PREDICTING USER ACCEPTANCE OF INFORMATION TECHNOLOGY: A QUANTITATIVE
ANALYSIS OF THE RUG’S NEW STUDENT PORTAL USAGE
Master thesis, MScBA, specialization Strategic Innovation Management
University of Groningen, Faculty of Economics and Business
January 18, 2016
ALBERTO LUBRANO
Student number: s2782251
Oosterweg 66a
9724 CK Groningen
E-mail: albertolubrano22@gmail.com
Supervisor
Dr. Q. (John) Dong
Co-assessor
Dr. Wilfred Schoenmakers
Abstract. Institutions of higher education are increasingly using web portals as a way to connect
with students and to provide them with all the didactic information and services they need
through a centralized point of access. Despite of the efforts made by universities to develop and
manage these portals, students may or may not be willing to use them. By adopting as its
theoretical basis the technology acceptance model (TAM), one of the most acclaimed IT utilization
models from IS literature, the research examines factors influencing the acceptance and use of the
student portal recently introduced in the university of Groningen. This paper explores the validity
of an extension of TAM based on the introduction of a novel construct, called Perceived
Innovativeness, in order to determine whether an expanded model would be able to better
predict user acceptance and use. Longitudinal data were collected by handing over to 102
students two distinct waves of survey, taken both prior and after the portal’s introduction. Results
from the PLS structural equation modeling show that perceived usefulness, perceived ease of use
and attitude (all belonging to the original TAM) had a significant contribution in explaining
behavioral intention and usage of the portal. Additionally, the integration of the novel construct of
perceived innovativeness was found to improve the fit of the model by the effect of its mediating
role between perceived usefulness/ease of use and attitude. Student attributions to these
perceptions are discussed and compared with findings from previous research. At last, limitations
and suggestions for future research are presented.
Keywords: technology acceptance model, student portal, e-portals acceptance, diffusion of
innovation, university of Groningen
2
Table of content
1. Introduction ………2
2. Theoretical Background ……….6
2.1 Theory of Reasoned Action and Theory of Planned Behavior ………..6
2.2 Technology Acceptance Model ………..9
3. Methodology ………..16 3.1 Data Collection ……….16 3.2 Measurements ……….17 3.3 Data Analysis ……….20 4. Discussion ……….27 5. Conclusion ………29
5.1 Limitations and Future Research Suggestions ……….30
1. Introduction
3 model of users’ acceptance and usage of technology, given that its measures (PU, PEOU and behavioral intentions) are highly reliable and may be used in various situations (King & He, 2006).
In their meta-analytic review of TAM, King & He (2006) group users into three categories: students, professionals and general users (non-students who were not using the system for work purposes). Most of the TAM-based researches were conducted on a business and company-level, by drawing attention to the interaction between white collar employees (or professionals) and the new systems. In this context, the effort in marketing a system is often as much as that required to develop it. Managers who fail to understand this tend to produce the poorest user acceptance, and consequently yield the poorest return on investment (Bartlett, 1989). It is important to notice that new Information Systems represent innovations for the target audience of potential adopters, thus their efficiency can also be analyzed through the theoretical lenses of the adoption and diffusion of an innovation (Agarwal & Prasad, 1997). Research moving in this direction highlighted different aspects that might be able to alter the degree of adoption, such as informal access to expertise and responses to social pressure (Frank, Zhao, & Borman, 2004), GDP per capita and access cost (Kiiski & Pohjola, 2002), or innovativeness related to the comprehension and the adoption of new products in a specific domain of consumer behavior (Goldsmith & Hofacker, 1991). Of particular relevance for this study is the adaptation of the innovativeness’ concept to the IT domain, which was conceived by Agarwal & Prasad (1998). The authors draw upon marketing researches and notions in order to apply innovativeness in the field of technology acceptance. Specifically, they created the construct of Perceived Innovativeness in the IT domain (PIIT), defined as “the willingness of an individual to try out any new information technology" (Agarwal & Prasad, 1998, p. 206). This construct is conceptualized as a trait, that is a relatively stable descriptor of individuals being invariant across situational considerations.
4
Table 1
Construct Definition
Relative Advantage The degree to which an innovation is perceived as being better than its precursor
Ease of Use The degree to which an innovation is perceived as being easy to use
Compatibility The degree to which an innovation is perceived as being consistent with the values, needs, and past experiences of potential adopters
Image The degree to which use of an innovation is perceived to enhance one's image or status in one's social system
Result Demonstrability The tangibility of the results of using an innovation
Observability The degree to which the results of an innovation are observable to others
Trialability The degree to which an innovation may be experimented with before adoption
Sources: Moore, G. C., & Benbasat, I. (1991). Development of an Instrument to Measure the Perceptions of Adopting an Information. Information Systems Research, 192-222
Agarwal, R., & Prasad, J. (1997). The Role of Innovation Characteristics and Perceived Voluntariness in the Acceptance of Information Technologies. Decision Sciences, 557-582
5 Previous research applied TAM within different contexts in order to cross-validate the results related to the model and to assess the validity and reliability of the original perceived usefulness and perceived ease of use instruments proposed by Davis. Nevertheless, previous research has paid little attention to the technology acceptance phenomenon within organizations such as universities. In this context, a relevant typology of Information System aimed to be used by students is the e-platform called web student portal. Generally speaking, web portals are gateways to information resources and different kinds of services. They are specially designed web sites that bring information together from diverse sources in a uniform way. E-portals are the equivalent of intranets, which are defined as “internal or private networks of an organization based on internet technology (such as hypertext and TCP/IP protocols) and accessed over the internet. An intranet is meant for the exclusive use of the organization and its associates (customers, employees, members, suppliers, etc.) and is protected from unauthorized access with security systems” (www.businessdictionary.com). Portals are commonly used in colleges and universities where prompt information and necessary updates must be readily available to a large number of students. For the practical reasons listed above, the university of Groningen has introduced its first student portal in mid-November, indicating its function as the new dashboard for students and as a start page with all links to all university information and applications. The purpose of the new portal is to gather information regarding multiple online services provided by the university such as Nestor, faculty intranets and infonet (My University).
6
2. Theoretical Background
In response to the preponderance of Information Technology in all the aspects related to the organizational and individual level, IT acceptance and usage have been the main concerns in IS research (Venkatesh, Morris, Davis, & Davis, 2003). Given the constantly growing investments in IT with the expectation of obtaining a positive output and enhanced productivity, an underutilized system could represent an issue for most of the organizations. Therefore, considering the elevated risk of failure of new information systems, organizations need to be able to more accurately predict the outcome of their IS development efforts (Szajna & Scamell, 1993). Understanding and creating the conditions under which information systems will be comprehended and embraced by the human aspect of an organization is a crucial issue if the organization wants to reap the benefits from the technology introduction (Venkatesh & Davis, 2000). For this reason, significant progress has been made over the last decades in explaining and predicting user acceptance of IT with regard to different types of software such as telemedicine software (Hu P. J., Chau, Liu Sheng, & Tam, 1999), configurational software (Gefen & Keil, 1998) or the use of debugging tools (Bajaj & Nidumolu, 1998). In particular, Davis (1989) established the Technology Acceptance Model, based on constructs and relationships in the theory of reasoned action by Fishbein and Ajzen (1975) (Karahanna, Agarwal, & Angst, 2006). The purpose of this research is the application of TAM to examine the impact of pre-adoption beliefs, expectations and intentions on post-adoption technology use of the new student portal introduced by the university of Groningen. The theory of reasoned action, followed by the theory of planned behavior, have been proved to predict how individuals will behave based on their pre-existing attitudes and behavioral intentions (Mathieson, 1991; Ajzen, 1991; Taylor & Todd, 1995; Armitage & Conner, 2001). Thus, these theories are considered as a valid starting point for the majority of TAM-based researches which are aimed at investigating the (direct or indirect) relationship between behavioral intentions to use a new IS and its actual usage. In the next paragraph I will propose the main differences between the two theories, and subsequently I will introduce TAM which will provide the bases for the hypotheses development.
2.1 Theory of Reasoned Action (TRA) and Theory of Planned Behavior (TPB)
7 Ajzen, 1975, p. 216); the second, Subjective Norm, has been defined as “the person’s perception that most people who are important to him think he should or should not perform the behavior in question” (Fishbein & Ajzen, 1975, p. 302). The theory postulates that behavioral intention, which is the main factor for the determination of actual behavior, is a “function of salient information or beliefs about the likelihood that performing a particular behavior will lead to a specific outcome” (Madden, Ellen, & Ajzen, 1992, p. 7). The intent to perform a behavior depends on the result of the measures of attitude toward behavior and subjective norm. A positive result will be the expression of behavioral intent (Glanz, Lewis, & Rimer, 1997). Davis et al. (1989) applied TRA to individual acceptance of technology and found that the variance explained was largely consistent with studies that employed TRA in the context of other behaviors (Venkatesh, Morris, Davis, & Davis, 2003).
The theory of planned behavior, as an extension of TRA, has its roots in behavioral intentions and how they can influence future usage. More specifically, the individual’s intention to perform a given behavior determines how hard people are willing to try, and how much of an effort they are planning to exert, in order to perform a behavior (Ajzen, 1991). TPB differs from TRA as it proposes that behavioral achievements depend jointly on motivation (intention) and the ability to perform a specific task. In addition to the constructs of attitude toward behavior and subjective norm, TPB also includes Perceived Behavioral Control (PBC) as an individual perception that can predict behavioral intentions and, in turn, the actual behavior. PBC refers to “people’s perception of the ease or difficulty of performing the behavior of interest” (Ajzen, 1991, p. 183). Figure 1 illustrates the TPB model. The addition of PBC was aimed at the prediction of
Figure 1 – TPB Model
8 behaviors that do not occur under complete volitional control. Therefore, while TRA can accurately predict behaviors that are not associated with an external locus of control (that is under voluntariness), the presence of an intention was proved to be not sufficient to determine behavior under circumstances in which the behavior involves mandatory requirements (Armitage & Conner, 2001). In fact, the formulation of TRA didn’t take into account the fact that the determinants of performing an action might be influenced by forces that are external to an individual’s voluntariness. By including a construct related to the perception of easiness or difficulty in exerting a behavior, TPB argues that behavioral intentions are also driven indirectly by control beliefs and perceived facilitation. According to Mathieson (1991) a control belief is a “perception of the availability of skills, resources, and opportunities”, while perceived facilitation is the “individual's assessment of the importance of those resources to the achievement of outcomes”. As well as control beliefs, behavioral and normative beliefs that are relevant to the behavior are also considered to be the prevailing determinants of a person’s intentions and actions. As shown by Table 2, each determinant of behavior is a function of salient information, or beliefs, relevant to the behavior. Hence, the theory of planned behavior is based on the principle that the more positive the
Table 2 – Determinants of Behavioral Intention in TPB
Attitude toward behavior Subjective Norm Perceived Behavioral Control bi = behavioral belief i
ei = outcome evaluation of belief i
n = number of salient outcomes
ni = normative belief about referent other i
mi = motivation to comply with referent other i
n = number of salient others
ci = control belief about availability of skill, resource, or opportunity i
pi = perceived facilitation of skill, resource, or opportunity i
n = number of salient skills, resources, or opportunities
Sources: Mathieson, K. (1991). Predicting user intentions: comparing the technology acceptance model with the theory of planned behavior. Information systems research, 173-191
Ajzen, I. (1991). The theory of planned behavior. Organizational behavior and human decision processes , 179-211
9 Hukkelberg, 2010). Recently TPB was used to predict individual behavior in several different situations, examining intentions related to tourism (Quintala, Leeb, & Soutar, 2010), healthcare (McEachana, Conner, Taylor, & Lawton, 2011), food purchasing (Alam & Sayuti, 2011), screening programs attendance (Cooke & French, 2008), use of social networks (Pelling & White, 2009) and many other categories of behavioral intentions. However, the Technology Acceptance Model has been proved by previous research to be able to predict a larger amount of variance about the relationship between intentions and behavior. According to Mathieson (1991), not only TAM explained more variance than TPB, but it also explained attitude towards using an IS much better than TPB. Hence, the author proposes TAM to be the model of choice when attitude toward behavior is a construct of particular interest within the investigation. Moreover, TAM was conceived by Fred D. Davis (1986) specifically for estimating intentions to use an Information Technology and it was introduced to analyze the behavior of IT users over a vast number of technologies and over different types of user characteristics. Furthermore, even though there is not a consistent number of studies that discuss about electronic portals acceptance, previous research relied on TAM in order to predict usage of such portals, especially those that are aimed to be used by students (Abuhamdieh & Sehwail, 2007; Chang, 2004; Presley & Presley, 2009). Thus, for the reasons listed above, TAM will be the reference model proposed for my research. In the next section I will present the model by illustrating in details the relationship between its constructs (hypotheses development) and the introduction of Perceived Innovativeness into the model.
2.2 Technology Acceptance Model (TAM)
10 perceived ease of use as independent variables, while attitude towards use and behavioral intention (BI) as dependent variables which will in turn determine the actual usage of a new IS. Figure 2 shows the original
Figure 2 – Early TAM
Source: Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1989). User Acceptance of Computer Technology: A Comparison of Two Theoretical Models. Management Science, 982-1003
Technology Acceptance Model, which has been widely used in the last decades to determine how well intentions to use a new technology can predict its actual use. As hinted before, several authors studied TAM and found that the model is able to predict a consistent proportion of the shared variance (in average approximate to 40%) in usage intentions and behavior, and that the model is compatible with alternative models such as TRA and TPB (Venkatesh & Davis, 2000).
Despite of its widespread notoriety, the model has been modified and implemented by researchers over time in terms of the relationship between its constructs. Particularly relevant for this research is the role of attitude towards use within TAM. Even though attitude has been proved to be very effective in predicting an individual’s behavior (as evidently explained by TRA and TPB), a stream of research on IT adoption has discounted the role of attitude in explaining technology acceptance behavior (Kim, Chun, & Song, 2009). This construct, as figure 2 displays, has been introduced in the original model with the purpose of mediating the impact of PU and PEOU on IS actual use. In contrast with their expectations, Davis, Bagozzi, & Warshaw (1989) discovered that attitude did little to help elucidate the linkages between beliefs and intentions. The authors found that attitude offers little value in predicting IS use, leaving the two independent variables of PU and PEOU as powerful and parsimonious predictors. Therefore, the construct has been gradually omitted from technology adoption studies. Some studies have excluded the attitude component from TAM because it did not fully mediate the effect of PU and PEOU (Venkatesh & Davis, 2000). However, other studies shed light on the importance of attitude in the Technology Acceptance Model. For example, results from the study conducted by Kim, Chun, & Song (2009) indicate that even in
11 the case of a group of future users with a weak attitude, the latter may not be ignored as its effect on BI is greater than that of PU. In addition, Mathieson (1991) demonstrated that TAM was even able to explain a higher percentage of variance in attitude than TPB (precisely 0.727 for TAM and 0.388 for TPB). Moreover, the A-BI relationship is fundamental to TRA, which is the theoretical grounding of TAM. This relationship implies that, all else being equal, people form intentions to perform behaviors toward which they have positive affect (Davis, Bagozzi, & Warshaw, 1989). Research in the field of social psychology verified that an individual's attitude towards the behavior (such as using an IS) has a higher predictive power in determining the actual behavior rather than the individual’s attitude towards objects involved in the behavior (such as the PEOU and PU of a new IS) (Mathieson, 1991).
In the context of IT acceptance, attitude toward the use of a new Information System refers to the evaluative judgment of adopting a technology which in the majority of cases will represent an innovation for the organization that seeks to introduce it. For this reason, some authors (Moore & Benbasat, 1991; Agarwal & Prasad, 1998; Zhou, 2008; López-Nicolás, Molina-Castillo, & Bouwman, 2008) decided to integrate the study of TAM with concepts related to the Diffusion of Innovations theory developed by Everett Rogers in 1962 and revised by the same Rogers in 1983 and 1995. Specifically, the purpose of this paper is to analyze to which degree the students’ beliefs about the introduction of new student portal will affect the actual usage of the new system. Before the introduction of the portal, students had different electronic platforms that provided different services. One of those platforms is My University, an intranet access point in which students could find specific information on their personal homepage called Dashboard. A small part of the Dashboard contains information for everyone, while the rest can be furnished to taste, using what are known as widgets. The second main e-service now incorporated into the student portal is called Nestor, a platform where students could get informed about courses’ material (such as course info, documents and announcements), schedule timetable and other features related to didactic and teaching such as search engine for academic journals and papers. Other services that are now part of the portal are ProgRESS (which provides information about grades and credits) and NEXT Career (which helps students to find the latest job offers and upcoming events associated with job offering).
12 which users consider the new system as more innovative than its predecessor. Moore & Benbasat (1991) study was the first to apply the concept of relative advantage to the field of technology acceptance. They argued that the construct was similar to Davis’ perceived usefulness, adopting its items for the measurement of the construct. Agarwal & Prasad (1997), in their investigation on innovation characteristics of IT and acceptance outcome, examined the notion of relative advantage and found a lack of significance of relative advantage in predicting current usage. Due to the fact that this result is in contradiction with prior relevant research (such as Moore & Benbasat in 1991 and Davis, Bagozzi, & Warshaw in 1989), and because the perceived characteristics of an innovation appear as key constructs in the domain of technology acceptance, the authors recommend to identify a different set of perceptions which can be able to capture how the perceived relative advantage influence the IS use. In addition, Carter & Bélanger (2005) found no correlation between relative advantage and intention to use, spurring future research to further clarify and investigate its role. Hence, the main contribution of this paper to the literature stream of technology acceptance is the introduction of a novel construct that, in line with relative advantage, is designed to comprise a set of perceptions and beliefs regarding the degree of innovativeness of a new IS perceived by a group of future users. This construct, called Perceived Innovativeness, has been defined as “the degree to which using the new student portal is perceived to be more stimulating than using its predecessors in terms of functionality, efficiency, layout interactivity and upgradedness”. An important distinction has to be made between PI and Personal Innovativeness in the IT domain (PIIT). In fact although they both measure individual perceptions regarding the use of IT, PIIT was developed to capture the individual skills and propensity toward using a given technology (Agarwal & Prasad, 1998), while PI’s purpose would be to capture the overall innovativeness of an IS when compared with its predecessors.
13 previous literature were not capable of establishing the link between RA and usage intentions, thus generating mixed empirical results. For example, Agarwal and Prasad (1997) found that compatibility, visibility, trialability, and voluntariness all had a direct and significant effect on Internet usage, whereas the effects of relative advantage and ease-of-use were not significant. As a consequence, the authors suggest future research to identify a different set of perceptions which could then be used as a basis for the construction of more complex models that include other explanatory variables. Mixed results, coupled with the ambiguous association between RA and PU, demand for the conceptualization of perceived innovativeness, which will be applied in this research to capture students’ perceptions of the portal’s innovativeness along four dimensions: functionality, efficiency, layout interactivity and upgradedness. The novel construct also differs from relative advantage in that it is measured with a completely new set of perceptions (discussed in the methodology section) instead of with items taken from previous research. Examples of such items are “the new Student Portal will have more functions than the current information systems” and “the layout design of the new Student Portal will be more interactive than the current information systems”, with a 7-level Likert scale response option ranging from strongly agree to strongly disagree (the complete list of items for the construct can be found in the Appendix section). Given that, as argued before, the new student portal can be considered as an incremental innovation for the university of Groningen, the attitude toward using it can be influenced directly through the effect of perceived innovativeness. The basic idea is that innovations that appear to be beneficial when compared to other methods, both current and previous, are more likely to be adopted. Perceived innovativeness, in turn, may be affected by perceived usefulness and perceived ease of use, which represent the most efficient determinants of attitude in the technology acceptance literature. Therefore, drawing on Diffusion of Innovations theory, I posit that students will adopt the new student portal only if they perceive a certain degree of innovativeness over previous information systems such as My University, Nestor and ProGRESS. This approach is consistent with the work of Choudhury & Karahanna (2008) who studied the influence of three dimensions (convenience, trust, and efficacy of information acquisition) of relative advantage on the adoption of electronic channels.
14 (Venkatesh, Morris, Davis, & Davis, 2003). The exclusion of attitude is due to the fact that the construct, defined as an individual positive or negative feeling about performing the behavior, has mainly been examined in the field of IT acceptance within mandatory settings. However, the academic portal of Groningen’s university includes both volitional and mandatory use, as it offers some functions that can be accessed only through the portal while other functions can be also accessed through other means. In this context intrinsic motivation, that is the self-desire to seek out new things and new challenges and to gain new knowledge, is expected to have as much of an impact as extrinsic motivation (Hsu & Chuan-Chuan Lin, 2008).
Considering that attitude may be able to affect intentions in the case of student portal acceptance, it is arguable that the degree of perceived innovativeness can determine whether a student will have a positive or negative feeling about using the new system. In other words, the extent to which the new student portal is perceived to be more innovative and functional than its predecessors can determine the student’s attitude toward usage. Besides, it can be argued that PU and PEOU will have a positive impact on perceived innovativeness. In fact, the more the student portal will appear to students as useful and easy to use, the higher level of advantages compared to its predecessors will be perceived. As indicated by Figure 3, the construct of perceived innovativeness
Figure 3 – Conceptual Model
15 predictions, prediction merely allows us to know the relationship between independent and dependent variables but it is not capable of explaining the relationships. One way to understand how or why the variables are associated in a certain manner is to investigate the mechanisms that underlie the relationships, by examining the presence of mediators in relationships among variables (Cheung & Lau, 2007). Since research in IS acceptance omitted the construct of attitude as Davis, Bagozzi, & Warshaw (1989) proved it to have a not significant correlation in its direct relationship with PU and PEOU, I propose an indirect impact of PU and PEOU on attitude through the mediating effect of perceived innovativeness. Given that Rogers (1995) in his Diffusion of Innovations (DOI) theory argues that the attribute of relative advantage is critical to the attitude an individual forms about a new technology, the purpose of this paper is to introduce this novel mediator (perceived innovativeness) which is in line with the relative advantage conceived by Rogers (1995). The conceptual model presented in Figure 3 is aimed at explaining how perceptions, attitudes and intentions may lead to actual behavior, hence examining psychological processes of the human mind. According to Shrout & Bolger (2002) “mediation models of psychological processes are popular because they allow interesting associations to be decomposed into components that reveal possible causal mechanisms” (p. 422). In addition, they argue that those models are useful in psychology research as they allow to develop and test theory as well as to identify possible opportunities for intervention in applied work (Shrout & Bolger, 2002). Therefore I will test at first the impact that perceived usefulness and perceived ease of use, considered as the main predictors of IT acceptance, have on perceived innovativeness. Secondly, I will look for a significant effect in the correlation between perceived innovativeness and attitude toward using the student portal. The subsequent step will be to examine the widely known and well-established relationship that constitute the basis of TRA and TAM, that is the one between attitude and intentions to use the new student portal. As last step, I will test the relationship between intentions and actual usage in order to verify whether the former has predictive power in determining the latter. Overall I will test, through the help of the SmartPLS software, the following four hypotheses:
Hypothesis 1a: The direct relationship between Perceived Usefulness and Attitude toward using the new Student Portal will be mediated by Perceived Innovativeness.
Hypothesis 1b: The direct relationship between Perceived Ease of Use and Attitude toward using the new Student Portal will be mediated by Perceived Innovativeness.
Hypothesis 2: Attitude toward using the new Student Portal will positively affect users’ intentions to use the system.
16 The purpose of this study is to provide a quantitative analysis of user acceptance of the new student portal recently introduced by the university of Groningen. The focal contribution of this research is the extension of the original Technology Acceptance Model to include the construct of perceived innovativeness, which represents a deeper elaboration of the concept of relative advantage created by Rogers (1995). The four hypotheses will be tested with the help of SmartPLS 2.0, a software with graphical user interface for variance-based structural equation modeling (SEM) using the partial least squares (PLS) method.
3. Methodology
3.1 Data Collection
In this research, I collected both primary and secondary data. Consistent with the majority of previous studies, primary data were collected through questionnaires which include items validated in prior research. The sample for the study is composed of 102 students of the university of Groningen in The Netherlands. Collection of data occurred in conjunction with the University’s implementation of a student portal whose purpose is to replace the extant e-services (course enrolling, course material, library etc.) with a single platform capable to gather all the essential information. Questionnaires were handed to students both manually and via email. Primary data were collected at the individual level and the respondents were asked to provide quantitative evaluations of their relationship with the new student portal. All the constructs in the model exclusively include items related to the individual level. On one hand, those items measure individual perceptions, attitudes, and intentions toward using the student portal, while on the other hand they capture the application of intentions that leads to the actual usage of the system. Secondary data were taken from prestigious and reliable sources; first, from IS journals such as Information & Management, MIS Quarterly, Journal of Strategic Information Systems and Management Science. Second, from journals and magazines related to social psychology and human resources, such as Journal of Applied Social Psychology and Human Resource Management Review.
17 time were divided in different sections, where each section was specifically dedicated to one construct and its items. The first wave of survey, spread out right before the launch of the student portal, was aimed at identifying perceptions and behavioral intentions of students toward the system’s usage. The second wave was collected one week after the portal’s introduction and differs from the first in that it also includes system’s actual usage. The final dataset was the result of melting the two waves of survey by matching each respondent’s perceptions and intentions to behave with his/her actual behavior (in order to analyze the degree of influence that pre-adoption beliefs have on post-adoption actual usage and satisfaction). Although the first survey reached 159 respondents, some of them didn’t complete the second one, leaving the number of the sample up to 102 students. The sample comprises of 56 males (54,9%) and 46 females (45,1%), with an education level ranging from bachelor student to PhD student, and an age ranging from 18 and 35 years. Table 3 presents the demographic of the research sample.
Table 3 – Demographics of the respondents
3.2 Measurements
Perceived Ease of Use and Perceived Usefulness. The first section of the questionnaire is dedicated to the two core constructs of TAM. According to Mathieson (1991) while measures of TPB’s beliefs (attitude, subjective norm, perceived behavioral control) need to be developed for each research context,
Profile Frequency Percentage
18 TAM is measurable in a standard way through the instruments established by Davis (1989). Thus we rely on items from prior research to analyze the core constructs of TAM. “The conceptual definitions of perceived usefulness and perceived ease of use were used to generate 14 candidate items for each construct from past literature” (Davis F. , 1989, p. 323). This section will contain measuring items such as “Using the system saves me time” and “Using the system enhances my effectiveness on the job” for perceived usefulness, and such as “The system is rigid and inflexible to interact with” or “I find it easy to get the IS system to do what I want it to do” for perceived ease of use (Davis F. , 1989).
Perceived Innovativeness. The second section is dedicated to this construct, which is aimed at capturing in an extensive way the perceptions held by students about the innovativeness of the new student portal. Although perceived innovativeness has never been used in prior research, it shares similarities with the construct of relative advantage developed by Rogers (1995) in his Diffusion of Innovation theory. Rogers (1995) defines relative advantage as “the degree to which an innovation is perceived as being better than the idea it supersedes” (p. 212). However, the vast majority of the studies about technology acceptance and innovation diffusion uses items that, consistent with Rogers (1995), resemble those of perceived psefulness (Agarwal & Prasad, 1997; Carter & Belanger, 2004) or that do not address the task of predicting behavioral intentions (O'Malley & McCraw, 1999; Carter & Bélanger, 2005). Therefore, the purpose of this study is to extend the understanding of this concept by giving the construct new dimensions of innovativeness, such as functionality, efficiency, layout interactivity and upgradedness. Perceived innovativeness has been defined as “the degree to which using the new student portal is perceived to be more stimulating than using its predecessors in terms of functionality, efficiency, layout interactivity and upgradedness”. Christensen (1997) argues that a sustaining or incremental innovation occurs when the organization improves existing product functionality for existing customers and markets. Efficiency improvements, in this case improvements in managing information and services provided by the portal, are considered to be one of the advantages that result from the deployment of an innovation (Benner & Tushman, 2003). As regards the third dimension of innovativeness, Ahn, Ryu, & Han (2007) argue that interactivity is an antecedent of playfulness, which refers to the entertainment value of a website and has been often associated with innovation by previous research (Woszczynski, Roth, & Segars, 2002). Therefore layout interactivity is supposed to impact on the overall quality of the incremental innovation represented by the portal. At last, since the portal is an advanced version of the previously available services, upgradedness is considered as an important element to assess the degree of innovativeness of the new system. The four items of this construct are displayed in the Appendix.
Attitude Toward Use. Attitude is represented in many ways such as evaluations, representations in
19 considered (Bajaj & Nidumolu, 1998). In line with previous research (Taylor & Todd, 1995; Moon & Kim, 2001), the construct of attitude comprises of items studied and validated by Davis, Bagozzi, & Warshaw (1989) in the context of technology acceptance (applied from the theory of planned behavior developed by Fishbein & Ajzen). Hence, attitude toward usage of the student portal is measured through items such as “I like the idea of using the new Student Portal” and “using the new Student Portal is pleasant”, in order to capture students’ positive or negative feeling (evaluative affect) about using the new system.
Behavioral Intention. The fourth section is dedicated to an essential construct that was conceived in
the first place by Fishbein & Ajzen (1975) to mediate the relationship between individuals’ perceptions about performing a behavior and the actual behavior. Similarly to attitude, the items developed by Fishbein & Ajzen (1975) for behavioral intention are still used nowadays to predict user acceptance of Information Technology (Venkatesh & Davis, 2000; Venkatesh, Morris, Davis, & Davis, 2003; Hsu & Chuan-Chuan Lin, 2008). Consistent with the extant literature, the items adopted for this study are, for instance, “I will try to use the new Student Portal in the forthcoming months” or “I plan to use the new Student Portal in the forthcoming months”.
Actual Usage. Usage is the dependent construct in the model. According to Bajaj & Nidumolu (1998), usage is the amount of time the user feels he/she has spent using the system. The authors used three measures for actual behavior: two were adapted from Davis, Bagozzi, & Warshaw (1989) and one from Leonard-Barton & Deschamps (1988). Examples of those measures are a 7-point Likert scale item, with very infrequent/very frequent as endpoints, and a 6-item checkpoint scale with categories: never tried it at all, tried it once but not since then, used it earlier but stopped now, used it for about ten percent of the time I use my computer, used it for between 10% and 50% percent of the time I use my computer, used it for more than 50% of the time I use my computer (Bajaj & Nidumolu, 1998, p. 218).
20 perceptions and intentions have on usage (Agarwal & Prasad, 1998), therefore the study also controls for this variable.
3.3 Data Analysis and Hypotheses Testing
The final dataset was imported in SmartPLS v. 2.0, a software with graphical user interface for variance-based structural equation modeling (SEM) using the partial least squares method (http://www.smartpls.de, 2015). The software can be adopted in empirical research, especially in quantitative approaches, to analyze collected data from surveys such as questionnaires and test the correlation among constructs. Two are the approaches to SEM analysis: while covariance-based analysis typically involves the method of maximum likelihood, variance-based analysis (also called component-based analysis) uses least square functions and is known as partial least squares path (PLS) modeling. PLS path modeling allows researchers to perform and observe path-analytics modeling with latent variables (variables that cannot be directly observed). SEM has been used to assess the complex relationship between independent and dependent variables. It is considered a second generation method of multivariate regression established to efficiently cope with limitations of first generation procedures such as multiple regression, discriminant analysis, and analysis of variance. SEM is capable of simultaneously assessing the reliability and validity of the measures of theoretical constructs and estimating the relationships among these constructs (Kijsanayotina, Pannarunothaib, & Speediec, 2009). SmartPLS allows to estimate both the measurement model and the structural model through the PLS analysis (Hansmann & Ringle, 2004). The measurement model is assessed by estimating the construct validity and reliability, while the structural model will measure the beta coefficients in order to understand the relationships between the constructs and thus to test the hypotheses.
Assessing the measurement model (Reliability and Validity)
21 attitude and behavioral intention are considered as reflective constructs (Petter, Straub, & Rai, 2007). Such constructs rely on observed measures which are affected by underlying unobservable construct. Thus, changes in the underlying construct are supposed to cause changes in the indicators. For example, respondent variations in the latent construct of perceived ease of use will make all of its measures reflect this change (Petter, Straub, & Rai, 2007).
According to Hair, Hult, Ringle, & Sarstedt (2013), reflective measurement models are assessed on their internal consistency reliability and validity. The first step for assessing the measurement model is to verify the internal consistency of the items to check whether they load to the same construct. A reliability analysis of the measures using Cronbach’s α and composite reliability was conducted. Table 4 illustrates the constructs’ descriptive statistics. As shown by the Table, the internal consistency reliability estimates (α’s) and the composite reliability values are all above the commonly used threshold value for acceptable reliability of 0.70, which is the minimum value recommended by Nunnally (1978) to indicate a reasonably high reliability of the research measures and constructs (Ifinedo, 2011). As regards convergent validity, a common measure to establish it at the construct level is the average variance extracted (AVE), while at the item level is necessary that the (standardized) outer loadings’ value is 0,7 or higher (Hair, Hult, Ringle, & Sarstedt, 2013). Therefore, convergent validity was primarily investigated with regards to the values of the AVE that, as proposed by Fornell and Larcker (1981), should be greater than 0,5. As Table 4 reports, all constructs produced values greater than 0.5, thus satisfying the criterion set by Fornell and Larcker. Secondly, standardized outer loadings were examined. With respect to individual item loadings (Table 5) values exceed 0.7, in line with the threshold of statistical significance established by Chin (1998). An exception is represented by the item pi1, whose value was 0,6607 but it was retained as it approximates to 0,7. High outer loadings on a construct indicate that the associated indicators have much in common, which is captured by the construct. Once ascertained that the criteria for convergent validity are satisfied, the
Table 4 – Construct Reliability and Convergent Validity
Construct Cronbach’s alpha (α) Composite Reliability Average Variance Extracted
Perceived Ease of Use (PEOU) 0,8791 0,9109 0,6722
Perceived Usefulness (PU) 0,9197 0,9398 0,7579
Perceived Innovativeness (PI) 0,8340 0,8905 0,6731
Attitude toward usage (ATT) 0,8875 0,9303 0,8165
Behavioral Intentions (BI) 0,9539 0,9701 0,9155
Actual Usage (U) 0,8013 0,8832 0,7163
22
Table 5 - Outer Loadings
ATT BI PEOU PI PU U a1 0,8916 0 0 0 0 0 a2 0,9408 0 0 0 0 0 a3 0,8772 0 0 0 0 0 bi1 0 0,9644 0 0 0 0 bi2 0 0,9505 0 0 0 0 bi3 0 0,9555 0 0 0 0 peou1 0 0 0,7468 0 0 0 peou2 0 0 0,8096 0 0 0 peou3 0 0 0,8034 0 0 0 peou4 0 0 0,8733 0 0 0 peou5 0 0 0,8601 0 0 0 pi1 0 0 0 0,6607 0 0 pi2 0 0 0 0,8564 0 0 pi3 0 0 0 0,8437 0 0 pi4 0 0 0 0,9002 0 0 pu1 0 0 0 0 0,9045 0 pu2 0 0 0 0 0,9167 0 pu3 0 0 0 0 0,8768 0 pu4 0 0 0 0 0,7766 0 pu5 0 0 0 0 0,8712 0 u1 0 0 0 0 0 0,8275 u2 0 0 0 0 0 0,8124 u3 0 0 0 0 0 0,8967
23 that discriminant validity is established in the model, thus proving that each construct is unique and captures phenomena not represented by other constructs in the model (Hair, Hult, Ringle, & Sarstedt, 2013).
Table 6 – Discriminant Validity
Assessing the structural model (Hypotheses testing)
Once the reliability and the validity of the construct measures has been established, the next step is dedicated to the assessment of the structural model results. This entails the examination of the model's predictive capabilities and the correlation among the constructs. After running the PLS-SEM algorithm, estimates such as path coefficients are obtained for the structural model relationships, which represent the hypothesized relationships among the constructs. The strength of the relationship is indicated by the β coefficients while the R2 highlights the percentage of variance in the model and gives an indication of its predictive power. The SmartPLS 2.0 results for the βs and the R2 are illustrated in Figure 4. The bootstrapping procedure allows to estimate the path significance levels. Whether a coefficient is significant ultimately depends on its standard error that is obtained by means of bootstrapping, which allows to compute the empirical t-values. When the empirical t-value is higher than the critical value, it can be affirmed that the coefficient is significant at a certain error probability (i.e., significance level). According to Hair, Hult, Ringle, & Sarstedt (2013), commonly used critical values for two-tailed tests are 1.65 (significance level= 10%), 1.96 (significance level = 5% ), and 2.57 (significance level = 1 % ). This means that in order to achieve a significance level of 0.05 for a two-sided test, which is the widely accepted minimum value for a significant effect between two constructs, the absolute value of the test statistic (t-value) must be greater than or equal to the critical value 1.96. Figure 5 illustrates the t-values computed by SmartPLS.
24
Figure 4 – The results from SmartPLS of the structural model analysis (βs and R2s)
PEOU
PI ATT BI USE
PU
Figure 5 – T-values results from the bootstrapping algorithm from SmartPLS
PEOU
PU
Notes: **Significant at p <0.01; *significant at p < 0.05
25 The figure shows that all the relationships among the constructs present values above 1.96, which is the threshold for significance accepted by most of the studies in the literature field. However, before establishing the confirmation of the hypotheses, the mediation effect of perceived innovativeness between perceived ease of use/perceived usefulness and attitude has to be verified. There are several ways in which a mediation hypothesis can be tested. A wide-know statistically accurate approach is represented by the Sobel test, which is capable of producing a more direct analysis of an indirect effect “by comparing the strength of the indirect effect of the independent variable on the dependent variable to the point null hypothesis that it equals zero” (Preacher & Hayes, 2004, p. 718). The Sobel test statistics should be with an absolute value greater than 1.96 and the two-tailed probability should be less than 0.05. After running the bootstrapping I inserted the output of the path coefficients provided by the SmartPLS software into the website of Dr. Daniel Soper (http://www.danielsoper.com) that is capable of calculating with an algorithm the Sobel t-statistic as well as the two-tailed probability. Figure 6 displays the values calculated by the website.
Figure 6 – Sobel test values
26 The test of the indirect effect is given by dividing ab by the square root of the above variance and treating the ratio as a Z test (i.e., larger than 1.96 in absolute value is significant at the 0.05 level) (Kenny, 2015). The figure demonstrates that the direct effect of perceived usefulness and perceived ease of use on attitude toward using the new student portal is mediated by perceived innovativeness. The SmartPLS bootstrap reports that the t-values of the direct effect of PEOU and PU on ATT without the mediating effect are respectively 5.731 and 10.042, while the t-values of the direct effect in presence of the mediator PI are 2.751 and 4.019. Full mediation, that is the maximum evidence for mediation, is present when the inclusion of the mediation variable in the model brings the correlation between the independent variable and dependent variable down to zero. However, it is very unlikely to happen. The most likely outcome for the inclusion of the mediating variable is that the relationship becomes a weaker, yet still significant path (Baron & Kenny, 1986). This is the case of partial mediation, in which the mediating variable accounts for some, but not all, of the relationship between the independent variable and dependent variable. Those considerations, along with the results of the Sobel test shown in Figure 6, prove support for Hypothesis 1a and Hypothesis 1b. Hypothesis 2 posits that attitude toward using the new student portal positively affects users’ intentions to use the system. The positive and statistically significant relationship between attitude and behavioral intention (β = 0.288; t = 2,555) supports H2 and is consistent with the majority of the studies that include attitude in TAM (Moon & Kim, 2001; Schepers & Wetzels, 2007; Hsu & Chuan-Chuan Lin, 2008; Park, 2009) and with the typical assumption of TRA and TPB that attitude to behave positively affects behavioral intention (Ajzen, 1991). Hypothesis 3 posits that users’ intentions to use the new student portal positively affect the actual use of the system. Again, the positive and statistically significant effect of behavioral intention on actual usage (β = 0,462; t = 4,605) provides support for Hypothesis 3.This result is in line with both TRA/TPB and TAM that postulates that behavioral intention is the major determinant of usage behavior (Davis, Bagozzi, & Warshaw, 1989). To sum up, results show that all the four hypotheses in the conceptual model have been satisfied at a confidence interval of 95% (at least).
4. Discussion
27 these conditions is necessary, but not sufficient, for the prediction of actual usage. The study not only confirms the theoretical importance (assessed by TAM) of perceived ease of use, perceived usefulness and behavioral intention in ultimately predicting the actual usage of a technology, but also demonstrates that, in the context of student portal acceptance, other particular beliefs about a technology such as perceived innovativeness can significantly influence students’ usage of the new system.
28 predict students’ usage of the new student portal, I included four dimensions of innovativeness that were hypothesized to influence the students’ attitude and, in turn, their intentions to use and the actual usage. By doing so, I demonstrated (through the confirmation of H1a and H1b) that the degree to which an IT user perceive the system to be more innovative than its predecessor(s) affects the positive feeling about using the technology and, ultimately, its usage. The significance of PI can be also deduced by its R2 value (0,509), which implies that 51% of the variability between the two independent variables (PU and PEOU) and perceived innovativeness have been accounted for. As the human mind is harder to predict than organizational processes, any field that attempts to predict human behavior, such as psychology, typically has R2 values lower than 50% (Frost, 2013). Therefore, this finding furtherly demonstrates the explanatory power held by the novel construct of perceived innovativeness.
Another result that is worth to mention consists in the relevance of the attitude variable in predicting behavioral intentions. Data analysis has brought to the confirmation of H1a, H1b and H2, indicating that the attitude variable is crucial to the relationship between beliefs and intention. This result is in line with some authors that included and investigated users’ attitude within the IS acceptance context (Porter & Donthu, 2006; Schepers & Wetzels, 2007; Hsu & Chuan-Chuan Lin, 2008). However Davis, Bagozzi, & Warshaw (1989), in a first attempt to examine TAM, discarded attitude from the model due to its little help in elucidating the causal linkages between beliefs and intention. In particular, they found a not significant effect of PU and PEOU on attitude toward usage. As a consequence, the majority of the studies about technology acceptance dropped the construct from the analysis because it was previously proved not to ultimately influence the actual behavior. A clear example comes from the well-known UTAUT study developed by Venkatesh, Morris, Davis, & Davis (2003), in which the authors even hypothesize (and then confirm) that attitude toward using a technology has not a significant influence on behavioral intention. However, those findings are in contradiction with the two theoretical groundings on which TAM poses, TRA and TPB, whose aim is to explain the relationship between attitudes and behaviors within human action by predicting how individuals will behave based on their pre-existing attitudes and behavioral intentions (Fishbein & Ajzen, 1975; Ajzen, 1991). Hence, this paper re-establishes the importance of the attitude-intention relationship; as it can be observed by the results of Hypothesis 2 testing, the relationship between attitude and behavioral intention is positive and significant at p <0.01, with a t-value of 2.555. In addition PU and PEOU, the two core perceptions included in every TAM, were found to have a significant effect on attitude through the mediating effect of perceived innovativeness. As the main focus of this study was to provide a contribution to the IT acceptance literature by applying an extended version of TAM to the e-portal usage in an academic context (university of Groningen), the findings fully validated this contribution.
29 usage, as suggested by Hypothesis 3. Results from the SmartPLS regression analysis indicate that behavioral intention positively and significantly influences the usage of the student portal, providing full support for Hypothesis 3. This support is provided by the strong path coefficient of 0.462 between intention and usage, which represents a high level of significance (t-value = 4.605; significant at p <0.01) in the relationship and denotes that the higher are students’ intentions toward the adoption of the new student portal, the higher will be their time spent using the new system. To sum up, four are the tested and supported relationships among the constructs that were hypothesized to have a positive influence on actual usage. This means that each of the factors (or antecedents) presented in the model (PU, PEOU, PI, ATT, and BI) has a positive and significant predictive effect on the final usage of the student portal by the students of the Groningen university.
5. Conclusion
30 Given the crucial role of portals in providing a centralized access to the wide variety of informational resources of an organization, it is important to evaluate their acceptance and their use patterns. This study integrates concepts from the Technology Acceptance Model and Diffusion of Innovation theory into an insightful model of e-portal adoption within the university of Groningen. Results indicate that TAM remains one of the most influential models for the prediction of user’s behavioral intention and, ultimately, the actual usage of information systems. Although examined within a different perspective, that is the mediation effect of perceived innovativeness, perceived usefulness and ease of use are validated and confirmed as the two main factors influencing decisions of users as they approach a new technology. Through the inclusion of perceived innovativeness, this study extended the work of other researchers by investigating the motivations that induce students to adopt an intranet system such as the new student portal.
5.1 Limitations and future research suggestions
31
Acknowledgements: I would like to thank my supervisor Dr. Q. (John) Dong for the time he spent explaining
the statistical procedures, his guidance of my research as well as many helpful comments resulting in great improvements of my work. I also want to thank all the students of the University of Groningen that participated to the study by filing the questionnaires, thus making this research feasible.
Bibliography
Abuhamdieh, A. H., & Sehwail, L. (2007). A Comparative Study of Campus Portal User Acceptance: Student and Faculty Perspectives. Journal of STEM Education, 40-49.
Agarwal, R., & Prasad, J. (1997). The Role of Innovation Characteristics and Perceived Voluntariness in the Acceptance of Information Technologies. Decision Sciences, 557-582.
Agarwal, R., & Prasad, J. (1998). A Conceptual and Operational Definition of Personal Innovativeness in the Domain of Information Technology. Information Systems Research, 204-215.
Ahn, T., Ryu, S., & Han, I. (2007). The impact of Web quality and playfulness on user acceptance of online retailing. Information & Management , 263-275.
Ajzen, I. (1991). The theory of planned behavior. Organizational behavior and human decision processes , 179-211.
Ajzen, I. (2002). Constructing a TPB questionnaire: Conceptual and methodological considerations.
Alam, S. S., & Sayuti, N. M. (2011). Applying the Theory of Planned Behavior (TPB) in halal food purchasing. International Journal of Commerce and Management, 8-20.
Armitage, C. J., & Conner, M. (2001). Efficacy of the theory of planned behaviour: A meta‐analytic review. British journal of social psychology, 471-499.
Bailey, J. E., & Pearson, S. W. (1983). Development of a tool for measuring and analyzing computer user satisfaction. Management science, 530-545.
Bajaj, A., & Nidumolu, S. R. (1998). A feedback model to understand information system usage. Information & Management, 213-224.
Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of personality and social psychology.
32 Benner, M. J., & Tushman, M. L. (2003). Exploitation, exploration, and process management: The
productivity dilemma revisited. Academy of management review , 238-256. Bowen, W. (1986). The Puny Payoff from Office Computers. Fortune, 20-24.
Boyson, S., Corsi, T., & Verbraeck, A. (2003). The e-supply chain portal: a core business model. Transportation Research Part E: Logistics and Transportation Review, 175-192.
Carter, L., & Belanger, F. (2004). The Influence of Perceived Characteristics of Innovating on e-Government Adoption . Electronic Journal of e-Government , 11-20.
Carter, L., & Bélanger, F. (2005). The utilization of e-government services: citizen trust, innovation and acceptance factors. Information Systems Journal, 5-25.
Chang, P. (2004). The Validity of an Extended Technology Acceptance Model (TAM) for Predicting Intranet/Portal Usage. University of North Carolina, 1-34.
Cheung, G. W., & Lau, R. S. (2007). Testing mediation and suppression effects of latent variables: Bootstrapping with structural equation models. Organizational Research Methods, 296-325. Chin, W. W. (1998). The partial least squares approach to structural equation modeling. Modern methods
for business research, 295-336.
Choudhury, V., & Karahanna, E. (2008). The relative advantage of electronic channels: a multidimensional view. MIS Quarterly, 179-200.
Christensen, C. M. (1997). The Innovator's Dilemma: The Revolutionary Book that Will Change the Way You Do Business. Collins Business Essentials.
Cooke, R., & French, D. P. (2008). How well do the theory of reasoned action and theory of planned behaviour predict intentions and attendance at screening programmes? A meta-analysis . Psychology and Health, 745-765.
Davis, F. (1986). A technology acceptance model for empirically testing new end-user information systems: Theory and results. Massachusetts Institute of Technology, 1-291.
Davis, F. (1989). Perceived Usefulness, Perceived Ease of Use, and User Acceptance of Information Technology. MIS Quarterly, 319-339.
Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1989). User Acceptance of Computer Technology: A Comparison of Two Theoretical Models. Management Science, 982-1003.
Del Aguila-Obra, A. R., & Padilla-Meléndez, A. (2010). Organizational Factors Affecting Internet Technology Adoption. Internet Research, 94-110.
Doll, W. J., & Torkzadeh, G. (1988). The measurement of end-user computing satisfaction. MIS Quarterly, 259-274.
33 Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable variables and
measurement error. Journal of marketing research, 39-50.
Frank, K. A., Zhao, Y., & Borman, K. (2004). Social Capital and the Diffusion of Innovations within Organizations: The Case of Computer Technology in Schools . Sociology of Education, 148-171. Frost, J. (2013). Regression Analysis: How Do I Interpret R-squared and Assess the Goodness-of-Fit? Tratto
da The Minitab Blog: http://blog.minitab.com/
Gefen, D., & Keil, M. (1998). The impact of developer responsiveness on perceptions of usefulness and ease of use: an extension of the technology acceptance model. Acm Sigmis Database, 35-49.
Glanz, K., Lewis, F. M., & Rimer, B. K. (1997). Linking theory, research, and practice. Health behavior and health education, 19-35.
Goldsmith, R. E., & Hofacker, C. F. (1991). Measuring Consumer Innovativeness. Journal of the Academy of Marketing Science, 209-221.
Hair, J. F., Hult, T. M., Ringle, C. M., & Sarstedt, M. (2013). A primer on partial least squares structural equation modeling (PLS-SEM). Sage Publications.
Hansmann, K.-W., & Ringle, C. M. (2004). SmartPLS manual. Hamburg: University of Hamburg.
Hsu, C.-L., & Chuan-Chuan Lin, J. (2008). Acceptance of blog usage: The roles of technology acceptance, social influence and knowledge sharing motivation. Information & Management, 65-74.
Hu, P. J., Chau, P., Liu Sheng, O. R., & Tam, K. Y. (1999). Examining the technology acceptance model using physician acceptance of telemedicine technology. Journal of management information systems, 91-112.
Hu, P. J., Chau, P., Liu Sheng, O., & Tam, Y. K. (1999). Examining the Technology Acceptance Model Using Physician Acceptance of Telemedicine. Journal of Management Information Systems, 91-112. Ifinedo, P. (2011). Internet/e-business technologies acceptance in Canada's SMEs: an exploratory
investigation. Internet Research, 255-281.
Ives, B., Olson, M. H., & Baroudi, J. J. (1983). The measurement of user information satisfaction. Communications of the ACM, 785-793.
Karahanna, E., & Straub, D. W. (1999). The Psychological Origins of Perceived Usefulness and Ease-of-Use. Information & Management, 237-250.
Karahanna, E., Agarwal, R., & Angst, C. M. (2006). Reconceptualizing Compatibility Beliefs in Technology Acceptance Research. MIS Quarterly, 781-804.
Kenny, D. A. (2015). Mediation. Tratto da http://davidakenny.net.
34 Kiiski, S., & Pohjola, M. (2002). Cross-country diffusion of the Internet. International Economics and Policy,
297-310.
Kijsanayotina, B., Pannarunothaib, S., & Speediec, S. M. (2009). Factors influencing health information technology adoption in Thailand’s community health centers: Applying the UTAUT model. International Journal of Medical Informatics, 404-416.
Kim, Y. J., Chun, J. U., & Song, J. (2009). Investigating the role of attitude in technology acceptance from an attitude strength perspective. International Journal of Information Management, 67-77.
King, W. R., & He, J. (2006). A meta-analysis of the technology acceptance model. Information & Management, 740-755.
Klopping, I. M., & McKinney, E. (2004). Extending the technology acceptance model and the
task-technology fit model to consumer e-commerce. Information task-technology learning and performance journal, 35-48.
Leonard-Barton, D., & Deschamps, I. (1988). Managerial influence in the implementation of new technology. Management science, 1252-1265.
López-Nicolás, C., Molina-Castillo, F. J., & Bouwman, H. (2008). An assessment of advanced mobile services acceptance: Contributions from TAM and diffusion theory models. Information & Management, 359-364.
MacKinnon, D. P. (2008). Introduction to statistical mediation analysis. Routledge.
Madden, T. J., Ellen, P., & Ajzen, I. (1992). A comparison of the Theory of Planned Behavior and the Theory of Reasoned Action. Personality and Social Psychology Bulletin, 3-9.
Mathieson, K. (1991). Predicting user intentions: comparing the technology acceptance model with the theory of planned behavior. Information systems research, 173-191.
McAfee, A. (2002). The Impact of Enterprise Information Technology Adoption on Operational Performance: An Empirical Investigation. Production and Operations Management, 33-53. McEachana, R., Conner, M., Taylor, N. J., & Lawton, R. J. (2011). Prospective prediction of health-related
behaviours with the Theory of Planned Behaviour: a meta-analysis. Health Psychology Review, 1-48. Moon, J.-W., & Kim, Y.-G. (2001). Extending the TAM for a World-Wide-Web context. Information &
Management, 217-230.
Moon, J.-W., & Kim, Y.-G. (2001). Extending the TAM for a World-Wide-Web context. Information & Management, 230.
Moore, G. C., & Benbasat, I. (1991). Development of an Instrument to Measure the Perceptions of Adopting an Information. Information Systems Research, 192-222.
Nunnally, J. (1978). Psychometric methods. New York: McGraw-Hill.
