• No results found

New technology adoption: Towards a comprehensive understanding of the factors influencing Mobile Payment adoption

N/A
N/A
Protected

Academic year: 2021

Share "New technology adoption: Towards a comprehensive understanding of the factors influencing Mobile Payment adoption"

Copied!
71
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

New technology adoption:

Towards a comprehensive understanding of the factors

influencing Mobile Payment adoption

(2)

New technology adoption:

Towards a comprehensive understanding of the factors

influencing Mobile Payment adoption

Master thesis

Author:

Simon Haijma

(1750070)

Rijksuniversiteit Groningen

Faculty of Economics and Business

Marketing Management

First supervisor: J. Berger

Second supervisor: P. van Eck

(3)

Management summary

Since 2000 onwards, research activities on Commerce have increased significantly. In the M-Commerce context, Mobile Payment (or M-Payment) has been acknowledged as a highly relevant subject. In this research, potential adoption factors of Mobile Payment are tested using the Technology Acceptance Model (TAM). This versatile and highly-used model is an useful instrument in finding out respondents behavioural intention towards the use of new technologies.

Based on an extensive literature discussion, several hypotheses are proposed and added to the original TAM model. Besides the constructs of Perceived Ease of Use and Perceived Usefulness, that are part of the original TAM, the constructs of Perceived Social Influence, Perceived Compatibility, Perceived Credibility, Perceived Costs, Perceived Mobility, Perceived Enjoyment and Perceived Self-Efficacy were added to the original model. This results in an integrative model that extends the original TAM with relevant constructs.

The proposed hypotheses were tested using survey data from 388 Rijksuniversiteit Groningen students. Based on the outcomes of this research, Perceived Compatibility has highest influence on the Intention to Use M-Payment. Besides Perceived Compatibility, the factors of Perceived Usefulness, Perceived Credibility, and Perceived Social Influence were also found positively influencing the Intention to Use M-Payment. A negative impact was found between Perceived Costs and the Intention to Use M-Payment. No empirical evidence was found for a relationship between Perceived Ease of Use and the Intention to Use M-Payment. Hence, Perceived Mobility, Perceived Self-Efficacy, and the hedonic adoption factor of Perceived Enjoyment were not found significantly related to Intention to Use M-Payment.

This study confirms that Perceived Compatibility has to be characterized as the most

influential adoption factor on Intention to Use M-Payment. This implies that M-Payment, in order for it to be a success, has to show a clear fit with people’s lifestyle and purchase needs. Secondly, Perceived Usefulness, or the extent to which a person believes that using M-Payment will enhance his or her job performance, is scored as a highly influential adoption factor, followed by Perceived Credibility, Perceived Social Influence and Perceived Costs. The lack of empirical evidence between Perceived Ease of Use and Intention to Use, implies that the “Easyness” or Perceived Ease of Use of trying new technologies may not be much of an issue for most students, who seem to be very confident on using new technologies and mastering them.

(4)

Table of content

§1

Introduction

...6

§2

Theoretical background

...9 §2.1 Literature review ...9 §2.2 Hypotheses ...16 §2.3 Conceptual Model ...25

§3

Research methodology

...27

§3.1 Measuring the constructs

...27

§3.2 Data collection procedure

...28

§3.3 Data processing procedure

...31

§4

Results

...33 §4.1 Representative sample ...33 §4.2 Distribution of survey ...34 §4.3 Missing values ...36 §4.4 Cronbach Alpha ...37

§4.5 Correlation between constructs

...38 §4.6 Multiple regression ...39 §4.7 Moderator analysis ...42

§5

Discussion

...47 §5.1 Discussion ...47 §5.2 Marketing implications ...52

§5.3 Limitations and future research

(5)

References

54

Appendix 1 : Online Questionnaire (distributed May, June 2011) 63

Appendix 2a: Mean Scores and standard deviations constructs 67

Appendix 2b: Kolmogorov-test

68

(6)

§1 Introduction

§1.1 Introduction

With the recent emergence of the wireless and mobile networks a new platform for trading products and services, known as M-Commerce, is beginning to gather attention from businesses. Slightly different than E-Commerce, which principle is based on using the network connectivity power of the internet, M-Commerce is using the principle of wireless connectivity in a mobile environment using

mobile devices (Wei et al.; 2009). From the 1990s onwards, their has been a great shift in methods of

doing business with the emergence of E-Commerce. Academics, businesses, and individuals have been focusing on this new way of conducting business online. E-commerce is conducted from a wired network to a wireless network. With M-Commerce, users conducting Commerce (such as E-Banking or purchasing of products), do not need to use a personal computer system, but can simply use mobile handheld devices (such as smartphones, tablets, PDA’s etc.). In the past, these mobile devices and technologies were seen as luxury, where nowadays these mobile devices are available for everyone, resulting in a significant growth on mobile technologies in the past years (Ngai; 2005). This growth is also recognized in the smartphone sale numbers. In 2009, the smartphone penetration in the US was 25% and worldwide, 14% of mobile device shipments were smartphones (Falaki et al; 2010). By 2011, smartphone sales are projected to keep increasing and to surpass desktop PCs (Selleck; 2010, Snol; 2009).

§1.2 Academic and practical relevance

(7)

To get a clear view on consumer adoption in the M-Payment field , it is essential to emphasize the role of the M-Payment technologies. M-Payment technologies have changed the last years, going from technologies such as SMS and WAP to contactless and easy to use technologies such as Near Field Communication (NFC) in recent years. This development, in combination with the increasing use of smartphones by consumers (with the use of the attached advanced mobile services), could seriously impact the role of M-Payment in day-to-day life. Therefore, it would be interesting to reverify the outcomes of M-Payment studies that used somewhat outdated M-Payment technologies (such as SMS and WAP). For example, in her study, Mallat (2007) finds out that several factors contribute to the adoption of Mobile Payments in Finland. In her study, several limitations and suggestions for future research are mentioned. First of all, her study is based on data from 2002 and therefore the technologies (such as SMS and WAP) relevant in that time period are being used in her research. Although some of these technologies could still be relevant, it would be interesting to test the proposed factors with the new contactless Mobile Payment technologies (such as NFC, derived from Zmijewska; 2005). In addition, it would be interesting to analyze whether their are social differences in the adoption of M-Payment technologies (Mallat; 2007, on M-Payment adoption, Mahatanankoon; 2007, Ngai; 2005, Liang & Wei; 2004, on M-Commerce adoption). A literature study by Dahlberg (2008) also shows that Mobile Payment research seems to focus on the Mobile Payment technologies, and the consumer perspective of Mobile Payments, thereby neglecting the social factors impacting Mobile Payment adoption.

(8)

§1.3 Aim of research

Given the high academic and practical relevance of Mobile Payment, this research tests acceptance factors by developing an extended Technology Acceptance Model (TAM), a model that is useful in explaining individual’s intention to use technology, and by involving the new contactless Mobile Payment technologies in the adoption process of Mobile Payments. Based on the literature review, several hypotheses will be proposed and implemented into an integrative model of factors determining consumers’ acceptance of M-Payment (Schierz. et al.; 2010).

§1.4 Structure of the thesis

(9)

§2

Theoretical background

First of all, the relevant concepts of this study are introduced in an extensive literature review. Second, the constructs under study are embedded in relevant literature and hypotheses are derived. And finally, the derived hypotheses are presented into a conceptual model.

§2.1 Literature review

§2.1.1 M-Commerce

M-Commerce is a relatively new area in research. Research activities on M-Commerce have increased significantly after 2000 (Ngai; 2005). Although some notable journals have been publishing on M-Commerce, and this increase in research activities in the field of M-Commerce is noteworthy, the topic of M-Commerce still remains a largely undiscovered topic in top journal publications.

From E-Commerce towards M-Commerce

In order to get an overview on the central topic of Mobile Payment, it is important to get a clear insight into the context of M-Commerce, in which Mobile Payment resides. To do so, it is essential to look at the differences between the concept of M-Commerce and the concept of E-Commerce, from which M-Commerce has ultimately emerged. First, the general similarities and differences between E-Commerce and M-E-Commerce will be discussed. Second, the concept of M-E-Commerce will be discussed by evaluating several relevant papers on M-Commerce.

General similarities and differences between E-Commerce and M-Commerce

Prior to the development of M-Commerce, E-Commerce was depended on costly infrastructure and equipments such as computers and fixed line networks. E-Commerce is conducted from a wired network to a wireless network using the connectivity of the internet, where M-Commerce is connected wirelessly in a mobile environment using mobile devices (Wei; 2009, Ngai; 2005). The difference between E-Commerce and M-Commerce is emphasized in the following statement:

“Going beyond the computer-mediated electronic commerce or e-commerce of the 1990s, this new type of mobile commerce – or m-commerce – is characterized by novel, location-based services delivered by a variety of handheld terminals”

(10)

With M-Commerce, users conducting E-Commerce (such as E-Banking or purchasing of products) do not need to use a personal computer system, but can simply use mobile handheld devices (such as smartphones, tablets, PDA’s etc.).

Several advantages of Commerce over E-Commerce are mentioned (Wei, 2009). First of all, M-Commerce offers more ubiquity and accessibility to users when compared to E-M-Commerce (Feng et al; 2006). In terms of accessibility, M-Commerce is an advantage over E-Commerce, as E-Commerce applications usually need a wired end-user device. Second, mobile devices are easy to carry because of their small (in size) and lightness (in weight). And thirdly, mobile devices are usually owned by individuals and not shared between different users, making it possible to target individual needs with interesting services (Schwiderski-Grosche and Knospe, 2002).

Theoretical perspective on M-Commerce

Academic sources do not consistently give one clear definition on M-Commerce (Liang; 2004). In literature, several definitions on M-Commerce are being found, some being broader in scope than others. According to most researchers, M-Commerce is a new type of E-Commerce that conducts business transactions through mobile devices (Abuelyaman; 2004), in other words, stating that M-Commerce can be seen as E-M-Commerce over the wireless devices (Varshney and Vetter, 2002). According to these stream of literature, M-Commerce can be viewed as a subset of E-Commerce (Kwon et al; 2004, Coursaris et al; 2002) and can be defined narrowly as “any transaction with monetary value that is conducted via a mobile network” (Clarke; 2001).

Other researchers adopt a broader perspective of M-Commerce, suggesting that M-Commerce is more than E-Commerce due to its different interaction style, usage pattern and value chain (Feng et al; 2006). Thereby implying that M-Commerce is a new and innovative business opportunity with its own unique characteristics and functions, such as mobility and broad reachability (Feng et al.; 2006). According to these sources, M-Commerce is defined broadly as services that involve communication, information, transaction, and entertainment (Schwiderski-Grosche and Knospe; 2002).

(11)

that mobile business is a broad definition that includes communication, transactions and different value-added services using various kinds of mobile terminals”. Because adding the term M-Business leads to unclarity and overlap, in this research the focus is laid on the concept of M-Commerce.

In this research, similar to the research done by Wei et al. (2009) on the adoption of M-Commerce in Malaysia, the definition of Tiwari and Buse (2007) is used, because it finds a compromis between the earlier mentioned sources on defining the concept of M-Commerce. Following the definition of Tiwari and Buse (2007, p. 33), the concept of M-Commerce will be defined as “any transaction, involving the transfer of ownership or rights to use goods and services, which is initiated and/or completed by using mobiles access to computer-mediated networks with the help of mobile

devices”.

§2.1.2 Mobile Payment

In this section, first, the theoretical relevance of M-Payment in the context of M-Commerce will be discussed. Second, based on relevant literature, a distinction will be made between the concept of M-Commerce and Mobile Payment (or M-Payment), resulting in a clear definition of M-Payment.

Theoretical relevance of M-Payment

In M-Commerce literature, the role of financial services such as M-Payment is well discussed.

Research activities on M-Commerce have increased significantly after 2000 (Ngai; 2007). After having reviewed the M-Commerce literature, Ngai (2007) finds out that the majority of M-Commerce applications were focussing on financial uses such as Mobile Payments and Mobile Banking. This shows that the topic of Payment is being acknowledged as a highly relevant subject in M-Commerce (Au; 2008).

M-Payment versus M-Commerce

(12)

Similar to literature on Commerce, no consistent definition is given on the concept of M-Payment. In the paper of Au (2008) a definition on M-Payment is mentioned based on Karnouskas (2004) and states that: “a mobile payment is any payment where a mobile device is used to initiate, authorize and confirm an exchange of financial value in return for goods and services.” A slightly different definition is mentioned by Pousttchi et al. (2009), in which M-Payments are defined as: “a type of payment transaction processing in which the payer uses mobile communication techniques in

conjunction with mobile devices for initiation, authorization, or completion of payment (Pousstchi;

2008). This last definition can lead to some unclarity when compared to the definition of Commerce formulated in the previous section. These definitions raise a vague boundary between Commerce and Payment, where both the term transaction is mentioned. In the previous section, M-Commerce was defined as: “any transaction, involving the transfer of ownership or rights to use goods

and services, which is initiated and/or completed by using mobiles access to computer-mediated networks with the help of mobile devices” (Tiwari and Buse; 2007, p. 33). Comparing both definitions,

M-Payment distincts itself from M-Commerce by focussing on a payment transaction.

To make this distinction between M-Payment and M-Commerce more clear; M-Payment differs from M-Commerce by focussing on the monetary value of the transaction. In other words, receiving (free) data on a mobile device, such as obtaining commercial information through QR-codes, or an application (app) download, can be considered as a form of M-Commerce (Tiwari and Buse; 2007). M-Payment can be characterized as a financial transaction with a mobile device for a product or service, with or without the help of an intermediary. Based on the discussion above, the definition of Mallat (2007) on M-Payment will be used:

“The use of a mobile device to conduct a payment transaction in which money or funds are transferred from payer to receiver”

Mallat (2007)

§2.1.3 Consumer adoption

(13)

why new Mobile Payments will or will not be used by their intended users (Mallat; 2007). Similar to prior research, the construction of “Intention to Use” is used as a proxy for consumer acceptance (Schierz et al.; 2010, Venkatesh and Davis; 2000). Empirical research from different authors (Wu; 2005, Sheppard et al. 1988) found that “Intention to Use” is an reliable predictor of later usage, which is proof of success from an IT adoption perspective (Venkatesh et al.; 2003).

§2.1.4 Technology Acceptance Model (TAM)

Several models on user acceptance in an information system (IS) context can be distinguished. A clear overview of these models can be found in the paper of Venkatesh et al. (2003, p428). In this research, the TAM model is chosen as the most appropriate model, because it offers the possibility to be improved by adding relevant constructs. The TAM, Technology Acceptance Model was designed to predict information technology acceptance and usage on the job. Unlike TRA (Theory of reasoned action), the final conceptualization of the TAM excludes the attitude construct in order to better explain intention parsimoniously. Schierz et al (2010) states that TAM can be considered as the most influential extention of TRA, replacing variables related to attitude and behavioural control with technology acceptance measures (Bagozzi; 2007). The TAM theorizes that an individual’s behavioural intention to use a system is determined by two beliefs: Perceived Usefulness, and Perceived Ease of Use (Venkatesh and Davis; 2000). Research from various authors (Venkatesh et al.; 2003, Taylor and Todd; 1995) found out that the constructs of “Perceived Usefulness” and “Perceived Ease of Use” are similar to the constructs of “Relative advantage” and “Complexity” of the Innovation Diffusion Theory (IDT). Together with Compatibility, they have found to be the most constant determinants of adoption (Mallat; 2009).

(14)

mentioned (Schierz et al.; 2010), such as its excellent measurement properties, conciseness, and empirical soundness (Pavlou; 2003). It is also applicable to a wide range of research questions, making it possible to employ the model in a variety of settings (i.e. on e-shopping, Ha; 2008, on mobile ticketing, Mallat; 2009, and also M-Payment, Schierz et al.; 2010).

It has been suggested that the model is too parsimonious and should be expanded by factors relevant to the specific technology under study (Venkatesh and Davis; 2000). Legris et al. (2003) found out that the TAM and TAM2 only explains 40% of system’s use. This suggests that significant factors are not included in the models. So, although the TAM is very useful in explaining behavioural intention (Mathieson; 1991), it is needed to apply certain constructs to the context under study, in this case M-Payment. In this research, a model is made in which several constructs new to the M-Payments field are added, mainly based on research in the field of E-Commerce and M-Commerce. This is done to get a broad view on the possible constructs that influence behavioural intention in M-Payment. In the proposed model, the construct of “Perceived Mobility” for example is based on Schierz et al. (2010) and Mallat (2009). In this particular study, besides constructs that were found significant in the field of Mobile Payment, several other constructs are added that were found significant in studies in the field of E-Commerce and M-Commerce. These constructs could well be relevant for Mobile Payment. The constructs that are added and were found relevant in the field of E-Commerce, M-Commerce or Mobile Payments involve the constructs of:

- Perceived Compatibility (Mallat; 2009, Wu; 2005),

- Perceived Social Influence (Schierz et al.; 2010, Lopez-Nicolas et al.2008)

- Perceived Enjoyment (Ha; 2008, Bruner; 2005)

- Perceived Credibility (Wang et al.; 2006, Luarn & Lin; 2005)

- Perceived Costs (Wei et al.; 2009),

- Perceived Self-Efficacy (Chen et al; 2009, Wang & Lin; 2006, Luarn & Lin; 2005, Wang

(15)

Besides utilitarian (such as “Perceived Usefulness”) constructs, the model also acknowledges that there needs to be room for hedonic motives (see “Perceived Enjoyment”) in using technology (Ha; 2008, Bruner; 2005). Because the construct of “Perceived Enjoyment’ is still untested in the M-Payment research field, this construct will be added to the model. Adding these concepts to the constructs from the TAM2, leads to the following schematic model:

(16)

§2.2 Hypotheses

The extended TAM in this research is based on academic work in the field of E-Commerce, M-Commerce and M-Payment. In this paragraph, the different constructs will be defined and hypotheses will be explained using additional literature. This paragraph concludes with the hypotheses applied to the proposed model.

§2.2.1 Perceived Ease of Use (EoU)

Ease of Use (or EoU) refers to “the degree to which a person believes that using a particular system would be free of effort” (Davis; 1989). This is in line with the definition of “ease”: “freedom from difficulty or great effort” (Davis; 1989). Practical examples in the context of M-Payment are for example, clear symbols and function keys, few and simple payment process steps, graphic displays, and help functions (Pagani and Schipani; 2005). Ease of Use is especially relevant in the M-Payment field, because M-Payment competes with established payment systems, and therefore needs to give benefits with regard to EoU. As a consequence, the construct of Ease of Use is incorporated in the proposed model.

As part of the original TAM, the construct of Ease of Use has been a topic for research in numerous fields. In the field of M-Payment, recent studies of Schierz et al. (2010) and Luarn & Lin (2005) found a significant positive relationship between Ease of Use and Usefulness. This relationship is based on a research by Venkatesh et al. (2003), proposing that the easier and more intuitive a Mobile Payment system is perceived to be, the more positive the usefulness will be assessed.

(17)

positive link between Perceived Ease of Use and Intention to Use. Therefore, the following hypothesis is presented:

H1: There is a positive relationship between the Perceived Ease of Use and the Intention to Use Mobile Payment

§2.2.2 Perceived Usefulness

Besides Ease of Use, Usefulness is the second construct derived from the original TAM. Perceived Usefulness can be defined as the extent to which a person believes that using the system will enhance his or her job performance (Venkatesh and Davis; 2000, Davis; 1989). An extensive amount of studies (e.g. Chen et al.; 2009, Schepers; 2007, Wu; 2005, Wang; 2003), including research in the field of M-Payment (Kim, Mirusmonov and Lee; 2010), prove that Perceived Usefulness has a direct positive influence on technology usage intention. The basic idea is that an individual who perceived a technology (in this case, M-Payment) as useful, is more intented to use that technology. The usefulness of the Mobile Payment NFC technology is stretched by Zmijewska (2005):

“People do not have to carry both their wallet and their mobile phone with them, but just one of them. When the user sees a poster with a movie they are interested in, they can simply touch the poster with their phone so the data from the ID tag in the poster is read into it; the service shortcut that was read to the phone allows the user to purchase tickets to the movie; it is also possible to book several tickets for friends.”

Zmijewska (2005)

Based on the above, the following hypothesis is proposed:

H2: There is a positive relationship between the Perceived Usefulness of M-Payment and the Intention to Use Mobile Payment

§2.2.3 Perceived Social Influence

(18)

characterized as an organization, on an individual. Therefore, the testing of “Social factors” and “Image” is more relevant for testing in an organizational context involving user acceptance (Venkatesh et al.; 2003). In this research, there is chosen for a consumer-centric setting using the TAM2 as a basis, in which Subjective Norm is more relevant for testing. In the TAM2 the construct of “Subjective Norm” accounts for denoting Social Influence (Lu. et al; 2003). Subjective Norm can be defined as: “the person’s perception that most people who are important to him think he should or should not perform the behaviour in question”. In other words, it refers to the perceived social pressure to perform or not to perform the behaviour (Ajzen; 1991).

The effect of Subjective Norm on Intention to Use is questioned in several studies. Direct effects of Subjective Norm on Intention to Use are not always found (Shin et al.; 2009, on Mobile Wallet), and sometimes an indirect relationship is proposed using the internalization mechanism, in which Subjective Norm influences technology acceptance through Perceived Usefulness (Lopez-Nicolás; 2008, Schepers et al.; 2007). Other academic sources state that Subjective Norm directly affects Intention to Use. Venkatesh and Davis (2000) for example hypothesized that Subjective Norm influenced Intention to Use, leading to the TAM2. More recent, also evidence was found that Social Influence, was a directly influence on the M-Commerce use intention in Malaysia (Wei et al.; 2009). In addition, Schepers et al. (2007) found that Subjective Norm had a larger impact on behavioural intention in Western studies than non-Western studies, making the likelihood of a direct positive relationship between Subjective Norm and Intention to Use even stronger. Therefore, the following hypothesis is proposed:

H3: There is a positive relationship between Perceived Social Influence and the Intention to Use Mobile Payment

§2.2.4 Perceived Compatibility

(19)

found out that adding the construct of Compatibility increased the predictive power of the TAM. Therefore the construct of Perceived Compatibility can be added to the model (Schierz et al.; 2010).

In his research on M-Commerce adoption, Wu (2005) found a direct positive significant link between Compatibility and the Intention to Use M-Commerce. This positive link was also found by Mallat in 2006 and 2009 on the adoption of Mobile Ticketing, stating that: “Compatibility was overwhelmingly the most important adoption determinant in our study”.

H4: There is a positive relationship between Perceived Compatibility and the Intention to Use Mobile Payment

§2.2.5 Perceived Credibility

Security and privacy issues play a role in the intention to use a new technology. Previous research used the constructs of Perceived Security (Schierz et al; 2010), mirrored by Perceived Risk (Wu; 2005), and Trust (Wei; 2009) to test these issues in the TAM. Gefen et al. (2003) states that trust can be considered as more crucial and complex in environments such as M-Commerce and E-Commerce, than general and traditional commerce, due to its uncertain environment (Cho et al.; 2007, Lu et al.; 2003) and information asymmetry (Cho et al.; 2007). Compared to E-Commerce, Lu et al. (2003) states that using strictly wireless technology like M-Commerce, is exposed to greater danger of insecurity than E-Commerce and therefore the importance of trust can be considered relatively higher in Commerce. Schierz et al. (2010) mentiones several potential causes for this uncertainty. First, M-Payment is a new technology and therefore consumers may be concerned with using this new electronic service that they never used (Bauer et al.; 2005a). Second, services (compared to tangible products) are more difficult to evaluate, leading to a consumers perceiving the service as risky (Gefen et al.; 2003). And finally, making a Mobile Payment is associated with a relatively high loss potential related to privacy, personal data, and the transaction itself (Bauer et al.; 2005b), leading to consumers perceiving the service as risky. In addition, in the online financial services market, Sathye (1999) found security concern as the “biggest obstacle” to adoption of online banking in Australia. Therefore, the constructs of Trust, and Perceived Security and Risk, are relevant for M-Payment.

(20)

Perceived Security, Risk, and Trust is rarely used simultaneously in the TAM (Shin et al.; 2009). This presumption is fueled by the lack of difference between the measurements items of Perceived Security and Trust (Schierz et al.; 2010 vs. Wei; 2009), making discrimination between the two constructs difficult. Therefore, in this research the construct of Perceived Credibility is used to cover the security and privacy issues concerning new technology usage. This relatively new construct was created by Wang et al. (2003). Wang et al. (2003) have conceptually distinguished the construct of Perceived Credibility from Perceived risks (e.g. Wu; 2005, Liao et al.; 1999) and Trust (e.g. Gefen et al.; 2003). Perceived Credibility can be defined as: “the extent to which a person believes that the use of Mobile Payment will have no security or privacy threats” (adapted from Luarn & Lin; 2005). Zmijewska (2005) indicated that in the case of using the NFC M-Payment technique, several security and privacy systems can be used to prevent malicious use:

“With NFC, communication takes place over a short distance (20 cm maximum). Using an ID number, or in some cases the handset’s fingerprint sensor, to authorize a transaction could prevent malicious use. In addition, techniques are in place that enable smartphones to be locked remotely by calling the handset from a preset phone number.”

Zmijewska (2005)

On his research on Internet Banking adoption, Wang et al. (2003) found a significant positive relationship between Perceived Credibility and Intention to Use. Research of Wang & Lin (2006), on Mobile Service adoption, and Luarn & Lin (2005), on Mobile Banking adoption, also showed a positive link between Perceived Credibility and the Intention to Use. Therefore, the following hypothesis is proposed:

H5: There is a positive relationship between Perceived Credibility and the Intention to Use Mobile Payment

§2.2.6 Perceived Costs

(21)

which a person believes that using Mobile Payment will cost money (adapted from Luarn & Lin; 2005). Although the use of the M-Payment technique would not necessary involve costs, obtaining a high-end Smartphone to actually use a M-Payment technique such as NFC, could lead to high expenses (Zmijewska; 2005, Zmijewska; 2004).

Several academic papers are published showing the relationship between Perceived Costs and Intention to Use. Wu (2005) and Wei (2009) found that Perceived Cost had a significant negative relationship on Intention to Use M-Commerce. Closely related to the field of M-Payment, this significant negative relationship was also found in the context of Mobile Banking (Luarn & Lin; 2005). Based on the above, the following hypothesis is proposed:

H6: There is a negative relationship between Perceived Costs and the Intention to Use Mobile Payment

§2.2.7 Perceived Mobility

Mobility refers to the movement of technologies, people, settings, etc. (Mallat; 2009). In this research, the role of people and their interaction with M-Payment services is emphasized. Therefore, the construct of Perceived Mobility is discussed from an individual perspective. Mallat (2009) states that, compared to E-Commerce, mobile computing provides access to information, communication, and services independent of time and place. Kakihara and Sørensen (2001) declare that people continuously frame their interaction with others, including their cultural background, situation and mood. In their research on M-Payment adoption, Schierz et al. (2010) also observed this general trend towards an increasingly mobile society, where there is still significant variance in the mobility of individuals. Other than traditional payment solutions, Mobile Payment services can be used anytime and virtually anywhere, leading to its advantage of ubiquity (Dahlberg et al. 2003, Kleinrock; 1996). Zmijewska (2005) stretches the role of mobility when using the NFC M-payment service:

“One feature that makes NFC m-payments truly mobile is the fact that the chip can even function when the phone is turned off. Therefore, when the battery is down or where there is no coverage, a payment can still go through.”

Zmijewska (2005)

(22)

literature, Mobility was found to be influencing Intention to Use. Direct and indirect positive relationships were found between Mobility and Intention to Use. Research of Mallat (2009) on mobile ticketing adoption, showed an indirect positive relationship between Mobility and Intention to Use, mediated by mobile use context. Besides this indirect relationship, Schierz et al. (2010) found a direct relationship between (individual) Mobility and Intention to Use. In their research on M-Payment adoption, Schierz et al. (2010) found that highly mobile persons will have a higher Intention to Use M-Payment services. Therefore, the following hypothesis is proposed:

H7: There is a positive relationship between Perceived Mobility and the Intention to Use Mobile Payment

§2.2.8 Perceived Enjoyment

Besides the usefulness of the product, refering to the utilarian path of using technology, it is also interesting to look at the hedonic path of using technology (Wei; 2009). The construct of Perceived

Enjoyment is added to the model to cover the hedonic motivations of individuals in adopting new

technologies. Teo et al. (1999) state that individuals may engage in a particular behaviour if it yields fun and enjoyment, implying that individuals may adopt technology because its use is enjoyable.

(23)

positive relationship between Ease of Use and Perceived Enjoyment, stating that the easier a system is to use, the more fun an user will derive from using the system. More recent research done by Ha (2008), on e-shopping acceptance, also found an indirect relationship between (shopping) enjoyment and Intention to Use.

In the specific field under study, the field of M-Payment, the direct influence of Perceived Enjoyment on Intention to Use has not yet been tested empirically. It is therefore uncertain if hedonic motivations play a role in performing M-Payment. Zmijewska (2004) states that enjoyment is considered instrumental for services primarily designed for entertainment, which does not apply to Mobile Payments. Contrary to this, Davis et al. (1992) already showed that Perceived Enjoyment can also have an positive influence on the Intention to Use a word processing program. The possibility exists that people prefer M-Payment over traditional payment services, because the usage of the new technology is more enjoyable. Therefore, it would be interesting to add the construct of Perceived Enjoyment to the model, and to test whether evidence can be found that shows that there is a direct relationship between Perceived Enjoyment and the Intention to Use M-Payment. The parsimonious design of the TAM enables direct testing between the construct of Perceived Enjoyment and Intention to Use. In this research, based on the discussion above, the construct of Perceived Enjoyment is defined as:

‘‘the extent to which the activity of performing M-Payment is perceived to be enjoyable, apart from any performance consequences that may be anticipated”

Adapted from Lee et al. (2005)

Based on the above, one can propose that when an individual’ perceives using M-Payment services as fun or enjoyable, the more likely it is that someone intents to use M-Payment. This leads to the following hypothesis:

(24)

§2.2.9 Perceived Self-Efficacy

According to the social cognitive theory, Self-Efficacy is the belief that one has the ability to perform a specific behaviour (Compeau & Higgins, 1995). Therefore, in this specific context, Perceived Self-Efficacy can be defined as the judgement of one’s ability to use M-Payment (adapted from Wang & Lin; 2006). In previous academic work, the construct of Perceived Self-Efficacy has proven to have a positive relationship with Perceived Ease of Use (Chen et al.; 2009, Luarn & Lin; 2005, Wang; 2003, Venkatesh et al.; 2003, Venkatesh and Davis; 2000), implying that the more an individual believes he or she has the required knowledge, skill or ability (Luarn & Lin; 2005) to use a technology, the more his or her’ Perceived Ease of Use will be.

In their research on the adoption of smartphone use, Chen et al. (2009) found a positive relationship between Self-Efficacy and behavioural Intention to Use. Luarn & Lin (2005) also found a significant positive relationship between Perceived Self-Efficacy and Intention to Use in the Mobile Banking context. In the Mobile Wallet or M-Payment context, this relationship was not evidenced to be significant by Shin et al. (2009), stating that Self-Efficacy had a moderating role on Perceived Ease of Use. However, Wang & Lin (2006) in their study on mobile service adoption, suggest that with the presence of Perceived Ease of Use in their model, Self-Efficacy was still a significant determinant of behavioural intention. Based on the discussion above, it is expected that Perceived Self-Efficacy influences Intention to Use, and the following hypothesis is proposed:

(25)

§2.3 Conceptual Model

Based on the literature discussion, several hypotheses were proposed and added to the original TAM model (Venkatesh and Davis; 2000). To explain the model presented in figure 2, the different linkages are shortly summarized. First, Perceived Ease of Use, was found to have a positive relationship on Intention to Use in the Mobile Payment context by Kim, Mirusmonov and Lee (2010), resulting in hypothesis 1. The construct of Perceived Usefulness also was found significant positive related to Intention to Use by Kim, Mirusmonov and Lee (2010), resulting in hypothesis 2. Perceived Social

Influence was found positively related to Intention to Use in M-Commerce (Wei; 2009), but the

relationship still needs testing in the field of M-Payment (H3). The construct of Perceived

Compatibility was found positively related in the field of M-Payment by Schierz et al. (2010),

resulting in hypothesis 4. Although constructs related to Perceived Credibility have been validated as positively related to Intention to Use in the field of M-Payment, such as Perceived Security (Shin et al.; 2009), the construct of Perceived Credibility still remains untested in the M-Payment field (H5).

Perceived Costs had a significant negative relationship on Intention to Use in M-Commerce (Wei;

(26)

Empirical evidence was found that the construct of Perceived Mobility (H7) had a positive relationship on Intention to Use in M-Payment (Schierz et al.; 2010). The following construct,

Perceived Enjoyment, still remains untested in the field of M-Payment, although research on handheld

internet devices (Bruner; 2005) and e-shopping acceptance (Ha; 2008) show that a positive relationship may exists between Perceived Enjoyment and Intention to Use (H8). At last, in their research on Mobile Wallet or M-Payment, Shin et al. (2009) found the construct of Self-Efficacy not significant positively related to Intention to Use. However, several other studies in the field of mobile services adoption (i.e. Chen et al.; 2009, Wang et al; 2006, Luarn & Lin; 2005) show that a relationship may exist between Perceived Self-Efficacy and Intention to Use (H9). In this research, the above summarized linkages form the basis of the conceptual model seen in figure 2. This results in an integrative model that extends the original TAM with relevant constructs. The green boxes represent validated results in the field of M-payment, where orange boxes represent still unvalidated or untested results in the field of M-Payment:

(27)

§3

Research methodology

In this chapter, the methods used in the research will be explained. First, the items that are needed to measure the constructs are described and the applied academic sources are given. Second, the data collection procedure is clarified. Finally, this chapter will conclude with explaining the appropriate method for processing the data.

§3.1 Measuring the constructs

To measure the constructs relevant to this research, it is adviced to adapt the measurement items for each construct to ensure content validity (Luarn & Lin; 2005). Therefore, in this research, measurement items that were found valid in previous studies are adapted to the field of Mobile Payment. The items that are used to measure the dependant variable of Intention to Use M-Payment are derived from previous work of Schierz et al. (2010) and are based on the original TAM (Gefen et al; 2003, Venkatesh and Davis; 2000, Davis; 1989). Besides the variable of Intention to Use

M-Payment, the other variables are in this research characterized as independant variables. The items to

measure Perceived Usefulness of M-Payment are derived from previous work of Schierz et al. (2010), and are based on academic work in the field of M-Commerce and E-Commerce (Luarn & Lin; 2005,

van der Heijden 2003, Devaraj et al.; 2002, Bhattacherjee; 2001).Being part of the original TAM, the

items to measure Perceived Ease of Use of M-Payment are based on research from several authors (Luarn & Lin; 2005, Venkatesh and Davis; 2000, Taylor and Todd; 1995, Davis; 1989), and are applied to the M-Payment field by Schierz et al. (2010). According to Luarn and Lin (2005), Perceived Usefulness and Perceived Ease of Use instruments: “show good convergent and discrimant properties (Davis; 1989), are internally reliable (Mathieson; 1991, Davis; 1989), and demonstrate predictive validity (Szajna; 1994)”. The construct of Perceived Social Influence, included in the TAM2, is measured by items derived from Schierz et al. (2010), based on academic work of Venkatesh and Davis (2000), Taylor and Todd (1995), and Mathieson (1991).

In addition to the constructs of the TAM2, several other constructs were added to the model proposed in this paper. The construct of Perceived Compatibility of M-Payment is measured by items mainly based on recent work of Schierz et al. (2010) and Kim, Mirusmonov and Lee (2010).

Perceived Credibility items are based on research done by Wang & Lin (2006) and Wang (2003). In

(28)

and privacy issues involving new technology usage, and may therefore add value to the construct of Perceived Credibility.

The Perceived Costs construct is measured by adapting several Perceived Costs items (Wei; 2009, Luarn & Lin; 2005, Wu; 2005) to the context of M-Payment. Next, to measure Perceived

Mobility, items from several papers (Schierz et al.; 2010, Mallat; 2009) were used. Perceived Enjoyment items are based on academic work of Ha (2008) and Bruner (2005), and were adapted to

the field of M-Payment. In addition, the item “I think using Mobile Payment would be more fun than alternative modes of payment (e.g. debit card, credit card, cash)” is added to the research. This item may add extra value to the construct of Perceived Enjoyment. And finally, the construct of Perceived

Self-Efficacy was measured by items derived from various sources, including Wang & Lin (2006) and

Compeau & Higgins (1995). A complete overview of the relevant constructs and accompanying measurement items and sources is given in table 1.

§3.2 Data collection procedure

Relevant data is collected using a questionaire distributed under all university students of the Rijksuniversiteit Groningen. The data is self-reported and the survey is distributed on the internet. The internet is an appropriate medium for this target group, because it fits well with their modern lifestyle.

Contrary to prior research (i.e. Shin; 2009), the role of actual users of a technology is not emphasized. The adoption of M-Payment is still in its preliminary phase in the Netherlands. Therefore, in this research, the emphasis is put on the potential consumers of M-Payment. Hence, a larger target group (university students) is chosen in which the potential adoption of M-Payment is tested. Preceding the survey an clear introduction is given on Mobile Payment. This is essential, to give the respondent a basic level of knowledge on the topic of M-Payment. Because M-Payment is still in its preliminary phase, it is critical to provide the respondent with information regarding the topic.

(29)
(30)

Besides gender and age, the survey will include a distinguishing between smartphone and non-smartphone owner. This is done, because the effort to use M-Payment (with NFC) is smaller for a smartphone user than for a non-smartphone user. In addition, it is likely that the influence of some constructs differ between smartphone/non-smartphone owners, i.e. the Perceived Costs for acquiring a smartphone with NFC-technology could play a considerable role for non-smartphone owners, compared to smartphone owners who already have the built-in NFC (and who are already used to buying expensive mobile devices). In practise anno 2011, the majority of smartphones do not have the technology of NFC built-in. However, some high-end smartphones (like the Samsung Nexus S) are provided with the technique, and Google’s mobile device software Android supports the technology. Google wants to provide more Android powered phones with the NFC technique in the coming period1.

Table 1: Measurement items

Each construct will be measured by several measurement items (see table 1). Respondents are given the opportunity to fill in answers on a seven point Likert scale, varying from strongly disagree to strongly agree (e.g. Wang & Lin; 2006). Likert scales, ranging from “strongly disagree” to “strongly

1

The Wall Street Journal, “Google sets role in Mobile Payment”

online.wsj.com/article/SB10001424052748703576204576226722412152678.html

Construct Item Source

Intention to Use M-Payment Given the opportunity, I will use Mobile Payment Schierz et al. (2010) I am likely to use Mobile Payment in the near future Gefen et al. (2003) I am willing to use Mobile Payment in the near future Venkatesh and Davis (2000) I intend to use Mobile Payment when the opportunity arises Davis (1989)

Perceived Ease of Use Learning to use Mobile Payment is easy for me. Luarn & Lin (2005) It would be easy for me to become skillful at using Mobile Payment Venkatesh and Davis (2000) I would find Mobile Payment easy to use. Taylor and Todd (1995), Davis (1989) Perceived Usefulness Mobile Payment is a useful mode of payment Luarn & Lin (2005)

Using Mobile Payment makes the handling of payments easier Van der Heijden (2003)

By using Mobile Payment, my choices as a consumer are improved (e.g., flexibility, speed) Devaraj et al. (2002), Bhattacherjee (2001) Perceived Social Influence Friend’s suggestion and recommendation will affect my decision to use Mobile Payment Schierz et al. (2010)

Family members/relatives have influence on my to use Mobile Payment Wei (2009),

I will use Mobile Payment if my colleagues use it Venkatesh and Davis (2000) Mass media (e.g. TV, newspaper, articles, radio) will influence me to use Mobile Payment Taylor and Todd (1995) I will use Mobile Payment if the service is widely used by people in my community Mathieson (1991) Perceived Compatibility Using Mobile Payment fits well with my lifestyle Schierz et al. (2010)

Using Mobile Payment fits well with the way I like to purchase products and services Kim (2010) I would appreciate using Mobile Payment instead of alternative modes of payment (e.g., credit card,

cash)

Perceived Credibility Using Mobile Payment would not expose my personal information Wang & Lin (2006) I would find Mobile Payment secure in conducting my transactions Wang (2003) I will use Mobile Payment if the security of the service is guaranteed by the service provider

Perceived Costs There are financial barriers (e.g., having to pay for handset) to my using Mobile Payment Luarn & Lin (2005) The cost of handset (i.e. smartphone) is high for me Wei (2009) I think the equipment cost is expensive of using Mobile Payment Wu (2005) It would cost a lot to use Mobile Payment

Perceived Mobility I could imagine having multiple jobs at a time Schierz et al. (2010) I would like to be able to keep in touch everywhere I am Mallat (2009) I would like to be able to coordinate my daily tasks everywhere I am.

I would like to be able to coordinate my daily tasks no matter what time it is

Perceived Enjoyment It is really fun to shop using Mobile Payment Ha (2008) I think it is fun to use Mobile Payment Bruner (2005) I think using Mobile Payment is enjoyable Karson (2000) I think using Mobile Payment would be more fun than alternative modes of payment

(e.g. debit card, credit card, cash)

Perceived Self-Efficacy I could use Mobile Payment…. Wang & Lin (2006), Luarn & Lin (2005) If I just had the (smartphone) build-in facility for assistance Pedersen (2005)

If I had seen someone else using it before trying it myself Wang (2003)

(31)

agree” are used for all questions, except for the items measuring Perceived Self-Efficacy (Luarn & Lin; 2005), which range from “not at all confident” to “very confident”. The Likert scale was selected based on validated results from prior studies (e.g. Schierz et al.; 2010, Mallat; 2009, Wang & Lin; 2006). The survey is distributed in English. This is done, because the items in previous academic work were also tested in the English language. Translation of these measurement items to i.e. the Dutch language gives room for variances in interpretation and hence, could seriously harm results. In addition, it is not expected that the target group has difficulty with the English language. The survey is shown in Appendix 1.

§3.3 Data processing procedure

In order to process the data derived from the survey, several tests are needed to test the data and to gain validated results. In this research, the survey uses multiple items in measuring a single construct. To process the data according to standards, the stepwise approach described in Wei (2009) is used as a framework. First, the reliability of the survey is tested using Cronbach Alpha. The reliability coefficients of all independent variables need to be above 0.70, suggested by Nunnally (1978).

The Cronbach Alpha score gives an indication whether the measurement items really affect one particular construct. If not, items can be dropped to improve reliability scores (Mallat; 2009).

Second, the construct validity is analyzed. Construct validity measures: “the degree to which a scale measures what it intends to measure” (Garver and Mentzer; 1999) In this case, convergent validity is used to check construct validity, by using measurement items that were found significant in prior studies.

(32)
(33)

§4

Results

In this chapter the results of the research are presented. An emphasis is put on the quantitative results derived from the data, in some occasions supported by some qualitative findings. First, to calculate the total number of respondents needed for a representative sample, a representative sample calculation is presented. Second, general statistics are given and when needed explained. Third, in order to get only the valuable and usable data, the approach concerning missing values is given. Fourth, to test whether the individual items in the survey combined measure underlying constructs, the Cronbach’s Alpha calculation is conducted. Fifth, to test the relationship between the different independent variables, a Pearson correlation test is conducted and explained in this paragraph. Sixth, to examine whether the independent variables really influence the dependent variable, a Multiple Regression analysis is executed. And last, to investigate whether moderating effects play a significant role in this research, a Moderator analysis is conducted.

§4.1 Representative sample

In order to get a representative sample it is important that an appropriate sample size is calculated based on the population. In this case, the students of the Rijksuniversiteit Groningen represent the whole population. There are a total of 27.699 students (October 1, 2010) attending a study at the Rijksuniversiteit Groningen. To calculate the appropriate sample size, the following equation is used:

n> = N x z ² x p(1-p)

z ² x p(1-p) + (N-1) x F ²

n = needed sample size

N = total population

z = standard deviation, when using 95% reliability level, standard deviation = 1,96.

p = chance of someone given an answer (variation between the answers), in most

cases 50%

(34)

Applied to this research, the following equation arisis:

n> = 27.699 x 1,96 ² x 0,50(1-0,50)

1,96 ² x 0,50(1-0,50) + (27.699-1) x 0,05 ²

n = 384,16

To conclude, a minimum of 385 respondents were needed to fulfill the needs for a representative sample.

§4.2 Distribution of survey

To achieve a representative sample, it was needed to give the online survey considerable attention and exposure to the student population. Distribution started in the beginning of May 2011, and the survey was closed in the end of June 2011. Distribution took place using a variety of instruments, gaining exposure to thousands of students across different faculties. First off, the survey was send to student groups (from various courses) through internal mail. Second, flyers were distributed among the different computer areas of the Rijksuniversiteit Groningen. Third, message boards, online and offline, were used to distribute the survey. And fourth, social media networks Hyves (mostly Dutch students) and LinkedIn (International and Business students) were used to contact students. At the point of closure in the end of June 2011, a total of 485 students had filled in the survey.

(35)

Table 2a: General statistics

To get a considerable amount of variety in the target population, the survey was distributed to students of different faculties of the Rijksuniversiteit Groningen. All the faculties are present in the sample. Most students that participated in this research are attending a study at the faculty of Economics and Business (48.5%). In addition, students of the faculty of Behavioural and Social Sciences account for 14.9% of total respondents. Also, the faculty of Mathematics and Natural Sciences (11.9%), and Arts (10.6) form a considerable group of respondents. From the 388 respondents that actually completed the questionnaire, the distribution over the faculties is as follows (see table 2b):

Table 2b: Distribution across faculties

When comparing the distribution of the sample with the distribution of the total population, the sample is not in proportion with the total population. Only the faculties Economics and Business (+26.8%),

Faculty Distribution of student faculty numbers (#) Percentage of sample (%) Total number of students (#) Percentage of total number (%)

Economics and Business 196 48,5 6016 21,7

Behavioral and Social Sciences 60 14,9 4190 15,1 Theology and Religious Studies 4 1,0 194 0,7

Arts 43 10,6 4975 18,0

Medical sciences 12 3,0 3712 13,4

Law 22 5,4 3798 13,7

Spatial Sciences 16 4,0 908 3,3

Philosophy 3 0,7 276 1,0

Mathematics and Natural Sciences 48 11,9 3630 13,1

(36)

Theology and Religious Studies (+0.3%), and Spatial Sciences (+0.7%) are well present in this research, compared to the percentages of the total population of students. The faculties Behavioral and Social Sciences (-0.2%), and Philosophy (-0.3%) are slighty underrepresented compared to the total population, where the remaining faculties of Mathematics and Natural Sciences (-1.2%), Arts (-7.4%), Law (-8.3%), and Medical Sciences (-10.4%), are even higher underrepresented. To conclude, the sample shows variety, but is not in exact proportion with the total population of Rijksuniversiteit Groningen students. The faculty distribution is not adding extra value to the sample. Therefore, the distribution of respondents over faculties is not used in further analysis.

One extra remark has to be placed at the above displayed table 2b. The total sample distribution of students (404) is higher than the number of respondents that actual completed the survey (388, see table 2a). This difference is easily explained, when one considers the possibility to participate in different studies over different faculties.

§4.3 Missing values

In this research, when a respondent accidentely forgot to answer an item, the missing value was estimated based on the average score of the relevant independent construct. To make this approach more clear, an example is mentioned based on scores of respondent X in table 3:

Table 3: Example Missing value calculation

According to the approach used in this research, PercCom3 is based on the mean scores of PercCom1 and PercCom2. By using this approach, not the mean of the total sample is used as a replacement for the missing value, but the individual scores of other more relevant items are taken as a benchmark. This approach was only used when a maximum of one item per construct was missing, with a

maximum of two missing items in total for the whole survey per respondent. In cases where more than two items were forgotten, the respondent was dropped from further analysis. In addition, when a respondent failed (or forgot) to answer a single item on the dependent variable Intention to Use, the respondent was immediately dropped from further analyses. In cases were the calculated mean score

Respondent Construct Item Score

X Perceived Compatibility PercCom1 4

PercCom2 4

PercCom3 ?

(37)

was in between two scores (1.5, 2.5, 3.5, etc.) scores were rounded off upwards. In total, the above formulated approach was used for 16 respondents, involving a total of 18 missing items.

§4.4 Cronbach Alpha

After the missing value approach was applied, the input of 388 respondents was taken into further calculations. To test for the reliability of the survey, correlation was tested using Cronbach’s Alpha. Like mentioned in the “Methods” section of this paper, the reliability coefficients of all independent variables need to be above 0.70, suggested by Nunnally (1978). Cronbach’s Alpha was used to see whether the individual items combined measure an underlying construct. Based on the Cronbach’s Alpha scores, all the items representing the independent constructs, involving Perceived Ease of Use (0.888), Perceived Usefulness (0.849), Perceived Social Influence (0,887), Perceived Compatibility (0.908), Perceived Credibility (0.773), Perceived Costs (0.876), Perceived Mobility (0.734), Perceived Enjoyment (0.944), Perceived Self-Efficacy (0.824), and the items representing the dependent construct of Intention to Use (0.948), showed good reliability. In addition, the items added to the constructs of Perceived Enjoyment and Perceived Credibility, added value. As seen in table 4a, deleting these items, did not explicitly result in to a better Cronbach’s Alpha score.

Table 4a: Cronbach Alpha adjusted constructs, including if deleted

Because all the independent (and dependent) items showed excellent reliability, means were calculated based on the item scores for the different constructs, and were used for further calculation. An overview of the Cronbach’s Alpha scores can be found in table 4b.

Construct Item Cronbach Alpha ….if item deleted

Perceived Credibility Using Mobile Payment would not expose my personal information 0,773 0,694

I would find Mobile Payment secure in conducting my transactions 0,625

I will use Mobile Payment if the security of the service is guaranteed

by the service provider 0,757

Perceived Enjoyment It is really fun to shop using Mobile Payment 0,944 0,923

I think it is fun to use Mobile Payment 0,916

I think using Mobile Payment is enjoyable 0,919

I think using Mobile Payment would be more fun than alternative

(38)

Table 4b: Cronbach’s Alpha scores

§4.5 Correlation between constructs

Now that the construct scores were calculated (see appendix 2a for an overview of the mean scores and standard deviations), it was necessary to calculate the relationship between the variables or constructs. Before using a correlation test, it was important to calculate whether all the scores were normally distributed. The Kolmogorov-Smirnov test can be used to test whether data is normally distributed (see appendix 2b). In this case, the majority of scores, including the dependent construct of Intention to Use, showed a non-normal distribution. However, some constructs seem to be normally distibuted (Perceived Credibility). Two tests can be used for testing correlation between variables. For normally distributed data, the Pearson correlation test is appropriate. For data in which the scores are not normally distributed, the “Spearman’s rho” correlation test is appropriate. Therefore, first a Pearson-correlation-test was conducted. Second, the “Spearman’s rho” test was applied for the same set of variables.

Like mentioned earlier in the “Method” section, based on Wei (2009), a correlation coefficient value range from 0.10 to 0.29 is considered weak, from 0.30 to 0.49 is considered medium, and from 0.50 to 1.0 is considered strong. From the Pearson correlation test, which is an excellent tool to correlate two or more variables on interval level, results showed that most associated pairs (except for cost in some cases) were significantly (P = 0.001) related to eachother. When comparing the two correlation tests (Pearson vs. Spearman’s rho), no striking differences were found concerning correlation values between correlation values. In addition, contrary to the normal Kolmogorov-Smirnov test, when observing the normality plots based on the Lillefors’s test (see appendix 2b), all the data seems to be normally distributed. Therefore, the Pearson correlation test is used as a proxy. From the Pearson correlation test, it was found that, as expected, Perceived Costs is negatively related to most variables (Mobility, Compatibility, Enjoyment, etc.), except for the variables Perceived Social

Construct/Variable Measurement Items (#) Cronbach's Alpha

(39)

Influence (+0.005) and Perceived Self-Efficacy (+0.046). Interesting to mention also is that the variable Perceived Usefulness showed strong relationships (>0.50) with most other variables. When only taking the independent variables as a benchmark, no multicollinearity occured. An overview of the Pearson correlation test results can be found in table 5. Green correlation scores indicate strong relationships (>0.50), where red scores may indicate multicollinearity (>0.80).

Table 5: Pearson correlation test results

§4.6 Multiple regression

In order to test the hypotheses that form the basis of this research the multiple regression test can be used. The (linear) multiple regression test offers the ability to analyze the relationship between a single dependent variable, in this case Intention to Use (M-Payment), and several independent variables (Hair et al; 2005). The regression equation can be formulated as follows:

Yi = b1X + b2X + b0 (constant)

When formulating an appropriate equation based on the multiple regression test, it is again important that the issue of multicollinearity is taken into account. To prevent multicollinearity issues, the tolerance indicator needs to be greater than 0.1, and their VIF values need to be less than 10. When performing a multiple regression, the results shown in table 6a were calculated.

Construct/Variable Ease of Use Usefulness Social Influence Compatibility Credibility Costs Mobility Enjoyment Self-Efficacy Intention to use

Ease of Use 1 Usefulness 0,39** 1 Social Influence 0,135** 0,399** 1 Compatibility 0,257** 0,686** 0,392** 1 Credibility 0,195** 0,571** 0,422** 0,627** 1 Costs -0,135** -0,13* 0,005 -0,225** -0,107* 1 Mobility -0,215** 0,425** 0,266** 0,508** 0,376** -0,241** 1 Enjoyment 0,186** 0,557** 0,457** 0,617** 0,478** -0,083 0,36** 1 Self-Efficacy 0,209** 0,506** 0,498** 0,448** 0,371** 0,046 0,316** 0,419** 1 Intention to Use 0,273** 0,744** 0,455** 0,806** 0,669** -0,237** 0,481** 0,591** 0,486** 1 ** = correlation significant at 0.01 level (two-tailed)

(40)

Table 6a: Multiple Regression scores

To analyze which independent variables are influencing Intention to Use, the (standardized) Beta’s

give insight in the “importance” of variables2. When using a p-value of 0.05, the independent variables

of, in order of influence, Perceived Compatibility (+0.406), Perceived Usefulness (+0.275), Perceived Credibility (+0.178), Perceived Costs (-0.087), and Perceived Social Influence (+0.065) all significantly influence the Intention to Use M-Payment. Perceived Mobility, Perceived Enjoyment, Perceived Self-Efficacy, and Perceived Ease of Use were not found significant.

A clear overview of the results found based on the hypotheses can be seen in figure 3. Green boxes indicate supported hypotheses, where orange boxes indicate rejected hypotheses.

** = significance found at 0.01 level (one-tailed) * = significance found at 0.05 level (one-tailed)

Figure 3: Hypotheses results

2

Adapted from Field (2009), from www.statisticshell.com/multireg.pdf

Construct/Variable (standardized) Beta t Sig Tolerance VIF

(Constant) -0,249 -0,765 0,445 Ease of Use -0,017 -0,599 0,549 0,834 1,199 Usefulness 0,275 6,898 0,000** 0,406 2,463 Social Influence 0,065 2,064 0,040* 0,649 1,540 Compatibility 0,406 9,578 0,000** 0,360 2,781 Credibility 0,178 5,141 0,000** 0,542 1,846 Costs -0,087 -3,200 0,001** 0,884 1,131 Mobility 0,027 0,901 0,368 0,699 1,431 Enjoyment 0,038 1,087 0,278 0,541 1,850 Self-Efficacy 0,049 1,525 0,128 0,614 1,630

Referenties

GERELATEERDE DOCUMENTEN

goal of this research question is to serve as the base to achieve solutions able to cope with different types of DDoS attacks.. To do so, we will use

In deze subparagraaf is onderzocht of het leveren van diensten die gerelateerd zijn aan de interne audit functie in combinatie met een audit de onafhankelijkheid van de

Met de invoering van de Wet dualisering gemeentebestuur in 2002 is de positie van de raad versterkt door middel van verschillende controle instrumenten. Daarmee zou ook de

Hence, big data tools and technologies can be used in monitoring the results of the decision, as well as in providing real-time or periodical feedback on the outcomes of

The presented prosthetic flexure-based finger joint is able to achieve 20N of contact force with an additional 5N of out of plane load over the entire 80˚ range of motion,

The goal of the current study was to compare the Theory of Planned Behavior to the Technology Acceptance Model in terms of their contribution to the understanding of mobile

The results of the current study provide novel factors influencing the adoption of eHealth innovations (co- developed vision, clinical integration, and ease of

While the lag number of reviews and overall rating introduced issues in the survival model it should not here as the modelling of post-adoption usage does not incorporate