• No results found

Blockchain technology : the identification of the main factors influencing perceived merchant adoption of cryptocurrencies as a viable payment solution

N/A
N/A
Protected

Academic year: 2021

Share "Blockchain technology : the identification of the main factors influencing perceived merchant adoption of cryptocurrencies as a viable payment solution"

Copied!
96
0
0

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

Hele tekst

(1)

BLOCKCHAIN TECHNOLOGY: THE IDENTIFICATION

OF THE MAIN FACTORS INFLUENCING PERCEIVED

MERCHANT ADOPTION OF CRYPTOCURRENCIES

AS A VIABLE PAYMENT SOLUTION.

Amsterdam, 26 June 2018

Wesley Verlinde

10632697

Supervisor: W. Dorresteijn

Blockchain Technology

(2)

ABSTRACT

This thesis presents a research model adapted from an extended technology acceptance model (TAM) replacing cost with perceived benefit. The proposed research model consists of five latent variables, being; perceived risk, perceived benefit, perceived compatibility, perceived usefulness and perceived ease of use. Behavioral intention to use was predetermined as the dependent variable of this study. The research model has been empirically tested using data collected from a survey of merchants. Reliability analysis, factor analysis and multiple linear regression has been used to examine the reliability and validity of the research model. The findings indicated this model to show a bad fit to the research variables. Results from a multiple linear regression indicated no direct and significant effects from any of the five latent variables towards behavioral intention to use. However, there were two significant empirical findings, being; familiarity with cryptocurrencies has a direct negative significant effect on perceived risk and perceived benefit has a direct positive significant effect on perceived usefulness. The limitations and implications of this study to researchers and merchants have been discussed.

(3)

STATEMENT OF ORIGINALITY

This document is written by Wesley Verlinde, who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(4)

ACKNOWLEDGEMENT

I would first like to thank my thesis supervisor, Mr. W.H. Dorresteijn, for his constructive criticism along the research period. Whenever running into an issue or question about my research or writing, he consistently steered me into the right direction. Remote assistance as well as face-to-face communication has been very pleasant and satisfying.

I would also like to acknowledge my gratitude towards the University of Amsterdam, for providing me with the knowledge and experiences I currently possess by receiving education in the field of Economics and Business.

Last but not least, I want to express my profound gratitude towards my family and especially my partner K. and my brother S. for providing me with unfailing support and continuous encouragement throughout my years of study. K. has been my source of inspiration, dedication and perseverance. This accomplishment would not have been possible without them.

(5)

TABLE OF CONTENTS (part I) Introduction ... 1 1. Background ... 3 1.1 Blockchain Technology ... 3 1.1.1 Cryptocurrencies ... 5 1.2 Money ... 10 1.3 SmartCash ... 13 1.3.1 Governance Model ... 13 1.3.2 Payment solution ... 14 2. Literature review ... 18

2.1 Research on the acceptance of new payment technologies ... 18

2.2 Current state of research on cryptocurrencies. ... 20

2.3 Research gap ... 22

3. Conceptual framework ... 23

3.1 Innovation Diffusion Theory ... 23

3.2 Criticism on IDT ... 24 3.3 Adaptation on TAM ... 24 3.4 Research Model ... 26 3.4.1 Hypotheses ... 27 4. Research methodology ... 30 4.1 Design ... 30 4.2 Sampling ... 33

5. Data analysis and results ... 36

5.1 Model measurement ... 37

5.2 Empirical analysis ... 39

5.3 Interview analysis ... 43

(6)

TABLE OF CONTENTS (part II)

Reference list... 53

Appendices ... 57

Appendix 1 – Interview Smartcash founder ... 58

Appendix 2 – Merchant Questionnaire ... 63

Appendix 3 – E-mail directed to merchants ... 70

Appendix 4 – List of merchant email addresses ... 71

Appendix 5 – Top 150 biggest cities in The Netherlands ... 76

Appendix 6 – Respondents frequencies ... 79

Appendix 7 – Reliability Analysis of all constructs ... 83

Appendix 8 – Factor analysis ... 87

Appendix 9 – Descriptive Statistics and Correlation ... 88

(7)

[1] Introduction

Blockchain as an emerging technology holds a significant potential to reshape our economy and a multitude of business processes. As Nakamoto (2008) states in his renowned paper, the idea behind Bitcoin is a currency to be used as a peer-to-peer exchange without the need for intermediaries. Blockchain technology and cryptocurrencies have recently seen a significant amount of attention in the press as well as amongst investors and trades.

During this research the status quo on behalf of the current literature has shown to prove the need for research on merchant adoption of cryptocurrencies. Extensive research has been done on Bitcoin and cryptocurrencies in general with respect to their speculative nature. However, there are numerous questions about Bitcoin and cryptocurrencies alike left to be answered. For instance, there is currently no accepted scientific model with a predictive power to provide answers related to cryptocurrencies and their respective use case(s), such as providing an alternative payment solution.

In regard to merchant adoption and alternative payment solutions, this thesis will be focussing on SmartCash. SmartCash is a peer-to-peer digital currency that can be used without the need of a third party. SmartCash’s drive for merchant adoption is reflected in their Point of Sale solution, including a (debit cryptocurrency) SmartCard, the Merchant Reader and InstantPay, which allows the merchant as well as the user to use SmartCash in a fast and secure manner without the need for a third party fee, unlike the current electronic payment solutions available to merchants.

However, for cryptocurrencies, such as SmartCash, to be widely adopted certain barriers will have to be overcome. Therefore, to overcome such barriers it is important to visualise the determining factors leading to merchant adoption. The research question of this thesis is as follows:

(8)

[2]

“What are the main factors influencing merchant adoption of

cryptocurrencies as a viable payment solution?”.

This research question will be answered via quantitative research based on a conceptual model adapted from the extended Technology Acceptance Model. Due to the fact that adoption (on a wide) scale has not taken place yet, the conceptual model of this thesis is aimed towards the behavioral intention of merchants to adopt and use cryptocurrencies as a payment solution.

(9)

[3] 1. Background

For a better understanding of the issue this thesis is focussing upon background information is needed on multiple subjects. Firstly, the concept of blockchain technology will be discussed. Subsequently the concept and characteristics of money will be described. Thirdly cryptocurrency and its features will be described. As stated before, this thesis will focus upon SmartCash and delineate on current integration as well as provide recommendations based on the research of this thesis on the account of merchant adoption. Therefore the design and features of SmartCash will be reviewed. Lastly, an acknowledgement and disclaimer of the binary intrinsic motivation for the subject of this thesis will be addressed.

1.1 Blockchain Technology

Mostly due to the popularity of Bitcoin the technology behind digital currency has attracted a lot of attention (Underwood, 2016). The blockchain technology can be considered as applicable to a multitude of businesses aside from cryptocurrencies (Romano, & Schmid, 2017). Blockchain technology makes it possible for contracts and other processes to act in a decentralized and distributed manner. It is evident that the decentralized functionality of the blockchain could become the economic layer the World Wide Web for example never had (Swan, 2015). This technology is spreading amongst companies, investors and mainstream media alike, due to the multitude of possible applications. Bitcoin for instance has been compared to the Dot-com bubble due to the enormous influx of investors’ money. Investors investing in Blockchain start-ups amounts to billions and this trend is ever increasing. However the potential of this technology is still fairly unknown throughout society (Mettler, 2016). Although researchers agree on the fact that Blockchain

(10)

[4]

has already proven to be a powerful tool, not only for minimizing costs but also able to bring forth major changes to the entire financial field (Nguyen, 2016). It is obvious that the emergence of this technology linked to the introduction of Bitcoin (Peters, & Panayi, 2016). However, as noted before, this technology is applicable to a wide array of fields.

Multiple use cases, aside from cryptocurrencies, have been identified. The application of blockchain technology in the energy sector has been extensively reviewed, however there are obstacles such as the non-existence of a legislative framework needed to overcome before this technology is ready to be introduced and put to work (Sikorski, Haughton, & Kraft, 2017). The applicability of blockchain indeed requires preparation on legal and policy aspects such as customer protection and management of capital and monetary policy. Said policies should be designed for society to benefit from blockchain technology while illegal use is prevented (Nguyen, 2016). Another industry presented to benefit from blockchain technology is healthcare. Since research has shown that the diffusion of innovation for example wearables in healthcare is still not common due to concerns about data security, there are initiatives investing in blockchain-based technology for storing, authenticating and verifying the integrity of medical personal data (Romano, & Schmid, 2017). One such example, is ‘Healthbank’, a Swiss digital health start-up, offering a platform where users can store and manage their personal health data in a secure database (Mettler, 2016). A decentralized database, whereas multiple parties are in the need of access to such information. The importance such a database lies in the sovereignty of personal data in the hands of the person this data belongs to.

In summary the innovativeness and importance of blockchain technology has been identified. Applicability to a wide array of industries besides cryptocurrency will be the main reason for this technology to gain

(11)

[5]

momentum and interest from a multitude of businesses. However obstacles such as the implementation of a legal framework and policies need to be overcome in order for this technology to be widely adopted and accepted.

1.1.1 Cryptocurrencies

Recently, cryptocurrencies received a lot of attention from mainstream media as well as amongst investors. Despite the recent popularity, research in the field of cryptocurrencies is scarce and thus there is no literal general accepted definition for cryptocurrencies. There are two basic features accepted amongst the cryptocurrency community, being; cryptocurrencies stem from a decentralized network and it uses public-to-private key cryptography. One definition is as follows: “Cryptocurrencies are physical precomputed files utilizing a public key / private key pairs generated around a specific encryption

algorithm. The key assigns ownership of each key pair, or ‘coin,’ to the person who is in possession of the private key.” (Ahamad, Nair, & Varghese, 2013). The ‘physicality’ of cryptocurrencies is debateable. Currently there are 1645 different cryptocurrencies, of which 852 (51.8%) are coins, and thus 793 (48.2%) are tokens or digital assets1. The difference is that a coin is an

independently operating cryptocurrency, whereas a token depends on another cryptocurrency as a platform to operate. For example Erc20 and NEP5-tokens, respectively Ethereum and Neo based tokens.

(12)

[6] Practice of cryptocurrencies

As Nakamoto (2008) described; the ideology behind cryptocurrencies is to provide the transfer of funds in a secure and easy manner with minimal transaction costs, while discarding intermediaries such as financial institutions. The system, or network behind Bitcoin and cryptocurrencies alike is decentralized, meaning there is no single point of failure. One important feature of cryptocurrencies is this decentralized system. This system is needed because transactions need to be tracked and recorded to prevent users from double spending (what Nakamoto refers to as a ‘double spending attack’). The idea behind blockchain came from Nakamoto, as he idealized the use of a chain where owners and users need to digitally sign every transaction and letting others in the network verify these transactions and signatures.

However, there is no central point of failure thus there is also no central point of verification. Therefore all transactions are to be verified and chained together (blockchain). In regards to Bitcoin and other cryptocurrencies using a POW consensus algorithm, these transactions need to be encrypted and broadcast across their respective network. Miners are users that are providing their computational power towards this network. By solving ‘puzzles’, computational problems, these miners prove that the transaction has been verified, processed and added to the chain. This concept is called ‘Proof of Work’. A lot has been said and written about Proof of Work and its implications, this will be discussed later on. This concept is crucial for this network to exist, because this guarantees the integrity of the entire network. If one would try to attack the network or try to change a transaction (double spend), this attacker would have to hack the entire blockchain. The distribution amongst miners depends on the overall difficulty of these ‘puzzles’. If the total processing power of the network increases, so does the difficulty. There is a maximum supply of

(13)

[7]

21 million Bitcoins, which are divisible by 8 decimals or 100 million units. One such unit (0.00000001 BTC) is called a satoshi. Each time a miner creates a block they are rewarded with a predetermined amount of Bitcoins. Due to Bitcoin’s protocol and the maximum supply there is a decreasing number of Bitcoins rewarded towards these miners.

On behalf of the concept of cryptocurrencies, a blockchain can be explained as a distributed ledger (database) which records all digital events (in the case of bitcoin; transactions) and shared between all the participants of this ‘digital network’ (Crosby, Pattanayak, Verma, & Kalyanaraman, 2016). Blockchain is a list of records of transactions which are linked and stored within blocks as well as secured via the use of cryptography, hence the name ‘cryptocurrencies’.

Fig. 1. Cryptocurrency transaction via a POW consensus algorithm.

A transaction consists of a cryptographic hash. A transaction will be verified by a consensus of participants in the system (Crosby et al., 2016). A transaction will be added to the distributed ledger. This transaction will be verified by the

(14)

[8]

nodes, any device such as a computer connected to the network, participating in this network. Once consensus has been reached, this new data is added to a block and this block will be added to the entire chain, hence ‘blockchain’ (Underwood, 2016). Fig. 1 visualises a cryptocurrency transaction being verified via a POW (Proof of Work) consensus algorithm.

Issues concerning POW – waste of resources

Although the necessity of miners within Bitcoin’s network has been stated, mining has certain important drawbacks and implications. The energy footprint and amount of wasted resources is one aspect caused by mining which has received significant attention, although the amount of research on this topic is scarce.

There has been an increasing trend in difficulty, for Bitcoin mining. This has been caused by the increase of resources put towards the entire network. Therefore the cost of Bitcoin mining hardware has seen exceeding costs, even beyond the value of the rewards (O’Dwyer, & Malone, 2014). As noted, the gain in popularity of Bitcoin started an arms race amongst miners, as well as a problem for the gaming industry. Bitcoin miners switched from CPU’s to GPU’s (graphic cards) due to the fact this offered higher performance. Data has suggested that miners (in 2018) currently use around 0.0012% of the entire energy consumption on this planet (Williamson, 2018). However this number seems rather small, research has shown that Bitcoin at this moment requires 80,000 times more electricity for one transaction than Visa (Williamson, 2018). The picture below (Digiconomist, 2018) shows the comparison of Bitcoin’s energy consumption compared to VISA.

(15)

[9]

Fig. 2. Bitcoin transaction energy consumption compared to VISA.

Fig. 2. shows that a single Bitcoin transaction uses 1017 kWh compared to 1 VISA transactions using 0.00169 kWh. Meaning that a single Bitcoin transaction uses 601,775 times more energy than a single VISA transaction ((1,017/169) x 100,000 = 601,775.15). Estimations on energy consumption, from mining, range from Irelands’ electricity consumption (O’Dwyer, & Malone, 2014), to 3-16 Petajoule compared to 2340 Petajoule used by the current banking industry (Vranken, 2017). It can be concluded that the energy consumption by Bitcoin is huge, however compared to the current energy consumption by the banking industry this seems relatively small.

(16)

[10] 1.2 Money

In order to use cryptocurrencies in (online) transactions, it should consist of the features and characteristics of ‘money’. A widely accepted definition of money is as follows: “Money is the stock of assets that can be readily used to make transactions” (Mankiw, 2015). There are two forms of money to be identified as of today; being (1) fiat money: which has no intrinsic value, for example the paper currency and (2) commodity money: which has intrinsic value, for example gold and silver. Intrinsic value means the actual value based on a perception of true value, which is in terms of both tangible and intangible factors. However the importance of money lies in the three functions it entails; medium of exchange, store of value and unit of account (Mankiw, 2015);

1. Medium of Exchange; when money is used to buy goods and services it acts as a medium of exchange with agreed upon terms between buyer and seller. The ideology behind Bitcoin is for it to be used as a currency in the form of a peer-to-peer exchange (Nakamoto, 2008). For bitcoin and cryptocurrencies alike to function as a medium of exchange it has to be accepted on a sufficient scale, because a user is only willing to accept money when he or she is confident that others will accept this in return (Lo, & Wang, 2014). Bitcoin is spreading to more mainstream vendors such as Amazon, which could be seen as a signal that the new technology and cryptocurrencies is perceived to be positive and thus has a future in payment solutions. One problem Bitcoin has to overcome to be widely used as medium of exchange is the difficulty of procuring new bitcoins, once transactions are made the user most now find a source of new bitcoins (Yermack, 2015). There is a lack of a ‘closed loop’ economy, people and businesses would need to get paid in cryptocurrencies as a primary payment method in order to withstand the

(17)

[11]

problem of procurement (Hileman, & Rauchs, 2017). Another problem which a lot of critics argue concerns the volatility of Bitcoin. As a result of this (high) volatility Bitcoin’s worth deflates and therefore limits the use of Bitcoin as a medium of exchange (Van Alstyne, 2014). However a user that controls the private key to their cryptocurrencies is able to use this asset as a speculative tool but more importantly as a medium of exchange since these funds cannot be seized and transactions can also not be censored (Hileman, & Rauchs, 2017).

2. Store of value; transferring purchasing power from the present to the future, or in other words storing wealth and maintaining its value without depreciation (Mankiw, 2015). Commodities such as gold and silver are perfect examples of stores of value, as it maintains its value in perpetuity. Again, volatility plays a huge role. As any holder or user of Bitcoins faces the risk of fluctuations (Böhme, Christin, Edelman, & Moore, 2015). At current market volume2 a trader seeking to trade a

certain amount of Bitcoin is unable to do so without affecting the market price. It must be known that Bitcoin was never meant to be a store of value, and thus price should not be of concern as long as it is traded against or backed by dollars or another currency. Currently Bitcoin is actually completely untethered to other currencies, therefore making it impossible to hedge against market risk for businesses as well as customers and therefore completely undermining the function store of value (Yermack, 2015).

2 dated May 18, 2018

(18)

[12]

3. Unit of account; this means the currency is considered as a common unit by which everyone measures prices and values (Mankiw, 2015). In economics this means a unit of account is a nominal unit of measure or currency which represents the real value of this item. Herein Bitcoin again faces certain obstacles. Retailers accepting the currency will need to recalculate automatically or at least on a sufficient frequency due to the volatility (Yermack, 2015). It is possible for customers to incur a psychological cost when they perceive the price in Bitcoin to fluctuate heavily, therefore the ability to serve as unit of account is again diminished (Lo, & Wang, 2014). Another negative psychological aspect influencing adoption might be that merchants are required to quote their prices into (at Bitcoin’s current price) five or even more decimal points. This is a practice that is highly likely to confuse buyers and sellers alike (Yermack, 2015).

To summarize, for being able to use cryptocurrencies as money it has been noted that it needs to fulfil the characteristics of three separate functions: medium of exchange, store of value, unit of account. In spite of criticism

concerning Bitcoin, it has already fulfilled characteristics in all of the different functions, with certain hurdles to overcome. Most of those hurdles and criticism seem to be based on (heavy) volatility, meaning fluctuations in price. Another important aspect, which can be identified as a barrier to adoption, is the psychology behind bitcoin transactions and the need for frequent price updates and price display of multiple decimal points.

(19)

[13] 1.3 SmartCash

SmartCash is a seed funded decentralized peer-to-peer digital currency able to be used without the need of intermediaries or putting trust in a third party. This means that a holder of this currency acts as his own bank. The focus of SmartCash lies in ease of use and being merchant friendly. Another aspect of SmartCash is the seed-funded governance model. SmartCash’s ultimate aim is to replace the centralized fiat currencies that are controlled by the so called ‘too big to fail’ banks. The governance ecosystem of SmartCash is self-funded and therefore allowing the founders not having to rely on outside investors who are in it for the profits. The main strength of SmartCash is their drive for merchant adoption and their current payment solutions such as the Point Of Sale solution, which includes the SmartCard, Merchant Reader and InstantPay. These features will allow merchants as well as users to use SmartCash in (online) transactions in a fast and secure manner without the need of a third party and thus there is no need for high transaction costs, such as credit-card fees.

1.3.1 Governance Model

The notion of governance entails all processes of governing, the process of interaction and decision making with relation to all actors which are involved, undertaken by any entity possible, such as governments, markets, business, families or social systems alike (Bevir, 2012). Governance implicates the manner to which norms and actions are to be regulated and the process of holding actors accountable. Governance takes on many forms such as a democracy, autocracy, oligarchy and many more. In many cases external actors, without absolute decision power, have the ability to influence the governing processes.

(20)

[14]

Speaking of governance, SmartCash’s SmartHive Governance Model is an unique aspect or feature of SmartCash’s design. Currently3 1.6 billion

SmartCash (out of 5 billion maximum supply) is in circulation. Based on the SmartCash’s richlist4, 52.61% (at the moment of writing) of the 1.6 billion

SmartCash in circulation (a total of 827 million) is locked up in the so called ‘SmartHive Community Address’. This balance holds the total ‘money’ available for funding potential projects. Everyone, even without holding SmartCash, is able to create a proposal with the goal of helping SmartCash in any way possible such as creating awareness, striving for merchant adoption or projects based on community support.

Each SmartCash holder holds the right to. One SMART equals one vote. Once a proposal’s deadline for voting is met, in order for the proposal to pass it needs at least 51% of ‘yes’ voting. The notion of SmartCash not being a democracy comes from the fact that relatively bigger holders of SmartCash have a higher say or a bigger amount to vote with, the reasoning behind this is that bigger stakeholders hold a relatively higher incentive for SmartCash to succeed and therefore their vote has a higher weight.

1.3.2 Payment solution

The core team of SmartCash has and is producing a huge amount of developments which make it able to integrate SmartCash into third parties. Their payment solutions will make merchants able to easily integrate SmartCash to their current payment systems and solutions, therefore easing the process of merchant adoption.

3 Dated May 18, 2018

(21)

[15] Smartcard

Consumers have used debit and credit card (payments) as the dominant payment method for recent years. Adoption of such cards have become simple for businesses and even smartphone-powered solutions are available today. Fig. 3. shows that transaction fees for the use of credit cards can amount to over 3% (ValuePenguin, 2018), which could lead to substantial costs for (small) businesses. In addition, buyers and merchants need access to major banks and their services (inducing extra costs) to facilitate these transactions.

Fig. 3. Credit card transaction fees.

SmartCash is developing the cryptocurrency alternative to debit cards, called the ‘SmartCard’. This card will use their respective blockchain to verify and initiate transactions. The fees are almost nihil, the cost of one transaction at the time of writing is fractions of a cent, regardless of the amount transacted. At the moment of writing 1 SMART equals €0.07635, therefore €100 equals

1310.62 SmartCash. Exemplary for current transaction costs, as visualized in Fig. 46, sending this amount to another user (equal to paying a merchant)

entails a transaction fee of 0.001 SMART, €0.0000763 or 1/131th of a cent. As the blockchain amongst cryptocurrency has proven already, transactions will

5 Data taken from: https://www.coingecko.com/en/price_charts/smartcash/eur; Retrieved

May 24, 2018.

(22)

[16]

be instant. Thus making the payment process convenient, as consumers are used to with regular card payments, however on a fraction of the respective cost. It should be noted that indeed there will be extra investment costs due to the need of the POS application for the merchant. However this trade-off would most likely fall off due to the incentive of significantly lower transaction fees on the merchant side. Comparing the credit card fees shown in Fig. 3, whereas the average transaction fee amounts to 2.23% compared to a SmartCash transaction with a 0.0000763% fee.

Fig. 4. SmartCash transaction fee.

Practice of POS (Point of Sale)

There are two applications needed for a transaction between merchant and customer with the use of SmartCash; (1) the SmartCard or the card app for the consumer and (2) the POS (Point of Sale) app for the merchant.

(1) The card app makes it able for a consumer to create their own SmartCard. Each card consists of a public address that can be used to load more funds onto (hence: debit card). Also, a QR code is included such that this can be scanned by the POS application by a merchant or any other user, from, for example, a phone screen. The workings of this debit card is similar to the currently adopted and used debit cards.

(23)

[17]

(2) A merchant only needs the Point of Sale application. The merchant only needs to specify the transaction amount in the respective native or local currency, the application automatically initiates the transaction in the equivalent amount in SmartCash. Subsequently the application connects to the SmartCash blockchain to broadcast this transaction. If previously set requirements (such as sufficient balance, correct pin-code) are met the transaction is instantly verified.

(24)

[18] 2. Literature review

In order to demonstrate the state of the field of research into cryptocurrencies, a critical and structured literature review is undertaken prior to defining important topics for further research. Although Bitcoin has been around for almost 10 years, in depth research is scarce. Research mostly focusses on the blockchain technology behind Bitcoin and the speculative aspects of cryptocurrencies. The following literature review will consist of three separate motives. Firstly, the current state of research on the acceptance of new payment technologies will be discussed. Subsequently the topic of cryptocurrencies will be discussed. Lastly, current literature has been examined to identify important topics for further research, leading to the research gap on the topic of merchant adoption of cryptocurrencies.

2.1 Research on the acceptance of new payment technologies

Research on currently adopted payment solutions could prove to be useful in identifying determining factors of adoption in the field of payment solutions. Extensive research has been undertaken on the subject of e-commerce as well as m-e-commerce. E-e-commerce, or electronic e-commerce is defined as the activity of buying and selling online (or over the internet). M-commerce, or mobile M-commerce, is defined as the activity of buying and selling through the use of wireless handheld devices, such as mobile (smart)phones and personal digital assistants (PDAs). This form of commerce is also known as ‘next-generation e-commerce’.

An important model that focusses upon the acceptance of new technologies can be found in TAM, Technology Acceptance Model (Davis, Bagozzi, & Warshaw, 1989). Visualisation of the model is shown below (Davis et al., 1989, p. 985).

(25)

[19]

Fig. 5. Technology Acceptance Model (TAM).

This model has received widespread attention, as well as numerous citations, in the field of Information Systems. This model places an emphasis on the roles of perceived ease-of-use and perceived usefulness on adoption decisions of new technologies (Plouffe, Hulland, & Vandenbosch, 2001a). TAM provides an explanation of determining factors of acceptance of new technologies. This model suggests that two ‘beliefs’ are the focal point of relevance for acceptance behaviours, which are; perceived usefulness and perceived ease of use. TAM determines Behavioral Intention to Use (BI) via two main factors being Attitude Toward Using (A) and Perceived Usefulness (U).

Research on m-commerce has strengthened the notion that perceived usefulness and perceived ease of use are the most relevant and strongest

predictors of intention to use (Kim, Mirusmonov, & Lee, 2010). However this notion is contradicted, as it has been shown that not perceived usefulness and perceived ease-of-use but perceived compatibility has the greatest effect on

intention to use (Schierz, Schilke, & Wirtz, 2010). As the authors state, this is an important finding because this factor is not included in the original TAM and therefore often overlooked. Compatibility has also shown to be an important factor in merchant adoption intention towards a new electronic payment system (Plouffe et al., 2001a). The relevance of TAM and the goal of current research is to extend this model to include other determinants and to understand how

(26)

[20]

the effect of those determinants change over time due to changing experiences and the targeted systems (Venkatesh, & Davis, 2000).

Additionally to providing frameworks of determining factors influencing intention to use or the acceptance of new technologies, research has provided us with (possible) barriers to adoption. One such barrier could include complexity. The complexity of payment systems as well as the lack of standardization could prove to put a strain on the adoption process (Mallat, & Tuunainen, 2008). Standardized solutions as well as proprietary solutions aimed towards cooperation between users and merchants is indeed needed for mobile payment services to succeed (Dahlberg, Mallat, Ondrus, & Zmijewska, 2008).

2.2 Current state of research on cryptocurrencies.

An article on Bitcoin Magazine (2018) suggests that Dutch citizens have a combined capital of 3 billion euros invested into the cryptocurrency market. In spite of the popularity there are signals from multiple stakeholders, such as the AFM and the Dutch Minister of Finance, that Bitcoin and cryptocurrencies alike are highly speculative tools from which investors should be protected (“Minister bekijkt aanpak speculatie in bitcoin”, 2018). It is suggested that the intrinsic value of Bitcoin is nihil and that the price of Bitcoin is indeed highly prone to speculation (Cheah, & Fry, 2015). The behaviour of Bitcoin reflects the characteristic of a speculative investment rather than a currency (Yermack, 2015). Payments are progressively becoming cashless, as suggested in an article on CNBC (Browne, 2017). Therefore blockchain technology in the form of digital currencies such as Bitcoin are increasingly being used as a payment solution. There are certain issues constraining Bitcoin to be used for (micro)payments, such as delay of transaction confirmation time (Bamert, Decker, Elsen, Wattenhofer, & Welten, 2013). These issues seem to be leading

(27)

[21]

to a rather low perceived ease of use among stakeholders due to transfer issues (Baur, Bühler, Bick, & Bonorden, 2015). Additionally a mechanism for the protection of the merchant as well as the customer should be in place. Bitcoin transactions pose a risk due to the absence of basic consumer protection, such as the provision of refunds (Yermack, 2015). This feature as a mechanism, if needed, to return money to the customer during a failed transaction is important (Bamert et al., 2013). The issue merchants face in the situation of a fault during a Bitcoin transaction and how to deal with returns needs to be addressed. Since cryptocurrency transactions, in the current state, cannot be cancelled but only be offset via a new transaction. This could possibly lead to a conflict between merchant and customer (Lo, & Wang, 2014).

Opposed to the negative aspects of cryptocurrencies, literature also reflects upon the benefits for the use and adoption of cryptocurrencies for merchants. Merchants can benefit from lower fees and avoid chargebacks, thus leading to cost-effective micropayments (Hileman, & Rauchs, 2017). The significance of cryptocurrency as a payment processing solution could indeed be noticeably low-cost for merchants (Böhme et al., 2015). The low-cost feature of cryptocurrencies seems to be the most important argument for merchants on behalf of adoption of this technology, again with regards to micropayments (Baur et al., 2015). However these authors also state that merchants seem to be focussed upon the importance of innovation which would be the most decisive factor to offer cryptocurrency payments. It is possible for Bitcoin to be used for cashless payments (Bamert et al., 2013). It has also been predicted that the legacy of Bitcoin will be remembered due to the innovation on the field of payment technology (Lo, & Wang, 2014). Research has shown that users indeed display and confirm the usefulness and convenience of Bitcoin (Baur et al., 2015). Positive as well as negative aspects of cryptocurrencies have been expanded upon. However, the usefulness, innovativeness and probable use

(28)

[22]

case are well established. Next step should be to identify the barriers or determining factors of widespread cryptocurrency adoption. The need for a certain level of computer knowledge is required for using bitcoins which could act as a barrier to adoption (Yermack, 2015).

2.3 Research gap

Current literature proves the need for research on merchant adoption of cryptocurrencies. There are many open questions about Bitcoin and other cryptocurrencies left to be answered. There is no scientific model with a predictive power to provide answers to questions about Bitcoin and related currencies on their use case (Bonneau, Miller, Clark, Narayanan, Kroll, & Felten, 2015). Developing a framework for understanding digital currency related issues along with the potential impact of this technology on commerce is acknowledged (Lo, & Wang (2014). Therefore, this thesis could contribute to the research upon this topic which is still needed for understanding the disruptive nature and adoption of cryptocurrencies (Baur et al., 2015). The reason that Bitcoin has value is partly due to the property of a medium of exchange. This is why merchant adoption of cryptocurrencies is an important and interesting topic of research, as the consequence of adoption is most likely that liquidity and thus the exchange of cryptocurrencies would increase (Van Alstyne, 2014).

(29)

[23] 3. Conceptual framework

3.1 Innovation Diffusion Theory

As stated before, the amount of in-depth research concerning the adoption of cryptocurrencies is scarce. In the process of researching adoption theories and theories related to the rate of adoption, one research stands out. Rogers (1995) ‘Innovation Diffusion Theory (IDT)’ has produced a theory on the adoption of innovations among individuals and organizations. This book has been cited often within this field of study. In his ‘Diffusion of Innovations Theory’, variables determining the rate of adoption of innovations are visualised as shown below (Rogers, 2010, p. 43).

Fig. 6. Variables Determining the Rate of Adoption of Innovations.

Rogers (1995) established five variables that determine the rate of adoption of innovations; Perceived Attributes of Innovations, Type of Innovation-Decision, Communication Channels, Nature of the Social System and Extent of Change

Agents’ Promotion Efforts. More importantly this theory determined five perceived attributes of innovation, being; Relative advantage, Compatibility,

(30)

[24]

Complexity, Trialability and Observability. Rogers (1995) has proven that;

perceived relative advantage of an innovation is positively related to its rate of adoption, perceived complexity is negatively related to the rate of adoption and perceived compatibility is positively related to the rate of adoption.

3.2 Criticism on IDT

With the extensive research amongst a multitude of disciplines on Diffusion of Innovation, during the years after the creation of IDT, this theory has been applied in different ways. This lead to a lack of cohesion in theory and made it difficult to be applied to new problems (Meyers, Sivakumar, & Nakata, 1999; Katz, Levin, & Hamilton, 1963). Not to mention, measuring what exactly leads to adoption or diffusion of a new technology is extremely difficult (Damanpour, 1996). Adding to that it has been shown that diffusion theories are unable to account for the entirety of variables and thus might overlook certain important predictors of innovation adoption (Plsek, & Greenhalgh, 2001). Consequently this has led to inconsistent results and therefore reducing the value of this theory (Downs, & Mohr, 1976). However the necessity of an heuristic model is contradicted by the suggestion that a one-way model is insufficient (Robertson, Swan, & Newell, 1996).

3.3 Adaptation on TAM

The conceptual model of this thesis is based on the extended TAM (Wu, & Wang, 2005), which integrates IDT, perceived risk and cost into the TAM. This model, visualised in Fig. 7., was used to research the determining factors on user acceptance of m-commerce (Wu, & Wang, 2005, p. 722). These authors describe that system use is recognized as an indicator of information systems success. Subsequently they suggest that acceptance and usage of mobile

(31)

[25]

commerce implies that consumers are embracing new technology innovations on payment solutions. The (relative) constructs in their model include behaviour prediction, user acceptance and innovation adoption.

Fig. 7. Proposed research model for mobile commerce acceptance.

Previous research has been critically reviewed in order to construct a list of measures (Wu, & Wang, 2005). Perceived usefulness, perceived ease of use, behavioral intention to use and actual use have been adopted from previous

research on TAM. The addition of perceived risk has been adapted from research of Pavlou and Eastin as well as the measurement cost has been derived from previous studies from Constantinides, Rupp and Smith (as cited in Wu, & Wang 2005). The authors suggest that numerous studies indicated that TAM needed the integration of additional variables in order to improve the predictive power of the model. Based on the behavioral decision theory, cost-benefit is significant to perceived usefulness as well as ease of use.

More importantly the results of this research have shown that perceived usefulness as well as perceived ease of use influence the actual use through

(32)

[26]

(Venkatesh, & Davis, 2000). However, the most important determining factor for behavioral intention to use is compatibility. Again, this result has already been described and supported (Plouffe et al., 2001a).

3.4 Research Model

Fig. 8. represents the conceptual model of this thesis. This model is an adaptation on the revised TAM constructed by Wu and Wang (2005). To illustrate, this construct consists of five dimensions; perceived risk, perceived benefit, compatibility, perceived usefulness, perceived ease of use. Perceived

risk can be described as the risks for merchants concerning the adoption of

cryptocurrencies, such as a possible conflict between merchant and customer (Lo, & Wang, 2014), for example the situation where a provision of refunds is needed (Yermack, 2015). Studies on the field of commerce have shown that perceived usefulness and perceived ease of use are the strongest predictors of intention to use (Kim et al., 2010). With the same reasoning this thesis will test these predictors on merchant adoption of cryptocurrencies.

(33)

[27]

Additionally this model consists of two primary changes. (a) Cost has been changed to Perceived Benefit. Research has shown that one of the most important factors cryptocurrencies consist of is the low-cost feature and can therefore be beneficiary as a cost-cutting payment solution for merchants (Hileman, & Rauchs, 2017; Böhme et al., 2015). (b) Actual Use has been taken out of the model, on account of the fact that there is little to no observable current use of cryptocurrencies as a payment solution in The Netherlands. Therefore it would be difficult to measure actual use. The objective of this thesis is to measure the factors determining perceived merchant adoption of cryptocurrencies. Therefore, measuring behavioral intention to use is sufficient. Subsequently the conceptual model of this thesis is translated into the following assumptions, in the form of hypotheses.

3.4.1 Hypotheses Perceived risk

Hypothesis 1: Perceived risk has a direct negative effect on behavioral intention to use. As Lim (2003) and Mitchell (1999) have shown, the perceived risk that is associated with innovations has received a significant increase in the field of consumer research on adoption of innovations (Schierz et al., 2010). Merchants are more likely to be concerned with financial risk than social risk. The importance of (perceived) risk is evident (Plouffe et al., 2001a). Although perceived risk has been suggested to positively influence behavioral intention to use (Wu, & Wang, 2005), this seems to be counter intuitive. Therefore the prediction is that perceived risk has a direct negative effect on behavioral intention to use.

(34)

[28] Perceived benefit

Hypothesis 2: Perceived benefit has a direct positive effect on behavioral intention to use. Although research on the adoption of innovations has mostly focused upon the (perceived) cost(s). This thesis focusses upon the perceived benefits. Considering the fact that this thesis expands upon the adaptation of TAM, a discrepancy needs to be noted. From the consumer perspective it has been shown that concerns related to cost is one of the significant issues concerning mobile commerce (Plouffe et al., as cited in Wu, & Wang, 2005). Further research, on the basis of the adaptation of TAM, has concluded that perceived cost has a lower influence on behavioral intention to use than perceived risk, compatibility and perceived usefulness (Wu, & Wang, 2005). This does not mean the importance of perceived cost(s) should be discarded, as it is evident that one of the major barriers of adoption is related to (investment) costs (Mallat, & Tuunainen, 2008). As cost has been changed to benefit, the prediction is that perceived benefit has a direct positive effect on behavioral intention to use.

Perceived compatibility

Hypothesis 3: Perceived compatibility has a direct positive effect on behavioral intention to use. Compatibility is an important aspect as this would most likely add value to the retail consumption experiences (Gatignon, & Robertson, as cited in Plouffe, Vandenbosch, & Hulland, 2001b). Perceived compatibility has the most significant effect on intention to use (Schierz et al., 2010). In addition to this notion, it has also been shown that compatibility is an important factor in merchant adoption intention towards new payment systems (Plouffe et al., 2001a). Therefore the prediction is that perceived compatibility has a direct positive effect on behavioral intention to use.

(35)

[29] Perceived usefulness

Hypothesis 4: Perceived usefulness has a direct positive effect on behavioral intention to use. Perceived usefulness needs no further defining. Perceived usefulness as well as perceived ease of use influences the actual use (adoption) through behavioral intention to use (Venkatesh, & Davis, 2000). Perceived usefulness is one of the strongest predictors in intention to use (Kim et al., 2010). Therefore the prediction is that perceived usefulness has a direct positive effect on behavioral intention to use.

Perceived ease of use

Hypothesis 5: Perceived ease of use has a direct positive effect on behavioral intention to use. Perceived ease of use is an aspect not to overlook. Adoption of an innovative new technology such as a new payment solution is supposed to produce a perceived ease of use amongst users and adopters, in order for adoption to spread. Although current payment providers advertise with better security, a higher ease of use and more importantly lower fees, none of the current payment providers have been able to provide all of these advertised features (Bauer et al., 2015). Perceived ease of use concerning cryptocurrencies, such as usability, mobile wallets, storage and much more is considered to be low (Bauer et al., 2015). Important to know is that especially early adopters value ease of use, moreover the empirical analysis on previous research has shown that perceived ease of use has a significant effect on intention to use mobile payment solutions (Kim et al., 2010). Not to be overlooked, a trade-off between security and ease of use exists (Dahlberg et al., 2008). Research has already determined that perceived ease of use has a significant indirect effect on behavioral intention to use (Wu, & Wang, 2005). Therefore the prediction is that perceived ease of use has a direct positive effect on behavioral intention to use.

(36)

[30] 4. Research methodology

4.1 Design

To verify this thesis is based on a comprehensive list of measures, previous research on the field of the acceptance of new technology has been reviewed. A model is introduced based on an adaptation of TAM. The purpose of this thesis is to develop a framework containing the determining factors that influence the merchant adoption of cryptocurrencies as a payment solution. Subsequently the research design of this thesis can be categorized as a multiple methods research design, specifically mixed methods research. This form of research design entails both quantitative and qualitative research combined into one research design (Saunders, Lewis, & Thornhill, 2012). Although there is no in-depth or an extensive volume of interviewing done for this research, the importance of triangulation is not to be overlooked. In multiple fields triangulation has and will allow for more confident interpretations, not just for testing but developing hypotheses as well. Additionally triangulation gives way for less unpredicted findings (Jick, 1979). This thesis’ mixed methods research consisted of three stages.

Firstly, a short remote individual interview was held with a founder of SmartCash. Structured questions7 have been electronically sent towards

Alexander, founder of SmartCash and Hive Coordinator. On the purpose of being able to explain SmartCash and its features in a deeper manner than the readily available information. Since this thesis focusses on SmartCash as an exemplary cryptocurrency, already identified issues by their core team can be described. Additionally the results of this research can be used as

7 Appendix 1: Interview

(37)

[31]

recommendations towards the future process of merchant adoption and integration of SmartCash.

Subsequently a questionnaire8 has been designed on the basis of the

conceptual model. The questionnaire consisted of two parts. The first part recorded the merchant’s gross profit and familiarity with cryptocurrencies. The second part consisted of 15 items to measure the five latent constructs as defined in Table 1.

Table 1

Definition of variables

Construct Definition Reference

Perceived risk Merchant’s subjective expectation (Schierz et al., 2010) of risks concerning cryptocurren- (Plouffe et al., 2001a) cies as a payment solution. (Wu, & Wang, 2005) Perceived benefit Merchant’s subjective expectation (Mallat, & Tuunainen, 2008)

of the benefits concerning crypto- (Wu, & Wang, 2005) currencies as a payment solution.

Perceived Merchant’s subjective expectation (Plouffe et al., 2001b) compatibility of cryptocurrencies being consis- (Schierz et al., 2010)

tent with the merchant’s existing experiences with current payment solutions and business processes.

Perceived Merchant’s subjective expectation (Venkatesh, & Davis, 2000) usefulness on the adoption of cryptocurrencies (Kim et al., 2010)

would enhance their performance.

Perceived Merchant’s subjective expectation (Bauer et al., 2015) ease of use on the ease of using cryptocurren- (Kim et al., 2010)

cies. (Dahlberg et al., 2008)

(Wu, & Wang, 2005) Behavioral intention Merchant’s intention to adopt and (Plouffe et al., 2001a) to use cryptocurrencies as a payment (Wu, & Wang, 2005).

solution. (Venkatesh, & Davis, 2000)

Table 2 shows these latent variables (excluding the first 2 variables from the first part of the questionnaire: GROSS and FAMIL). These items and measurements correspond with the questions on the survey questionnaire sent

8 Appendix 2: Questionnaire

(38)

[32]

out to merchants. Items are measured on either a five point Likert-scale, ranging from 1 (strongly agree) to 5 (strongly disagree) or a two pointed scale where ‘1’ stands for ‘yes’ and ‘2’ stands for ‘no’. These so called items are latent, hidden from the surface, perceptions and therefore subjective.

Table 2

Measurements Measures

Item Measurement

Gross Gross profit

FAMILb I am familiar with cryptocurrencies

P_RISK1a I believe using cryptocurrencies in monetary transactions poses a risk for me

as a merchant.

P_RISK2a I believe using cryptocurrencies puts my privacy as a merchant at risk.

P_RISK3a I believe using cryptocurrencies could lead to possible conflict between me

as a merchant and my customers.

P_BENEF1a I believe using cryptocurrencies as a payment solution could cut my

transaction costs.

P_BENEF2a I believe using cryptocurrencies could enhance my overall financial

performance.

P_COMP1a I believe using cryptocurrencies as a payment solution is compatible with

most aspects of my online transactions.

P_COMP2a I believe using cryptocurrencies as a payment solution fits with the goods

that I am selling.

P_USEFUL1a I believe using cryptocurrencies would improve my performance.

P_USEFUL2a I believe using cryptocurrencies would increase my productivity.

P_USEFUL3a I believe using cryptocurrencies would increase my effectiveness in payment

solutions.

P_USEFUL4a I believe using cryptocurrencies would be very useful for me as a merchant.

P_EOU1a I believe learning to use cryptocurrencies is easy.

P_EOU2a I believe becoming skilled in the use of cryptocurrencies is easy.

P_EOU3a I believe using cryptocurrencies is easy.

BIb If I had access to cryptocurrencies as a payment solution, I intend to use it. a Measured on a five pointed Likert scale (1= fully agree – 5= fully disagree)

b Measured on a two pointed scale (1=yes, 2= no)

Lastly, after determining and producing the questionnaire, the following action consisted of distributing these questionnaires among a multitude of retailers with the aim of answering the research question and hypotheses in a statistical

(39)

[33]

manner. The target market was decided for logistical reasons to be the Dutch retail market. In order to be able to run this research, a statistically significant sample size of respondents (retailers) needed to be determined. In an article “Minder etalages, meer beeldschermwinkels” (2017) it was suggested that in 2017 there were 95,630 physical stores and around 32,000 web shops. This research focusses on merchant adoption with respect to SmartCash’s payment solutions for merchants. Therefore only physical stores were questioned. The number of questionnaires needed were calculated via the sample size formula as shown in Fig. 9 (Hinkle, & Wiersma, 2009).

Fig. 9. Sample size formula

The proportion of the population being a merchant is unknown so 0.5 has been chosen as the respective factor. With respect to a desired confidence level of 95% (implying a Z-score of 1.96), with e as a margin of error of 5% and N as the population size of 95630, this leads to a number of 383 respondents needed.

4.2 Sampling

Sampling technique used for this thesis could be described as binary probability sampling, meaning a mixture of simple random sampling and convenience sampling. As the population size can be described as being substantial. Therefore a considerable sample size is needed. The sampling approach should have a likely outcome of resulting in a large enough and representative

(40)

[34]

sample. Maximizing response rate has been tried to be achieved via two different approaches towards data collection.

Firstly, the top 150 biggest cities in The Netherlands9 (in amount of

citizens) have been identified. Ten of these 150 cities were randomly chosen from the list (1= Breda, 2= Groningen, 3= Dordrecht, 4= Tilburg, 5= Den Haag, 6= Amsterdam, 7= Maastricht, 8= Terneuzen, 9= Waddinxveen, 10= Woerden). E-mail addresses from merchants were assumed to be easily available and therefore the assumption was made that on average 50 contact addresses would be pulled from VVV-sites or other shop databases per city. Eventually Waddinxveen and Woerden fell off due to no such available database or other readily available information website. Eventually 440 e-mail addresses were collected, however 27 of those were removed due to identified duplicates and/or errors received during the delivery towards multiple email addresses. Eventually, 413 merchants10 received an email11 with an invitation to the

questionnaire via sharing an anonymous link created via Qualtrics.

Subsequently, due to the necessity of convenience, two cities were picked from which physical stores would be visited to invite the merchants in filling out the questionnaire on paper. They were instructed on their right to privacy and they were given the option to use prepaid envelopes to send their questionnaires back. The chosen cities were Amsterdam and Bergen op Zoom, respectively 30 and 20 merchants were invited to fill in the questionnaires. Eventually all of the data was analyzed on their respective response rate, as seen in Table 3.

9 Appendix 5: Data taken from: http://statline.cbs.nl/StatWeb/publication/?DM=SLNL&PA

=80706ned&D1=50&D2=a&HDR=T&STB=G1&VW=T; Retrieved 16 May, 2018.

10 Appendix 4 11 Appendix 3

(41)

[35] Table 3

Sampling, # of respondents and response rate percentage

Sampling Total Effective Total Returns Response Questionnaires Percentage Outgoing Source: Merchants 440 429 7 1.63% electronically invited via Qualtrics Source: Merchants 30 30 10 33.33% visited in Amsterdam Source: Merchants 20 20 4 20.00%

visited in Bergen op Zoom

Total 490 479 21 4.38%

The overall response rate has been truly unsatisfactory, which is mainly due to the extremely low net response rate of 1.63% from the electronically invited merchants. This is presumably caused by; (1) merchants have responded electronically stating they do not have the time or do not wish to cooperate in this research and (2) merchants have responded with having no idea what cryptocurrencies are about and therefore discarding the invitation. The low response rate might lead to problems concerning internal validity and the empirical analysis. Unlike general research such as research concerning consumers where there is an available presence of incentivizing them to cooperate, there is little to no incentive attractive enough for merchants to pull them towards cooperation.

(42)

[36] 5. Data analysis and results

Twenty-one returned questionnaires were received in total. Leaving this research with only 21 sets of data to be used for the statistical analysis.

That aside, data12 has shown that 52.4%13 of all included respondents

were familiar with cryptocurrencies, which is a substantial and unexpected amount. This could be due to the fact that respondents chose to cooperate with this research on the basis of their already existing interest towards cryptocurrencies. This is speculating and would require further investigation. Moreover data indicated that 57.1%14 responded with a behavioral intention to

use cryptocurrencies such as Bitcoin as a payment solution.

Table 4

Merchant attributes on Gross Profit, Familiarity and Behavioral int. to use.

Frequency Percentage Cumulative (%) percentage (%) n=21 Gross profit (€) 0 – 50.000 4 19.0 19.0 50.000 – 100.000 8 38.1 57.1 100.000 – 250.000 5 23.8 81.0 250.000 – 500.000 3 14.3 95.2 > 500.000 1 4.8 100.0 Familiar with cryptocurrencies

Yes 11 52.4 52.4 No 10 47.6 100.0 Behavioral intention to use

Yes 12 57.1 57.1 No 9 42.9 100.0 12 Table 4 13 Appendix 6 14 Appendix 6

(43)

[37] 5.1 Model measurement

Data obtained was tested for reliability and validity. The model included 14 items used to describe five latent constructs being perceived risk, perceived benefit, perceived compatibility, perceived usefulness and perceived ease of use. All items with respect to their construct were checked to verify they are stated in the same direction, which was the case.

Subsequently the internal consistency, or scale reliability, for each construct was estimated via a reliability analysis, including inter-item correlations15. The reliability analysis presents us with a Cronbach alpha, a

measure of internal consistency, or in other words an estimate of how closely related the set of items are with one construct. Three of five constructs have evident issues, one of which is severe. A desired cut-off point of ≥ 0.7 on Cronbach alpha has not been met for three constructs. The most severe issue concerns Perceived Compatibility with a negative Cronbach alpha, as Table 5 shows. The reason behind this negative alpha is evident when looking at the data16, suggesting a negative inter-item correlation (-0.074) in this construct.

Table 5

Constructs and internal consistencies (Cronbach alpha)

Construct No. of items Internal Internal Consistency Consistency* Perceived Risk 3 0.55 0.80**

Perceived Benefit 2 0.64 0.64 Perceived Compatibility 2 -0.16 -0.16 Perceived Usefulness 4 0.63 0.63 Perceived Ease of use 3 0.68 0.79*** * Internal consistency improved by deleting items

** P_RISK1 deleted *** P_EOU1 deleted

Perceived compatibility needs to be dropped from the model due to the negative Cronbach alpha. Reliability analysis for Perceived Benefit and

15 Appendix 6

(44)

[38]

Perceived Usefulness provides us with an alpha between 0.60 and 0.70, although this is insufficient it has to be accepted by default. Perceived Risk and Perceived EOU can be improved by deleting the first item for both constructs, respectively providing those constructs with an improved alpha of 0.8017 and

0.7918. Deleting items should only be done when Cronbach alpha significantly

(+0.05) improves, which is the case in both constructs.

Table 6

Factor analysis – Eigenvalue and Variance explained

Construct Item Eigenvalue Variance Cumulative explained (%) (%) Perceived Risk P_RISK1 1.677 55.92 55.92

P_RISK2 1.116 37.18 93.10 P_RISK3 0.207 6.90 100.00 Perceived Benefit P_BENEF1 1.476 73.78 73.78

P_BENEF1 0.524 26.22 100.00 Perceived P_COMP1 1.074 53.72 53.72 Compatibility P_COMP2 0.926 46.28 100.00 Perceived Usefulness P_USEFUL1 1.912 47.79 47.79

P_USEFUL2 1.058 26.44 74.23 P_USEFUL3 0.688 17.19 91.42 P_USEFUL4 0.343 8.60 100.00 Perceived EOU P_EOU1 1.858 61.92 61.92

P_EOU2 0.848 28.28 90.20 P_EOU3 0.294 9.80 100.00

Additionally to computing the alpha coefficient of reliability, the dimensionality of the scales has been investigated via a Factor Analysis19. There is no single

accepted agreement on the necessary sample sizes, research shows that factor analysis performed within a study containing a sample size 20 times the number of factors, will result in stable solutions (Arrindell, & Van der Ende, 1985). It is evident that this study does not fit this criteria. Although the default

17 Appendix 7 (1) Reliability analysis for Perceived_Risk 18 Appendix 7 (5) Reliability analysis for Perceived_EOU 19 Appendix 8

(45)

[39]

solution in most statistical software is to retain all factors with Eigenvalues greater than 1.0 (Costello, & Osborne, 2005), this will be discarded due to the previous statements. Therefore Factor Analysis on the model of this thesis should be viewed as inadequate.

However, this analysis is done to show evidence of unidimensionality. As can be seen in Table 6 for all constructs, except Perceived Compatibility, the eigenvalue for the first factor is significantly higher than the other factors. Combined with a higher explained variance for the first factor of all four constructs, it can be concluded that the only construct with non-unidimensional scale items is Perceived Compatibility. Next to the negative Cronbach alpha, this is another reason for dropping P_COMP out of the model.

Table 7

Descriptive statistics – Mean, Standard deviation and Correlation

M SD 1 2 3 4 5 6 1. Gross Profita 2.48 1.12 2. Familiarityb 1.48 0.51 0.195 3. P. Riskc 2.60 1.43 -0.388 -0.441* (0.80) 4. P. Ben.c 2.81 0.93 0.019 -0.063 -0.165 (0.64) 5. P. Usef.c 3.14 0.95 0.237 0.135 -0.423 0.562** (0.63) 6. P. EOUc 2.98 1.24 0.421 0.255 -0.267 0.289 0.278 (0.79) 7. Behav. I.b 1.43 0.51 0.15 0.138 -0.301 0.288 0.358 0.415

* Correlation is significant at the 0.05 level (2-tailed) ** Correlation is significant at the 0.01 level (2-tailed)

a 1 = [0 - €50k], 2 = [€50k - €100k], 3 = [€100k - €250k], 4 = [€250k - €500k], 5 = [> €500k]. b 1 = yes, 2 = no.

c 1 = strongly agree, 2 = agree, 3 = undecided, 4 = disagree , 5 = strongly disagree, Reliabilities (α) are displayed on the diagonal

5.2 Empirical analysis

Before answering the hypotheses of this research, the four left over constructs (perceived risk, perceived benefit, perceived usefulness and perceived ease of use) with their respective items have been transformed into a summated scale,

Referenties

GERELATEERDE DOCUMENTEN

The definition of employee engagement is described as follows: “A collection of psychological forces that determine the direction of a person’s behavior in an organization

In this research, UTAUT and its extension are used in order to allow us to describe a positive and direct influence of several factors on the intention to use a

The criteria used are based upon the case study of An- gelou taking the rules of dependencies, project cycles, and intuitively designing the structure in mind.. Criteria

This study researched which requirements non-early adopters have for a dashboard in an electronic vehicle to ease and/or preclude range anxiety.. For this a

In addition to measuring the head circumference, clues in the neurological examination that indicate a secondary cause would include depressed or altered level

In the first study, we asked experts and laymen in design to evaluate the familiarity, perceived ease of use, and attractiveness of a set of more and less conventional graph

The model shows an adjusted R-square of 0.749, which means that 74,9% of the variance in the dependent variable Intention to Use, can be explained using the

Indicates that the post office has been closed.. ; Dul aan dat die padvervoerdiens