journal of Marketing (46), 1982, pp. 81-93. Smith, R.E., and Swinyard, W.R. "Attitude-Behavior Consistency: The Impact of Product Trial Versus Advertising," Journal of Marketing Research (20), 1983, pp. 257-267. Smith, R.E., and Swinyard, W.R. "Cognitive Response to Advertising and Trial: Belief Strength, Belief Confidence and Product Curiosity," Journal of Advertising, (Vol. 17:Issue 3,), 1988, pp. p3,12p. Steuer, J. "Defining Virtual Reality: Dimensions Determining Telepresence," Journal of Communication (42:4), 1992, pp. 73-93. Suh, K.-S., and Lee, Y . E . "Effects of Virtual Reality on Consumer Learning: An Empirical Investigation in Web-based Electronic Commerce," (working paper, Yonsei University), 2004. Summers, J.A. "Effect of Interactivity upon Student Achievement, Completion Intervals, and Affective Perceptions," J. Educational Technology Systems (19:1), 1990-91, pp. 53-57. Tan, S.J. "Strategies for Reducing Consumers' Risk Aversion in Internet Shopping," Journal of Consumer Marketing (16:2), 1999, pp. 163-180. 129 References Taylor, S.E., and Thompson, S.C. "Stalking the Elusive "Vividness" Effect," Psychological Review (89:2), 1982, pp. 155-181. Teo, H. , Oh, L . , Liu, C , and Wei, K. "An Empirical Study of the Effects of Interactivity on Web User Attitude," International Journal of Human-Computer Studies (58), 2003, pp. 281-305. Thompson, J.G., and Jorgensen, S. "How Interactive is instructional technology? Alternative models for looking at interactions between learners and media," Educational Technology), 1989. Todd, P., and Benbasat, I. "An Experimental Investigation of the Impact of Computer Based Decision Aids on Decision Making Strategies," Information Systems Research (2:2), 1991, pp. 87-115. Urban, G.L., Hauser, J.R., Quails, W.J., Weinberg, B.D., Bohlmann, J.D., and Chicos, R.A. "Information Acceleration: Validation and Lessons From the Field," Journal of Marketing Research ( X X X I V : 1), 1997, pp. 143-153. Urban, G.L., Weinberg, B.D., and Hauser, J.R. "Premarket Forecasting of Really-New Products," Journal of Marketing (60:1), 1996, pp. 47-60. Venkatesh, V . "Determinants of Perceived Ease of Use: Integrating Control, Intrinsic Motivation, and Emotion into the Technology Acceptance Model," Information Systems Research (11:4), 2000, pp. 342-365. Venkatesh, V. , and Davis, F.D. "A model of the antecedents of perceived ease of use: development and test," Decision Sciences (27:3), 1996, pp. 451-481. Walsh, K.R., and Pawlowski, S.D. "Virtual Reality: A Technology in Need of IS Research," Communications of the Association for Information Systems (8), 2002, pp. 297-313. Wann, J., and Mon-Williams, M . "What Does Virtual Reality Need?: Human Factors Issues in the Design of Three-Dimensional Computer Environments," International Journal of Human-Computer Studies (44), 1996, pp. 829-847. Webster, J., Trevino, L .K. , and Ryan, L . "The Dimensionality and Correlates of Flow in Human-Computer Interactions," Computers in Human Behavior (9), Winter 1993, pp. 411-426. Wolfinbarger, M . , and Gilly, M.C . "Shopping Online for Freedom, Control, and Fun.," California Management Review, (43:2), 2001, pp. 34-55. Wright, A . A . , and Lynch Jr., J.G. "Communications Effects of Advertising Versus Direct Experience When Both Search and Experience Attributes Are Present," Journal of Consumer Research (21:4), 1995, pp. 708-718. Yokokohji, Y . , Hollis, R.L., and Kanade, T. "WYSIWYF Display: A Visual/Haptic Interface to Virtual Environment," Presence (8:4), 1999, pp. 412-434. Yoshikawa, T., and Nagura, A . "A Touch/Force Display System for Haptic Interface," Presence (10:2), 2001, pp. 225-235. 130 References Zhang, P. "The Effects of Animation on Information Seeking Performance on the World Wide Web: Securing Attention or Interfering with Primary Tasks?," Journal of the Association for Information Systems (2: Article 1), 2000. 131 

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.831.1-0091783/manifest

Comment

Related Items