Exploring User Acceptance In Implementing Social Computing Systems

Transcription

1 Exploring User Acceptance In Implementing Social Computing Systems Alfred Chan A thesis submitted in partial fulfillment Of the requirements for the degree of BACHELOR OF APPLIED SCIENCE Supervisor: Professor David Zweig Department of Mechanical and Industrial Engineering University of Toronto March, 2008

2 Abstract The Information Technology industry has seen a shift in the development of software applications since the arrival of the Internet. With the increased technological capabilities and the changing social nature of online users, a new generation of applications known as Web 2.0 has become prominent. These applications focus on tailoring their design on the users and encourage users to generate contents to power them. The increasing popularity of these applications such as MySpace, Facebook, YouTube and Wikipedia has defined a new era of software. Consequentially, business enterprises are looking into integrating Web 2.0 technologies into their organizations. The resulting applications are known as Enterprise 2.0. One of the major areas of concern for businesses in implementing Enterprise 2.0 is whether its workforce will acceptance and adopt the technology to use. In this thesis, the traditional Technology Acceptance Model (Davis, 1989) is used as the basis in understanding user s intention to use an innovation. It has been extended to include the concepts of Diffusion of Innovation and Network Effects to more closely relate to the social nature of Enterprise 2.0. Finally, goal-orientation, self-efficacy and technology adoption measures are collected in a study to determine if they are able to better predict attitude towards use and behavioral intention to use in the context of two groupware systems, a traditional information repository and a wiki system. From the results, it is expected that insights into measures which can better predict user acceptance can be found and to determine whether Enterprise 2.0 technologies have an influence on user s intention to use the innovation. Exploring User Acceptance in Implementing Social Computing Systems i

3 Acknowledgement In the creation of this thesis, I would like to extend a sincere appreciation to my thesis supervisor, Professor David Zweig of the Department of Management at the University of Toronto, for his guidance. The project has been a useful learning experience and the exposure to the work of a thesis has increased my interests in future research opportunities. His expert advice has greatly enhanced the overall learning experience of this thesis. I would also like to thank: Professor Michael Carter of the Department of Mechanical and Industrial Engineering at the University of Toronto for allowing the study to take place in his class My roommate, Alvin Chick, for putting with the piles of dishes that have accumulated over the course of thesis work My family and friends for their continual support throughout my undergraduate experience Sincerely, Alfred S.M. Chan Industrial Engineering, University of Toronto 0T7 + PEY Exploring User Acceptance in Implementing Social Computing Systems ii

4 Table of Contents 1. INTRODUCTION BACKGROUND Social Computing Technological Factors Social Factors Social Identities in Communities Consumer Space The Long Tail of Social Computing MOTIVATION OF THESIS AND THEORETICAL BACKGROUND Technology Acceptance Model Diffusion of Innovation Network Effect Network Technology Acceptance Model (NTAM) Network Technology Acceptance Model Early Adoption Network Technology Acceptance Model Late Growth OBJECTIVE OF THESIS METHODOLOGY OF STUDY METHOD PARTICIPANTS MEASURES Goal Orientation Self-efficacy Technology Adoption Attitude Towards Use and Behavioral Intention to Use STUDY SCOPE RESEARCH QUESTIONS ANALYSES RESULTS PREDICTING BEHAVIORAL INTENTIONS PREDICTING INTENTIONS TO USE ELECTRONIC VS. TRADITIONAL COMMUNICATIONS CHANNELS PREDICTING ATTITUDES TOWARDS USE DISCUSSION Exploring User Acceptance in Implementing Social Computing Systems iii

5 List of Figures Figure 1: Two Forces Driving Social Computing (Charron, Favier & Li, 2006)...3 Figure 2: The Long Tail of Market Demand...8 Figure 3: The Long Tail of User-Generated Content...8 Figure 4: Technology Acceptance Model (TAM) (Davis, 1989)...11 Figure 5: Innovation-Decision Process Model...13 Figure 6: Cumulative Penetration of Innovation (Rogers, 1995)...14 Figure 7: Technology Adopter Distribution Bell Curve (Rogers, 1995)...15 Figure 8: Network Effect of Telephones...17 Figure 9: Network Technology Acceptance Model Early Adoption...20 Figure 10: Network Technology Acceptance Model Late Growth...21 Figure 11: Two Groupware Systems, OfficeZilla and PBWiki...26 Figure 12: Questionnaires Screenshots...26 Exploring User Acceptance in Implementing Social Computing Systems iv

6 List of Tables Table 1: Number of Year of Experience with Electronic Communication Tools...24 Table 2: Technology Adoption Questionnaire Items...29 Table 3: Rotated Component Matrix for 6-item Technology Adoption Measure...29 Table 4: Attitude Towards Use Questionnaire Items...30 Table 5: Behavioral Intention To Use Questionnaire Items...30 Table 6: Study Means, Standard Deviations, Reliabilities and Intercorrelations...35 Exploring User Acceptance in Implementing Social Computing Systems v

7 List of Appendices Appendix A: Original Questionnaire...I Appendix B: Goal Orientation Items III Appendix C: Computer Learning Self-efficacy Items IV Appendix D: Original Technology Adoption Items..V Exploring User Acceptance in Implementing Social Computing Systems vi

8 1. Introduction The Internet has had a major impact on the planning, design, deployment and maintenance of software applications. For example, search engines and s have completely transformed the way users search for information and communicate with one another. Most importantly, the nature of users has been changed dramatically with the popularity of the Internet. Users demand faster, more dynamic and more interactive applications than the traditional stand-alone applications such as word-processors or spreadsheets (O Reilly, 2005). Currently, a new wave of computer applications is emerging that utilizes the connecting capabilities of the Internet to power their uses. These new applications, also known as Web 2.0, incorporate existing computing functions with shared computer resources over any network, allowing for the generating, sharing and refining of information over the Internet (McAfee, 2006). These applications will likely change the way computer users interact socially with one another, but more importantly, they have underlying effects on how work can be done within business enterprises. The intersection of computing and social interaction is called Social Computing and will be discussed in more detail in Section 1.1. Due to the potential influence of this technology on information systems in business enterprise, this thesis will try to gain a better understanding of its effects. The purpose of the thesis is to investigate the emergence of social computing, what influences users adoption of this technology and ultimately, the practical implications for businesses that utilize it in their enterprises. Exploring User Acceptance in Implementing Social Computing Systems 1

9 1.1 Background Social Computing With the increasing demand from users for computational capabilities from software applications, there have been many information technology developments that look to provide the required needs. Technologically, developers have been trying to adapt to the users need by introducing customizable and components-based applications that have increased flexibility compared to traditional software such as word processors. For example, software such as widgets and toolbars has seen increasing popularity on the Internet in recent years. Moreover, these Web 2.0 applications are accessible over a network such as the internet or an intranet. They are popular due to their ubiquity, or their ability to be quickly and easily deployed on any computers with the use of a web browser. The widespread use of Web 2.0 has been linked to a concept called Social Computing (Charron, Favier & Li, 2006). Social Computing describes the area of computer science that is concerned with the intersection of social behavior and computational systems. It is any type of computing application in which software serves as an intermediary or a focus for a social relation (Schuler, 1994). The marketing research firm Forrester defines social computing as A social structure in which technology puts power in communities, not institutions. (Charron, Favier & Li, 2006). To understand this statement, the emergence of social computing needs to be considered first. Exploring User Acceptance in Implementing Social Computing Systems 2

10 There are two main driving forces that fuel social computing. On one hand, there are the technological factors which provide the tools to enable social computing. On the other hand, social factors bring about a change in user behaviors and how they interact with computers. These two forces work together to create social computing as seen in Figure 1. Figure 1: Two Forces Driving Social Computing (Charron, Favier & Li, 2006) Exploring User Acceptance in Implementing Social Computing Systems 3

11 1.1.2 Technological Factors Technology has improved in an increasing pace in the last years. Hardware has become cheaper to produce and this has lead to an increase in processing power, memory, and storage space for the masses (Charron, Favier & Li, 2006). Furthermore, wireless connections have been popularized by the increasing number of Wi-fi hotspots, the decreasing cost to access them and the introduction of powerful portable devices, such as laptops, cell phones, Blackberries and iphones. The increase in accessibility has freed users to tap into the power of the network anywhere, anytime. This phenomenon is called always-on connectivity (Charron, Favier & Li, 2006). It is especially evident for knowledge workers and business users, who require constant access to the Internet for their work. Another technological advancement that contributes to social computing is connectivity software. Software developments such as open-source languages, dynamic scripting languages, open protocols and cross-platform standards have accelerated the growth of connectivity software. Concepts like Internet as a Platform and Software as a Service (SaaS) encourage developers to create applications that allow componentization of software and create more client-server based software. When compared to traditional standalone applications, these applications are accessible anywhere, are more dynamic and perform faster (O Reilly, 2005). Lastly, all of these applications would not be possible without specialized databases that allow them to generate dynamic contents. Database management is a key to freeing up traditional applications. By having databases behind applications, it allows users to read, write and delete information from data storage that is connected to a network. The Exploring User Acceptance in Implementing Social Computing Systems 4

12 freedom resulting from database management allows user to share resources much more quickly. For example, one user can update a record in the database and another user can read from the record immediately following the update elsewhere on the network. Altogether, the technological factors mentioned enable the possibility of social computing (Charron, Favier & Li, 2006) Social Factors The social factors in modern society are also contributing to social computing. The growing number of diverse users on the Internet is one such social change. As the number of users increases, smaller communities on the network are developed based on the different needs of users. The existence of such communities is often brought about by people who share similar goals, beliefs, or values with such commonality forming the basis of an agreement to form and sustain a virtual existence (Figallo, 1998). The ability to locate any sort of niche on the Internet has allowed many unique communities to be established. Another social factor is the acceptance of virtual representations of physical activities (Fjeld et al., 2002). The major example of this change is e-commerce. E-commerce capitalizes on allowing users to shop online instead of having to travel to a mall. The physical experience of going to a mall to shop is virtualized onto online shopping sites (Nilan & D Eredita, 2004). Virtual artifacts such as the notion of shopping carts become the manifest of physical objects that are associated with the activities. Another example is with inboxes and trash cans being virtualized to represent physical mailing activities. When sending an to another person, users conceptualize their mails being delivered to the inbox of the receiving users. Exploring User Acceptance in Implementing Social Computing Systems 5

13 Lastly, the belief of information commons on the Internet strengthens social computing. A commons, simply understood, is a resource or a facility that is shared by a community of producers or consumers (Oakerson, 1992). Today, with the Internet, citizens have vastly greater opportunities to access the information and ideas (Kranich, 2004). The commons allow users to become more than just consumers of information. It enables users to research, gather, produce and share information and ideas with others. These social changes shape the development of technology and the two forces of technological and social changes converge to form social computing Social Identities in Communities Forrester s (2006) definition of social computing describes a fundamental shift in social behaviors that changes the way information is managed. Information is no longer passed on from institutions to individuals. Every individual is free to produce and share information with one another. Those who share related information will form communities to allow them to better reach others with similar interests. The freedom to express or to communicate ideas gives individuals power to influence others within their community. Controls and innovation will be in the hands of individual users. Institutions will no longer be able to enforce changes that are not accepted by individuals. With this transfer of power, there is a shifting focus to the rights, needs and responsibilities of these individuals as citizens (Bollier, 2003). The choice of individuals to decide if they want to be part of the collective group becomes a decision of whether they want to become a citizen and to associate themselves with the network that has been formed. The virtualization of individual profile and social identify affects a user s decision in whether or not to use technology that enables them to join these online social structures. Exploring User Acceptance in Implementing Social Computing Systems 6

14 1.1.5 Consumer Space So far, social computing influences have been mainly in the consumer space, where the masses have chosen to use the social structure formed to connect with one another. The popularity of blogs, wikis and social network has been spreading at an increasingly fast pace. The power of these tools to enable users to collaborate, share information and facilitate creativity increases their popularity. The technological advancement in support of these tools creates a second-generation of web-based applications that is commonly known as Web 2.0. Web 2.0 is mainly focused on providing a transition of websites from isolated read-only information silos to interactive computing platform that allows users to download and more importantly, upload information. Web 2.0 is also centered on the specific technologies such as XML and Ajax that enable the transitions. The success stories of Web 2.0 in consumer space such as YouTube, MySpace, Wikipedia and Blogger have sparked interest into how this technology can be used elsewhere to create the same values as it has in the consumer space The Long Tail of Social Computing One of the reasons for the phenomenally fast spread of Web 2.0 is the amount of user-generated content freely available to users. The contribution efforts of users can be described by the statistical concept of the Long Tail (McAfee, 2006). Coined by Anderson (2004), the Long Tail describes the supply of customers in retail niche strategy that buys the hard-to-find items. Based on the Long Tail, the cumulative sales of many low demand items (the long tail ) will be greater than the total sales of the few high demand items. In Figure 2, the right of the graph represents the long tail of low demand users that make up approximately 80% of the total demand, while the high demand items (to the left) might have more users per items, but only make up 20% of total demand. Exploring User Acceptance in Implementing Social Computing Systems 7

15 Figure 2: The Long Tail of Market Demand Similarly, in terms of user-generated content contributions, the Long Tail can also be used to describe the users willing to produce (McAfee, 2006). For example, the number of contributors for Wikipedia only makes up for a small percentage (< 2%, based on usage statistics on Wikipedia) of all Internet users that access Wikipedia. The size of the group of contributors decreases as the level of contributors or the number of edits of each contributor increases. This means that the group of most active contributors is the smallest group. This phenomenon can also be modeled by the Long Tail as in Figure 3. Figure 3: The Long Tail of User-Generated Content Exploring User Acceptance in Implementing Social Computing Systems 8

16 1.1 Motivation of Thesis and Theoretical Background Since the development of Web 2.0 applications has been very rapid, little academic research has been conducted to observe and to measure the effects of the technology. Moreover, there has been increasing interest by business enterprises to understand the values, implications and issues surrounding social computing. The term, Enterprise 2.0, has been used to describe applications developed for businesses based on Web 2.0 principles (McAfee, 2006). Since the key component of any social computing systems is the users, the most important decision relating to implementing Enterprise 2.0 is whether the technology will be accepted and in turn, adopted by users. As Enterprise 2.0, like Web 2.0, requires a social structure to work effectively, it is critical to have actual system use by users in the enterprise to power the applications. Without user acceptance and participation to create user-generated contents, any sort of IT investment into Enterprise 2.0 will likely fail. Enterprise 2.0 with low user acceptance and participation is more likely to fail due to the size of the supply of users willing to contribute content. This can be understood by referring to the concept of the Long Tail of user-generated content. Since the fraction of Internet users that are contributors are very, very small and if the same fraction is applied within an enterprise, the number of users that will generate content will become minimal (McAfee, 2006). This is why user acceptance and participation have to be strongly encouraged in the organization and are the keys to the success of any attempt to utilize Enterprise 2.0. Exploring User Acceptance in Implementing Social Computing Systems 9

17 This thesis is aimed at determining measures of user acceptance of social computing systems based on the existing Technology Acceptance Model (TAM; Davis 1989). A hypothetical extension of the TAM has been proposed that tries to capture the social aspect of the new technology. The social aspect is suggested to be broken down into two parts, Diffusion of Innovation and Network Effect. Diffusion of Innovation is a theory that determines the rate that new technology spread through cultures (Rogers, 1962). It differentiates among users who select are early adopters (EA) of new technologies and late adopters (LA) of new technologies. The Network Effect (NE) is the value for a user based upon the number of contribution of prior users within the network (Liebowitz &Margolis, 1998). It is a characteristic that causes a good or service to have a value to a potential customer which depends on the number of other customers who own the good or are users of the service Technology Acceptance Model The Technology Acceptance Model introduced by Davis (1989) is an adoption of a social psychology model called Theory of Reasoned Action (TRA; Fishbein & Ajzen, 1975). According to TRA, a person s performance of a specified behavior is determined by his/her behavioral intention (BI) to perform the behavior and BI is jointly determined by the person s attitude (A) and subjective norms (SN) concerning the behavior (Fishebein & Ajzen, 1975). BI is a measure of the strength of one's intention to perform a specified behavior. A is defined as an individual's positive or negative feelings (evaluative affect) about performing the target behavior. Subjective norm refers to "the person's perception that most people who are important to him think he should or should not perform the behavior in question". Exploring User Acceptance in Implementing Social Computing Systems 10

18 TRA has been used to cover a wide area of subject areas. Specifically tailored to user acceptance of information systems, the Technology Acceptance Model (TAM) provides explanation of the acceptance determinants and user behavior for computing technologies (Davis, Bagozzi & Warshaw, 1989). The model is shown in Figure 4. It describes user acceptance using TRA s factors. Behavioral intention to use (BI) is a function of attitudes (A) and perceived usefulness (U), while perceived ease of use (EOU) influences both U and A. The model is summarized by: BI = A + U A = U + EOU U = EOU + External Variables EOU = External Variables Figure 4: Technology Acceptance Model (TAM) (Davis, 1989) TAM s two main beliefs, perceived usefulness and perceived ease of use, are the primary determinants of user acceptance. Perceived usefulness (U) is defined as the prospective user's subjective probability that using a specific application system will increase his or her job performance within an organizational context. It depends on the extent to which an application contributes to the user s performance enhancement. Perceived ease of use (EOU) refers to the degree to which the prospective user expects Exploring User Acceptance in Implementing Social Computing Systems 11

19 the target system to be free of effort, in order to take advantage of the application (Davis, Bagozzi & Warshaw, 1989). Studies have been conducted to validate the model and it is still being used to determine user acceptance of new systems. The focus in the past has been measuring usefulness and ease of use to predict actual system use. However, with the development of social computing systems and the concept of the network effect, the experience and competency of users, (i.e. technology adoption competency) must also be measured to determine how they affect the adoption and use of these systems. TRA does not specify the beliefs that are operative for a particular behavior. In order for researchers to target a specific behavior under investigation, salient beliefs for subjects regarding the behavior under investigation must be determined (Davis, Bagozzi & Warshaw, 1989). Due to time restriction of this thesis, salient belief elicitation specific to the systems being tested could not be performed. Instead, drawing on research by Zweig and Webster (2004), existing items created for the purpose of determining the orientation of students toward learning how to use and apply a new technology (e.g., their goal orientation), and the level of efficacy students have toward the use of new technologies, were modified to the specific target (information repository or wiki system) within the specific context (the MIE case study course). These measures captured the psychological variables through which users attitude and intentions to use are influenced. Exploring User Acceptance in Implementing Social Computing Systems 12

20 1.1.2 Diffusion of Innovation The spreading of technology trends, in which changes are promoted within a social system through the domino effect, is based heavily on diffusion theories. More specifically, Rogers (1995) Diffusion of Innovations provides a scientific approach to understand the spread of new technologies. Rogers (1995) describes diffusion as the process by which an innovation is communicated over time among the members of a social system. For example, the main diffusion channel for popular social network is invitations by existing users. When a new innovation is introduced to a social system such as an organization, each member of the social system bases his/her own innovation-decision (the decision to adopt innovation or not) in a 5-step process (see Figure 5; Rogers, 1995). The user becomes aware of an innovation (Knowledge) and forms a favorable or unfavorable attitude towards it (Persuasion). The user will engage in activities with the innovation that leads to a choice to adopt or to reject the innovation (Decision). Lastly, the user puts the innovation to use (Implementation) and will continue to evaluate that decision until a new innovation replaces it (Confirmation). Figure 5: Innovation-Decision Process Model Exploring User Acceptance in Implementing Social Computing Systems 13

21 One of the most important features of the theory discussed by Rogers is the social effects of the innovation-decisions made by other members of the system. Empirically, Rogers (1995) modeled the cumulative penetration of innovation by the S-Curve (see Figure 6). This curve shows that initially the number of users, approximately 10-25%, that adopt the innovation will take a relatively long time. However, a rapid increase in adoption by the remaining members will follow. Lastly, those that hold-out on the technology will eventually adopt the innovation. Figure 6: Cumulative Penetration of Innovation (Rogers, 1995) Another way of looking at the technology acceptance by the different members is to look at adopter s distribution bell curve (see Figure 7; Rogers, 1995). This figure shows that the number of members that adopt the technology early on (Early Adopters, EA), including the innovators, early adopters and part of the early majority, will make up an increasing number of members. The late adopters (LA) who will make the innovationdecisions to adopt the technology later have the highest number of members. The laggards, who will eventually join the innovation, make up a small percentage of the members and are those that do not feel comfortable with new technology until later on. Exploring User Acceptance in Implementing Social Computing Systems 14

22 Figure 7: Technology Adopter Distribution Bell Curve (Rogers, 1995) As illustrated in both of Rogers curves, the innovation-decisions made by the LA are heavily influenced by the adoption of the EA. There are two main factors that create the phenomenon observed and they are related to the two evaluations that all users conduct for the innovation-decision, cost-benefit analysis and risk assessment (Rogers, 1995). One of the major obstacles in deciding whether to adopt a new innovation is uncertainty. People will adopt an innovation if it will enhance their ability to perform. By conducting a cost-benefit analysis, the user can weight in the relative advantage of using the innovation against the cost to adopt it (Beach & Mitchell, 1978). By observing others who may adopt the innovation, the uncertainty of the costs and benefits of the system can be reduced. Exploring User Acceptance in Implementing Social Computing Systems 15

23 Ultimately, this leads to the second evaluation that users perform as part of their innovation-decision. Risk assessment evaluates the likelihood that by engaging in a certain behavior, the probability for that behavior to result in a failed outcome. If a user is able to determine that the risk of adopting a new innovation is low, he/she will more likely adopt (Rogers, 1995). On average, people are risk-averse and will postpone their decision until further evidence can be gathered (Orr, 2003). This leads to some users taking a longer time to evaluate the risk of a new innovation. However, the risk tolerance of individuals is different and is largely characterized by individual differences such as personality, education and experience. The diversity in individuals is what allows diffusion to occur (Orr, 2003). Based on the diverse risk tolerance of individuals, the EA can be viewed as members of the systems who are more likely to take risk and have higher self-efficacy. The internal psychological decision process to adopt a technology by the EA will differ from those of the LA. Also, LA s decision for technology adoption is largely influenced by the EA s adoption (Rogers, 1995). Lastly, it should be noted that the factors which influence user s innovation-decisions discussed in Diffusion of Innovations can be closely paired with the TAM. User s costbenefit analysis can be mapped to the belief of perceived usefulness. In the same sense, user s risk assessment is similar to perceived ease of use. The close relationship between these two theories is not surprising as they both try to capture the individual s decision to adopt innovation or not. Exploring User Acceptance in Implementing Social Computing Systems 16

24 1.1.3 Network Effect Another factor that influences the spread of innovation adoption is network effect. Network effect or network externalities describe the change in benefit that an agent derives from a good when the number of other agents consuming the same kind of good changes (Liebowitz and Margolis, 1998). One of the prominent examples of network effect is the telephone. As the number of users that uses the telephone increases, the benefit for the next user that adopts telephone will increases because the number of connections that can be made increases (see Figure 8). Figure 8: Network Effect of Telephones Exploring User Acceptance in Implementing Social Computing Systems 17

25 Metcalfe (1980), founder of Ethernet, mathematically expressed the number of unique connections in a network of a number of nodes (n) as n(n-1)/2, implying that there are n 2 unique connections (Metcalfe, 1980). The expression is called Metcalfe s Law. The law measure the potential number of contacts that can exists in a network. One of the key considerations of the Metcalfe s law is to distinguish between is the potential number and the actual social utility of a network. The social utility of a network depends on the number of nodes that are in contact or linked (Hendler & Golbeck, 2007). This means that even though n 2 models the total possible connections, only those nodes that are connected will create utility. With social computing systems, this is more easily achieved because of the social constructs encouraged within them (Hendler & Golbeck, 2007). The interactions between users and the links that are created between them generate the network effect. Specific technologies such as tagging, folksonomies, RSS feeds, permalinks and other Web 2.0 artifacts make the social connections even stronger (O Reilly, 2005). The prevalence of users who generate content allows the utility for other members to increase. This effect is especially more predominant within a business enterprise (Liebowitz and Margolis, 1998). Employees within a firm will be more likely to adopt specific innovations if it is compatible with the rest of the firm. Liebowitz and Margolis (1998) state that compatibility within the firm is of greater importance than compatibility with external organizations. This point emphasizes the network effect of social computing system within business enterprises. Exploring User Acceptance in Implementing Social Computing Systems 18

26 Another driving motivation behind the concept of network effect within businesses is the concept of social capital. Social capital is a social construct that highlights the importance of networks of strong interpersonal relationships for trust, cooperation and collective action (Daniel, Schwier & McCalla, 2003). It encourages members that are participating in a social computing system to be more willing to generate content that will enhance other s performance, as the success of the overall organization will create mutual benefits for all members. In general, network effect is a concept that is difficult to capture empirically. However, based on Metcalf s Law and the concept of social capital, network effect can be defined loosely as the utility or value generated for any new adopter based on the contributions and usages of existing users of a system. The amount of system usage by any existing users will likely contribute to the perceived usefulness of the system for a new user. Network effect is likely to have a major influence in social computing systems because of their emphasis on create social communities and user-generated contents. Exploring User Acceptance in Implementing Social Computing Systems 19

27 1.1.4 Network Technology Acceptance Model (NTAM) The difficulty in extending the TAM to include the network effect is that early adopters of a social computing system will have no network effect to begin with as there have not been any prior users on the system. This is where the Diffusion of Innovation comes into play. As described earlier, the rate of adoption for social computing system will have slow growth in the opening phase. However, once more users have adopted the technology and created network effect, the late growth of adoption will be much more rapid (Rogers, 1995). To differentiate between how the technology is adopted in the two phases, the extended TAM must represent the different phases, which will be called early adoption and late growth. The proposed extension of TAM will be called Network Technology Acceptance Model (NTAM) to describe the network nature of the technology being modeled. Figure 9 and 10 illustrates the NTAM based on the two phases Network Technology Acceptance Model Early Adoption In the early adoption phase, the TAM will resemble the original TAM where perceived usefulness and perceived ease of use will be used to predict actual system use. The measures for actual system use will be used to represent the Network Effect (N) created by early adopters Figure 9: Network Technology Acceptance Model Early Adoption Exploring User Acceptance in Implementing Social Computing Systems 20

28 1.1.6 Network Technology Acceptance Model Late Growth In the late growth phase, the Network Effect is one of the characteristics used to predict actual system use because of its influence on perceived usefulness of the system. If there are existing users on the social computing system, then the value of the system may increase as more users are contributing information and resources to it. This will likely influence perceived usefulness and thus, lead to increases in actual system use of by the late growth user. Conversely, if the number of users on the social computing system is decreasing, the network effect will also decrease causing user s perceived usefulness of the system to lower. Hence, the network effect and the number of active users for the system have become key characteristics of user acceptance for social computing systems. The network effect will continue to affect perceived usefulness for the system until it becomes obsolete or it no longer has any active users. Figure 10: Network Technology Acceptance Model Late Growth Exploring User Acceptance in Implementing Social Computing Systems 21

29 1.2 Objective of Thesis One of the objectives of the thesis is to investigate the effects of social computing on business enterprise and to study with the technology s influences on the Technology Acceptance Model (TAM). A hypothetical extension to the TAM has been proposed to determine if measures of social aspects - Diffusion of Technology and Network Effect - will have an influence on the TAM. The extension is named the Network Technology Acceptance Model (NTAM). Measures of goal orientation, self-efficiency and technology adoption are used to determine if they will have any effects on attitudes towards use, behavioral intentions to use and intention for actual usage. The measures will be tested in a study that compares traditional enterprise groupware to social computing groupware. The study will help determine what factors affect user acceptance of the two systems. The results will be able to help Information Managers and System Designers in predicting the effectiveness of social computing systems compared to traditional tools. Exploring User Acceptance in Implementing Social Computing Systems 22

30 1.3 Methodology of Study A study was conducted to test the hypothetical Network Technology Acceptance Model (NTAM). The nature of the study was based upon the original field test conducted for the original TAM and any subsequent field tests based upon the model (Davis, 1989). In order to assess the model, data was gathered from 30 full-time students in the University of Toronto s Department of Mechanical and Industrial Engineering during one of their case study courses. Two separate systems were introduced for use by students in support of one of their case studies. A traditional information repository system was introduced to one group of students, while the other group was given a social computing wiki system. The two groupware systems were selected for this study because they both require social interaction and member participation to make them useful. The details of both systems will be described later in Section 2.2 Setup of Study. At the beginning of the study, students were asked to complete a brief questionnaire (see Appendix A) measuring their goal orientation, self-efficacy and technology adoption (early vs. late). Following the initial questionnaire, the students were given video tutorials that provide orientation to the two software systems. The students were allowed to interact with the systems for four weeks. Finally, a second questionnaire was given to measure the attitudes, intention to use and perceived usefulness of the tools used by the students. Exploring User Acceptance in Implementing Social Computing Systems 23

31 2. Method 2.1 Participants Thirty four participants enrolled in a Mechanical and Industrial Engineering case study course completed a questionnaire assessing their goal orientation, self-efficacy and technology adoption (early vs. late). The participants were split into two groups to test the traditional repository system and the social computing wiki system. The two groups were determined based on the case study group already organized for the class. For the traditional information repository, fourteen participants took part in the study. For the wiki system, twenty participants were included in the study. As part of the same sample, thirty participants completed measures of attitudes, behavioral intentions to use the technology and perceived usefulness measures after having used both systems for approximately four weeks. Four participants did not continue to participate in the study after the initial questionnaire. Thirteen participants remained for the information repository, while seventeen remained for wiki. Females constituted 33.4% of the final sample. 80% of participants were in the age range of 21-23, 16.7% were between age of and only 1 participant was older than 30. The degree of experience with electronic communication tools of both groups participants are summarized in Table 1 below. Year of Experience Traditional Information Repository Social Computing Wiki Under 1 year Table 1: Number of Year of Experience with Electronic Communication Tools Exploring User Acceptance in Implementing Social Computing Systems 24

32 2.2 Setup of Study The study was setup using two free online groupware products, OfficeZilla ( and PBWiki ( These two software programs were chosen because of their free licenses, quick setup, easy accessibility and available training resources. Also, these two software programs have been widely used by many different types of organization as noted on their respective websites. In both systems, the focus is to allow users to create and collaborate with other members of a group. OfficeZilla is a free virtual office that provides free message board, file sharing, calendar, knowledge base system and many other features. In the study, all of the features except for message board and knowledge base were disabled to ensure that the participants can focus on using those features only. The focus of the traditional groupware has been posting information and using search engines to query for the necessary information. The message board allowed participants to start topic threads and reply to them to discuss the case study work that they were assigned. While the knowledge base allowed them to post specific information relating to specific topics in a searching database. PBWiki is a free hosted wikis provider that offers organization their distinct wiki systems to manage their information. It provides an easy to use editor that allows pages to be created, edited and linked together. PBWiki offers customizable pages and widgets to be deployed on its wiki and information from other sources such as MS Office suites can be pasted onto it. The wiki offers content management that keeps track of changes made. The system encourages users to create and edit pages by providing edit functions on every page. Figure 11 shows a screenshot of both systems. Exploring User Acceptance in Implementing Social Computing Systems 25

33 Figure 11: Two Groupware Systems, OfficeZilla and PBWiki Two questionnaires that were used to collect the measures were administered to participants as online questionnaires. Poll Daddy ( was used to generate the online questionnaires (see Figure 12). Individual questionnaire items were assessed using Likert scales ranging from 1 = Strongly Disagree to 7 = Strongly Agree.. Figure 12: Questionnaires Screenshots Exploring User Acceptance in Implementing Social Computing Systems 26

34 2.3 Measures Goal Orientation To assess individual difference factors that might influence attitudes and intentions to use the technologies, goal orientation measures from Zweig and Webster (2004) were used. Goal orientation is a motivational variable that describes an individual s orientation toward learning and applying new constructs. Based on an individual s implicit theory of intelligence (Dweck, 1986), Zweig and Webster (2004) created a measure to assess three different factors. Performance approach is characterized by a desire to attain favorable judgments. Performance avoidance is characterized by a desire to avoid unfavorable judgments. Finally, learning orientation is focused on increasing competency. These different goal orientations have been demonstrated to have an influence on learning, training effectiveness, and job performance (Zweig & Webster, 2004). Twenty-one items were employed to capture goal orientation in the initial questionnaire (see Appendix B). The internal consistency reliabilities of the items were as follows: Learning orientation (α =.69), performance approach orientation (α =.72), and performance avoidance orientation (α =.70), Self-efficacy Self-efficacy is the belief in one s capacity to perform a task. Further self-efficacy beliefs have been demonstrated to have a strong influence on actual performance (e.g., Zweig & Webster, 2004). The most valid measures of self-efficacy are task specified and in the context of this study, computer learning self-efficacy (CSE) was measured. The items in the measure were reworded to the target of electronic communication tool to refer to either the information repository or the wiki system. A 7-item measure from Zweig and Webster (2004) was once again adopted and placed into the initial Exploring User Acceptance in Implementing Social Computing Systems 27

35 questionnaire (see Appendix C). The internal consistency reliability of CSE for the study was Technology Adoption As discussed above, people who are more likely to adopt new technologies quickly may help create a network effect. Technology adoption (TA) is designed to measure technical confidence of the user to determine if he/she is more likely to be an early adopter or a late adopter. The subjective technical confidence (STC) belief items developed by Beier (1999) were referenced in the construct of technology adoption items. Beier (1999) states STC as an individual belief in one s own ability to solve technical problems and is measured by the STC-Questionnaire developed by him. The modified questions were created to determine the technology adoption measure. An 8- item measure based on Beier (1999) was created for the initial questionnaire (see Appendix D). To assess the factor structure of the items, a factor analysis was conducted using principal components and varimax rotation. The analysis revealed that two separate factors loaded onto the constructs (early adopters vs. late adopters). Two of the items were discarded because of cross-loading on both factors. The final 6-item measure is shown in Table 2. Exploring User Acceptance in Implementing Social Computing Systems 28

36 Technology Adoption (TA) Early Adoption I like to read about new technologies (TA3) I like to recommend new technologies to others (TA5) If the technology is right for me, I would wait in line to be one of the first to buy it (TA6) I don t want to risk missing out on a new technology by waiting to use it (TA8) Late Adoption I prefer to wait until the bugs have been worked out before I adopt a new technology (TA2) I depend on recommendations from friends before I decide to try a new technology (TA4) Table 2: Technology Adoption Questionnaire Items Table 3 shows the principal component analysis. The factor analysis revealed that TA item 3, 5, 6 and 8 is loaded as early adopters, while item 2 and 5 is loaded as late adopters. The reliabilities of the two factors were 0.61 for early adopters and 0.63 for late adopters, respectively. Component 1 2 TA TA TA TA TA2 R TA4 R Table 3: Rotated Component Matrix for 6-item Technology Adoption Measure *Extraction Method: Principal Component Analysis *Rotation Method: Varimax with Kaiser Normalization Rotation converged in 3 iterations R The items were reversed when conducting the factor analysis as noted in Table 5. Exploring User Acceptance in Implementing Social Computing Systems 29

37 2.3.4 Attitude Towards Use and Behavioral Intention to Use Eight-item adapted from Zweig and Webster (2002) assessing attitudes towards the use of electronic communications tools (A) and behavioral intention to use electronic communications tools (BI) were used in the second questionnaire after participants used the two systems for four weeks. All of the items were rated on seven-point Likert scales where 7 = positive responses and 1 = negative responses. The reliability of the attitudes measures was 0.74 and the reliability of behavioral intentions was The items are shown in Table 4 and Table 5. Behavioral intentions were also assessed using two dichotomous items forcing respondents to use electronic communications tools versus more traditional (e.g., faceto-face) methods of communication. Attitude Towards Use (A) What is attitude towards the use of this electronic communication tool? Do you feel that this electronic communication tool would be more useful than speaking with others to learn about new things? Do you feel that this electronic communication tool would be useful in accomplishing tasks for this job? Table 4: Attitude Towards Use Questionnaire Items Behavioral Intention to Use (BI) How willing would you be to use this electronic communication tool in the future? How often would you likely use this electronic communication tool in the future? To what extent do you feel that you would like this electronic communication tool in the future? To what extent would you look forward to using this electronic communication tool in the future? If you wanted to learn about a new procedure, process, or technical approach, would you be most likely to: choice between using traditional methods of the tools If a team member wanted to learn about a new procedure, process, or technical approach that you have knowledge about, would you rather: choice between using traditional methods of the tools Table 5: Behavioral Intention To Use Questionnaire Items Exploring User Acceptance in Implementing Social Computing Systems 30

38 2.4 Study Scope Initially, the NTAM model was to be explored within a large telecommunications firm that had both traditional and wiki-based communication tools in place. However, due to limitations in time, resources and availability of data, some of the measures in the proposed Network Technology Acceptance Model could not be explored. These measures were not identified due to time and resource constraints of the thesis: Network effect is not measured in the study due to the complexity in monitoring each individuals behavior on the system over the duration of the study Actual performance following the use of the system was not collected because of the constraint in accessing the results in time for the completion of the thesis The actual distinction between early adopters and late adopters because of the lack of a monitoring system in place to track actual usage by the participants These constraints necessitated a change in the scope of the research project that limited the ability to test all of the factors proposed in the NTAM. Despite the difficulties encountered in collecting data with the telecommunications firm as originally intended, I was able to explore the influence of individual difference factors (e.g., goal orientation, computer self-efficacy) and rate of adoption (late vs. early) on attitudes and intentions to use two different communications tools. Exploring User Acceptance in Implementing Social Computing Systems 31