Factors influencing the use of e payment

Findings – The results showed significant results between social influence, TAM factors, and security with intention to use e-payment for online shopping in Malaysia.. Thus, this study w

Purpose – The main objective of this paper is to investigate determinants that influence the

intention to use e-payment system for online shopping among young adults in Malaysia

Design/methodology/approach – A survey was distributed to 200 respondents in Klang Valley

area with 165 valid responses used as sample with the respond rate of 83% Multiple linearregression and independent t-test were conducted to analyze the data

Findings – The results showed significant results between social influence, TAM factors, and

security with intention to use e-payment for online shopping in Malaysia On the contrary, trust,quality of internet connection, and gender were found not to be significant in this study

Research limitations/implications – Small sample size that only covered specific purpose, region

and demographic – Other usable factors were not assumed The findings could be the evidence foronline transaction facility providers to continually enhance their e-payment services and it mustalso meet customer expectation to have a positive growth rate Future research can use this study as

an illustration to conduct similar studies in other region/countries or further develop actual use

Originality/value – This study narrows the use of e-payment into more specific context, which is

for online shopping and has not been studied in Malaysia yet – Combine both TAM and TRAfactors to study the intention to use e-payment, and uses additional factors of gender and quality ofinternet connection that has not previously studied

CHAPTER 1: INTRODUCTION

1.1 Background of the study

We are living in the time when everything is available through the internet With suchexposure, many things can be done promptly by anyone, anywhere and anytime Education started

to utilize online systems gradually with e-books and e-learning (Sun et al., 2008; Chen, Chen, andWang, 2009) Social networks are the biggest trend for young generation as a form ofcommunication (Ellison, Steinfield, and Lampe, 2007; Pempek, Yermolayeva, and Calvert, 2009).Businesses also start to move towards online with the introduction of e-commerce in early 1990s(Monsuwe et al., 2004; Horrigan, 2008) and even banking transactions can be done instantly withinternet banking (Tan and Teo, 2000; Lee, 2009)

According to BNM payment statistics, the growth of e-payments in Malaysia has increasedimmensely due to the rapid development of e-commerce They are both interconnected as e-payment systems are capable of making online money transfer available once customers made theirpurchases via Internet Specifically, a recent online survey by Wong (2014) revealed that 91% ofthe Internet users in Malaysia shop online regularly Over half (54%) of them confessed to shop atleast once a month online, 26% shops once a week, and 7% shops online everyday The NielsenCompany also revealed that Malaysians spent RM1.8 billion shopping via Internet in 2010 (Ho,2011) This is the reason why the development of e-payment is crucial in order to accommodate thedemand in virtual shopping behaviour

Despite the growth, many consumers are still reluctant to adopt e-payment systems,especially in emerging countries because they think it is more convenient to use physical money(cash) More than half (60%) of citizens make payment by cash (The Nielsen Company, 2014a) orcheques as a means of payment, even to the young adults who practically like everything to befaster and prefer using high technology system Moreover, there is still a large group of customerswho refuse to adopt such services due to their risk perception (uncertainty and security concerns) indoing their transactions online Wong (2014) showed that the remaining citizens of Malaysia do notshop online due to the lack of trust on the sellers and security issues 52% of the shoppers confessthey do not trust giving their credit card information to the virtual world (The Nielsen Company,2014b) Additionally, another study has indicated customers’ apprehension or lack of interest inusing online payment methods (Johar and Awalludin, 2011)

2

1.2 Research objectives and questions

From the derived rationale and background above, this study is trying to investigate both ofthe positive and the negative side of the intention to use e-payment system Thus, this study willinclude the mainly used factors of e-payment usage such as Technology Acceptance Model, withthe negative/resistance factors considered as perceived risk Furthermore, this research would like

to study at a greater length by finding the relationship between social influence, quality of internetconnection and demographic (gender) in the intention to use e-payment system to conductpurchases online

Based on the past theoretical studies conducted with the aim of study to discover theintention to use e-payment, the research questions are proposed as follows:

1 Does Technology Acceptance Model (perceived usefulness and perceived ease of use)integrated with the intention to use electronic payments?

2 Do perceived risks (security and trust) influence the intention to use electronic payments?

3 Does quality of internet connection could potentially influence the intention to useelectronic payments?

4 Does subjective norm factor (i.e social influence) could potentially influence the intention

to use electronic payments?

5 Is there any difference among gender (male and female) in the intention to use electronicpayments?

1.3 Significance of the study

The significance of this study is to create awareness among all Malaysian consumers toconduct payment electronically as it would create benefits to them; i.e: convenience, efficient, andflexible (He and Mykytyn, 2007) Moreover, it is to promote more e-payment system usage amongcustomers in Malaysia as it is believed that efficiency of financial system could be achieved if e-payment systems are developed Cost of transaction would be reduced, enhanced liquidity, andbetter allocation of financial resources to everyone participated in the financial sector (BNM,2015b)

1.4 Research Design

A quantitative method in a form of survey with questionnaires (offline) was conducted forthe research It was based on purposive sampling technique in intention to assess specifically on e-payment usage A total of 200 questionnaires were distributed in Klang Valley area with 165 validresponses The main targeted respondents were young adults with age range from 18 to 26 andabove

For the reliability of this research, Cronbach’s Alpha test was used to score the validity forall the constructs in the questionnaire, with an addition of face validity To test the hypothesis andsee the significance between the constructs, Multiple Linear Regression was carried out on thefactors extracted in the factor analysis For the demographic variable, an Independent t-test wasconducted

1.5 Findings

After the regression analysis was conducted to test 6 hypotheses, four of them showedsignificant results with p-value below 0.05 The significant factors are Perceived Ease of Use,Perceived Usefulness, Social Influence, and Security The remaining factors (Quality of InternetConnection and Trust) showed insignificant results In addition, independent t-test for gender (H7)showed an insignificant result with p-value above 0.05 Therefore, out of 7 hypotheses proposed,four of them were supported whereas three were rejected

1.6 Research limitations and future research

Unequal ratio of respondents that might affect overall significance level, the small samplesize effect in relation to specific respondents and geographical area does not cover populationgeneralizability and only creates conceptual generalizability Moreover, this research only selectedparticular factors that are deemed to be important in the researcher’s own perspective Therefore,some suggestions that could be given for future studies are consideration of larger sample size thatwill cover more range of population in broader area and more important factors that have been done

by past researchers should also be taken into consideration in order to enhance the results andgeneralisation

4

1.7 Conclusion

After the overall of this study is introduced, this study will be organized as follows:Chapter 2 will highlight literature review, along with the theoretical framework and hypotheses inaccordance with research questions The research methods are developed in Chapter 3 Chapter 4will elaborate on the results tested And lastly, Chapter 5 will provide implications based on thefindings, along with conclusion of this study with some recommendations that include the researchlimitations as well as future studies

CHAPTER 2: LITERATURE REVIEW

2.1 Electronic Payment

2.1.1 Background of the study

Presently, electronic payment (e-payment) has become a popular means for paying onlinepurchases due to the growth of e-commerce This market is expected to grow by 18.1% from 2010per year until 2014 (with an estimated total of 34.8 billion and a value of $1,792.4 billion).Subsequently for e-payments, as shown by the trends in e-commerce, are expected to grow by18.1% yearly to a total of 34.8 billion transactions (Capgemini, 2013) The expansion of bothsectors is one of the vital factors to economic growth For instance, the immense escalation of creditand debit cards use as means of e-payment has added $983 billion to 56 countries’ GDP between

2008 and 2012 (Visa, 2012) This has contributed to 0.8% increase in GDP in emerging marketsand 0.3% increase in developed markets; creating an equivalent of 1.9M more jobs (Moody’sAnalytics, 2013)

In Malaysia, payments by cards and e-money remained as the mostly used method (71.6%),followed by payment through mobile and internet banking (13.2%), and with others being directdebit and e-cheques (BNM, 2012) It has continued to expand, with almost 90% of the 1.8 billioncash non-payments, amounting to RM 17.1 trillion are now made electronically compared to 52.6%

a decade ago (BNM, 2012) Particularly, 80% of retail payment transactions were conductedelectronically in 2012 and have it has lifted the transactions amounting RM49.5 trillion (Teoh et al.,2013)

Additionally, Bank Negara Malaysia (BNM) with the association is currently conducting aroadshow with objectives of creating awareness of e-payment systems and educating the public one-payment services The roadshow will be held by BNM for major financial institutions andpayment system operators to showcase their products on card acceptance and online banking Theycould also offer various incentives and promotions for businesses to accept payment card and useonline banking As Malaysia is one of the biggest and most successful emerging countries, the use

of e-payment is essential to achieve higher economic growth and keep up with the competitive edgeamong the developed countries (BNM, 2015b)

6

2.1.2 Definition of e-payment

The most basic definition of e-payment is an action of paying something (payer) orsomeone or of being paid (payee), electronically Likewise, electronic Payment SystemsObservatory (ePSO) defines e-payment as electronic transfer of payment from the payer to thepayee through the use of an e-payment instrument E-payment could also be described astransferred monetary value between two entities as compensation for the receipt of goods andservices through electronic channel (Tan, 2004)

Many of e-payment context are associated to banking/financial sector (e-banking) Itmainly includes paying electronic bills and tax, electronic fund transfer (EFT), online credit/debitcard payments, and reload/top-up using stored-value money (Özkan, Bindusara and Hackney,2010) Bank accounts are accessible and manageable remotely by customers through web-basedinterfaces, that is e-payment services (Weir, Anderson and Jack, 2006; Lim, 2008)

Aside from e-banking, many have associated e-payment with e-commerce AmericanUniversity (2004) described e-payment as a subset of e-commerce to carry out an electronicpayment from the goods and services to be bought or sold via Internet Moreover, e-payment isdefined as a financial exchange between buyers and sellers with digital financial instrument (such

as encrypted credit card numbers, electronic cheques or digital cash) It is done through onlineenvironment, assisted by banks, intermediaries, or legal tenders (Kalakota and Whinston, 1997;Fok, 2013)

This study is specified on the use of e-payment system for online shopping Therefore, thisstudy adopts the definition of e-payment as a manner of delivery payments from customers tovendors after the customers have decided to pay for the particular products or services Thepurchases made will be via Web browser and paid using credit or debit card (Bitpipe, 2006, cited

by Teoh et al., 2013) It is believed that e-payment is the most effective, efficient, and problem-freeway to pay for those purchases (Abrazhevich, 2004)

2.1.3 E-payment as a subset of e-commerce

In a broad sense, many people associated e-commerce with “shopping via Internet”

based on its entities are as follow: B2B (business to business), B2C (business to customer), C2C(customer to customer), and B2G (business to government) As the focus of this study is a paymentmade through customers’ web shopping experience, B2C and C2C is the suitable approach.According to Schneider (2012), C2C can be categorized into B2C because even though C2C onlyincludes one party selling to another party (purchases made to each other such as online auction), it

is still acting as a business Effective in 2014, the most frequently items to purchase (or intended topurchase) online in order are airline tickets and reservation, clothing and accessories, tours andhotels reservation, and hardcopy books (The Nielsen Company, 2014b)

Since e-payment refers to financial exchange, it is regarded as one of the main role ofonline shopping payment mechanism E-payment systems are pivotal for e-commerce futurebecause the e-commerce growth will be depending on the development of e-payment systems(Abrazhevich, 2004) In addition, banks have tried to keep up with the competitive development ofe-commerce activities by introducing several products that could enhance such actions They alsotook advantages of the growth by expanding their services exclusively for e-commerce purposes(Wenninger, 2000)

2.1.4 Types of e-payment (E-payment instruments)

In general, e-payment can be classified into five categories (Lawrence et al., 2002;Abrazhevich, 2004; Dai and Grundy, 2007; Schneider, 2012) Yet, for e-commerce purposes, themost widely used instruments are electronic cash, debit cards, and credit cards (Tsiakis andSthephanides, 2005)

Electronic cash (e-cash) is online cash based system that is used as a payment methodthrough electronic currency exchange (Kim et al., 2010) The amount of money to be used isidentified with the unique identification number How it works is customers must purchaseelectronic digital cash from the issuing company (Abrazhevich, 2004); and those digital cash will

be transferred through computers or other telecommunications channels The development of this cash has now been considered as one of the alternatives to the credit card spending on onlinepurchases of goods or services This paying method does not require a high cost Thus, it isconvenient for micro-payment (Lawrence et al., 2002; Kim et al., 2010)

e-Credit card uses server authentication along with bank verification from customers to verifythe sufficient funds needed to purchases goods or services It is an account based system that worksboth online and offline The amount will be charged to the customers’ account and it is payable in a

8

form of credit card bills (Kim et al., 2010) Due to this system, credit card is usually used for heavypurchases; and it might not be suitable for small-value transactions (less than a dollar) (Kalakotaand Whinston, 1997) Additionally, the system makes it available for customers to track theirtransaction record

The substitute of credit card and the most used method for e-payment is debit card Banksusually issue debit card where credit cards are not accepted (Abrazhevich, 2004) The differencebetween this and credit card is customers must have a positive balance in their bank account inorder to proceed with their purchases The money from customers’ account will be automaticallydeducted once they finish their debit transactions (Kim et al., 2010)

2.1.5 Antecedents of e-payment

The use of e-payment system has become popular throughout the year; proven by quite anumber of previous research for e-payment for e-commerce purposes, including Abrazhevich(2004); Tsiakis and Sthephanides (2004); Kim et al (2010); Aw, Ab Hamid and Eaw (2011);Özkan, Bindusara, and Hackney (2010); Lin and Nguyen (2011); and Teoh et al (2013) All ofthese researchers found that most consumers have adopted e-payment to do their onlinetransactions These researches were done based on the technological factors and users’ acceptance.Most of these studies emphasize on trust and security issues, but in a different manner It is said thatthese 2 factors are the key of why consumers’ are still hesitant and have a negative feeling aboutusing e-payment

Kim et al (2010) focused on using technical protections, transaction procedures, securitystatements, and e-payment objective dimension as their variables to study the significance betweentrust and security to e-payment system Tsiakis and Sthephanides (2004) suggested that a high level

of mechanism should be used in order to increase the confidence in security and trust Thismechanism involves cryptography that has both symmetric (secret) form and asymmetric (public)form Aw, Ab Hamid and Eaw (2011) emphasized on risk perception in adoption of e-payment fore-commerce purposes The models used were based on five different risk perceptions, which arephysical risk, performance risk, psychological risk, time-loss risk, and financial risk Others studieshave adopted the widely used Technology Acceptance Model (TAM) (Teoh et al., 2013; Lin andNguyen, 2011; Özkan, Bindusara, and Hackney, 2010) Additionally, further variables have beentested such as anonymity, reliability, traceability, and convertibility (Abrazhevich, 2004)

2.2 Frameworks used in investigating the intention to use electronic payments

(research model and gap)

This study will integrate factors from several existing models and theories, to indicate theintention to use e-payment This includes factors from both Technology Acceptance Model (TAM)and Theory of Reasoned Action (TRA) It is believed that these two models are the main factors toexplain the reasons behind the behavioural intention, which is e-payment in this context Moreover,the significance of these models has been tested by Özkan, Bindusara, and Hackney (2010) andTeoh et al (2013) and it shows a positive relationship between TAM and TRA factors with the use/adoption of e-payment Factors used derived from TAM will be Perceived Usefulness (PU) andPerceived Ease of Use (PEOU), adapted from the origin of Davis, Bagozzi, and Warshaw (1989)

This study also intends to investigate the negative factors on the use of e-payment such asperceived risks (security and trust) (Venkattesh and Bala, 2008; Aw, Ab Hamid and Eaw, 2011;Kim et al., 2010) Perceived risk in this study is defined as the perceived risk in transaction context

as it is the main objective behind the use of e-payment

Other additional factors would be used are demographic (gender), subjective norm which isderived from TRA, and quality of internet connection Not many studies have incorporated thesefactors As Pikkarainen et al (2004) tried to include quality of internet connection in their study, butthe factor did not make it in their factor analysis, hence rejected Moreover, several studies haveused role of gender as the mediator of factors influencing adoption of e-payment (Black, 2005;Olalekan, 2011; Liébana-Cabanillas, Sánchez-Fernández and Muñoz-Leiva, 2014)

10

2.3 Technology Acceptance Model

Technology Acceptance Model (TAM) is a theoretical model derived from the Theory of

Reasoned Action by Fishbein and Ajzen (1975) that explains how users come to accept and use

technology (Davis, 1989) Due to its prominent model on technology, TAM has been extensively

incorporated as a methodology to measure attitude towards technology adoption from users in

financial domain The origin TAM constructs consisted of Perceived Ease of Use (PEOU) and

Perceived Usefulness (PU) Both of these constructs will ultimately determine the Attitude towards

Using (ATU), Behavioural Intention to Use (BI), and Actual System Use (AU) (Davis, Bagozzi and

Warshaw, 1989)

Figure 2.2: Technology Acceptance Model (Davis, Bagozzi and Warshaw, 1989)

Several researchers have included and further developed the two TAM constructs into their

studies PEOU and PU were conceptually extended to have an in-depth understanding of electronic

communication (Karahanna and Straub, 1999; Gefen and Straub, 1997) It has also combined with

consumer behaviour, and psychology constructs to study online commerce (Koufaris, 2002)

Furthermore, TAM constructs are studied to ascertain customers’ acceptance of online banking and

their satisfaction (Adamson and Shine, 2003; Chau and Lai, 2003; Pikkarainen et al., 2004)

From the model above, Venkatesh and Davis (2000) proposed TAM2 that incorporates

extended theoretical constructs to PU, which are social influence processes and cognitive

instrumental processes These additional constructs were further tested and significantly influenced

user acceptance TAM2 was soon developed again by Venkatesh and Bala (2008, pp 278-282),

which results in the start of TAM3 It is the combination of TAM2 and the model of the

determinants of PEOU; namely computer self-efficacy, perception of external control, computer

anxiety and playfulness, perceived enjoyment, objective usability It also proposed new

relationships (moderated by experience) between PEOU to PU, computer anxiety to PEOU, and

PerceivedUsefulness

ActualSystemUse

BehaviouralIntention toUse

AttitudetowardsUse

External

Variables

PerceivedEase of Use

Since online shopping utilizes innovative technology systems, and e-shopping activities(e.g., browsing, transaction, etc.) is a type of consumer usage system, TAM provides a usefulfoundation to investigate the intention to shop online that leads to the use of e-payment (Ha andStoel, 2009) Many researchers argued that PU and PEOU are the main reason behind customersshifting to e-payment Time and cost effective are the perceived important advantages of e-paymentsystem (Leong, Ewing and Pitt, 2003; Özkan, Bindusara and Hackney, 2010)

2.3.1 Theory of Reasoned Actions (TRA)

Theory of Reasoned Actions (TRA) is defined as behaviour that is determined by aperson’s intention to perform and their surroundings influences (i.e subjective norm) (Davis,1989) In other words, the rationale of persons doing activities is solely in their own control withoutany influenced by unconscious incentives

In terms of applicability, TAM explained about the usage acceptance of new technologyand mainly on IS as proven by many researchers However, TRA explains better on any kinds ofintention to use based on the person’s willingness to behave It is supported by Legris, Ingham, andCollerette (2003) who testified that the original model of TAM only interpreted 40% of consumerbehaviour intention, with nearly half of the other important components were not explained,whereas TAM2 that includes social influences; which is the epitome of TRA shows more than 60%explanation as evidence (Venkatesh and Davis, 2000)

While TAM is the most used factors, it does not entirely explain about the behaviouralintention of e-payment usage (Özkan, Bindusara, and Hackney, 2010) This statement is alsosupported by Abrazhevich (2004) who justified that TRA suits more to explain how customers’behaviour influence their acceptance of payment technology for online transactions because TRAapplies to a wider range of situations Moreover, unlike TAM, TRA takes subjective norms (i.e.social influences) into account and understands various factors surrounding the usage of e-payment

As a general basic behaviour model, TRA proposes that a consumer’s Actual Behaviour(AB) (whether to perform or not to perform) is determined by his or her Behavioural Intention (BI),both of which are influenced by personal Attitude towards Behaviour (AT) and Subjective Norm(SN) (Fishbein and Ajzen, 1975) The proposed model is as follow:

12

Figure 2.3: Theory of Reasoned Action (Fishbein and Ajzen, 1975)

2.3.2 Perceived Usefulness (PU)

Perceived Usefulness is presumed could enhance a person’s job performance by taking andusing the advantages of particular system (Davis, 1989; Lin and Nguyen, 2011) This factor hasbeen deemed to be one of the important factors of usage intention along with its acceptance Therelevant application could be seen from online/e-banking perspective since e-banking is highlydependent on technological systems Users will believe that the system is useful if it providesseveral benefits (e.g: less time and financial loss to increase efficiency and productivity) to them,hence increasing their intention to adopt (Chong et al., 2010) Several features that are perceived asuseful in online banking adoption are transactions speed, user-friendliness, and accuracy (Liao andCheung, 2002) Likewise, PU is a critical factor in attracting online shoppers in terms of thebenefits compared to normal retailing (Ramayah and Ignatius, 2005)

Since e-payment is the subset of e-banking and also incorporated in online shopping, PUwould be a significant factor to study In the general context of e-payment, PU is valid if theavailability of information in the system could greatly help customers in every circumstances untilthey fully understand the usage and intend to integrate with e-payment for their daily activities (Linand Nguyen, 2011) This usefulness can then obtain the potential users’ attention as an innovation

of alternatives in performing same action, but in a more highly effective and efficient way thatcould lead to better productive and satisfaction

This study will define PU as the degree in which the persons believe that using e-paymentsystem could enhance their job performances and hence become more efficient This will

Stimulus

Conditions

Beliefs andEvaluation

Normativebelieves andMotivation tocomply

Attitude

Subjective Norm

BehaviouralIntention

ActualBehaviour

use (Koufaris, 2002; Pikkairainen et al., 2004; Teoh et al., 2013).Therefore, by combining thesetheories, the proposed hypothesis for this factor for this study is:

H 1 : Perceived usefulness (PU) will exert a significant influence on intention to use e-payment

systems

2.3.3 Perceived Ease of Use (PEOU)

In earlier research on the original TAM factors, Davis (1989) suggested that “ease of useoperates through usefulness” But later on, it is found that although customers may believe thegiven application is useful, at the same time they might think that the system is difficult to use(Davis, 1989) Pikkarainen et al (2004) also argued that PEOU only cause slight impact ontechnology acceptance than PU because when more users learn about PEOU, the more vivid theimpact is In this study however, the link between PU and PEOU will be ignored instead, thedifferences are highlighted

PEOU is presumed must be effortless when a person is using a particular system (Davis,1989) Similar to PU, PEOU is another important factor for IS and technology acceptance hence thesame relevant applications could be used In e-banking, different features on banks’ website thatmeets customers’ needs in their usability has in impact on adoption of e-payment These featuresare described as user-friendly interface that has clear and visible steps/commands, suitable contentand graphical layouts, helpful navigations, and easy to understand error messages (Gerrard andCunningham, 2003; Lin and Nguyen, 2011)

Particularly in e-commerce, Vijayasarathy (2004) emphasizes on the customers’ free ofeffort in purchasing online Therefore, the structure of the website and its content has impacts onthe ease of use In a different respect, Gefen and Straub (2000) did a research about the impact ofPEOU to intention to use IT Their study supported the hypothesis of ease of use has a significantrelation to intentions to use the site for inquiries but not for purchases

From all of these previous studies, this study is trying to develop an establishment bytaking it a step further: Based on effortless and content on companies’ websites on how to conductthe e-payment must be clear and precise By having these, customers will assume this as ease of useand that will also increase their intention to use e-payment system This statement is supported byAbrazhevich (2001) who concluded that a successful design of e-payment systems from the users’

14

point of view is important to attract users’ acceptance toward e-payment Therefore, the proposedhypothesis is:

H 2 : Perceived ease of use (PEOU) will exert a significant influence on intention to use e-payment

systems

2.4 Perceived Risk

Perceived risk is defined as the uncertainty associated with a consumer’s action that couldlead to a certain outcome/consequences (Aw, Ab Hamid and Eaw, 2011) This study will solelyfocus on one risk, which is financial risk Financial risk is defined as unrefundable transactionsafter performing the payment, therefore cause financial loss (Aw, Ab Hamid and Eaw, 2011) Thisfinancial risk could possibly caused by two issues, which are security and trust

In online transactions, perceived risk and trust are closely related, as proposed byFeatherman and Pavlov (2003) that level of risk would decline if customers trust parties involved inthe transactions On the other hand, risk could be the mediation of trust and willingness to purchaseonline (Jarvenpaa, Tractinsky, and Vitale, 2000)

It is possible that customers believe that disclosing their credit card information is risky,and they have no control over it afterwards (Aw, Ab Hamid and Eaw, 2011) The risks involved aremostly on credit card fraud and hacking, in which categorized as security issues Therefore, theresistance to use e-payment services rises and customers are less motivated to adopt e-paymentsystem as they perceived it to be riskier than traditional payment method (Jarvenpaa, Tractinsky,and Vitale, 2000)

2.4.1 Security

Security is steps taken to verify the information source and guarantee the integrity andprivacy of the information (Tsiakis and Sthephanides, 2005) In e-payment context, security may bedefined as protection of customers’ transaction details from a possible internal and externalfraud/criminal usage/breach Customers would feel anxious that their personnel information might

be stolen if they pay online (Lim, Lee, and Kurnia, 2006; Özkan, Bindusara, and Hackney, 2010).Security in e-payment will considered as confidential as long as customers are satisfied with the

technical perspective, online payment facility providers should ensure the customers’ integrity,confidentiality, and authentication with profound security measurements (examples in e-banking:encryption, digital signature, etc.) (Flavian and Guinaliu, 2006).

Security is significantly a challenge to online banking implementation, which could hinderthe use of e-payment systems (Sathye, 1999) The hinder heavily lies on weak confidence intechnology that controls customers’ data Customers wish to have control over their own data thathas been collected by banks and they want to know the purpose of the data collection before bankscould process it As online banking is closely related to e-payment system, this could be the reason

of customers’ decision before starting to utilize e-payment (Abrazhevich, 2004)

In online shopping context, companies must ensure their payment systems are free from theinternal/external harm in the Internet environment This is due to the nature of Internet that has nocentralized control and barriers (Abrazhevich, 2004) Therefore, it could be an easy target forcriminals Online purchases are presumed to be more unsafe than conventional ways since there is

no human factor involved in the purchases and it is done virtually (Whiteley, 2000) and this willraise customers’ suspicions Hence, this brings up researchers to study the security concerns ofusers and the effect on the intention to use e-payment systems (Kurnia and Benjamin, 2007) Thisstudy will adopt security definition from Tsiakis and Sthephanides (2005) and look into proceduresinvolved in security measurement will affect e-payment usage The proposed hypothesis is:

H 3 : Security will exert a significant influence on intention to use e-payment systems.

2.4.2 Trust

From customers’ point of view, trust is defined as ‘‘a psychological state leading to thewillingness of customers to perform payment transaction over the internet and expect the paymentplatform fulfilling its obligations, irrespective of customer’s ability to monitor or control paymentplatform’s actions.’’ (Mayer, Davis, and Schoorman, 1995) This definition mainly representstrusting beliefs and attitudes in online payment Being one of the most important factors inconsumers’ intention to do transactions and carry out the payment electronically, many researchershave come out with their theories and findings in regards (Gefen, Karahana, and Straub, 2003;Black, 2005; MacInnes, 2005; Chong et al., 2010; Zhou, 2014)

Kniberg (2002) explained that customers’ perception of trust in e-payment channels could

be based on their individual psychological condition, which is their own mental and emotionalreasoning to use those channels Thus, different perception might occur whether positive or

16

negative This leads to trust and mistrust in e-payment channels Adoption is said to be valid ifcustomers react positively (i.e trust e-payment channels).

As e-commerce activities are conducted in an internet environment, it is difficult toguarantee the vendors intention and behaviour as they could possibly conduct undesirable,unethical, and opportunistic behaviour easily (e.g: unfair pricing, presenting inaccurate information,distributing personal data) (Gefen, 2000) Moreover, online environment does not provide a face-to-face interaction Physical interaction is essential to convince the customer and that leads to anincrease in trust (Gefen and Straub, 2003) As such, the importance of trust is crucial in e-paymentdue to the high degree of uncertainty and risk present in most online transactions (Zhou, 2014).This study will adopt the definition of trust as the extent to which online businesses can build trustsignificantly influences the willingness of concern to make e-payment purchases (MacInnes, 2005).The proposed hypothesis is:

H 4 : Trust will exert a significant influence on intention to use e-payment systems.

2.5 Quality of Internet Connection

As this internet is also something to do with information system and technology, this factor

is worth noticing because it is the fundamental aspect of electronic-based activities Specifically, inorder to carry out online shopping, or even electronic payment, internet connection is much neededfor consumers to complete their transactions

Hoffman and Novak (1996) suggested that download speed is another notable determinant

of users’ satisfaction Slow response time caused by sluggish connection is frequently the case ofcustomers’ resistance to conduct online transactions This is due to disruption/delay occurred in theservice delivery that makes customers unsure about whether or not the transaction is completed(Jun and Cai, 2001; Pikkarainen et al., 2004) Without proper and stable internet connection, it isextremely difficult for consumer to conduct online shopping, and therefore makes it not possible toconduct e-payment In addition, Sathye (1999) used Internet access as one of the factors affectingthe adoption of online banking in her research

Quality of Internet connection could also be the determinant of transactions speed With theavailability of Internet, customers now demand a fast transactions speed whereby they could reduce

delivery time In other words, time-saving was an essential consideration for users of conventionale-banking facilities like ATMs (Liao and Cheung, 2002).

This study is focusing on whether quality of internet connection affects customers’willingness to conduct online shopping and use electronic payment system Again, as e-payment is

a subset of online banking, the theory above would still be suitable Therefore, the proposedhypothesis is:

H 5 : Quality of internet connection will exert a significant influence on intention to use e-payment

systems.

2.6 Subjective Norm

In subjective norm, it is presumed that individuals will behave accordingly if the peoplewho are important to them think they should/should not perform the behaviour (Fishbein & Ajzen,1975; Venkatesh & Davis, 2000) It is a direct determinant of behavioural intention based on TRAmodel and is the equivalent of social and environmental influences This direct determinant is based

on the fact that a person’s decisions to perform behaviour may changes overtime as they areunaware of the consequences and outcomes In addition, their behavioural references could comeunder people/peers from social networks where it has become popular platform to consult foropinions and shares experiences (Arvidsson, 2014)

Referents from social networks could form a social image, in which aligned with JoséLiébana-Cabanillas, Sánchez-Fernández and Muñoz-Leiva (2014) who stated that social image is

an influential factor to drive social influence, especially to introduce and spread newinnovations/products

Since e-payment in this context is conducted in the Internet environment, the reputation andimpression of the systems could be communicated straight forward to other users/customers viaonline communities (i.e social networks), creating yet another social impact on the system Hence,social influences, e.g opinions and behaviour of other users, family and friends, and reputation ofcompanies and online payment facility providers involved should be taken into account forintention to use e-payment systems (Abrazhevich, 2004)

This study will adopt the theories above to analyze how social influence could affect aperson’s behavioural intention to use e-payment Therefore, the proposed hypothesis is:

18

H 6 : Social Influence will exert a significant influence on intention to use e-payment systems.

2.7 Demographic (Gender)

In a general situation of IS and technology, there are evidences from various studiessuggesting that difference between male and female in adopting technology is apparent Male aremore likely to adopt a new technology and are more technology-oriented than their femalecounterpart Women who adopted it use it at a lower degree compared to men (Hamza and Shah,2014)

Prior studies showed the impact of gender on attitudes towards online shopping Femaleperceived shopping as social activity compared to male (Slyke, Comunale, and Belanger, 2002).Moreover, females are not particularly familiar with how internet works and therefore would spendless time there This explains a lower likelihood of female to conduct online purchases andpayments (Miyazaki and Fernandez, 2001; George, 2002) However, this is not always the case.Study showed that female may generally make fewer accesses to Internet and online purchasesbecause they are always certain on what items to purchase (Black, 2005)

Differences among male and female on how they perceived e-banking services are alsoapparent Male customers utilize e-banking due to its in-depth technical involvement whilst femalecustomers utilize it just because it is faster and more accurate (Olalekan, 2011) In addition, JoséLiébana-Cabanillas, Sánchez-Fernández and Muñoz-Leiva (2014) used gender as the moderatingeffects from TAM (PEOU and PU) to the attitude towards the use of mobile payment system

This study is trying to extend the existing literatures by analyzing whether there is adifference between male and female in their intention to use e-payment systems in relation to theirintention to purchase goods online Therefore, the proposed hypothesis is:

H 7 : Intention to use e-payment systems differs between male and female.

Overall, the proposed conceptual framework is as below:

Figure 2.1: Conceptual Framework

20

PerceivedUsefulnessTechnology

Acceptance Model

Perceived Ease ofUse

Intention to use payment system foronline purchases

e-SecurityPerceived Risk

Trust

Quality of InternetConnection

Subjective Norm(social influence)

Gender

2.8 Conclusions

This chapter covers the background and literatures which will be used to study the intention

to use e-payment systems for online shopping among young adults in Malaysia The backgroundhighlighted the importance to study this field from the economic side Secondly, definitions andtypes of e-payments are focused on the payment made after purchases via web browser using e-payment instruments (i.e: debit/credit cards, e-cheques, or e-money.) It is completed withillustration of previous studies on e-payment The literatures concentrate on both e-banking andonline shopping, but the emphasis is still on the e-payment systems from these two activities It isused because e-payment is the subset of e-banking and this study is centred on online shopping.Moreover, the literatures will provide the factors used, which are TAM factors (PEOU and PU),perceived risks (security and trust), quality of internet connection, subjective norms (i.e socialinfluence) which is derived from TRA, and lastly demographic (gender) Based on the literatures,conceptual framework is developed Finally, all of these constructs under conceptual frameworkthat needs to be tested will be elaborated in Chapter 3

CHAPTER 3: RESEARCH METHODOLOGY

This chapter will further discuss on the methodologies used in order to test determinants ofe-payment usage for online shopping It contains the explanation of research design and samplingdesign to conduct the research, measurement assessment to analyze the reliability and validity ofeach construct; complete with its list of initial constructs used and the respective sources, andhypothesis testing/data analysis to see the significance of each independent constructs with thedependent constructs based on research questions and hypothesis discussed in literature review

3.1 Sampling Design

Research strategy in a form of survey is the most appropriate one to be conducted to study

the possible reasons among variables to produce models of relationships Survey strategy providesseveral advantages First, it allows large collection of data from substantial population.Subsequently, survey allows the findings of data collection to be generalized and be representative

of the whole population in a highly economical way rather than collecting data for the actualamount of population Second, it is likely to be easier to explain and understandable by people ingeneral Lastly, it gives the freedom to control the research process (Saunders, Lewis, andThornhill, 2009) Limited capacity/range of measurement unlike any other research strategies andtime constraints due to the dependency of information would be the disadvantages of this researchstrategy (Saunders, Lewis, and Thornhill, 2009)

A mono method quantitative is implemented in this study because this study only focuses

on using one quantitative collection; that is primary data survey in a form of questionnaire withquantitative analysis procedures; that are numerical data, graphs and statistics using SPSS program

For this study, a quantitative method of survey in a form of questionnaire was conducted The

anonymous questionnaire was intended to target 150-200 respondents to be the representative of thepopulation From the sum of 200 questionnaires distributed, 198 were collected with 2 beinguncollectible Since this study only specified the respondents who have the actual experience onusing online payment transactions to conduct their online shopping, the final sum of respondentsthat were selected after the screening was 165; with the remaining 33 invalid respondents; i.e: noexperience The questionnaires were administered face to face; as such the response rate wassatisfactorily high And finally, all of the 200 questionnaires were randomly distributed aroundKlang Valley area to ensure the randomness and avoid bias issue possible

22

For a faster responds due to limited time constraints during this study, non-probability

method in a form of purposive sampling techniques was used Known as judgemental sampling,

this allows a judgement for specific cases that will best answer the research questions and meet theobjectives (Saunders, Lewis, and Thornhill, 2009) It is normally best to use with small samples;and this sample should be considered as the “representative” of the population As maximum of 200respondents were required, purposive sampling was also used for conceptual generalisability of thepopulation (Battaglia, 2008) Additionally, mall-intercept data collection technique was used for abetter convenient

The questionnaire consisted of 3 sections with a mixture of both open and close endedquestions The first section was for the determinants of e-payment for online shopping usage; inwhich the questions were built based on adaption of several existing theories and self-developmentfor the particular study Each determinant had 6-9 items to measure with five-point Likert scalefrom strongly disagree (1) to strongly agree (5) Second section was trying to evaluate therespondents’ intention to use online payment transaction with 7 items and the same five-pointLikert scale Lastly, third section was for demographic purposes such as gender, age, programme ofstudy, and income/allowance level; with addition of respondents’ experience using e-paymentsystems for online shopping

The reliability test used was focused on internal construct consistency It involves the

correlation between each question in the questionnaire along with the responses Hence, theconsistency of responses across all questions is highly measured (Saunders, Lewis, and Thornhill,

2009) One of the most frequently used internal consistency method is Cronbach’s Alpha The

reliability coefficient for it generally ranges between 0 and 1 A closer coefficient to 1 indicates thegreater internal consistency of items in the questionnaire (Gliem and Gliem, 2003) Ranges ofacceptable Cronbach’s Alpha that is widely used to measure the reliability by Nunnally (1978) are

Table 3.1: Cronbach’s Alpha Value

The first step of testing the validity of questionnaire is by face validity Face validity looks

at how valid the measurements are based on its face value; that is when it appears logically to

reflect accurately what it was intended to measure (Saunders, Lewis, and Thornhill, 2009) For this

study, measurement of face validity was done through adapting the existing theories and extensive

review of literature as listed below:

Constructs Initial number of questions Sources: Adaption

Teoh et al., 2013

Abrazhevich, 2004Pousttchi and Wiedemann, 2005Kim et al., 2010

Abrazhevich, 2004Kim et al., 2010Teoh et al., 2013

Sánchez-Fernández and Muñoz-Leiva, 2014

Table 3.2: Development of questionnaire

After face validity was ensured, the next step of the test was determining the construct

validity Construct validity is used to measure the actual constructs that are intended to be measured

before running the factor analysis The measurements used here are Kaiser-Meyer-Okin measure of

sampling adequacy (KMO) and Bartlett’s Test of Sphericity

KMO signifies the ratio of the squared correlation to the squared partial correlation

between its respective variables; whether individual or multiple variables Moreover, KMO can

signal in advance whether the sample size is large enough to reliably extract factors The closer its

range to 1, patterns of correlations become more solid Therefore, the factors are distinct and

reliable when the factor analysis is done (Field, 2009) The acceptable KMO values are as follows:

24

Table 3.3: KMO’s value range Source: Field, 2009 pp 647

Bartlett’s Test measures whether the data comes from multivariate normal distribution withzero covariance The significant level of Bartlett’s test is when the associated is less than 0.05 Thismeans there are significant correlations among variables Ultimately, If KMO value is at most inbetween 0.7 – 0.8, and Bartlett’s Test of Sphrecity shows a p-value of less than 0.05, the data aresuitable for factor analysis (Hair et al., 2010)

Factor analysis is used to test the construct validity; that is to identify the core structure in the analysis (Hair et al., 2010) For this study, it is done through factor extraction, with principal component method due to data reduction as a main concern that considers total variance derived.

The interpretability of factors can be improved through rotation, which is used to reduce

ambiguities of each factor Rotation method used is Direct Oblimin to see if there is a possibility that the constructs may be linked to each other Once it is tested, eigenvalues must be greater than 1

to show valid constructs (Hair et al., 2010) Next, individual items from the constructs are tested

based on factor loading with Pattern Matrix The significant level for factor loading varies within

the respective sample size Regardless of the stated satisfactory level, a factor loading of ±0.50 peritems for 165 respondents is used for a rigid result (Hair et al., 2010) Afterwards, the final

reliability test will be conducted For items that are both valid and reliable, summated scales of the

items were conducted for data analysis

The next step was Hypothesis Testing/ Data Analysis, which would be based on research

questions Therefore, the respective constructs along with its hypotheses were listed down below:

Dependent Constructs

Intention to Use online payment transactions (EPAYMENT)

Scale

Independent Constructs

Perceived Usefulness (USE)Perceived Ease of Use (EASE)

Scale (H1- H2); will answer RQ1Security (SECURITY)

Trust (TRUST)

Scale(H3 – H4); will answer RQ2Quality of Internet Connection

Social Influence (SOCIAL) Scale (H6); will answer RQ4

RQ5

Table 3.4: Constructs for Hypothesis Testing

3.3 Multiple Linear Regression (MLR)

Since all the independent constructs except GENDER are continuous data and the

dependent construct (EPAYMENT) is also continuous, Multiple Linear Regression was conducted.

MLR assesses the strength of the relationship between one dependent variable and two or moreindependent variables (Saunders, Lewis, and Thornhill, 2009) In this study, the relationships to beassessed are e-payment usage factors (determinants) and intention to use e-payment system foronline shopping Thus, MLR is used to test H1 – H6 based on RQ1 – RQ4 In respect to theconstructs, the regression equation is:

EPAYMENT = α+β1USEi+β2EASEi+β3SECURITYi+β4TRUSTi+β5INTERNETi+β6SOCIALi+ε

3.3.1 Test of Assumption

There are three tests of assumption for multiple linear regressions to test the relationshipbetween the independent and dependent constructs, which are linearity, constant variance(homoscedasticity), and normality Firstly, the dependent construct must show a normally

26

distributed skewt Next, the scatter plot of residual/error term should not show any fixed pattern.And lastly, the normal P-P plot should lie very closely to the 45o line from the origin.

3.3.2 Model Fit

Coefficient of determination (R2) and Adjusted R2 are used to assess the goodness of fit in theregression equation and see the proportion of variation in a dependent variable that can beexplained by the independent variable(s) The more detailed outcome of variation could be seen inthe Adjusted R2 as it takes into account on the number of independent variables along with samplesize in the regression equation Thus, it avoids overestimating the impact of adding an independentvariable on the amount of explained by the estimated regression equation (Saunders, Lewis, andThornhill, 2009) Moreover, the significance level provided in the regression must show of a p-value <0.05 to reject the null hypothesis Another important determinant is the analysis of variance(ANOVA), in which F-value along with its significance should be taken into consideration to findout relationship between constructs occurred by chance

3.3.3 Test of Significance

Table of Coefficient was used to assess the test of significance Firstly, t Value measures

the significance of the partial correlation of the variable reflected in the regression coefficient; that

is the coefficient is not equal to zero Regression coefficient (B-value) reflects the change in the

dependent measure for each unit change in the independent variable The standardized coefficient

(β)) shows the level of importance in each significance of independent variable Lastly, collinearity statistics reflects the impact of collinearity on the independent variables in the regression equation

(Hair et al., 2010)

3.4 Independent sample t-test

Meanwhile for gender, since it is a discrete data with only two groups, Independent T-test

spread of scores For this study, independent t-test was used to test the relationship betweenGENDER and EPAYMENT (H7 addressed in RQ5).

3.4.1 Test of Assumption

Before testing the significance level, test of assumptions was conducted using Levene’s Test of Equality of Variance Levene’s test tests the null hypothesis that the variances in different

groups are equal (Field, 2009) The hypotheses for equal variance are:

H0: Equal variance is assumed (σ2 male = σ 2 female)

H1: Equal variance is not assumed (σ2 male ≠ σ 2 female)

If the F-statistics displays/results in a p-value of greater than 0.05, H0 is supported.Therefore, the first line of the SPSS output in regards to the Levene’s test would be read as theequal variance is assumed However, if the result states otherwise, (i.e: F-statistics in p-value of

<0.05), H0 is rejected and the second line would be read as the equal variance is not assumed

3.4.2 Test of Significance

The test would show a significant result; (i.e the probability of difference occurred by only

chance is relatively low) if the data has a large t statistics with a probability (p-value) of less than

0.05 (Saunders, Lewis, and Thronhill, 2009) Thus, H0 is rejected The hypotheses for test ofsignificance are:

H 0 : μ male = μ female (there is no difference between male and female in the intention to use payment for online shopping)

e-H 1 : μ male ≠ μ female (there is a difference between male and female in the intention to use payment for online shopping)

e-28

3.5 Conclusion

This chapter covers the research methodologies used to conduct the study’s test analysis.Firstly, this research implemented a quantitative method, in which a survey in hand-outquestionnaire form is utilized as this study tried to examine the respondents’ intention to use e-payment Purposive sampling with mall-intercept data collection technique is used due to limitedtime constraints and better convenient The total respondents after the last screening are 165 Forreliability test, Cronbach’s Alpha is used Face validity and construct validity (using KMO andBartlett’s Test of Sphericity) are conducted before running factor analysis Factor analysis is runwith factor loading in Pattern Mix After the reliability and validity is ensured, Hypothesis testing isconducted for all constructs MLR is implemented for H1-H6 along with its respective regressionequation, whilst independent t-test is for H7 in which test of assumptions and significance are testedfor both Afterwards, the results based on measurements above will be explained in Chapter 4

CHAPTER 4: ANALYSIS FINDINGS

This chapter will further highlight the findings of the measurement assessment and dataanalysis as discussed in Chapter 3 It consists of enlightenment for descriptive statistics,justification of reliability and validity tests for both test of assumptions and tests of significance,results of factor analysis, and justification of hypothesis testing / data analysis

4.1 Descriptive Statistics

From the total of 200 questionnaires, the response rate was 99% However, based on thefindings for demographic, only 83% rate with a total of 165 respondents was taken as this studyonly specify on e-payment usage Therefore, the demographic presented are all online paymentusers The majority of the respondents are practically even between the range of 18-23; with 46.7%are 18-20 years old and 43% are at the age of 21-23 The remaining 10.3% consist of age range of24-26 (7.9%) and >26 (2.4%) As for the gender, female are the majority of respondents, with therate at 66.1% whilst male are only at 33.9% In terms of income/allowance level, the respond rate isonly 99.4% as there is 1 missing data 27.3% of the respondents earned <RM500 per month, 33.9%earned between RM500 and RM1000 per month, 29.1% earned between RM1001 and RM2000 permonth, and lastly, 9.1% earned >RM2000 per month Most of the respondents are majoring inBusiness, with 47.9% rate Since the category for programme of study is limited, 34.5% of themchose “others”; in which they were specified as several different major such as architecture, design,mass communication, pharmacy, etc The remaining ones are 12.1% of hospitality students and5.5% of engineering students Out of 165 respondents, 89.7% of them use credit/debit card as themean of online transaction payment The other 9.7% are split between E-Money/Cash users andelectronic cheque users

Table 4.1: Demographic Profile

Table 4.2: KMO Value and Bartlett’s Test

KMO and Bartlett's Test

Chi-Square

3369.195

4.2.1 Reliability for independent and dependent constructs

The highest value of reliability is INTERNET, with an excellent value of 0.886; followed

by TRUST and SECURITY with alpha of 0.87 and 0.866 respectively The remaining independentand also the dependent constructs have Cronbach’s Alpha above 0.7 except for USE, which wasonly at 0.683 Therefore, all of the constructs used are satisfactorily reliable and there is an internalconsistency between the questions and the responses

Table 4.3: Reliability Test

Constructs Scale Mean Scale Standard Deviation Cronbach’s AlphaQuality of internet

4.2.2 Validity of the constructs

Based on the result of factor analysis, factor extraction with principal component methodwas done in order to obtain the best linear combination of variables From the initial number ofitems for each constructs, it could be seen that there are a substantial extraction for perceivedusefulness and an extraction for both trust and perceived ease of use After the extraction, the datashows Eigenvalue greater than 1 for all constructs, with the highest of 8.761 for quality of internetconnection and the lowest of 1.501 for perceived usefulness Therefore, it can be concluded that allfactors are now significant

As for the percentage of cumulative variance explained, the total variance from the datauntil the last extracted construct is 56.80% Total variance of 60% (or even less at times) for asocial science is considered as satisfactory (Hair et al., 2010) Therefore, this study has anacceptable cumulative variance explained and is practically significant

32