Virtual goods, real goals: Exploring means-end goal structures of consumers in social virtual worlds

Information & Management 51 (2014) 520–531

Contents lists available at ScienceDirect

Information & Management journal homepage: www.elsevier.com/locate/im

Virtual goods, real goals: Exploring means-end goal structures of consumers in social virtual worlds Yoonhyuk Jung *, Suzanne D. Pawlowski Ulsan National Institute of Science and Technology (UNIST), School of Business Administration, UNIST-gil 50, Ulsan 689-798, Republic of Korea

A R T I C L E I N F O

A B S T R A C T

Article history: Received 25 November 2012 Received in revised form 22 January 2014 Accepted 7 March 2014 Available online 29 March 2014

The aim of this study is to understand consumer goals for virtual consumption in social virtual worlds. Using a means-end chain approach, interviews with 93 users were analyzed to produce a hierarchical goal map, revealing that virtual consumption enables a diversity of goals beyond the previously identified experiential goals. The article’s theoretical contributions include: (1) extending our understanding of user goals/goal relationships for virtual activities and (2) developing virtual liminoid theory, which describes transitions between users’ real-world and virtual identities and virtual consumption as a transition catalyst. The study illustrates the value of a goal focus for future research on user behavior in virtual worlds. ß 2014 Elsevier B.V. All rights reserved.

Keywords: Virtual consumption Virtual goods Virtual worlds Goal structure Means-end chain Laddering interview Virtual liminoid theory

1. Introduction

With more of life lived online, spending on things that don’t exist seems more normal. . . Consuming things made of bits might sound weird, but actually it offers many of the same attractions that make people consume things made of atoms. (p. 28) - The New York Times Magazine [92] One of the major developments in the arena of social computing over the last decade has been the introduction of the traversable, three-dimensional cyberspaces called virtual worlds. The virtual world industry has experienced dramatic growth in the number of users, the establishment of new worlds and economic activity. By 2012, the number of virtual world users was 1.9 billion, three times the corresponding figure for 2009 [40]. As an example, Habbo Hotel, regarded as the largest teen virtual world, increased from 200 million users in 2010 to 270 million in 2012 [79]. The popularity of virtual worlds is also reflected in the explosion of new virtual worlds, such as Blue Mars, FarmVille, and Zynga, with the estimate that the number of virtual worlds would have reached more than

* Corresponding author. Tel.: +82 52 217 3122; fax: +82 52 217 3102. E-mail address: [email protected] (Y. Jung). http://dx.doi.org/10.1016/j.im.2014.03.002 0378-7206/ß 2014 Elsevier B.V. All rights reserved.

900 by 2013 [41]. Real-money trades, including user-to-provider and user-to-user transactions in virtual worlds, have similarly experienced dramatic growth. In the popular virtual world Second Life, total transactions among users for virtual goods reached $3.2 billion during the decade since its launch in 2003 [46]. In gaming virtual worlds (GVWs) (e.g., Everquest, World of Warcraft), user activities depend on pre-defined themes and plots imagined and produced by the game designers. Their activities typically have the aim of a quest or level-ups rather than social interactions with other users. In contrast, social virtual worlds (SVWs) (e.g., Second Life, Habbo Hotel) endow users with the ability to personalize their virtual experiences, which induce various social interactions, under minimum constraints. Virtual worlds have economic systems similar to real-world economic systems [17], and the minimum rules nature of SVWs leads to a wide range of economic activities. As SVWs support transactional systems (i.e., virtual currency, internal virtual markets) and allow users to create and own virtual property, users’ activities are extended to the production and real-money trading of virtual goods and property. Such a virtual economic system, replicating the real-world economy, is a favorite attraction of virtual worlds [17] and a compelling lure for SVW users due to their support of a laissez-faire economic system consisting of user-to-user transactions. The study reported in this paper focuses on an essential aspect of user behavior and experience in virtual worlds and a vital segment of the virtual economy - virtual consumption in social virtual worlds (SVWs).

Y. Jung, S.D. Pawlowski / Information & Management 51 (2014) 520–531

One of the predominant user behaviors in SVWs is the in-world purchasing of virtual goods, or virtual consumption. Virtual consumption is primarily supported by in-world real-money trading. The consumption of virtual goods is not limited to SVWs, but has become widely diffused in diverse online environments such as gaming virtual worlds, social networking sites, and social gaming. In 2012, virtual consumption was estimated at $2.9 billion in the US [62] and predicted to grow to $5 billion in 2016 [33]. Why do the users of Second Life and other SVWs, or residents as they are called, purchase rainstorms or the sounds of barking? Gaining insights into this question is key to our understanding of the user experience in SVWs. While the products consumed may be ‘virtual,’ the economic systems are not. Though often dubbed the ‘virtual economy,’ this is a real economy with links to other realworld economic systems. The research described in this paper examines the question of why users consume virtual goods in SVWs. This question is important not only to researchers interested in user behavior in cyberspace, but it is also fundamental to the creators and sellers of virtual goods and virtual world operators. By understanding the needs and goals that can be satisfied by purchasing virtual objects such as Tudor-style homes or a waterfall for a living room, we can also gain a greater understanding of the essence of the social virtual world experience from the user perspective. This understanding is crucial, as the metaverse of public virtual worlds continues to evolve, entrepreneurs are adopting virtual goods as a business model, and corporations are developing business strategies based on enterprise virtual worlds. To investigate consumer goals related to virtual consumption in SVWs, we employed means-end chain analysis, which is a method for eliciting individuals’ goal structures (i.e., goals and their relationships) for an object or event [55]. The goal-centric approach is a novel approach for exploring user behavior in cyberspace, providing a different perspective than previous studies based on a user-adoption framework (e.g., [1,25,48]). In addition, prior studies investigating user goals or motivations in cyberspace have produced sets of separate, individual user goals (e.g., [71,93,95]) but offer little explanation for the relationships among goals. This limitation may preclude the ability to develop a more comprehensive picture of user goals. The means-end chain analysis approach can be used to gain a more comprehensive understanding through eliciting and placing goals within a structure that includes a hierarchical system of interrelated goals [60]. The hierarchical goal map, which is an outcome of the analysis, provides richer information and offers more nuanced understandings of virtual consumption in social virtual worlds. We begin the paper with an overview of social virtual worlds and the virtual economy, followed by a description of the meansend chain approach, which provided the methodological framework for the study. Next, we describe the research method, including the laddering interview technique used to elicit user goals for virtual consumption, and an analysis of the data to generate the hierarchical goal map. The map is described in detail and interpreted in the results section that follows. Finally, the implications of goal-directed behavior in virtual consumption highlighted by the map are discussed, including preliminary implications for theory and practice. 2. Research background 2.1. Prior research on virtual consumption An important SVW behavior is ‘‘consuming things made of bits,’’ or virtual consumption. SVWs allow users to create and to sell custom virtual objects. Virtual goods production and consumption has created burgeoning economies within SVWs. Real money and

521

increasing transaction volumes have added to the complexity of and issues pertaining to SVWs such as security and taxation regulation concerns [12]. Despite its importance, there is a paucity of knowledge concerning various aspects of virtual consumption [48]. Some research on the virtual economy focuses on legal aspects and economic analysis. One issue engaging researchers in this area is whether real-money trading in these environments should be regarded as real transactions in the legal sense [42]. From an economic perspective, other researchers have performed analyses of virtual goods using concepts such as demand curves, price flexibility, or macroeconomic indicators [16]. While legal and economic research is valuable in understanding the overall environment of the virtual economy, it rarely reflects the perspective of the economic subject, that is, users’ understandings of the virtual economy. More recently, researchers have begun to pay attention to the users of SVWs. This body of research primarily examines factors affecting user adoption of SVWs. In these studies, utilitarian, hedonic, and social factors have been found to have a significant influence on adoption. Based on the technology acceptance model (TAM), Shen and Eder [77] demonstrated that perceived usefulness and enjoyment affect users’ intention to use Second Life for business. Verhagen et al. [90] found that both experiential factors (escapism and entertainment) and economic value play an important role in users’ satisfaction with an SVW. The importance of experiential factors in user adoption of SVWs is confirmed in Holsapple and Wu [31], who report a positive effect of fantasy and enjoyment on users’ acceptance of an SVW. Research also reveals that social factors affect user adoption of SVWs [20]. Finally, studies by Zhou et al. [95] and Jung and Kang [32] provide additional evidence that utilitarian, hedonic, and social factors are the primary motivations for SVW use. While research investigating users’ overall adoption behavior has been conducted, there has been little research on specific user behavior in SVWs such as virtual consumption. Given the growing importance of SVWs from a business perspective, it is important to look beyond adoption to explore and gain a deeper understanding of the varied types of user behavior in SVWs. Virtual consumption, which is our target behavior in this study, is a key behavior in that it is widespread in SVWs [1]. To date, very few studies investigate virtual consumption from the user’s perspective. Prior studies on virtual consumption are largely an extension of investigations of online shoppers, in which primarily utilitarian and hedonic factors are examined to understand shopper behavior. These studies have demonstrated that both utilitarian and hedonic factors are significantly influential in attitudes towards or intention to purchase online [13,18,39]. Animesh et al. [1] revealed that experiential factors (i.e., a sense of presence, flow) have a significant impact on users’ intention to purchase virtual items in an SVW. Another study by Guo and Barnes [25] explored purchase behavior in an SVW from the perspective of intrinsic and extrinsic motivations. In addition to utilitarian and hedonic factors, some studies reveal a significant influence of social factors on users’ consumption of virtual goods. Through scrutinizing virtual item attributes in virtual worlds, Lehdonvirta [43] classified key drivers of purchasing virtual goods into three types: functional, hedonic, and social. Similar results were reported in Ma¨ntyma¨ki and Salo [48], demonstrating that users’ perceptions of usefulness and enjoyment and network size (the degree of interaction with others in a virtual world) influenced virtual consumption. While providing initial insights into virtual consumption behavior, prior research regarding virtual consumption also has a significant limitation. While these studies have examined the primary factors affecting purchase intention in a confirmatory manner, they do not address a wide range of user goals for virtual consumption. More precisely, these studies examine factors based

522

Y. Jung, S.D. Pawlowski / Information & Management 51 (2014) 520–531

on users’ assessments of purchase environments but do not scrutinize the motivations or goals for the consumption of virtual goods. Research on consumption values supports there being a diversity of motivations for consumption. For example, Sheth et al. [78] classify consumption values into five categories (functional, social, emotional, epistemic, and conditional), and Sweeney and Soutar [81] suggest four value dimensions of consumption (performance, social, emotional, and value-for-money). Various motivations for virtual consumption are completely unexplored in previous studies. This deeper level of understanding is fundamental to our understanding of the drivers of the virtual economy. By employing a goal hierarchy approach, the current study contributes to this understanding by identifying diverse user goals for virtual consumption and elucidating the relationships among goals. 2.2. Goal hierarchy approach A goal hierarchy approach has been used extensively as a framework to understand humans’ decision making by relating their activities to their self-knowledge (e.g., goals, values) [19,55,63]. The approach draws upon theoretical perspectives from psychology, including personal construct theory [34], human values [72], and cognitive structure [69,76]. The central concept of the approach is the notion of a hierarchical goal system. A goal is a desired outcome of an action [47]. The view that goals exist within a hierarchical system is a common assumption in research on goals (e.g., [4,63]). A goal is located within the hierarchical structure between its superordinate and subordinate goals; furthermore, each goal is a means to achieve its superordinate goal [38,53,60]. The goal hierarchy approach posits that the attributes of activities or objects represent the means by which individuals achieve benefits and important personal values (i.e., ends) such as happiness and self-esteem [26,55]. The hierarchical structure of goals is widely accepted in various fields, such as marketing [3], organizational behavior [88], and hospitality [36]. In particular, recently some studies adopt a goal-hierarchy approach in exploring users’ goals in online realms, such as virtual worlds [32,45], social networking sites [57], and e-tailers [50]. In the following section, means-end chain analysis, which is one method of goal hierarchy approaches, is introduced. 3. Method 3.1. Overview of means-end chain analysis The aim of means-end chain analysis (MECA) is to understand the way in which the respondent perceives the world [61]. As an example, means-end chains for a sample of consumers revealed the following sequences of inferences regarding organic food: no chemicals ! avoid illness ! long life [24]. The traditional MECA approach assumes that consumer knowledge and sensemaking concerning a product or service is hierarchically organized into three levels of abstraction: attributes, consequences and values [67]. Attributes refer to the physical or observable properties of a product or service; consequences are the psychosocial benefits associated with experiencing or using the product or service; and values imply highly abstract motivations that guide usage behavior [36]. An attribute-consequence-value chain is typically expressed through a hierarchical map, which consists of nodes (i.e., attributes, consequences, and values) and the relationships among them. Although there are some variants, all means-end chain analyses essentially consist of three steps: laddering interviews, content analysis, and generating the hierarchical structure. MECA typically depends on a laddering interview technique to elicit hierarchical constructs. The laddering technique has been

extensively applied in consumer research to understand consumers’ preferences regarding products or services (e.g., [36,70,91]). To help respondents elicit lower or higher levels of abstraction for the concepts, researchers typically conduct the laddering interview by presenting three questions in the following order: the attribute question (What attribute makes the product (or service) attractive to you?), the consequence question (Why is that attribute important or desirable to you?), and the value question (Why is that (the subject’s response) important to you?). The laddering procedure can be conducted in the form of a paperpencil version (i.e., a laddering questionnaire) (e.g., [11,61]) or in an interview format. While the laddering questionnaire technique can be a cost-effective method for data collection, an advantage of the laddering interview method is that it enables the researcher to have in-depth interaction with respondents and ask questions to clarify understanding. The next step in MECA is content analysis, the aim of which is to aggregate and generalize the means-end data across respondents [21]. The majority of studies employing MECA employ a multicoder strategy to establish the reliability of coding responses. Two or more coders first work independently to classify responses into the three levels of abstractness (i.e., attribute, consequence, and value) and then classify the responses into a set of codes, which are primarily derived from words or phrases used by respondents during the interviews. Finally, the results of each coder are compared, and disagreements are resolved through discussion. For example, in Klenosky [36], responses were classified into the three levels of abstractness and assigned to content categories that the first coder developed inductively. A second coder then recoded the interviews using these categories; 75% of coding assignments reflected consensus, and disagreements were resolved via discussion. The final step of MECA is to summarize the means-end data and generate a hierarchical structure. The means-end data are summarized in an implication matrix, which includes information on how many times an element (content category) is used as a means and an end. The matrix is ultimately expressed through a hierarchal value map indicating ‘‘a graphical representation of means-end structures’’ ([23], p. 24). In generating a hierarchical value map, nodes are positioned based on their level of abstractness and linkages are made based on information on means-end chains in the matrix. There are two issues concerning the creation of a hierarchical map: (a) how to determine an element’s level of abstractness and (b) which linkages to include in the hierarchical map. All MECA approaches assume that knowledge is organized hierarchically on the basis of levels of abstraction [67]; however, approaches can vary with respect to assumptions regarding what the different levels represent. As explained above, the original approach assumes three levels of abstractness (i.e., attribute, consequence, and value). In an alternative MECA methodology employing network theory (see [75]), the abstractness of each element is computed and used to determine the position of the element in a hierarchical map rather than a strict specification of three levels of abstraction [3,15,61]. This method allows researchers to identify the relationships among elements without having to devote additional effort to classifying the elements into three pre-specified levels. It also addresses a critique of the traditional MECA approach, namely, where a subject’s response to an elicitation question frequently does not correspond to the level targeted by the question (e.g., a consequence or value response to the attribute question). Studies that employ network theory reduce this limitation because each element has a level according to its abstractness without needing to conform to a priori categories or a three-fold categorization of knowledge. Another advantage of this approach is that it is adaptable to a variety of contexts and objects and not limited to use with physical products

Y. Jung, S.D. Pawlowski / Information & Management 51 (2014) 520–531

523

[61]. In a hierarchical map based on this alternative MECA approach, elements (nodes) are positioned according to their individual values of abstractness. For the study presented in this paper, we employed the alternative, less directive adaptation of the MECA approach [3,61] as described above rather than the rigid attribute-consequencesvalues sequence of questions and analytical approach. Because of the exploratory nature of our study, we adopted an inductive methodological approach, specifically means-end chain analysis, to elicit user goals and determine their relationship structure. After introducing our participants in the next section, we explain the three stages of the MECA approach used in the study: laddering interviews, content analysis, and drawing a hierarchical goal map. Fig. 1. Screenshot of interview.

3.2. Study participation The target social virtual world for this study was Second Life, which is regarded as a leading SVW in terms of the number of registered users and technological advancement [58]. Second Life users have their own avatar, a visual representation and embodiment of themselves in the media interface. Second Life users may also elaborate a personal avatar’s face, hair, and body and clothe it. They can also create 3D objects (e.g., chairs, buildings, waterfalls) using basic shapes (e.g., squares, triangles, cubes) and pieces of code called script; additionally, they can perform virtual tasks with these items or sell them to other members. Via personal avatars, users can enjoy synchronous chatting at the park or on the beach, dancing at nightclubs, or taking a class. Three approaches were used to recruit participants. First, we sent the leaders of a wide diversity of groups emails asking them to distribute our message to their members to solicit participation in the study. To participate, subjects made appointments with us by email or Second Life IM (Instant Message). Second, we directly recruited logged-in group members via Second Life IM. Finally, we visited various public places in Second Life (e.g., beaches, parks, sandboxes) and directly solicited participation from users who were in those places. Ninety-three Second Life users participated in a laddering interview, and each participant received a five-dollar reward in the form of Linden dollars (the currency in Second Life) or PayPal, depending on their preference. Participants were heterogeneous in demographics (see Table 1). Approximately half of the participants were older than 34 years of age and had undergraduate or graduate degrees, indicating that people of all ages and

educational backgrounds use Second Life. Over 30% of the participants had a premium membership, which allows members to have virtual territory in Second Life. The majority of the participants can be regarded as highly attached users, in that approximately 80% of the participants logged into Second Life almost daily and 72% of the participants purchased virtual goods at least once a week. 3.3. Laddering interviews To capture a more detailed goal structure through interaction with respondents, we employed an interactive laddering interview rather than a laddering questionnaire. All interviews were performed in a comfortable location in Second Life where other avatars would not disturb the interview (see Fig. 1). The participants were first asked to complete a web survey including demographic and supplemental questions. This was followed by the interview, which was conducted by text chatting and took approximately 20 min on average. The interview consisted of three questions to elicit consumers’ goal structures. Based on the adapted laddering interview style described by Pieters et al. [61], we asked three open-ended questions to elicit users’ goals for virtual consumption: (1) What kinds of virtual goods or property have you bought in Second Life? Could you give me three examples?; (2) (first probing question) Why did you buy this specific item? or Why do you use this specific item in Second Life?; (3) (second probing question) Why is this reason important to you?

Table 1 Demographics of participants. Frequency

Percent

Age

18–24 25–34 35–44 45–54 55 or older No answer

20 26 22 17 7 1

21.5 28.0 23.7 18.3 7.5 1.1

Gender

Male Female No answer

39 52 2

41.9 55.9 2.2

High school Community college Undergraduate Graduate No answer

33 15 17 26 2

35.5 16.1 18.3 27.9 2.2

Free Premium No answer

64 29 0

68.8 31.2 0.0

Education

Type of account

Frequency

Percent

Tenure as a Second Life user

1 month or less 2 to 6 months 7 to 12 months 1 year to 2 years Longer than 2 years No answer

10 20 15 24 23 1

10.8 21.5 16.1 25.8 24.7 1.1

Login frequency

Once a month Once a week Several times a week Almost daily No answer

0 2 17 74 0

0.0 2.2 18.3 79.6 0.0

SL purchase frequency

Never Once a month Once a week Several times a week No answer

3 22 27 40 1

3.2 23.7 29.0 43.0 1.1

Y. Jung, S.D. Pawlowski / Information & Management 51 (2014) 520–531

524

Table 2 Cumulative number of new codes by time-ordered groups.

Number of new codes Cumulative

Group 1

Group 2

Group 3

Group 4

Group 5

Group 6

Group 7

Group 8

Group 9

First 10 samples

Next 10 samples

Next 10 samples

Next 10 samples

Next 10 samples

Next 10 samples

Next 10 samples

Next 10 samples

Next 13 samples

31 31

8 39

5 44

1 45

2 47

0 47

1 48

1 49

0 49

For example, an interviewee might state virtual clothing, virtual accessories, and a virtual house in response to the first question. Then, for each item, we asked the second question (e.g., Why did you buy virtual clothing?), followed by final question (e.g., Why is that reason important to you?). The responses to these questions, for example, might be ‘‘to look better’’ and ‘‘so that I can be popular.’’ Because this was a text-chat interview, we could capture and save the interview content. The interviewer halted an interview when he felt that an interviewee offered the same answers repeatedly in four or more questions, indicating that there might be no additional laddered concepts. 3.4. Content analysis The first step in the data analysis was to code the responses to the interview questions. One of the researchers coded the data using an open coding procedure in which codes were not predetermined but rather emerged from the data. In instances in which the response contained more than one subject, multiple codes were assigned. For example, the response ‘‘A house is for changing my avatar’s clothing and for having parties with friends’’ was assigned two codes – Privacy and Socialization. Forty-nine detailed codes were identified from the interview responses. The initial coding proceeded in parallel with the data collection process. To confirm theoretical saturation, indicating that no additional new codes appear during data collection [80], we checked the number of new codes over time (Table 2). Theoretical saturation assumes that the more data are collected, the fewer new codes emerge, and finally no new codes emerge; that is, saturation. Thus, over time, the cumulative number of new codes should not rise once the saturation point is reached. Our data confirmed this when 93 interviews had been completed. A second coder, the other researcher, independently re-coded the data using the set of codes identified by the first coder. The two raters were in agreement on 469 of the 545 codes assigned (Cohen’s Kappa = 0.83), indicating an acceptable level of inter-rater reliability [21]. Inter-rater disagreements were then reconciled through discussion. Finally, associated codes were grouped into 19 topics (content categories) through discussion between two raters, as shown in Table 3. 3.5. Generating the hierarchical goal structure The next step in the analysis was to generate the goal hierarchy for virtual consumption. The coded responses to the three interview questions generated a means-end chain, or a ladder of meanings; that is, responses to the first question regarding what virtual property was purchased in Second Life served as a starting point, and responses to the second question corresponded to the means of attaining the goal reflected in the response to the last probing question. For example, if the coded responses were Clothing (first question), Uniqueness (second question), and Selfexpression (last question), then the following hierarchical chain would be created: Clothing ! Uniqueness ! Self-expression. All relationships were summarized in an Implication Matrix (Table 4), which depicts the number of times each product/topic (content category) led to each other topic in the responses [36]. For

instance, T15 (Role-playing) led to T1 (Amusement) 9 times, to T4 (Creativity) 3 times, to T5 (Escapism) 5 times, etc. As stated previously, the current study employed an alternative method for analyzing laddering interviews based on network analysis and proposed by Bagozzi and Dabholkar [3] and Pieters et al. [61]. Instead of classifying responses into three pre-defined categories, this approach produces a hierarchical structure by comparing the number of times each topic is mentioned as a means versus an end. The approach uses out-degrees and in-degrees to estimate the abstractness level of each element. The out-degree of a particular topic refers to the number of times the topic serves as the source or origin (means) of linkages to other topics (i.e., the row sum of the element in an implication matrix); whereas the indegree of the topic indicates the number of times the topic serves as the object or end of linkages with other topics (i.e., the column sum of the element in an implication matrix) [61]. The abstractness of an element is the ratio of the in-degree divided by the in-degree plus out-degree of the element and ranges from 0 to l [61]. Elements with high abstractness scores are primarily regarded as ends, while those with low abstractness scores are primarily considered means. Based on this alternative approach, we calculated the abstractness score of each element (see Table 5). Additionally, for informative analysis, we calculated centrality of each element, which represents the degree to which the element occupies a central role in the structure [37]. Centrality is calculated by dividing the ratio of the in-degree plus out-degree of a particular element by the sum of all active cells in the implication matrix (sum = 281 in the current study). The next step was to generate the hierarchical goal map according to the information in the implication matrix. During this stage, the important point was to determine what linkages would be included in the hierarchical goal map. Because including all linkages can decrease a map’s usefulness and informativeness, we did not incorporate all linkages and instead decided to employ a cutoff level [68]. Researchers consider several cutoff levels and select the level that provides a balance between the complexity and interpretability of a map. As a guideline, one recommendation is that selected linkages should correspond to more than twothirds of all linkages in an implication matrix [23]. For instance, in Bagozzi and Dabholkar [3], a cutoff level of four (indicating that a linkage appears at least 4 times in an implication matrix) explained 62% of all relationships among elements; in Pieters et al. [61] a cutoff of four explained 72% of all connections. We constructed Table 5 to select the cutoff level. The selected cutoff level was four, indicating that only the relationships with values four or greater in the Implication Matrix were included in the map. This cutoff level represented 23.9% of the active cells and 51.2% of the active linkages, which corresponds to a measure of variance [23]. According to the cutoff criterion, T6 (Exploring) was excluded from the map because it had no linkages with a value of four or greater. The hierarchical goal map in Fig. 2 provides a graphical summary of the means-end structure pertinent to virtual consumption in a social virtual world. In the map, the topics are placed relative to their abstractness scores; the more abstract a topic, the higher it is located in the map. To avoid complexity in the map, linkages between virtual property and goals are depicted

Y. Jung, S.D. Pawlowski / Information & Management 51 (2014) 520–531

525

Table 3 Topics (Content categories). Topics

Definition

Examples

T1

Amusement

The positive emotional state of being amused

T2

Better avatar appearance

The positive appraisal of the visual appearance of one’s own avatar

T3

Business

T4

Creativity

T5

Escapism

Engaging in the exchange of goods and services for one another or money in a SVW The psychological state of transcending traditional ideas, patterns, relationships, or the like The psychological state of avoiding reality

 It just enhances my enjoyment of the SL experience.  I get pleasure out of looking at them (art–photos/sculpture) and some have inspired me.  To present a professional appearance.  I thought most of the human avatars looked really terrible; I could be a squirrel here. . .I thought it would be cute and better than other humans.  Manage and resell land in smaller pieces to other residents.  The building is a place for me to sell my goods.  Building stuff is the creative aspect of SL.  This is a virtual world where the imaginative part is the most important one.

T6

Exploring (removed)

Travel in or through unfamiliar parts of a SVW

T7

Fantasy

T8

Knowledge acquisition

Expression of what the user wants to happen, especially one that is unlikely to happen in the real world Gaining information and skills, particularly knowledge relevant in a SVW

T9

Like shopping

T10

My own place

T11 T12

Projection of real-world convention/activity Decorative activity

T13

Realistic experience

T14

Relaxing

Make less tense and anxious

T15

Role-playing

Creation of a persona with a background story and interaction with others

T16

Self-esteem

Confidence and satisfaction with one’s self

T17

Self-expression

T18

Socialization

Expression of one’s own personality, feelings and ideas Association with or in the company of others in a SVW

T19

Uniqueness

A fondness or preference for browsing and/or purchasing goods The perception of having virtual spaces that are under one’s control Simulation of real-world conventions/activities To beautify virtual environments with ornamental objects Perception of things in the SVW as similar to the way that they are in the real world

The trait of pursuing differences relative to others

separately in the lower part of Fig. 2. The goal map is described more fully in Section 4. 4. Results The hierarchical goal map in Fig. 2 clearly reveals what goals users seek through virtual consumption and how these goals are related. The map indicates that Amusement, Decorative activity, Selfexpression, and Socialization are the predominant goals for virtual consumption in this SVW measured by the centrality of the topics. These four goals also correspond to 49.3 percent of all linkages in the implication matrix (Table 4). Avatar clothing/accessories/ animation and Virtual estate/furniture jointly correspond to 85% of virtual goods mentioned in the implication matrix. They are also important means, connected to 17 of the 19 goals. The key findings are summarized below. Decorative activity, which is supported by three categories of virtual property (Building tools/materials, Avatar clothing/accessories/animation, and Virtual real estate/furniture), leads to several of

I forget about my real life issues. When I am in SL, I let go of real life and truly get into the experience. I sometimes take friends on tours of Second Life on the ship. I can just hop on a motorcycle and drive around. My hair in real life is more unmanageable than I care for. I can take off this constraint in SL.  In SL I have the freedom to do so; in real life I do not.  It is useful for practicing skills like building.  Our sim members share common interests and they are very helpful. Advice on SL problems (like how to do something, or where to go to find something, or where a party is going to be, or where a contest is).  I like to shop.  I love walking into a store and everything fits.  Just space and a bit of privacy.  I got tired of hanging out in clubs and public places and wanted a place to go and be able to do what I wanted and think and to have a place for me.  In real life we wear clothing. We treat SL similar to real life on many levels.  It is a spot that I and my real-world boyfriend can meet virtually.  I enjoy beauty, I like to create avatars.  Enjoy interior decorating.  The clothes here seem very realistic.  To me SL is not a game and it really is a life. It gives me a sense of being settled down.  Sense of peace.  A place to hang out and relax. . .some of my friends and I used to get together and watch movies here.  I need them to fight when I role-play.  Since I lived in Caledon (a sim), I needed nice Victorian/steam punk style clothes.  It does give you more attention.  Someone asks me where I got my hair, and then It just makes me feel good and I get complimented.  I’ve spent a lot of time customizing it.  This is an inner look to my inner personality what I want to look like.  Spend time with friends and make parties.  People tend not to interact as quickly with someone in Linden-supplied clothing the word noob (newbie) would be best to describe what they would call them.  Makes it more personal, unique.  It makes me stand out amid an ocean of other avatars.     

the higher-level goals. First, Decorative activity is a primary vehicle to achieve two central goals of virtual consumption: Amusement and Self-expression. The strong association with Amusement implies that decorating and creating virtual objects is highly intrinsically motivated (e.g., ‘‘See how your look turns out. Play with the many features you can tamper with’’; ‘‘Decorating the reef. . .because it looks lovely’’) and the consumption of virtual goods can be regarded as a ‘‘playful consumption’’ experience ([29], p. 728). This finding is in line with hedonic consumption theory [14,28], which has been widely used in investigations of consumers of video or online games (e.g., [29,49]). Decorative activity also contributes to Self-expression that, in turn, supports Amusement. Virtual goods for Decorative activity enable users to present themselves in certain ways (e.g., ‘‘It gives others a sense of your taste and who you are’’; ‘‘To show off’’). Virtual consumption contributes to the generation of a self through the images and styles conveyed by virtual property, primarily through Avatar clothing/accessories/animation and Virtual real estate/furniture. Avatars themselves can be regarded as an extended form of self

Y. Jung, S.D. Pawlowski / Information & Management 51 (2014) 520–531

526 Table 4 Implication matrix. Topics P1. Avatar clothing/ accessories/animation P2. Building tools/materials P3. Gadgets/vehicles/ musical tools P4. Virtual real estate/ furniture P5. Weapons T1. Amusement T2. Better avatar appearance T3. Business T4. Creativity T5. Escapism T6. Exploring T7. Fantasy T8. Knowledge acquisition T9. Liking shopping T10. My own place T11. Projection of RL conventions/activity T12. Decorative activity T13. Realistic experience T14. Relaxing T15. Role-playing T16. Self-esteem T17. Self-expression T18. Socialization T19. Uniqueness In-degrees Abstractness Centrality

T1

T2

5

48

T3

T4

T5

T6

T7

T8

9

2 1

2 3

3

6

T9

T10

8

T11

T12

T13

T14

T15

T16

T17

T18

T19

Outdegrees

4

18

9

1

17

14

34

13

20

200

19 2

4

1

1 1

3

25 23

25

3

3

5

16

92

1

5

8

2 4

1 3 1

2

1

1 5 2

2

1

3

1

14

2

8

1 4

3 2

2 5

2

3 2

1

1 1

2

1 2

3

1

1

2

2

2

1

1 4 2

1 1

9 4 1 9 1 6 6

6

2

5

2 3 1 1 1

52 0.91 0.20

2 5 1 2

3

2

6 2 1 4 3 1 1

2

1

4 1

1 4

2 1

5

1

2 5

1 5

1 1 1

2

1 1

3

4 11 10 15 0 28 11 2 0.11 0.65 1.00 1.00 0.00 0.68 0.92 0.29 0.14 0.06 0.04 0.05 0.01 0.15 0.04 0.02

2

3 16 0.12 0.89 0.09 0.06

1 1 1

3 2 2

1 4 3

5 12 22 0.09 0.39 0.67 0.21 0.11 0.12

13 2 1 1 2

2 2 7

3 1 4 4 4 2 3

4 13 34 33 0.11 0.48 0.56 0.60 0.13 0.10 0.22 0.21

1

4

5 34 6 0 0 3 13 1 5 23 2 52 19 11 32 14 27 22 12

6 281 0.33 0.06

Out-degree: the number of times the topic serves as the source or origin (means) of linkages with other topics. In-degree: the number of times the topic serves as the object or end of linkages with other topics. Abstractness = (In-degrees)/(In-degrees + Out-degrees). Centrality = (In-degree + Out-degree)/the sum of all active cells. We regard all virtual property as initial origins; thus, virtual property has no in-degree value.

because the user’s avatar is regarded as the manifested virtual self in SVWs, and virtual consumption can provide resources for selfexpression (‘‘To show how one feels inside, a way to show the inner self. . .I choose to be a rabbit here’’). For example, in their study of the link between individuals and their avatars in virtual worlds in terms of identity, Parmentier and Rolland [58] found that virtual consumption supported four types of identity positioning in SVWs: (1) duplication, (2) improvement, (3) transformation, and (4) metamorphosis. Self-expression extends beyond the presentation of self through an avatar’s appearance to virtual possessions. Just as material possessions can be an extension of the self and displaying them affirms their owner’s identity [8,83], users in SVWs can exploit virtual possessions in a similar manner (‘‘A place to display photos. . .it is personalized’’). This implies, for example, that virtual houses acquire a ‘‘second skin’’ status in SVWs, as do houses in the real world [7]. Uniqueness (e.g., ‘‘You can look different from everyone else’’) is also strongly linked to the higher-level goal of Self-expression. This

relationship reflects the view that the self consists of the individual and the affiliative [35], and the expression of individual uniqueness can be one way to express the self. Real-world consumption is occasionally motivated by the human desire to be distinguished from the crowd [89]. Without distinctive clothing or accessories, an individual’s avatar will appear similar to many other avatars in a SVW. Consumers’ need for uniqueness is a key construct in consumption theory, defined as ‘‘the trait of pursuing differences relative to others through the acquisition, utilization, and disposition of consumer goods for the purposes of developing and enhancing one’s self-image and social image’’ ([84], p. 52). Another important aspect of this theoretical perspective is the dialectical tension between the need for uniqueness and the need for social inclusion [74]. That tension, however, may be relaxed to a large extent in SVWs (e.g., ‘‘Unlike real life you can be as unique as you want to be, without censure or shame’’). Decorative activity in SVWs occasionally reflects real-world conventions, activities or relationships (Projection of real-world

Table 5 Statistics for determining a cutoff value. CutoffNumber of active cells in thePercentage of active cells at or aboveNumber of active linkages in the implication matrixPercentage of active linkages at or above level the cutoff level (%) the cutoff level (%) implication matrix 1 2 3 4 5

153 101 65 47 33

100.0 66.0 42.9 31.4 21.6

623 572 500 446 387

100.0 91.8 80.3 71.6 62.2

Y. Jung, S.D. Pawlowski / Information & Management 51 (2014) 520–531

527

Fig. 2. Hierarchical goal map for virtual consumption.

convention/activities), and virtual goods can be used to create or support these aspects of the real world. For example, one user commented that she decorates her SVW house with furniture and sculptures because an empty virtual house depresses people, as does an undecorated house in the real world; another user mentioned that she dresses up to meet her real-world boyfriend, who lives far away from her in the real world. Additionally, some users build virtual objects for sale and decorate their stores to better attract users (Business); users’ decorative activities stimulate their Creativity (e.g., ‘‘It allows me to be creative’’) and enact Fantasy experiences (e.g., ‘‘I used to be a jazz singer; I am now an invalid, in a wheelchair and can’t really leave my house much. SL (Second Life) gives me the opportunity to live a virtual life that would be my real life if I could’’). Users’ expressions of their creativity and their efforts to realize their dreams are critical to the evolution of SVWs because they are cyberspaces imagined and created by their users [10]. Role-playing is the creation of a persona with a background story and interaction with others [93]. As indicated in the goal map, role-playing is not only another stimulus for the consumption of virtual property, particularly Weapons and Avatar clothing/accessories/animations, but is also a means to various ends including Socialization, Knowledge acquisition, Amusement, Fantasy, and Escapism. Role-playing provides users with opportunities for meeting different sorts of individuals (Socialization) and can contribute to learning (Knowledge acquisition) (e.g., ‘‘I have met some wonderful people who have the same interests, and we learn about our past history’’). Role-playing has an essential part in users’ hedonic consumption of virtual goods. In addition to enjoyment, hedonic consumption can be driven by role projection, fantasy, and escapism [28]. Following the rules of a role-playing community

provides users with challenging assignments that can produce enjoyment. Additionally, Role-playing allows users’ imaginings to be realized in the virtual world (Fantasy) (e.g., ‘‘It is my fantasy world’’) and temporarily takes them away from real-world problems (Escapism) (e.g., ‘‘Soothes me. I forget about my real life issues’’). Users may accomplish certain virtual consumption goals, such as Fantasy, Self-esteem, and Self-expression, via a Better avatar appearance. First, Better avatar appearance can also be a means for realizing dreams, for example to resemble a famous movie star. Second, Better avatar appearance supports Self-esteem, which leads to Socialization. When users feel that their avatars are visually appealing, they develop self-confidence that ultimately encourages social interaction or Socialization (e.g., ‘‘I get to have a pretty avatar and the satisfaction in knowing that I chose the selections, and that others think my avatar is pretty so that means I have reasonably good taste’’). Several interview participants stated that they recognize ‘newbies’ (i.e., newcomers) in SVWs based on avatar appearance and avoid interaction with them (e.g., ‘‘Newbies basically get discriminated against. It is like social class status’’). This understanding implies that Better avatar appearance may function as a ritual separating us from them [82] and is a means for social inclusion, which situates the self within the world [73]. Moreover, Better avatar appearance can also be a means for expressing the self. Virtual consumption helps users to generate a unique and desired self through their avatars’ appearance. Accordingly, Better avatar appearance enables users to establish us (i.e., a social identity) and is simultaneously a means to demonstrate me (i.e., a personal identity). While users strive to be socially included through socially acceptable avatar appearance, they make their personal identity salient through desired self-

528

Y. Jung, S.D. Pawlowski / Information & Management 51 (2014) 520–531

images. This finding confirms the argument that a person regulates the self between social identity and personal identity [35,65]. Virtual consumption reflects this dynamic nature of self-regulation. In SVWs, virtual property, particularly Avatar clothing/ accessories/animations for Better avatar appearance, is used to portray users’ simultaneous desire to be a part of the world and for independence. The tension between social virtual identity and personal virtual identity is one of the underlying drivers of virtual consumption. Virtual real estate offers My own place that an owner can control for Socialization and Relaxation. Popular locations where numerous avatars meet not only generally suffer from a lag problem, such as slow-motion avatars, but are also where a user may encounter hooligans. Having control over virtual real estate can serve as an alternative to solve these problems (e.g., ‘‘I got tired of hanging out in clubs and public places and wanted a place to go to and be able to do what I wanted and think and have a place for me and my partner to be together’’). It can also provide a place where users can hold social events such as parties. Additionally, users can make their own virtual real estate a comfortable place for Relaxing. Because in many cases users relax with their friends on virtual real estate, Relaxing has an association with Socialization (e.g., ‘‘A place to hang out and relax. . .Some of my friends and I used to get together and watch movies here in SL’’). 5. Discussion While the findings of this study resonate with those of prior studies exploring users’ general adoption or consumption of virtual goods, the findings provide a much more detailed picture of user goals for virtual consumption in social virtual worlds, including the relationships among those goals. As the first comprehensive investigation of the motivations for virtual consumption, the most important value of this study is its informative merits and the foundational knowledge it provides for the development of theory on virtual consumption. The study also serves as a reminder to researchers of the importance of the topic by confirming that virtual consumption is associated with a wide range of activities and values in SVWs. Although the underlying technologies will continue to evolve and new SVWs will emerge in the future, it is not unreasonable to posit that users will continue to seek these same types of goals in these worlds. In the research exploring users’ participation in virtual worlds, Eisenbeiss et al. [20] found that virtual world users are primarily motivated by socializing, creativity, and escape. Those motivations overlap with the goals of virtual consumption revealed in our findings from the present study, with the implication that virtual consumption is an essential means for users to fulfill their objectives for participation in virtual worlds. Accordingly, researchers attempting to understand the behavior of SVW users need to bear in mind that virtual consumption is an essential component of users’ virtual lives. Underestimating virtual consumption as merely a set of peripheral activities could omit an important element in understanding user behavior in the virtual realm. 5.1. Implications for theory The main theoretical implication of this study is to propose potential theoretical frameworks for future research on virtual consumption and user adoption research. As discussed further below, specific contributions include: (1) an elucidation of the experiential aspect of virtual consumption, (2) an exploration of users’ liminal transitions between real-world and virtual identities and the development of virtual liminoid theory, and (3) proposing the means-end chain approach (i.e., user-goal oriented) as an alternative for adoption research.

In their seminal paper, Holbrook and Hirschman [30] advocated consumer behavior theory that moved beyond the ‘‘information processing model,’’ in which consumers logically process information on products in an attempt to make rational decisions and maximize utility, to the ‘‘experiential view’’ of consumption: ‘‘Consumption has begun to be seen as involving a steady flow of fantasies, feelings, and fun encompassed by what we call the ‘‘experiential view.’’ This experiential perspective is phenomenological in spirit and regards consumption as a primarily subjective state of consciousness with a variety of symbolic meanings, hedonic responses, and esthetic criteria.’’ (p. 132) Experiential values are an essential aspect of virtual consumption [48]. Our findings confirm that diverse experiential values (e.g., amusement, fantasy, decorating) are the primary purpose of virtual consumption. Furthermore, the findings reveal that these values are realized through various activities involving virtual items, that is, engaging in different practices to obtain a variety of experiential values. Based on these findings and given the nonmaterial nature of virtual goods, the stream of research based on the phenomenological approach using the experiential lens, with its emphasis on fantasy and the emotive facets of consumption, is particularly appropriate in the context of SVWs. Thus, an intensive investigation (e.g., ethnographic or qualitative approaches) of user fantasies and aesthetic behavior will provide a better understanding of virtual consumption. The findings of this study suggest another theoretical framework that may prove useful in understanding virtual consumption: the notion of liminal transitions and their relationship with symbolic consumption developed by Noble and Walker [54].1 This model of symbolic consumption activities and psychological phenomena integrates van Gennep’s [87] framework of ‘‘liminal transitions’’ with Belk’s [7] notion of the ‘‘extended self’’ and strongly resonates with the findings of the study. Liminality indicates the ambiguity, instability and suspended identity that can occur during transitions from one significant role to another [87]. During liminal periods, individuals stand at the threshold between their prior and new status, and as a result, their identities are characterized by ambiguity [85]. Belk’s [7] notion of the extended self provides insights into how individuals use possessions to cope with and expedite passages through liminal periods. Possessions help to define identity, and studies have shown how the extended self is manipulated through symbolic consumption during liminal periods to help consumers complete the transition to a new role and redefined self [9,51]. The Liminal Transition Model proposed by Noble and Walker [54] describes how, during liminal periods, individuals may place high value on two types of transitional objects: (a) possessions symbolizing past relationships and (b) possessions symbolizing the new role to which they are transitioning. Both types of symbolic objects can ease the psychological discomfort of liminal states. Applying Noble and Walker’s [54] observations on symbolic consumption during liminal periods to virtual consumption in SVWs offers interesting insights into the user experience and the role that virtual consumption may play for certain users. In the virtual world context, as users log in to SVWs and their avatars represent them, an environment for liminality is spatially and temporally formed. Users decorate their virtual houses and maintain their avatars’ appearance by consuming virtual goods. In so doing, they experience a transition or liminal state during which real-world rules and status are dissolved and virtual-world conventions take their place. Users create their virtual identities based on their real-world identities or oftentimes generate quite 1 Our thanks to the anonymous reviewer who suggested liminality as a useful concept to interpret the study’s findings.

Y. Jung, S.D. Pawlowski / Information & Management 51 (2014) 520–531

different ones, including ‘trying out’ future roles they may desire. In other words, their liminal personae are often a combination of their actual and imagined status. Users then enact their virtual identities and perform activities with them in SVWs using virtual objects. However, when users log off of SVWs, their virtual identities are separated, they lose access to their virtual possessions and they return to real-world status. Therefore, we cannot say that users’ new (virtual) identities are re-incorporated into themselves, that is, this is an incomplete reincorporation. Turner [86] proposed the concept of liminoid to denote piecemeal transition experiences that do not lead to fundamental identity changes, and this concept may be more appropriate in describing liminality in the SVW context. While the liminal process can lead to a significant shift in prior status through various social or religious rituals, the liminoid state indicates transient change experiences through play and entertainment (e.g., movies, festivals). Consequently, we propose a virtual liminoid theory, arguing that SVWs are spheres for liminoid activities and virtual consumption facilitates those activities in these worlds. More specifically, we posit that: (1) SVWs offer environments appropriate to liminoid activities by separating users from real-world space and time, (2) users’ real-world status are transiently dissolved and transformed into virtual identities, primarily by consuming virtual goods, and (3) their virtual identities are not incorporated into their actual status. Through decorating their virtual houses and avatars, having fantasies and role-playing, users have liminoid experiences in SVWs. Virtual goods serve as facilitators to support liminoid experiences in these virtual domains. Based on the insights of Noble and Walker [54] regarding symbolic consumption during liminal transitions, we also posit that SVW users may also place high value on virtual objects during partial liminoid transitions that symbolize their real-world identities and virtual possessions symbolizing the new role they are enacting in the virtual world. Virtual liminoid theory represents a good starting point to explain the user experience in SVWs, including the virtual consumption of symbolic objects that represent a combination of a user’s realworld identity and the new roles expressed by their virtual persona. It is possible that these symbolic objects may be used to ease the psychological discomfort of liminoid states as users ‘try on’ new identities. Third, elements of hierarchical goal theory that provide the theoretical foundation for MECA can also be drawn upon to advance our understanding of user goals related to technology adoption. The hierarchical structure of the goal map not only identifies means-end chain relationships, but it also allows us to identify proximal, intermediate and distal goals. Proximal goals (instrumental goals, at lower levels in means-end chains) regulate immediate motivation and action, while distal, or high-order, goals define more abstract, enduring aspirations (e.g., avatar clothing (proximal) ! uniqueness (intermediate) ! self-expression (distal)) [5,19]. Each type of goal drives motivated behavior. Proximal goals are instrumental toward higher-level goals, and distal goals represent the deep layer of motivation [61,68]. How might this perspective be useful in the context of technology adoption theory, which is a critical aspect of IS research? While this is also a question to be explored in future research, a few avenues for consideration are offered. If higher-order goals are more enduring over time, then it is important to identify the types of higher-order user goals that technology can support (e.g., self-expression, creativity). In turn, these may become higher-order constructs in technology adoption theories. A related question that arises is the level(s) of goals that users have in mind when making the adoption decision. Consumer behavior research, for example, has revealed that the highest-level goal in a means-end chain may not be what the individual is thinking of when making a consumption decision [27]. For some technologies, the adoption decision may be based on

529

goals lower in the chain, while the continued use of a technology over the long term may be influenced by the extent to which it supports higher-level goals. Additionally, the MECA approach would provide a structured means of identifying commonalities and differences in goal structures for different types of technology, groups of users, and so on. As a final observation, it is noteworthy that the hierarchical goal map includes goals that are seemingly in opposition such as fantasy/realism, a need for uniqueness/following social conventions, and socialization/a place to be alone. Clearly, there is no universal set of goals sought by SVW users via virtual goods. This observation suggests that the adoption and diffusion of emerging technologies such as social virtual worlds may be influenced more by the range of capabilities provided by the technology that enable the user to pursue self-determined goals and values (‘‘perceived adaptability’’) than the predefined functionality envisioned by the designers. Moreover, although user adoption has been one of the central research topics in the IS discipline, its importance may become less relevant for the new generation of users, who are familiar with emerging user-empowering technologies [94]. To these users, the adoption of new technology may be largely taken for granted. As a consequence, in the future, the central research concern may shift from adoption to the development of theory centered on the goals/values pursued through technology use. This study, then, may presage a paradigm of user-goal-centric research in the IS field beyond traditional user adoption research. 5.2. Implications for practice The study also has preliminary implications for the producers/ sellers of virtual goods, firms and organizations utilizing SVWs, SVW service providers, and the designers of 3D cyberspaces. The overall implication for service providers is that their service models need to carefully negotiate a balance between a bazaar standard and cathedral standard. The bazaar standard indicates a user-created approach or the principle of laissez-faire; the cathedral standard indicates a service- or product-providergoverned approach [64]. In the context of virtual worlds, Second Life, by emphasizing user autonomy, corresponds to the bazaar model; World of Warcraft and Cyworld, where all systems, including the production and sales of virtual items, are governed by service providers [2] corresponds to the cathedral model. Our finding that users pursue diverse activities and goals through virtual consumption in Second Life suggests that the principle of laissez-faire works well in Second Life. However, we found that virtual consumption is inextricably interwoven into users’ virtual lives and can be considered a daily activity. For producers/sellers, the hierarchical goal map for virtual consumption highlights the goals users pursue via their virtual goods purchases and which goals are strongly associated with specific types of virtual property. This information can be useful in determining what types of virtual goods to produce, selecting the specific attributes to include/emphasize in product design, and the content of the messages that will be effective in marketing them to SVW users. Similarly, the hierarchical goal map is useful for companies and organizations utilizing SVWs to achieve their goals. The map is informative in terms of determining ‘‘user requirements’’ for enterprise virtual worlds, for example, by highlighting the goals sought by users and the types of experiences they seek, as well as some of the mechanisms that can be used to support those goals and provide those experiences. The study’s findings can also better inform those envisioning and designing future virtual world environments and 3D Web applications of the important goals and values that these technologies can provide users, from uniqueness and selfexpression to fantasy and creativity.

530

Y. Jung, S.D. Pawlowski / Information & Management 51 (2014) 520–531

Finally, growing virtual markets driven by the increasing volume of virtual consumption entails another concern related to price systems. As the transactions of virtual goods increase, the ambiguity of prices may arise. An extreme case is the situation in which certain users sell virtual property they obtained for free elsewhere in that virtual world and the buyer could also have acquired at no cost. This type of information asymmetry may distort the virtual market and create a distrustful virtual society. To enhance market clarity, service providers might add price comparison systems and user review systems for sellers, an approach that has proven workable in B2C e-commerce. 5.3. Limitations There are some limitations of the study that are important to note. First, we examined user goals in the context of a single class of technology, social virtual worlds, and a single instance of this technological environment, Second Life. We could observe different patterns of goal hierarchies in other technologies and systems. Nevertheless, as user-empowered cyberspaces such as Second Life are becoming the dominant part of the Web, our findings and interpretation can provide valuable knowledge for exploring the consumption of virtual goods in similar cyberspaces. Furthermore, a comparison of our findings with results obtained from other types of virtual worlds represents an important direction for future research. Other limitations arise from the samples of users. First, the study’s samples may be biased because participants were recruited over the course of three weeks and the majority of participants were users who frequently log in to Second Life. Practically, it was difficult to include transient users in our samples, and they are not represented. Second, because the participants responded to the survey voluntarily, our results may be biased towards highly motivated SVW users. Third, the study did not control for variance among participants due to culture. Second Life is a worldwide cyberspace with users from various countries, and culture may function as a crucial variable in fully understanding SVW users’ goals for virtual consumption. Because generalization was not a concern for this study, the main impact of these potential biases is that the goals of some users may not have been included in the goal map. Finally, another potential issue is that this study was limited to virtual consumption through purchase and did not include virtual items that users can acquire costlessly (‘‘freebies’’) in Second Life. However, because the purchase of an item is likely to be a stronger expression of value to the user than the use of free items, this was not judged to be a significant limitation given the objective of the study. 6. Conclusion This study provides both theoretical and methodological contributions for future research. The theoretical lenses for virtual consumption proposed in this study (e.g., the experiential perspective, virtual liminoid theory) can provide useful frameworks for future research on user behavior in virtual worlds. From a methodological perspective, a goal hierarchy approach uncovering the ‘reasons for the reasons’ for virtual activities can be used to examine issues such as the relative importance of goals or identify inhibitory relationships between lateral elements in the goal map (e.g., between competing goals or competing means) (e.g., [38]). The approach can also be used to develop a taxonomy of user goals for virtual consumption (e.g., [22]) or identify satisfaction paths/ user segments (e.g., [56]). Virtual worlds and associated technologies are one of the next major developments of the future Web, as 3D virtual environments and Web-as-participation are poised to become the dominant

features of the Internet environment. SVWs may have a radical impact by changing the means of communication on the Web by offering a more interactive and immersive environment than current, flat websites [52]. In this promising social cyberspace, virtual consumption has become a crucial component of virtual life. Because user retention is a critical issue for SVWs, an important question concerns the stability and longevity of user goals and the ability of different types of virtual goods to support those goals over time. The hierarchical goal map of virtual consumption generated in this study provides key insights into virtual consumption. With the continued expansion of virtual consumption (in terms of commercial significance) and the metaverse of virtual worlds, insights into virtual consumption will become all the more important.

References [1] Animesh, A. Pinsonneault, Y. Sung-Byung, O. Wonseok, An odyssey into virtual worlds: exploring the impacts of technological and spatial environments on intention to purchase virtual products, MIS Quart. 35, 2011, pp. 783–789. [2] R. Arakji, K.R. Lang, The virtual cathedral and the virtual bazaar, Adv. Inf. Syst. 38, 2007, pp. 33–39. [3] R.P. Bagnozzi, P.A. Dabholkar, Consumer recycling goals and their effect on decisions to recycle: a means-end chain analysis, Psychol. Mark. 11, 1994, pp. 313–340. [4] A. Bandura, Self-regulation of motivation and action through internal standards and goal systems, in: L.A. Pervin (Ed.), Goal Concepts in Personality and Social Psychology, Lawrence Erlbaum Associates, Hillsdale, NJ, 1989, pp. 19–85. [5] A. Bandura, Self-Efficacy. The Exercise of Control, Freeman, New York, 1997. [7] R.W. Belk, Possessions and the extended self, J. Consum. Res. 15, 1988, pp. 139– 168. [8] R.W. Belk, Extended self and extending paradigmatic perspective, J. Consum. Res. 16, 1989, pp. 129–132. [9] R.W. Belk, Moving possessions: an analysis based on personal documents from the 1847–1869 Mormon migration, J. Consum. Res. 19, 1992, pp. 339–361. [10] S.K. Bonsu, A. Darmody, Co-creating second life: market-consumer cooperation in contemporary economy, J. Macromark. 28, 2008, pp. 355–368. [11] G. Botschen, E.M. Thelen, Using means-end structures for benefit segmentation, Eur. J. Mark. 33, 1999, pp. 4–58. [12] D.A. Bray, B. Konsynski, Virtual worlds: multi-disciplinary research opportunities, Data Base Adv. Inform. Syst. 38, 2007, pp. 17–25. [13] E. Bridges, R. Florsheim, Hedonic and utilitarian shopping goals: the online experience, J. Bus. Res. 61, 2008, pp. 309–314. [14] C. Campbell, The desire for the new: its nature and social location as presented in theories of fashion and modern consumerism, in: R. Silverstone, E. Hirsch (Eds.), Consuming Technologies: Media and Information in Domestic Spaces, Routledge, London, 1992, pp. 48–64. [15] D. Capozza, R. Falvo, E. Robusto, A. Orlando, Beliefs about Internet: methods of elicitation and measurement, Pap. Soc. Represent. 12, 2003, pp. 1.1–1.14. [16] E. Castronova, Synthetic Worlds, The University of Chicago Press, Chicago, IL, 2004. [17] E. Castronova, Virtual worlds: A first-hand account of market and society on the cyberian frontier. CESifo Working Pap Series 2001. No. 618, available: http:// papers.ssrn.com/abstract=294828. [18] T.L. Childers, C.L. Carr, J. Peek, Hedonic and utilitarian motivations for online retail shopping behavior, J. Retail 77, 2001, pp. 511–535. [19] R.S. Cropanzano, D. James, M.A. Citera, A goal hierarchy model of personality, motivation and leadership, in: L.L. Cummings, B.M. Staw (Eds.), Research in Organizational Behavior, JAI, Greenwich, CT, 1993, pp. 267–322. [20] M. Eisenbeiss, B. Blechschmidt, K. Backhaus, P.A. Freund, The (real) world is not enough: motivational drivers and user behavior in virtual worlds, J. Interact. Mark. 26, 2012, pp. 4–20. [21] J.L. Fleiss, Statistical Methods for Rates and Proportions, John Wiley & Sons, New York, 1981. [22] M.A. Ford, Motivating Humans: Goals, Emotions, and Personal Agency Beliefs, Sage, Newbury Park, CA, 1992. [23] C.E. Gengler, T.J. Reynolds, Consumer understanding and advertising strategy: analysis and strategic translation of laddering data, J. Advert. Res. 35, 1995, pp. 19–33. [24] K.G. Grunert, T. Bech-Larsen, Explaining choice option attractiveness by beliefs elicited by the laddering method, J. Econ. Psychol. 26, 2005, pp. 223–241. [25] Y. Guo, S. Barnes, Purchase behavior in virtual worlds: an empirical investigation in Second Life, Inf. Manage. 48, 2011, pp. 303–312. [26] J. Gutman, A means-end chain model based on consumer categorization processes, J. Mark. 46, 1982, pp. 60–72. [27] J. Gutman, Means-end chains as goal hierarchies, Psychol. Mark. 14, 1997, pp. 545–560. [28] E.C. Hirschman, Aesthetics, ideologies and the limits of the marketing concept, J. Mark. 47, 1983, pp. 45–55.

Y. Jung, S.D. Pawlowski / Information & Management 51 (2014) 520–531 [29] M.B. Holbrook, R.W. Chestnut, T.A. Oliva, E.A. Greenleaf, Play as a consumption experience: the roles of emotions, performance, and personality in the enjoyment of games, J. Consum. Res. 11, 1984, pp. 728–739. [30] M.B. Holbrook, E.C. Hirschman, The experiential aspects of consumption: consumer fantasies, feelings, and fun, J. Consum. Res. 9, 1982, pp. 132–140. [31] C. Holsapple, J. Wu, User acceptance of virtual worlds, ACM SIGMIS Database 38, 2007, pp. 86–89. [32] Y. Jung, H. Kang, User goals in social virtual worlds: a means-end chain approach, CHB 26, 2010, pp. 218–225. [33] Juniper Research, Virtual Goods – Real Revenues on Mobile, 2011. [34] G.A. Kelley, The Psychology of Personal Constructs, Norton, New York, 1955. [35] S.S. Kleine, R.E. Kleine Iii, C.T. Allen, How is a possession me or not me? Characterizing types and an antecedent of material possession attachment J. Consum. Res. 22, 1995, pp. 327–343. [36] D. Klenosky, The pull of tourism destinations: a means-end investigation, J. Travel Res. 40, 2002, pp. 385–395. [37] D. Knoke, R.S. Burt, Prominence, in: R.S. Burt, M.J. Minor (Eds.), Applied Network Analysis, Sage Publications, Beverly Hills, CA, 1982, pp. 195–222. [38] A.W. Kruglanski, J.Y. Shah, A. Fishbach, R. Friedman, W.-Y. Chun, D. SleethKeppler, A theory of goal systems, in: M.P. Zanna (Ed.), Advances in Experimental Social Psychology, Academic Press, New York, 2002, pp. 331–379. [39] M. Koufaris, Applying the technology acceptance model and flow theory to online consumer behavior, Inf. Syst. Res. 13, 2002, pp. 205–223. [40] KZero, Virtual worlds: Industry & user data, universe chart for Q1 2012, May 2012, available: http://www.slideshare.net/nicmitham/kzero-universe-q1-2012. [41] KZero, Growth forecasts for the virtual worlds sector, 2010, available: http:// www.kzero.co.uk/blog/growth-forecasts-for-the-virtual-worlds-sector. [42] F.G. Lastowka, D. Hunter, The laws of the virtual worlds, Calif. Law Rev. 92, 2004, pp. 3–73. [43] V. Lehdonvirta, Virtual item sales as a revenue model: identifying attributes that drive purchase decisions, Electron. Commerce Res. 9, 2009, pp. 97–113. [45] H.-W. Lin, Y.-L. Lin, Digital educational game value hierarchy from a learners’ perspective, CHB 30, 2014, pp. 1–12. [46] Linden Lab, SL10B Infographic: Ten Years of Second Life, 2013 available: http:// community.secondlife.com/t5/Featured-News/SL10B-Infographic-Ten-Years-ofSecond-Life/ba-p/2053857. [47] E.A. Locke, G.P. Latham, A Theory of Goal Setting and Task Performance, Prentice Hall, Englewood Cliffs, NJ, 1990. [48] M. Ma¨ntyma¨ki, J. Salo, Purchasing behavior in social virtual worlds: an examination of Habbo Hotel, Int. J. Inf. Manage. 33, 2013, pp. 282–290. [49] B.A.S. Martin, Using the imagination: consumer evoking and thematizing of the fantastic imaginary, J. Consum. Res. 31, 2004, pp. 136–149. [50] S. Matook, Conceptualizing means-end chains of user goals as networks, Inf. Manage. 50, 2012, pp. 24–32. [51] R. Mehta, R.W. Belk, Artifacts, identity, and transition: favorite possessions of Indians and Indian immigrants to the United States, J. Consum. Res. 17, 1981, pp. 398–411. [52] D. Milbank, Goofy characters and weird people – Sounds like a hearing, The Washington Post, April 2, 2008. [53] A. Newell, H. Simon, Human Problem Solving, Prentice-Hall, Englewood Cliffs, NJ, 1972. [54] C.H. Nobel, B.A. Walker, Exploring the relationships among liminal transitions, symbolic consumption, and the extended self, Psychol. Mark. 14, 1997, pp. 29–47. [55] J.C. Olson, T.J. Reynolds, Understanding consumers’ cognitive structures: implications for marketing strategy, in: L. Percy, A.G. Woodside (Eds.), Advertising and Consumer Psychology, Lexington Books, Lexington, MA, 1983, pp. 51–57. [56] C. Orsingher, G.L. Marzocchi, S. Valentini, Consumer (goal) satisfaction: a meansends chain approach, Psychol. Mark. 28, 2011, pp. 730–748. [57] P. Pai, D.C. Arnott, User adoption of social networking sites: eliciting uses and gratifications through a means–end approach, CHB 29, 2013, pp. 1039–1053. [58] G. Parmentier, S. Rolland, Consumers in virtual worlds: identity building and consuming experience in Second Life, Rech. Appl. Mark. 24, 2009, pp. 43–55. [60] L.A. Pervin, Goal Concepts in Personality and Social Psychology, Lawrence Erlbaum, Hillsdale, NJ, 1989. [61] R. Pieters, H. Baumgartner, D. Allen, A means-end chain approach to consumer goal structures, Int. J. Res. Mark. 12, 1995, pp. 227–244. [62] B. Plumer, The economics of video games, The Washington Post, September 28, 2012, available: http://www.washingtonpost.com/blogs/wonkblog/wp/2012/09/ 28/the-economics-of-video-games. [63] W.T. Powers, Behavior: The Control of Perception, Aldine, Chicago, 1973. [64] E.S. Raymond, The Cathedral and the Bazaar, O’Reilly Press, Sebastopol, CA, 1999. [65] S.D. Reicher, Crowd behaviour as social action, in: J.C. Turner, M.A. Hogg, P.J. Oakes, S.D. Reicher, M.S. Wetherell (Eds.), Rediscovering the Social Group: A SelfCategorization Theory, Basil Blackwell, Oxford, England, 1987, pp. 171–202. [67] T.J. Reynolds, C.E. Gengler, D.J. Howard, A means-end analysis of brand persuasion through advertising, Int. J. Res. Mark. 12, 1995, pp. 257–266. [68] T.J. Reynolds, J. Gutman, Laddering theory, method, analysis, and interpretation, J. Advert. Res. 28, 1988, pp. 11–31. [69] T.J. Reynolds, J.C. Olson, Understanding Consumer Decision Making: The Means.end Approach to Marketing and Advertising Strategy, Mahwah, NJ, Erlbaum, 2001 [70] T.J. Reynolds, J.P. Rochon, Consumer segmentation based on cognitive orientations: the ChemLawn case, in: T.J. Reynolds, J.C. Olson (Eds.), Understanding

[71] [72] [73] [74] [75] [76] [77] [78] [79]

[80] [81] [82]

[83]

[84] [85] [86] [87] [88] [89] [90]

[91] [92] [93] [94] [95]

531

Consumer Decision Making: The Means-End Approach to Marketing and Advertising Strategy, Lawrence Erlbaum, Mahwah, NJ, 2001, pp. 283–298. C.M. Ridings, D. Gefen, Virtual community attraction: why people hang out online, J. Comput. Mediat. Commun. 10, 2004. M. Rokeach, The Nature of Human Values, Free Press, New York, 1973. D.W. Rook, The ritual dimension of consumer behavior, J. Consum. Res. 12, 1985, pp. 251–264. A. Ruvio, Unique like everybody else: the dual role of consumers need for uniqueness, Psychol. Mark. 25, 2008, pp. 444–464. J. Scott, Social Network Analysis: A Handbook, Sage Publications, London, 1991. W.A. Scott, Structure of natural cognitions, J. Pers. Soc. Psychol. 12, 1969, pp. 261– 278. J. Shen, L.B. Eder, Exploring intentions to use virtual worlds for business, JECR 10, 2009, pp. 94–103. J.N. Sheth, B.I. Newman, B.L. Gross, Why we buy what we buy: a theory of consumption values, J. Bus. Res. 22, 1991, pp. 159–170. Slulake, Habbo Hotel hits 200 million registrations, Sulake.com, January 20, 2011, available: http://www.sulake.com/press/releases/habbo-hotel-hits-200-millionregistrations. A.L. Strauss, J.M. Corbin, Basics of Qualitative Research: Techniques and Procedures for Developing Grounded Theory, Sage, Thousand Oaks, CA, 1998. J.C. Sweeney, G.N. Soutar, Consumer perceived value: the development of a multiple item scale, J. Retail. 77, 2001, pp. 203–220. H. Tajfel, J.C. Turner, The social identity theory of inter-group behavior, in: S. Worchel, W.G. Austin (Eds.), Psychology of Intergroup Relations, Nelson-Hall, Chicago, 1986, pp. 7–24. C.J. Thompson, E.C. Hirschman, Understanding the socialized body: a poststructuralist analysis of consumers’ self-conceptions, body images, and self-care practices, J. Consum. Res. 22, 1995, pp. 139–153. K.T. Tian, W.O. Bearden, G.L. Hunter, Consumers’ need for uniqueness: scale development and validation, J. Consum. Res. 28, 2001, pp. 50–66. V. Turner, The Ritual Process: Structure and Anti-Structure, Aldine De Gruyter, New York, 1969. V. Turner, Liminal to liminoid, in lay, flow, and ritual: an essay in comparative symbology, The Rice Univ. Stud. 60, 1974, pp. 53–92. van Gennep, The Rites of Passage, Routledge & Kegan Paul, London, 1977. J. van Rekom, C.B.M. Van Riel, B. Wierenga, A methodology for assessing organizational core values, J. Manage. Stud. 43, 2006, pp. 175–201. T. Veblen, The Theory of the Leisure Class: An Economic Study of Institutions, Macmillan, New York, 1899. T. Verhagen, F. Feldberg, B. van den Hooff, S. Meents, J. Merikivi, Satisfaction with virtual worlds: an integrated model of experiential value, Inf. Manage. 48, 2011, pp. 201–207. R. Voss, T. Gruber, I. Szmigin, Service quality in higher education: the role of student expectations, J. Bus. Res. 60, 2007, pp. 949–959. R. Walker, Immaterialism, The New York Times Magazine May 9, 2009, p. 28. N. Yee, Motivations for play in online games, CyberPsychol. Behav. 9, 2006, pp. 772–775. Y. Yoo, Computing in everyday life: a call for research on experiential computing, MIS Quart. 34, 2010, pp. 213–231. Z. Zhou, X.-L. Jin, D.R. Vogel, Y. Fang, X. Chen, Individual motivations and demographic differences in social virtual world uses: an exploratory investigation in Second Life, Int. J. Inf. Manage. 31, 2011, pp. 261–271. Yoonhyuk Jung is an assistant professor in the School of Business Administration at Ulsan National Institute of Science and Technology (UNIST). He earned his Ph.D. in information systems from Louisiana State University. His work focuses on user adoption and sensemaking of emerging technologies in areas of web services, wireless technology, and healthcare. His publications have appeared in European Journal of Information Systems, Information Systems Journal, Communications of the Association for Information Systems, among others.

Suzanne Pawlowski is a Lecturer in Information Systems at California Polytechnic State University, San Luis Obispo. She recently retired from Louisiana State University as Faculty Emeritus. She earned her Ph.D. inCIS from Georgia State University and M.B.A. and B.A. degrees from the University of California,Berkeley. Professional experience prior to her academic career includes 20 years as a systems developer and IT manager at Lawrence Livermore National Laboratory. Her current research interests focus on cybersecurity from a social psychology perspective. Scholarly publications include papers in MISQ, JAIS, Information & Management, and Database, among others. She is currently member of the editorial board for Information & Organization.