Mobile banking
Rens Jansen op de Haar 1476335
University of Groningen
Faculty of Economics and Business Msc Business Administration February, 2013
Grianestraat 5d 1055 EX Amsterdam +31642604885
rens.jansenopdehaar@gmail.com
The Determinants of Consumers’ Intention to Use Mobile Banking:
An Integrative Framework of TAM2 and Perceived Risk
Title: The Determinants of Consumers’ Intention to Use Mobile Banking:
An Integrative Framework of TAM2 and Perceived Risk Author: Rens Jansen op de Haar
Department: Marketing Qualification: Master Thesis Completion Date: 18 February 2013 Supervisors: dr. ir. M.J. Gijsenberg
Y.C. Ou
Management summary
This study integrates the perceived risk construct into the extended technology acceptance model (TAM2) of Davis and Venkatesh (2000) in order to investigate which factors determine consumers’ intention to use mobile banking. Where TAM2 originally focuses on the social influence and cognitive instrumental processes in a work environment, this study aims to discover whether these processes, and thus TAM2, are applicable in the mobile banking context as well. Moreover, the perceived risk construct is implemented into TAM2 as a negative utility facet because TAM2 only addresses the positive utility gains of technology adoption.
Three conclusions could be drawn. Firstly, it is found that consumers’ intention to use mobile banking is based on a trade-off between perceived usefulness (positive utility) and perceived risk (negative utility).
Secondly, this study discovers that perceived usefulness is only predicted by relevance and perceived ease-of-use. In contrast to TAM2, perceived usefulness is not influenced by subjective norm, image, output quality and result demonstrability in the mobile banking context.
Thirdly, perceived ease-of-use also appears to be a negative predictor of perceived risk. In other words, perceived ease-of-use indirectly affects consumers’ intention to use in a positive direction via both perceived usefulness and perceived risk.
Preface
Whilst writing my master thesis I gradually started to realize that I was approaching the end of an era. In this period of time I made a couple of friendships for life. Moreover, it was a time in which I was able to develop myself in various (human) facets of life.
Even though this may have caused some delay, I dare to say that the experience and memories of this era will definitely help me during the rest of my life. I would like to take the opportunity to thank my parents Herman and Joke, who gave me all the opportunities in life so far. I think you do not realize how thankful I am for your support.
Moreover, without Charlotte I would never be the person that I am today. You made me realize what is really important in life and you definitely was a great support during my master studies.
Last but not least, I would like to thank Maarten Gijsenberg for all the supportive comments that he gave me. Your clear feedback helped me a lot during the process of writing this thesis.
Table of content
Management summary ... 3
Preface………….. ... 4
Table of content ... 5
Chapter 1. Introduction ... 7
Chapter 2. Literature review ... 10
2.1 Technology acceptance ... 10
2.2 Conceptual model ... 12
2.3 Basic Technology Acceptance Model ... 13
2.3.1 Behavioral intention ... 13
2.3.2 Perceived usefulness ... 13
2.3.3 Perceived ease-of-use ... 14
2.4 Extended Technology Acceptance Model ... 15
2.4.1 Social influence processes ... 15
2.4.1.1 Subjective norm and intention to use ... 16
2.4.1.2 Subjective norm, image and perceived usefulness ... 16
2.4.1.3 Social influence processes and experience ... 18
2.4.2 Cognitive instrumental processes ... 19
2.4.2.1 Relevance ... 20
2.4.2.2 Output quality ... 21
2.4.2.3 Result demonstrability ... 22
2.5 Perceived risk ... 23
2.5.1 Perceived risk and trust ... 23
2.5.2 Theory of perceived risk ... 23
2.5.3 Facets of perceived risk ... 24
2.5.3.1 Performance risk ... 24
2.5.3.2 Financial risk ... 25
2.5.3.3 Privacy risk ... 25
2.5.4 Perceived risk and perceived ease-of-use ... 25
Chapter 3. Research methodology ... 27
3.1 Data collection ... 27
3.2 Structural Equation Modeling... 27
3.3 Individual construct definitions ... 28
3.4 Pre-test ... 29
3.5 Missing data ... 30
3.6 Sample size ... 30
Chapter 4 Data analysis and Results ... 31
4.1 Descriptive statistics ... 31
4.2 Measurement model ... 32
4.3 Structural model ... 35
4.4 Hypotheses testing ... 36
4.5 Data driven post-hoc analysis ... 38
4.5.1 Experience versus Non-Experience ... 41
Chapter 5. Discussion ... 43
5.1 Predictors of intention to use ... 43
5.2 Predictors of perceived usefulness ... 44
5.3 Predictor of perceived risk ... 45
5.4 Predictor of image ... 46
5.5 Moderators of perceived usefulness ... 46
5.6 TAM2 compared to IDT in the mobile banking context ... 46
5.7 Experienced versus Non-experienced customers ... 47
Chapter 6. Implications ... 48
6.1 Managerial implications ... 48
6.1.1 Experience ... 50
6.2 Academic implications ... 51
Chapter 7. Conclusion ... 53
7.1 Limitations and future research ... 54
Appendix A. Questionnaire ... 56
References……. ... 58
Chapter 1. Introduction
At the end of 2011, there were approximately 6 billion mobile subscriptions worldwide according to the International Telecommunication Union (ITU). This number is equivalent to almost 87 percent of the total world population. Compared with the number of subscriptions worldwide in 2010 (5.4 billion) and 2009 (4.7 billion), this means a tremendous increase. Portio Research Ltd. (2012) even predicts 6.5 million subscriptions at the end of 2012. Focusing on active mobile-broadband subscriptions, the ITU estimates 1.2 billion subscriptions in the world, which stands for 17 percent of total world population. This number has grown 45 percent annually over the last four years (ITU, 2011).
These numbers represent a trend for the forthcoming years in which the mobile phone will become more and more prominent. Plenty of tasks which were normally done by personal computer (PC) or in-store in the past, will be done by mobile phones or tablets in the future.
According to the ITU, who made a global top-25 of broadband economies at the beginning of 2011, something noteworthy occurs regarding the Dutch market. The Netherlands appears to be top leader considering the percentage of fixed-broadband subscriptions. However, it is absent on the list of top-25 economies with the highest number of active mobile-broadband subscriptions per 100 inhabitants. The annual report of the OPTA in 2011 shows similar figures. Where the Netherlands tops the international rankings when it comes to the use of fixed broadband, it ranks eighteenth with only 44 mobile broadband connections per every 100 residents. However, this report shows an enormous increase in the use of mobile-broadband connections in the Netherlands – from 6.3 million in 2010 to 8.2 million in 2011. More recently in November 2012, OPTA published new figures showing that the number of mobile broadband connections has exceeded 10 million in the Netherlands. Apparently there is an enormous growth of mobile broadband users in the Netherlands during the last two years.
Consequently with this trend, mobile banking turns out to be a popular alternative way of banking. Research provided by Berg Insight (April 2010) suggests that there will be 894 million users of mobile banking and related services like money transfers worldwide in 2015. A report of Global Industry Analysts, Inc. (February 2010) even predicts a global customer base of 1.1 billion by the year of 2015. This growth is driven by efforts of operators and banks in developing countries (particularly in Asia) to bank the unbanked. If considering Europe and North America, mobile banking is an extension of online banking since banks respond to the growth of mobile Web. In other words, mobile banking is in its early stages in Europe and is especially driven by convenience and value-add.
In the Netherlands, mobile banking has been introduced by the three leading banks (ING, ABN AMRO and Rabobank) at the end of 2011. Nine months after the introduction already 35 percent of the population has downloaded the mobile banking application on their mobile phone or tablet (ING,
2012). More recently, at the end of 2012, these banks published figures showing that in the meantime over three million unique people have really used a mobile banking app.
In conclusion, it is interesting to investigate which factors drive people to make use of mobile banking as a new technology in the Dutch market. In recent studies, the most accepted and robust model for explaining user acceptance of a new technology is the Technology Acceptance Model (TAM) by Davies (1989). Many recent research articles have utilized TAM as a base model to predict adoption (Teo et al., 1999; Gefen and Straub, 2000; Moon and Kim, 2001; Pavlou, 2001). According to TAM this adoption or intention to use is influenced by two determinants: perceived usefulness and perceived ease-of-use.
Venkatesh and Davis (2000) extend this model with both social influence processes and cognitive instrumental processes in the work environment. Therefore, their model is called TAM2. Compared with TAM, TAM2 includes some antecedents of perceived usefulness. Since perceived usefulness has been found to be the most important predictor of intention to use , TAM2 is more sophisticated than TAM. The constructs applicable for the social influence processes are subjective norm, voluntariness and image. The cognitive instrumental processes consists of the constructs job relevance, output quality, result demonstrability and perceived ease-of-use. Both processes significantly influence user acceptance by employees in a work environment. Extending to their research, this study will focus on consumers of mobile banking. One goal of this study is to find out which constructs of the TAM2 influence the intention to use mobile banking.
Where the TAM2 primarily focuses on the positive utility gains, it leaves out negative utility (potential losses) attributable to technology adoption. However, research on online purchase transactions shows that consumers have reluctance to complete these transactions particularly due to risk concerns (Pavlou, 2001). Therefore, Featherman and Pavlou (2003) implemented the perceived risk construct into the TAM in order to include negative utility as well. Where their research only combines the original TAM constructs (perceived usefulness and perceived ease-of- use) and perceived risk, this study combines the TAM2 constructs and perceived risk in order to find out which factors influence the intention to use mobile banking.
Although some constructs in TAM2 seem not that relevant in a mobile banking context beforehand, all constructs will be tested in order to find out to what extent the TAM2 is applicable in the mobile banking context. In other words, this study tests the usefulness and applicability of TAM2 in a context different from the work environment. Moreover, this study would like to discover to what extent the perceived risk construct is applicable for implementation into the TAM2 in the mobile banking context, since TAM2 leaves out negative utility constructs.
In conclusion, this results into the following research questions:
Main question for managerial implications:
- Which factors determine consumers’ intention to use mobile banking?
Sub-questions for academic implications:
- To what extent do the TAM2 constructs explain the intention to use mobile banking?
- To what extent does perceived risk explain the intention to use mobile banking beyond perceived usefulness and perceived ease-of-use?
Chapter 2. Literature review
This chapter will elaborate on the TAM2 constructs and perceived risk in further detail. Although TAM2 is the building block of the conceptual model, the basic TAM will be introduced first to
understand the principles of TAM2. First of all, a short introduction of technology acceptance in general delivers a short overview of recent research.
2.1 Technology acceptance
Over the decades, various theories have been developed in order to identify which factors cause people to accept and make use of Information Systems (IS). In 1989, Davis proposed the technology acceptance model (TAM) to explain the potential user’s behavioral intention to use a technological innovation in the work environment. Davis suggests that among the many variables that may influence system use, two determinants are especially important. First, potential users are more willing to use a new technology if they believe it will increase their job performance. He refers to this first variable as perceived usefulness. Second, even if people conclude that a given new technology is useful to them, it is still possible they refuse to use it because they believe the technology is too complicated to use. In this situation the performance benefits of usage are exceeded by the effort of using the system. So, according to TAM, usage behavior is also influenced by perceived ease-of-use.
Venkatesh (1999) demonstrates that TAM consistently explains the intention to use with approximately 40 percent of the variance. Other behavioral models like TRA score lower on their predictive power of intention to use. For that reason and also because of its understandability and simplicity (King and He, 2006), TAM has become one of the most widely used models in explaining the intentional use of IS.
Across the more than 20 years of research on TAM, perceived usefulness has been the strongest and most consistent predictor of usage intention. Therefore, Venkatesh and Davis (2000) introduce the extended TAM, also known as TAM2 in order to include additional constructs which represent the key antecedents of perceived usefulness and usage intention. TAM2 inserts both social influence processes and cognitive instrumental processes, which are expected to be the drivers of perceived usefulness. The constructs applicable for the social influence processes are subjective norm, voluntariness and image. The cognitive instrumental processes consists of the constructs job relevance, output quality, result demonstrability and perceived ease-of-use.
Besides these new constructs, Venkatesh and Davis (2000) implement experience as a moderator of both the relationship between subjective norm and perceived usefulness, and the relationship between subjective norm and intention to use.
Another well-known theory being used for relevant IT and IS studies is the Innovation Diffusion Theory (IDT) of Rogers (1995). This theory mentions five significant innovation characteristics:
relative advantage, compatibility, complexity, trialability and observability. These five characteristics are used to predict the rate of adoption. A study by Agarwal and Prasad (1998) suggests that only relative advantage, compatibility and complexity are consistently related to innovation adoption.
Wu and Wang (2005) combine both TAM and IDT in order to investigate which factors determine user mobile commerce acceptance. Based on the study of Chen, Gillenson and Sherrel (2002), who implement the compatibility characteristic into TAM to evaluate and explain consumer behavior, Wu and Wang (2005) suggest that TAM and IDT are extremely similar in some constructs and supplement one another. They state that the constructs of TAM are fundamentally a subset of the perceived innovation characteristics. For that reason, the literature review of this study briefly compares some of the constructs used in TAM2 with the characteristics used in IDT in order to be able to determine to what extent these two models show similarities and are able to explain users’ acceptance of new technologies.
The literature review is divided into three parts. In the first part, the literature review will discuss the relationship between perceived usefulness, perceived ease-of-use and intention to use as described in the basic TAM. In the second part, the social influence processes and cognitive instrumental processes will be elaborated on in further detail. Moreover, the moderating role of experience in the social influence processes is investigated. These two parts are separated because the original TAM is found to be the most reliable predictor of intention to use. However, TAM2 has only been used for a decade. For a reason, one of the main purposes of this study is to find out whether the TAM2 is applicable in the mobile banking context. Most probably, the basic TAM is applicable in this context. Therefore, the basic TAM will be elaborated on first in order to build an understandable basis for TAM2. Finally, in the third part, the perceived risk construct will be discussed in order to exhibit the possible loss when pursuing a desired result by using a new technology.
Where previous studies have used construct definitions which are applicable for the work environment, the construct definitions of this study are translated into the mobile banking context in order to form appropriate hypotheses. Furthermore, as already mentioned in the introduction, the TAM2 constructs will be briefly compared with the IDT characteristics during the discussion of the specific construct definitions.
2.2 Conceptual model
As noted earlier in the introduction, the conceptual model of this study integrates perceived risk into the TAM2 model in order to determine which factors influence customers’ intention to use mobile banking. A graphical overview of the conceptual model is presented in figure 2. The conceptual model consists of a TAM, TAM2 and perceived risk component. These components will be elaborated on step-by-step in the following paragraphs of the literature review.
Figure 2. Conceptual model.
H3 H1
H2 H9
H12 H11
H8 H4
H7
H10 H5
H13
Image H6
Subjective Norm
Relevance
Result Demonstrability Output Quality
Perceived Ease-of-use
Perceived Usefulness Intention to Use
Experience
Perceived Risk
2.3 Basic Technology Acceptance Model
Figure 3. Basic Technology Acceptance Model (TAM).
2.3.1 Behavioral intention
Behavioral intention is an indicator of the strength of people’s subjective probability that they will perform a specified behavior (Fishbein and Ajzen, 1975). TAM postulates that behavioral intention is the major determinant of usage behavior. Furthermore, TAM states that behavioral intention mediates the relationship between usage behavior on the one side and perceived usefulness and perceived ease-of-use on the other hand (Davis et al., 1989). In the mobile banking context, behavioral intention entails the probability that a person will make use of mobile banking now or in the future. In other words, behavioral intention is translated into intention to use in this study.
TAM posits that the adoption of a new information system is determined by users’ intention to use the system, which in turn is determined by the user’s beliefs about the system. Furthermore, TAM suggests that two beliefs are instrumental in explaining the variance in users’ intention (Luarn and Lin, 2005). Those two beliefs, perceived usefulness and perceived ease-of-use are elaborated on in further detail in the next two paragraphs. The basic TAM is graphically presented in figure 3.
2.3.2 Perceived usefulness
Perceived usefulness is defined as ‘the extent to which a person believes that using a particular system would enhance his or her job performance within an organizational context’ (Davis, 1989).
Therefore, a system is perceived as being useful if the user believes in the existence of a positive use- performance relationship (Teo, Lim and Lai, 1999). In other words, the user believes that the use of such a system would yield positive benefits for task performance.
This definition is used for an organizational context, but could be easily translated into a mobile banking context. In that case, perceived usefulness refers to the extent to which consumers believe that performing mobile banking would enhance the way they normally perform banking.
Wu and Wang (2005) indicate that perceived usefulness is similar to relative advantage, which is mentioned in IDT. According to Rogers (1995), relative advantage is the degree to which an innovation is perceived as better than the idea it supersedes. Tornatzki and Klein (1982) find that relative advantage is an important factor in determining adoption of new innovations.
H3
H2 Perceived Usefulness H1 Intention to Use Perceived Ease of Use
Perceived usefulness has also been closely linked to outcome expectations, instrumentality and extrinsic motivation (Davis et al., 1989). Many studies have shown that perceived usefulness is a strong determinant of user acceptance, adoption and usage behavior (Davis et al., 1989; Mathieson, 1991; Taylor and Todd, 1995). Therefore,
H1: Perceived usefulness will have a positive direct effect on intention to use.
2.3.3 Perceived ease-of-use
Perceived ease-of-use refers to ‘the degree to which a person believes that using a particular system would be free of effort’ (Davis, 1989). Radner and Rothschild (1975) state that effort is a finite resource that a person may allocate to the various activities he or she is responsible for. For that reason, Davis (1989) states that all else being equal, an application that is perceived as more easy to use is more likely to be accepted by users.
Regarding IDT, perceived ease-of-use could be linked with complexity (Wu and Wang, 2005).
Complexity is defined as the degree to which an innovation is perceived as relatively difficult to understand and to use (Rogers, 1995). This study should focus on the ‘difficulty to use’ component of the definition of complexity because this explains the extent to which a system is easy or difficult to use. For that reason, the ‘difficult to understand’ component of the definition should be neglected in this comparison. Moreover, it is noteworthy that the definition of perceived ease-of-use is positively formulated, where the complexity definition is negatively formulated. According to Rogers (1995), complexity is negatively related to the rate of adoption. In other words, TAM and IDT approach the relationship between ease-of-use and intention to use in the opposite direction, but they argue the same.
Davis et al. (1989) state that perceived ease-of-use is a direct determinant of perceived usefulness because, all else being equal, the less effortful a system, the more it can increase performance in a given period of time. Furthermore, there is extensive empirical evidence that perceived ease-of-use is significantly linked to intention both directly and indirectly via its impact on perceived usefulness (Davis et al., 1989; Venkatesh, 1999). Apparently, perceived ease-of-use has a direct effect on intention and an indirect effect on intention to use via perceived usefulness. In other words, if a system is more easy to use, it is more likely that people will use it. Therefore,
H2: Perceived ease-of-use will have a positive effect on perceived usefulness.
H3: Perceived ease-of-use will have a positive direct effect on intention to use.
2.4 Extended Technology Acceptance Model
Since the basic technology model has been elaborated on in the previous section, it is possible to focus on the relevant extended technology acceptance model. As mentioned before, the key determinants of perceived usefulness are divided into two categories, namely the social influence processes and the cognitive instrumental processes. Firstly, the social influence processes will be discussed. Those processes are graphically presented in figure 4.
2.4.1 Social influence processes
Figure 4. Social influence processes.
Social influence refers to the degree of interaction among people in their social context (Kim and Kim, 2003). Ventatesh (1996) states that social influence helps to determine whether technologies are adopted and whether products are purchased. For that reason, social influence may also affect the use of mobile internet services like mobile banking. Downes and Mui (1998) argue that the telecommunications industry is designed to facilitate social interactions. Since mobile Internet services are part of this telecommunication industry, social influence is likely to play a very important role in the adoption process of new technologies such as mobile banking. The social system consists of those individuals, organizations, or agencies that share a common ‘culture’ and are potential adopters of an innovation (Mahajan and Peterson, 1985).
Subjective norm is defined as ‘a person’s perception that most people who are important to him or her think that he or she should or should or should not perform the behavior in question’ (Fishbein and Ajzen, 1975). Davis et al. (1989) find that subjective norm has no significant effect on behavioral intention and therefore leave it out of the basic TAM. However, they acknowledge the need for additional research in order to investigate the impact of social influence processes like subjective norm on usage behavior. Apparently, Venkatesh and Davis (2000) find proof for the influence of subjective norm on usage behavior.
Another social influence process implemented into TAM by Venkatesh and Davis (2000) is image.
The rationale behind image is that a person’s position or status within a group could be enhanced if
H6
H4
H7 H5
Experience
Perceived Usefulness
Subjective Norm Image
they adopt the new technology (Moore and Benbasat, 1991). Both constructs and their consequences on the social influence processes will be adapted to the mobile context in the next two paragraphs.
2.4.1.1 Subjective norm and intention to use
Unlike TAM, subjective norm is included as a direct determinant of intention to use in TRA. The reason for this is that Fishbein and Ajzen (1975) argue that people sometimes choose to perform some behavior, even if they do not prefer to perform this behavior or its consequence. They only perform this behavior because one or more referents expect them to behave this way. Apparently, these referents posses the power to force people to perform this behavior. Recent studies provide mixed evidence for the relation between subjective norm and intention to use. Some studies show significant results (Mathieson, 1995), where other studies reject the relationship between subjective norm and intention to use (Taylor and Todd, 1995).
According to the meta-analysis of the technology acceptance model by Schepers and Wetzels (2007) there are large correlations between subjective norm and behavioral intention and between subjective norm and perceived usefulness. However, just like Venkatesh and Davis (2000), they find an interesting moderating effect of compliance. Compliance refers to the condition that the technology use is mandatory. Both studies show that subjective norm only has a significant effect on intention to use in mandatory settings. For that reason, Venkatesh and Davis(2000) implement voluntariness as a construct into TAM2 in order to distinguish between mandatory and voluntary usage settings. Voluntariness is defined as the extent to which potential adopters perceive the adoption decision to be non-mandatory (Hartwick and Barki, 1994).
However, currently mobile banking is in its early stages and could be considered as a voluntary alternative way of banking. Therefore, contrary to TAM2, the conceptual model of this study excludes the moderating effect of voluntariness on the relationship between subjective norm and intention to use. Moreover, the relationship between subjective norm and intention to use is expected to be absent in the conceptual model because of the voluntary setting in the mobile banking context.
2.4.1.2 Subjective norm, image and perceived usefulness
According to Venkatesh and Davis (2000) there are two theoretical mechanisms whereby subjective norm can influence perceived usefulness. In this case, subjective norm has an indirect effect on intention to use via perceived usefulness. The first mechanism deals with the relationship between subjective norm and perceived usefulness and the second mechanism deals with the indirect relationship between subjective norm and perceived usefulness via image.
The first mechanism of subjective norm is called internalization. Internalization refers to the phenomenon of incorporating a referent’s belief into one’s own belief structure (Warshaw, 1980). In other words, this social influence causes people to accept information from a referent as evidence about reality (Deutsch and Gerard, 1955). For that reason, Deutch and Gerard (1995) refer to internalization as informational social influence. In the mobile banking context, this mechanism means that people are sensitive to the behavior of people who are important to them. If their referents make use of the mobile banking alternative, those people are more prone to adopt this technology either because they presume that mobile banking should be useful. French and Raven (1959) compare internalization with expert power, which means that a person accredits expertise and credibility to influencing people based on their. Therefore,
H4: A positive subjective norm will have a positive direct effect on perceived usefulness.
The second mechanism of subjective norm, called identification, refers to the phenomenon that if important members of a person’s social system believe that this person should perform a specific behavior, then performing this behavior incline to elevate his or her position within the social system (Pfeffer, 1982). For that reason, Deutsch and Gerard (1995) refer to identification as normative social influence. In the mobile banking context, this means that important members of a person’s social group believe that this person should use mobile banking. In return, this person will receive more respect and a higher position within the social group if he or she performs the same behavior. For that reason, TAM2 theorizes that subjective norm will positively influence image.
However, where TAM2 is tested in the work environment, this research focuses on a new technology used by customers instead of employees. So the social system and the hierarchical structure in society differs from the hierarchical structure in organizations because the interrelationships are not the same in both situations. Where in the work environment technology acceptance will be based on hierarchical grounds, technology acceptance will be primarily based on demographics in real life society (Rogers, 1995).
Therefore, one aspect of the innovation diffusion theory of Rogers (1995) is more applicable as a hierarchical categorization. Rogers (1995) defines diffusion as the adoption of an innovation ‘over time by the given social system’ and specifies five adopter categories, namely innovators, early adopters, early majority, late majority and laggards. This categorization is based on their innovativeness and these categories do have some demographic characteristics. Early adopters of technological innovations are often stated to be relatively young, to have higher incomes, to be better educated and to have higher social status occupations (Rogers, 1995). Since they are more
technology-ready and sensitive to trends, younger people are more easily influenced by technology characteristics and peer opinions than older people (Scheper and Wetzels, 2007).
Especially the qualification that this age group is sensitive to trends explains why they belong to the early adopters. This clarifies that only a small portion of this group, the so-called innovators, has such an enormous influence on what is trendy among other young people. The same argument holds for the use of new technologies. Therefore,
H5: Subjective norm will have a positive effect on image.
As Pfeffer (1982) states, if people perform behavior that is consistent with group norms, ‘they achieve membership and the social support that such membership affords as well as possible goal attainment which can only through group action or group membership’. According to French and Raven’s (1959) taxonomy, instead of expert power, which forms the basis of internalization, referent power is perceived as being the pillar of identification. The stronger the identification of a person with a group, the greater the referent power of this group. In other words, if a person is highly attracted to another person or group, this person will strongly have the desire to become closely associated with this other person or group. Remarkably, sometimes people may be unaware of this referent power.
Venkatesh and Davis (2000) state that a person may perceive that using a system will lead to improvements in the job performance indirectly due to image enhancement over and above any performance benefits directly attributable to system use. Translating this into mobile banking by consumers, this entails that only the belongingness to a group of users will give people the feeling that mobile banking is useful. Therefore,
H6: Image will have a positive effect on perceived usefulness.
2.4.1.3 Social influence processes and experience
According to Venkatesh and Davis (2000), system experience has a moderating effect on both the relationship between subjective norm and perceived usefulness, and between subjective norm and behavioral intention. Hartwick and Barki (1994) suggest that before people have actually used a system, they have to rely on the opinions of others as a basis for their intentions due to the fact that their knowledge about the system is insufficient. After they have experienced the system themselves, this informational social influence disappears. The same argument goes for the effect of subjective norm on perceived usefulness, already mentioned as internalization (Venkatesh and Davis, 2000). Doll and Ajzen (1992) show that internalization declines after experience has increased
because greater experience provides concrete sensory information, which overrules the reliance on social cues as a basis for perceived usefulness. In other words, if people become more experienced, the effect of subjective norm on perceived usefulness declines.
However, subjective norm only has an effect on intention to use in mandatory settings.
Therefore, there is no moderating effect of experience on this relationship in the mobile banking context, since mobile banking is perceived as a voluntary technology at the moment.
On the other hand, Venkatesh and Davis (2000) find that subjective norm does have a significant effect on perceived usefulness regardless the fact that the system use is mandatory or voluntary. For that reason, the moderating role of experience on the relationship between subjective norm and perceived usefulness is applicable for the mobile banking context. Therefore,
H7: Experience will have a negative moderating effect on the relationship between subjective norm and perceived usefulness.
2.4.2 Cognitive instrumental processes
Figure 5. Cognitive instrumental processes.
Besides the effect of social influence processes on perceived usefulness and intention to use, TAM2 denominates four cognitive instrumental determinants of perceived usefulness, namely job relevance, output quality, result demonstrability and perceived ease-of-use. This is graphically represented in figure 5.
The relation between perceived ease-of-use and perceived usefulness will not be elaborated on again, since it has already been discussed. The general philosophy behind cognitive instrumental
H10
H2 H8
H9
Perceived Usefulness Relevance
Output Quality
Perceived Ease of Use Result Demonstrability
processes is that people cognitively judge perceived usefulness by comparing the capability of a system or technology with the things they need to get done.
Venkatesh and Davis (2000) argue that the cognitive instrumental processes are derived from three main areas: work motivation theory, action theory from social psychology and task-contingent decision making from behavioral decision theory. Especially the task-contingent decision making from behavioral decision theory seems to fit well in the mobile banking context.
Within behavioral decision theory, Beach and Mitchell (1996, 1998) have conducted an image theory. According this image theory, there exists two distinct decision stages during the adoption decision of a new technology. In the first stage a compatibility test is conducted to withdraw the incompatible options in relation with decision standards. Subsequently, the best option will be determined by conducting a profitability test in the second stage. Both the compatibility test and the profitability test are based on cognitively assessing the match between the characteristics of the trajectory image and the perceived consequences of the alternative action plans.
Besides perceived ease-of-use, TAM2 suggests that people determine the perceived usefulness based on the relevance, output quality and result demonstrability of the technology. Therefore, those three constructs will be elaborated on in further detail in the next three sections.
2.4.2.1 Relevance
Venkatesh and Davis (2000) describe job relevance as being a key component of the matching process between job goals and the consequences of using a specific technology. If translating job relevance into the relevance of mobile banking for customers, relevance could be defined as the individual’s perception of the degree to which the target system is applicable. In other words, relevance is a function of the importance of the set of tasks that the system or technology is capable of supporting.
Constructs which are comparable to relevance have been used in recent literature to explain user acceptance, such as involvement. There exists diverse definitions and measures of involvement due to the different applications of involvement. In the user acceptance context, involvement is defined as a person’s perceived relevance of the object based on inherent needs, values and interests (Zaichkowsky, 1985). Barki and Hartwick (1989) define involvement as a subjective psychological state, reflecting the importance and personal relevance of a specific system to a user.
The definition of compatibility, which is the degree to which an innovation is perceived as consistent with existing values, past experiences, and needs of potential adopters (Rogers, 1995), covers a large part of the definition of involvement by Zaichkowsky (1985). Furthermore, according to Zaichkowsky (1985) and Barki and Hartwick (1989), the definition of involvement is primarily based on a person’s perceived relevance. Apparently, relevance, involvement and compatibility are
highly interchangeable constructs. Therefore, the relevance construct of TAM2 and the compatibility construct of IDT are very similar.
Venkatesh and Davis (2000) come to the same conclusion because they conceptualize perceptions of relevance to be part of the compatibility test within the context of Beach and Mitchell’s (1996, 1998) image theory since systems above a threshold value of perceived relevance will be screened for further adoption consideration. In other words, a technology which is perceived as more relevant will be perceived as more useful as well. Therefore,
H8: Relevance will have a positive effect on perceived usefulness.
2.4.2.2 Output quality
Besides the capability of simply performing certain tasks by a system or technology and to which extent those tasks match goals, the consideration of people on how well the system or technology performs these tasks is at least as important. TAM2 refer to this consideration as output quality.
In the context of the image theory of Beach and Mitchell (1996, 1998), output quality judgments are made by performing a profitability test in order to choose a system or technology that delivers the highest output quality (Venkatesh and Davis, 2000). Davis, Bagozzi and Warshaw (1992) find empirical evidence for the relationship between output quality and perceived usefulness. The results of the study conducted by Venkatesh and Davis (2000) imply that technology usefulness judgments are affected by people’s cognitive matching of their goals with the consequences of technology use, which is also known as relevance. Moreover, Venkatesh and Davis (2000) find that output quality takes on greater importance in proportion to a technology’s relevance. In other words, there is an interaction effect between relevance and output quality in determining perceived usefulness. Davis et al. (1992) and Goodhue (1995) come to the same conclusion. Apparently, the relevance of a (new) technology has a larger effect on perceived usefulness if people perceive the output quality as higher.
It is remarkable that output quality or another synonym is missing in IDT, since it has been found to have a moderating effect on the relationship between relevance and perceived usefulness. For that reason, in contrast to IDT, this study explicitly includes the output quality construct into the conceptual model.
One could argue that the compatibility test is both a compatibility and profitability test.
According to the image theory, a system will be evaluated on certain needs and values during the first stage of the decision process. Subsequently, a system will be eliminated from one’s choice set if the system do not pass the capability test. Even though the conceptual model of this study implements output quality as a moderating construct, this could argue that the compatibility
construct in IDT covers both the compatibility test and the profitability test, which in turn represent the relevance and output quality constructs of TAM2 respectively.
H9: Output quality will have a positive moderating effect on the relationship between relevance and perceived usefulness.
2.4.2.3 Result demonstrability
Even though a given system or technology contains certain advantages, it is possible that it fails to gather user acceptance because people are not able to attribute these advantages to the technology. TAM2 refers to this phenomenon as result demonstrability. Moore and Benbaset (1991) define result demonstrability as the tangibility of the results of using the innovation.
Venkatesh and Davis (2000) theorize that result demonstrability will directly influence perceived usefulness. Apparently, people perceive a certain technology as more useful when there is an obvious covariation between the usage of a technology and achieving positive results. However, sometimes this covariation is less obvious. In that situation people will face the desired positive results but do not realize that these results are due to the technology.
Whereas Rogers (1995) describes five significant innovation categories in IDT, Agarwal and Prasad (1998) suggest that only three categories are applicable for innovation adoption. These three categories are already mentioned and compared with the constructs of TAM, namely: relative advantage, complexity and compatibility with perceived usefulness, perceived ease-of-use and relevance respectively.
Rogers (1995) defines observability as the degree to which results of an innovation are visible to others. This definition is very similar to result demonstrability. However, the difference between both definitions is that result demonstrability focuses on the demonstrability to yourself and observability focuses on the demonstrability to others.
This study focuses on the demonstrability to others in order to find out whether people are capable of explaining the advantages of mobile banking to others. However, people should be aware of those advantages by themselves in the first place. In that case, these people could act as ambassadors of mobile banking. For that reason, the definitions of observability and result demonstrability are highly comparative. Therefore,
H10: Result demonstrability will have a positive effect on perceived usefulness.
2.5 Perceived risk
Figure 6. Perceived risk and TAM.
2.5.1 Perceived risk and trust
In recent literature there has been some confusion about the difference between trust and perceived risk. Mayer, Davis and Schoorman (1995) define trust ‘as the willingness to take risk’ and perceived risk ‘as the likelihood of both positive and negative outcomes’. Kim and Prabhakar (2000) describe risk as a combination of relative advantages and negative consequences and suggest that there is a trade-off between trust and perceived risk when it comes to consumers’ adoption of internet banking. In other words, consumers will adopt this technology if the level of trust exceeds the level of perceived risk. However, this study only implements the perceived risk construct into the conceptual model.
2.5.2 Theory of perceived risk
According to Bauer (1967), consumer behavior involves risk because purchase actions will produce consequences which consumers cannot anticipate on with any certainty and some of the consequences are at least likely to be unpleasant. As a result, the more risk consumers perceive, the less likely they will purchase. Perceived risk is powerful at explaining consumer behavior because more frequently consumers are motivated to avoid mistakes than to maximize utility in purchase situations (Mitchell, 1999). In the same vein, it is found that symmetric changes in good and bad outcomes are not equal in their effects on judged risk (Coombs and Lehner, 1981). Cunningham (1967) conceptualizes perceived risk in terms of two similar components, namely the amount that would be lost (i.e. that which is at stake) if the consequences of an act were not favorable and the individual’s subjective feeling of certainty that the consequences will be unfavorable. The amount at stake is a function of the importance or magnitude of the goals to be attained, the seriousness of the penalties that might be imposed for non-attainment and the amount of means committed to achieving the goals (Cox, 1967).
H11 H13 H1
H2
H3 H12
Perceived Usefulness Perceived Risk
Perceived Ease of Use
Intention to Use
2.5.3 Facets of perceived risk
Cunningham (1967) distinguishes two categories of perceived risk, namely performance and psychosocial risk. Furthermore, he classifies perceived risk as having six dimensions: (1) performance, (2) financial, (3) time, (4) psychological, (5) social and (6) safety loss. However, safety loss is not likely to occur in a mobile banking context because consumers will not face any physical harm during the use of the service. For that reason, Featherman and Pavlou (2003) have replaced safety risk by privacy risk.
Moreover, Featherman and Pavlou (2003) ranked those six risk facets in order of importance and found that performance, financial and privacy risk are by far the most important determinants of perceived risk in the Internet banking context. Therefore, since these three facets are expected to be most applicable in the mobile banking context as well, only performance, financial and privacy risk will be discussed during this literature review. In other words, the perceived risk construct in this study consists of only these three determinants because time, psychological and social risk are found to be too marginal determinants of perceived risk in the internet banking context. The relationship between perceived risk and the TAM2 constructs are presented in figure 6.
2.5.3.1 Performance risk
Perceived performance risk is the possibility that the purchased product does not work as it should work or can only be used for a short period of time (Jacoby and Kaplan, 1972). In other words, performance risk is the probability that a product or service does not work as the consumer would like it to work. Grewal, Gotlieb and Marmorstein (1994) define performance risk as the possibility of the product malfunctioning and not performing as it was designed and advertised for and therefore failing to deliver the desired benefits.
Consumers are often apprehensive that a breakdown of system servers or a disconnection from the Internet will occur whilst conducting online transactions because these situations may result in unexpected losses (Lee, 2009). Littler and Melanthiou (2006) state that malfunctions of online banking websites would reduce consumers’ willingness to use online banking. Moreover, Featherman and Pavlou (2003) find that a high frequency of website breakdowns and disconnections affects perceived usefulness. The same could be expected for mobile banking. Therefore,
H11a: Performance risk will have a negative effect on perceived usefulness.
H12a: Performance risk will have a negative effect on intention to use.
2.5.3.2 Financial risk
Financial risk refers to the potential for monetary loss due to transaction errors or bank account misuse (Lee, 2009). Kuisma, Laukkanen and Hiltunen (2007) state that many consumers reject using online banking because they fear losing money whilst performing transactions or transferring money over the Internet. Whereas clerical personnel verifies whether the account number and the amount of money to transact is correct, such safeguards regularly lack in online banking situations. This could cause feelings of insecurity and uncertainty. Therefore,
H11b: Financial risk will have a negative effect on perceived usefulness.
H12b: Financial risk will have a negative effect on intention to use.
2.5.3.3 Privacy risk
According to the study of Featherman and Pavlou (2003) privacy risk is the most important risk factor in the online banking context next to financial risk. Lee (2009) defines privacy or security risk as a potential loss due to fraud or a hacker compromising the security of an online bank user. The importance of security, which protects consumers against financial and privacy losses, has been noted in many banking studies regarding the adoption of online banking (Roboff and Charles, 1998;
Sathye, 1999, Tan and Teo, 2000; Howcroft, Hamilton and Hewer, 2002). In the online banking context, security and privacy were found to be significant obstacles to the adoption of online banking in Australia (Sathye, 1999).
Phishing is probably the youngest technique to fraudulently acquire sensitive information, such as usernames, passwords and credit card details. Phishers pretend to be a trustworthy entity in an electronic communication (Reavley, 2005). Both fraud and hacker intrusion not only lead to a financial loss but also violate the users’ privacy. Furthermore, it has been stated in many studies that the greatest challenge to the electronic banking sector will be winning the trust of consumers over the issues of privacy and security (Bestavros, 2000; Furnell and Karweni, 1999). Therefore,
H11c: Privacy risk will have a negative effect on perceived usefulness.
H12c: Privacy risk will have a negative effect on intention to use.
2.5.4 Perceived risk and perceived ease-of-use
Moore and Benbasat (1991) identify that the complexity of a system reduces system evaluation and adoption intention. Apparently, products or services that are perceived as being complicated to use are also likely to be evaluated as being more risky to use. Services that are perceived as being complex, problematic and not free of effort may also be perceived as being plagued with
performance problems and usage uncertainties (Featherman and Pavlou, 2003). In contrast, the opposite holds for systems that are evaluated as being easy to use. In other words, the more user- friendly the system interface of a mobile banking application, the more likely that people will adopt the service because they expect that the system is free of performance problems. Therefore,
H13: Perceived ease-of-use will have a negative effect on perceived risk.
Chapter 3. Research methodology
3.1 Data collection
Data for this study is collected from the business-to-consumer (B2C) context. In other words, the respondents represent personal bankers. Hence, this study does not include private or commercial bankers.
In order to gather information about people’s incentives to use mobile banking, an online survey in Dutch was sent out to approximately 450 users or potential adopters between the age of 16 and 30 years old. This age group seems to represent the early adopters regarding their demographic variables (Rogers, 1995). Moreover, even people who do not use mobile banking were supposed to be able to answer the questionnaire because a short introduction video about mobile banking is inserted at the beginning of the survey research.
To encourage participation, a gift card worth 25 Euros was raffled among the respondents. At the end of the questionnaire was explained how the respondent could get access to the raffle. The respondents had to send their e-mail address in order to participate. Consequently, the data is not linked to the respondent. Moreover, respondents had the opportunity to neglect the raffle. Finally, an online raffle website was used to ultimately choose the lucky winner from the list of e-mail accounts.
The data is collected within week 46 of 2012 and in order to increase the response rate a reminder was sent out after four days.
3.2 Structural Equation Modeling
In order to analyze the collected data, the structural equation model (SEM) approach is chosen.
SEM is a family of statistical models that seeks to explain the relationships among multiple variables (Hair, Black, Babin and Anderson, 2009). SEM has become a popular multivariate approach in a relatively short period of time because it provides a conceptually attractive way to test theory (Herschberger, 2003). SEM is such a conceptually attractive method is because it is able to assess how well the theory fits reality as represented by the data after the literature review has been expressed in terms of relationships among measured variables and latent constructs.
The origin of SEM stems from two familiar multivariate techniques, namely confirmatory factor analysis (CFA) and multiple regression analysis. Therefore, the SEM approach is divided into two parts. Moreover, these two parts are subdivided into a total of six stages.
The first two stages concern defining the individual constructs and the development of the overall measurement model. Especially those stages are of great importance because SEM should never be attempted without a strong theoretical basis for specification of both the measurement and
structural models (Hair et al., 2009). Although theory is important for all multivariate procedures, it is particularly important for SEM because it is considered to be a confirmatory analysis. So, it is useful for testing and potentially confirming theory. Stage three deals with designing a study to produce empirical results. The validity of the measurement model is assessed during the fourth stage. If supported by theory, the measurement model will be respecified in order to accomplish a better model fit. Here the CFA part ends. In stage five, the structural model comes into play. The structural model draws dependence relationships between the constructs based on the literature study. Finally, in stage 6, the structural model can be modified, but again, only if this is supported by theory.
Because of the dependence relationships between the constructs, stage five and six are considered to represent the multiple regression analysis part.
3.3 Individual construct definitions
The reliability and validity of the hypotheses tests involving the structural relationships among constructs mainly depends on how well those constructs are implemented in the measurement model. First of all, the theoretical definitions of the constructs needs to be secured. The definitions of the constructs are summarized in table 1.
Construct Definition Reference
Intention to Use
Perceived Usefulness
Perceived Ease-of-use
Subjective Norm
Image
Relevance
Output Quality
Result Demonstrability
The user’s likelihood to engage in mobile banking.
The extent to which a person believes that using mobile banking would enhance the task performance.
The extent to which a person believes that mobile banking would be free of effort.
The individual’s perception that most people who are important to him think he should or should not perform mobile banking.
The degree to which mobile banking is perceived to enhance one’s status in one’s social system.
The individual’s perception regarding the degree to which mobile banking is applicable.
The individual’s perception regarding the degree to which mobile banking performs well.
The tangibility of the results of mobile banking.
Wu and Wang (2005)
Davis (1989)
Davis (1989)
Fishbein and Ajzen (1975) Moore and Benbasat (1991) Venkatesh and Davis (2000) Venkatesh and Davis (2000) Moore and Benbasat (1991)
